Still Feeding That Baby

Still feeding that baby

A cluster randomised controlled trial, with training of general practice nurses, to

examine motivational support for ongoing breastfeeding.

Megan Elizabeth Elliott-Rudder

MBBS DRANZCOG GradDipRuralGP FRACGP CertIVBreastfeedEd(Comm,Couns)

Submitted for the degree of Doctor of Philosophy Rural Clinical School, Faculty of Medicine The University of New South Wales Australia

30 March 2012

ORIGINALITY STATEMENT

‘I hereby declare that this submission is my own work and to the best of my knowledge it contains no materials previously published or written by another person, or substantial proportions of material which have been accepted for the award of any other degree or diploma at UNSW or any other educational institution, except where due acknowledgement is made in the thesis.

Any contribution made to the research by others, with whom I have worked at UNSW or elsewhere, is explicitly acknowledged in the thesis.

I also declare that the intellectual content of this thesis is the product of my own work, except to the extent that assistance from others in the project's design and conception or in style, presentation, and linguistic expression is acknowledged.’

Signed ……………………………………………......

Megan Elliott-Rudder

Date ……………………………………………......

THE UNIVERSITY OF NEW SOUTH WALES Thesis/Dissertation Sheet

Surname or Family name: Elliott-Rudder

First name: Megan Other name/s: Elizabeth

Abbreviation for degree as given in the University calendar: PhD

School: Rural Clinical School Faculty: Medicine

Title: Still feeding that baby: A cluster randomised controlled trial, with training of general practice nurses, to provide motivational support for ongoing breastfeeding.

Abstract 350 words maximum: AIM: To increase exclusive breastfeeding and any breastfeeding of infants aged four months and six months compared to controls and to evaluate trial processes. INTRODUCTION: Despite its known importance, continuation of breastfeeding in Australia is lower than recommended. Cross-disciplinary research identifies the dilemma that mothers face: they are questioned regarding when, where, how and why they are breastfeeding. Protection and promotion of breastfeeding are inadequate. This complex interpersonal behaviour needs more support. Systematic literature review suggests that training in breastfeeding management and counselling skills, within a theoretical framework, are needed for effective health professional support. METHOD: Using a motivational interviewing framework The Conversation Tool flowchart was developed. Ten hours of training prepared the intervention practice nurses to use the Tool during infant immunisation appointments, in a cluster randomised controlled trial. Independent mail recruitment and telephone data collection for breastfeeding outcomes were conducted with mothers. Statistical analysis was conducted in STATA, adjusted for clustering. Periodic telephone contact was maintained with practice nurses. RESULTS: The trial involved 15 general practices, 19 intervention practice nurses, and 330 mothers. Training improved practice nurse breastfeeding knowledge (t 7.27(14), p=.000 2-tailed) and attitudes (t 6.25(14), p=.000 2-tailed). Full breastfeeding OR 1.95 (95%CI 1.03-3.69) p=0.041 and exclusive breastfeeding OR 1.88 (95%CI 1.01-3.50) p=0.047 were increased at four months (24 hour recall) compared to control, adjusted for return to work/study plans. Duration of breastfeeding was not significant compared to controls. Mixed methods process evaluation included analysis of 43 telephone interviews with practice nurses. Despite community resistance, workplace restrictions, and their own learning curves, support for ongoing breastfeeding was satisfying for practice nurses and mostly non-threatening. DISCUSSION AND RECOMMENDATIONS: This successful clinical translational research demonstrates a rural preventive health intervention and improved breastfeeding outcomes. The motivational interviewing framework and this training model are appropriate. Policy commitment to address structural barriers is needed. Future research is awaited on support for ongoing breastfeeding with women from disadvantaged groups and on improved maternal access to support.

Declaration relating to disposition of project thesis/dissertation

I hereby grant to the University of New South Wales or its agents the right to archive and to make available my thesis or dissertation in whole or in part in the University libraries in all forms of media, now or here after known, subject to the provisions of the Copyright Act 1968. I retain all property rights, such as patent rights. I also retain the right to use in future works (such as articles or books) all or part of this thesis or dissertation. I also authorise University Microfilms to use the 350 word abstract of my thesis in Dissertation Abstracts International (this is applicable to doctoral theses only).

……………………………………………… …………………………..…………… ………………. Signature Witness a Date

The University recognises that there may be exceptional circumstances requiring restrictions on copying or conditions on use. Requests for restriction for a period of up to 2 years must be made in writing. Requests for a longer period of restriction may be considered in exceptional circumstances and require the approval of the Dean of Graduate Research.

FOR OFFICE USE ONLY Date of completion of requirements for Award:

THIS SHEET IS TO BE GLUED TO THE INSIDE FRONT COVER OF THE THESIS

3 [this page is blank]

ACKNOWLEDGEMENTS

This thesis is dedicated to my mother Trixie Rudder (1937-2008). She loved learning, and encouraged each of her children to continue their education.

Research I thank my supervisors, Professor Louis Pilotto and Associate Professor Ellen McIntyre, for mentoring me with such skill and warmth. I have benefitted enormously from their expertise, generosity, and unfailing faith in my ability to complete this doctorate.

I thank my PhD colleagues, Dr Ginny Sargent and (soon to be Dr) Jane Anderson-Wurf for their academic support in our isolated rural campus. I cherish their friendship.

I thank those who guided and supported my research, by sharing ideas, insights, and wise advice, and those who contributed time and effort so that recruitment, meetings, and writing could be achieved. Dr Lisa Amir, Dr Nina Berry, Dr Wendy Brodribb, Dr Karleen Gribble, Joel Porter, Dr Julie Smith and other researchers and experts have responded kindly to my questions over the past five years. My work stands on the shoulders of so many others. I thank friends and medical colleagues for their interest in my path of research capacity building.

I thank Jen Byrne and Australasian Lactation Courses for providing the kit used in practice nurse training, Breastfeeding Management in a Baby Friendly Health Service: An 8 hour structured course for health professionals working with mothers and babies in maternity settings 5th Ed.

I thank each of the intervention practice nurses; the outcomes of this study are due to their work. I am grateful for their enthusiasm, their interest in learning, and their care for each patient who walks through the door. The valuable role of practice nursing deserves respect and support from the rest of the team.

I thank all the women who participated. Without the generous gift of their time and interest, this study could not have been conducted. I wish them long and fulfilling relationships with their children.

Personal I thank past and current fellow volunteers in the Australian Breastfeeding Association (formerly Nursing Mothers') for their enthusiastic, irreverent, and talented support for breastfeeding and parenting. I am grateful for our shared stories of mothering and mother-to-mother support. From the tears and embraces of my early motherhood to the typing of my thesis, their welcome continues to motivate me.

Finally, I thank my husband, four children, and my husband's parents for their tolerance and interest over the past five years of my research. I will always be grateful for my husband Glenn's commitment to fathering, his enjoyment of our family, and his constant support in my own journey of mothering (and breastfeeding) our children.

Grants 2009 UNSW Postgraduate Student Support Scheme for conference presentation 2007 to 2009 Department of Health and Ageing PhD scholarship

DISSEMINATION

Peer-reviewed publications Elliott-Rudder M. Researcher networking drives change: an autoenthnographic narrative analysis from medical graduate to primary health researcher. Australian Journal of Primary Health 2010;16:108–115 (Two citations)

Smith J, Dunstone M, Elliott-Rudder M. Health professional knowledge of breastfeeding: Are the health risks of infant formula feeding accurately conveyed by the titles and abstracts of journal articles? Journal of Human Lactation 2009;25(3): 350-58. (Seven citations)

Smith JP, Dunstone MD, Elliott-Rudder ME. ‘Voldemort’ and health professional knowledge of breastfeeding – do journal titles and abstracts accurately convey findings on differential health outcomes for formula fed infants? Canberra: Australian Centre for Economic Research on Health, Working Paper No. 4 2008; December. (Six citations)

Other publications Elliott-Rudder M. Why don’t more women breastfeed? In: Zest for Life. Wagga Wagga: Riverina Division of General Practice and Primary Health, Winter 2008

Elliott-Rudder M. Support for ongoing breastfeeding: A randomised clinical trial endorsed by the APNA Policy & Research Committee. Primary Times 2008;1:16-17

Invited workshops 19 and 26 February 2012: ACT Breastfeeding Strategy/ACT Medicare Local, Support for ongoing breastfeeding. Two 5-hour training workshops. (Eight practice nurse participants) 21 and 28 February 2012: ACT Breastfeeding Strategy/ACT Medicare Local, Support for ongoing breastfeeding. Two 5-hour training workshops. (11 practice nurse participants)

Invited Presentations 23 March 2010: Global Online Lactation Discussion (GOLD10). Presentation title: Still breastfeeding that baby? How, what and when to ask 18 June 2009: Breastfeeding Information Evening, Wodonga Regional Health Service, VIC. Presentation 1 title: Taking on board new skills, support for ongoing breastfeeding Presentation 2 title: A systematic literature review, support for ongoing breastfeeding 1 May 2009: Australian Practice Nurses Association Visionary Conference, Melbourne VIC. Presentation title: Taking on board new skills, support for ongoing breastfeeding Poster title: Discovery, exploration and uptake of community resources, support for ongoing breastfeeding 13 September 2007: Wodonga Hospital Perinatal Meeting, Wodonga VIC. Presentation title: Enhancing women’s motivation to breastfeed, is change possible? 5 August 2006: NSW Lactation College for the Annual Conference, Sydney NSW. Presentation title: Motivation to breastfeed: Our choice of words, her choice of action, is change possible? 26 March 2004: NSW Lactation College rural seminar, Wagga Wagga NSW. Presentation title: Enhancing women’s motivation to breastfeed Other Presentations 5 September 2009: Australian Lactation Consultants Association Conference, Melbourne VIC. Presentation title: Development of a Conversation Tool: Support for Ongoing Breastfeeding 15 July 2009: General Practice & Primary Health Care Conference, Melbourne VIC. Poster title: Research networking drives change, a narrative analysis 5 October 2007: RACGP for 50th Annual Scientific Convention, Sydney NSW. Presentation title: Are you still feeding that baby? The role of General Practice in supporting ongoing breastfeeding. A literature review 2 August 2007: Australian Breastfeeding Association Hot Milk! National Conference, Melbourne VIC. Presentation title: Motivation to breastfeed: Our choice of words, her choice of action, is change possible? Innovation The Conversation Tool v3, registered with NewSouth Innovations 2012

Above and right: 2012 workshops, Canberra. Lower right: 2009 APNA Conference.

ABSTRACT

AIM: To increase exclusive breastfeeding and any breastfeeding of infants aged four months and six months compared to controls, and to evaluate trial processes. INTRODUCTION: Despite its known importance, continuation of breastfeeding in Australia is lower than recommended. Cross-disciplinary research identifies the dilemma that mothers face: they are questioned regarding when, where, how and why they are breastfeeding. Protection and promotion of breastfeeding are inadequate. This complex interpersonal behaviour needs more support. Systematic literature review suggests that training in breastfeeding management and counselling skills, within a theoretical framework, are needed for effective health professional support.

METHOD: Using a motivational interviewing framework The Conversation Tool flowchart was developed. Ten hours of training prepared the intervention practice nurses to deliver the Conversation Tool intervention during infant immunisation appointments, in a cluster randomised controlled trial. Independent mail recruitment and telephone data collection for breastfeeding outcomes were conducted with mothers. Statistical analysis was conducted in STATA, adjusted for clustering. Periodic telephone contact with practice nurses enabled mentoring and mixed methods process evaluation

RESULTS: The trial involved 15 general practices, 19 intervention practice nurses, and 330 mothers. Training improved practice nurse breastfeeding knowledge (t 7.27(14), p<0.001 2-tailed) and attitudes (t 6.25(14), p<0.001 2-tailed). Full breastfeeding OR 1.95 (95%CI 1.03-3.69) p=0.041 and exclusive breastfeeding OR 1.88 (95%CI 1.01- 3.50) p=0.047 were increased at four months (24 hour recall) compared to control, adjusted for return to work/study plans. Duration of breastfeeding was not significantly different from controls. Process evaluation included 43 practice nurse interviews; despite community resistance, workplace restrictions, and their own learning curves, support for ongoing breastfeeding was satisfying for practice nurses and mostly non- threatening.

DISCUSSION AND RECOMMENDATIONS: This successful clinical translational research demonstrates a rural preventive health intervention and improved breastfeeding outcomes. The motivational interviewing framework and this training model are appropriate. Policy commitment to address structural barriers is needed. Future research is awaited on support for ongoing breastfeeding with women from disadvantaged groups and on improved maternal access to support.

ABBREVIATIONS AND GLOSSARY

1, 2, 3, 4, 5, 6, 12 Numerals used for the age of infants in months, in this thesis 24-hour recall Current practice dietary assessment by a trained interviewer regarding all foods and drinks consumed in the previous 24 hours. ABA Australian Breastfeeding Association AF Artificial feeding, infant formula as a breastmilk substitute ALC Australasian Lactation Courses AMS Aboriginal Medical Service, community-controlled health organisation ANHS Australian National Health Survey APNA Australian Practice Nurses Association BF Breastfeeding, breastfed, includes ever fed breastmilk, may substitute any feeds with non-human milk, solids or other drinks BFHI Baby Friendly Health Initiative CATI Computer assisted telephone interviews Division Division of general practice, within Australian General Practice Network, linking GPs in a local geographical region EBF Exclusive breastfeeding, only additional medicines FBF Full breastfeeding, only additional medicines/water-based drinks Formula Non-human milk, commercially processed with additives, usually a powdered unsterile product, reconstituted for infant feeding General practice Family medical practice with unrestricted non-referred access Gestation Weeks of pregnancy, from last menstrual period to term (40) GP General practice/Medical doctor's office; or general practitioner/medical doctor/family physician HREC Human Research Ethics Committee Infant <12 months old. "Children" are 12 months or older NHMRC National Health and Medical Research Council, Australia NSW New South Wales, Australia NSW Health The Ministry of Health, New South Wales state government PN Practice Nurse; registered nurse employed in general practice Primary care Primary health care, community-based vs. tertiary, hospitals-based Since birth recall Retrospective dietary assessment regarding all foods and drinks ever consumed since birth Solids Soft, semi-solid, or solid food including thickened fluids, spoon-fed or handheld for infant feeding. Does not include water-based drinks Substitute Food or drink given as a replacement for breastmilk The Code World Health Organisation Code of Marketing for Infant Formula UK United Kingdom UNSW The University of New South Wales USA United States of America WHO World Health Organisation WIC Supplemental Nutrition Program for Women, Infants, and Children for disadvantaged women in USA

TABLE OF CONTENTS Originality statement ...... 2 Thesis/Dissertation Sheet ...... 3 Acknowledgements...... 5 Dissemination ...... 7 Abstract...... 10 Abbreviations and glossary...... 11 List of tables...... 20 List of figures...... 21 THESIS INTRODUCTION ...... 24 Research question ...... 24 Thesis plan...... 24 BACKGROUND LITERATURE REVIEW...... 26 1.1 THE IMPORTANCE OF BREASTFEEDING ...... 26 1.1.1 Importance as a biological imperative...... 26 1.1.2 Importance as a health issue ...... 28 Risks of formula use ...... 30 1.1.3 Importance for environmental and economic outcomes ...... 30 1.1.4 Importance as a human right ...... 31 1.1.5 Definition and measurement controversies...... 32 Measurement and analysis...... 33 Exposure to breastmilk substitutes...... 33 Current use of breastmilk substitutes ...... 34 Recall time period...... 34 Describing the data...... 35 1.1.6 Recommended breastfeeding practices ...... 36 1.2 INFANT FEEDING IN AUSTRALIA ...... 37 1.3 PROTECTION OF BREASTFEEDING ...... 38 1.3.1 Protection against unethical marketing of formula ...... 39 1.3.2 Protection in paid employment...... 42 1.3.3 Protection in public...... 44 1.4 PROMOTION OF BREASTFEEDING ...... 45 1.4.1 Promotion against cultural norms ...... 46 1.4.2 Promotion of the visible use of breasts ...... 48 1.4.3 Promotion and infant behaviour historically ...... 50 1.4.4 Promotion with a research bias of "best" and "benefits" ...... 52 1.4.5 Promotion methods of health professionals...... 53

12 Knowledge and attitudes...... 53 Communication skills ...... 55 1.5 SUPPORT FOR BREASTFEEDING ...... 55 1.5.1 Personal support needs to be offered ...... 56 1.5.2 Support for the embodied experience ...... 57 1.5.3 Support from family ...... 59 1.5.4 Support from peer organisations ...... 61 1.5.7 Support for choice ...... 61 1.5.5 Support from health professionals...... 63 1.6 OVERVIEW OF SYSTEMATIC REVIEWS...... 65 1.7 DEVELOPING THE RESEARCH QUESTION ...... 66 1.7.1 The need ...... 66 1.7.2 Choosing the theoretical framework...... 67 1.7.3 Choosing the setting and methods...... 67 1.7.4 A potential role for practice nurses...... 67 The nature of practice nursing ...... 68 Challenges to scope of practice for practice nurses ...... 69 Enhanced nursing roles in preventive health care ...... 69 1.8 CONCLUSION TO CHAPTER 1 ...... 70 SYSTEMATIC LITERATURE REVIEW OF RANDOMISED CONTROLLED TRIALS.. 71 INTRODUCTION...... 72 The need for breastfeeding support ...... 72 Intervention provider training...... 72 2.1 AIM ...... 74 2.2 METHOD...... 75 2.2.1 Search strategy ...... 75 2.2.2 Inclusion criteria ...... 76 2.2.3 Exclusion criteria ...... 77 2.2.4 Analytic approach...... 77 Organisation of results ...... 77 Assessment of quality ...... 78 2.3 RESULTS...... 78 2.3.1 BFHI provider training ...... 80 2.3.2 Trained lactation consultants...... 82 2.3.3 Other specified provider training...... 85 2.3.4 Provider training not specified ...... 88 2.4 DISCUSSION OF SYSTEMATIC LITERATURE REVIEW...... 98

2.5 CONCLUSION TO CHAPTER 2...... 100 METHODOLOGY ...... 102 3.1 RANDOMISED CONTROLLED TRIAL PROCESSES ...... 104 3.1.1 Objectives and outcome measures...... 104 3.1.2 Sample size calculations...... 105 3.1.3 Participant eligibility ...... 105 Exclusion criteria ...... 105 Figure 3.2 The CONSORT methods flowchart ...... 106 3.1.4 Participant recruitment process during pregnancy...... 107 Poster ...... 107 Mail questionnaire ...... 107 Postcard reminder ...... 108 3.1.5 Participant local birthing hospitals...... 108 3.1.6 Midwifery research assistance...... 109 Duties ...... 109 Data management ...... 109 3.1.7 Setting...... 110 Rurality ...... 110 3.1.8 General practice recruitment...... 110 3.1.9 Randomisation ...... 112 3.1.10 Statistical methods...... 113 3.2 INTERVENTION PROCEDURES...... 113 3.2.1 Literature review: motivational interviewing for health promotion ...... 114 Background MI literature...... 114 Systematic MI literature search with structured review...... 115 Results...... 116 3.2.2 The Conversation Tool to guide the intervention ...... 119 Development of the Tool...... 119 Figure 3.4 The Conversation Tool (Version 2)...... 120 Format of the Tool ...... 121 Engagement panel ...... 121 Response and action panel...... 121 Moving out panel...... 122 Testing of the Tool ...... 123 3.2.3 Training workshop aims ...... 123 3.2.4 Resources provided at the training workshop...... 124 Breastfeeding Confidence booklet ...... 124

14 Starting family foods leaflet...... 125 Balancing breastfeeding and work leaflet...... 126 Local services directory ...... 126 Posters ...... 127 3.2.5 Delivery of the training sessions...... 127 3.2.6 Curriculum for the training sessions ...... 130 Breastfeeding management...... 130 Communication skills...... 131 Motivational interviewing...... 133 Mother-to-mother support...... 134 Reflective Practice...... 136 3.2.7 Piloting the training delivery and curriculum...... 138 Content and delivery...... 138 Summary of evaluation...... 138 3.2.8 Alternative brief format training ...... 139 3.2.9 Supporting practice nurse to hold new conversations...... 139 Workplace visits ...... 140 Workplace telephone calls ...... 140 Evening education update meetings...... 140 3.3 DATA COLLECTION...... 141 3.3.1 Primary outcome measures...... 141 3.3.2 Secondary outcome measures...... 142 3.3.3 Potential confounders...... 142 3.3.4 Development of data collection instruments...... 143 3.3.4.1 Mothers’ baseline questionnaire...... 143 3.3.4.2 4 month and 6 month breastfeeding outcomes (CATI) ...... 143 3.3.4.3 Practice nurse knowledge and attitudes quiz...... 144 3.3.4.4 Practice nurse semi-structured interviews...... 145 3.5 MIXED METHODS ...... 146 3.5.1 Use of mixed methods in this thesis...... 146 3.5.2 Use of mixed methods in practice nurse data ...... 147 3.5.3 Collection of discourse data ...... 147 3.5.4 Analysis of discourse data...... 148 3.6 CONCLUSION TO CHAPTER 3 ...... 150 INNOVATION...... 151 ETHICS ...... 152 ENGAGEMENT...... 153

RESULTS OF THE RANDOMISED CONTROLLED TRIAL...... 155 4.1 INTRODUCTION TO CHAPTER 4 ...... 156 4.2 RESPONSE RATE ...... 156 4.3 ASSIGNMENT ...... 158 4.4 PARTICIPANT FLOW...... 158 4.5 DATA ...... 159 4.5.1 Baseline data ...... 159 4.5.2 Demography ...... 159 4.5.3 Maternal environment ...... 160 4.5.4 Planning ...... 160 4.5.5 Perinatal factors ...... 162 4.6 ANALYSIS ...... 162 4.6.1 Intention to treat ...... 162 4.6.2 Adjustment for clustering ...... 163 4.6.3 Outcome measures...... 163 4.6.4 Points of measurement ...... 164 4.6.5 Definitions for reporting...... 165 Any breastfeeding...... 165 Full breastfeeding ...... 165 Exclusive breastfeeding...... 166 4.7 OUTCOMES ...... 166 4.7.1 Any breastfeeding at 4 months ...... 166 4.7.2 Full breastfeeding at 4 months (24-hour recall) ...... 167 4.7.3 Exclusive breastfeeding at 4 months (24-hour recall)...... 167 4.7.4 Exclusive breastfeeding to 4 months (since birth recall)...... 168 4.7.5 Any breastfeeding at 6 months ...... 169 4.7.6 Full breastfeeding at 6 months (24-hour recall) ...... 169 4.7.7 Exclusive breastfeeding at 6 months (24-hour recall)...... 170 4.7.8 Exclusive breastfeeding to 6 months (since birth recall)...... 170 4.8 SUMMARY OF RCT RESULTS...... 171 4.8.1 Primary outcome measures ...... 171 4.8.1.1Duration ...... 171 4.8.1.2 Exclusivity: 4 month data collection ...... 171 4.8.1.3 Exclusivity: 6 month data collection ...... 172 4.8.2 Comparison of outcomes for three alternate exclusivity definitions ...... 173 4.8.3 Use of infant formula...... 173 4.8.4 Outcome of background analyses ...... 173

16 4.8.5. Comparison to national data ...... 174 PROCESS EVALUATION...... 176 5.0.1 Results of discourse data collection ...... 177 5.0.2 Results of descriptive data collection ...... 179 5.0.3 Results of training effects data collection ...... 179 5.0.4 Structure of the discourse data analysis...... 179 5.1 PART 1 FEASIBILITY ...... 180 5.1.1 Summary of practice nursing literature analysis...... 180 5.1.2 Thematic analysis: context ...... 181 5.1.3 Nursing context: providing care...... 181 Patient care ...... 181 Working ...... 182 Research...... 183 5.1.4 Mothering context: work ...... 184 Becoming a mother ...... 184 Workload ...... 184 Accessing support...... 185 Research...... 186 5.1.5 The community context: out there ...... 188 5.1.6 General practice context: part of a team ...... 189 Quality nursing ...... 189 Supporting the nurse...... 190 5.1.7 Research context: to use and to adapt...... 191 Engaging, opening the conversation...... 192 Coloured boxes for motivation...... 193 Resources, moving out from the conversation...... 194 5.1.8 Learning context: trained in whatever we do...... 196 New knowledge, skills, resources ...... 196 Poor health professional knowledge or skills ...... 196 5.2 PART 2: FIDELITY ...... 198 5.2.1 Personal background descriptive analysis ...... 198 5.2.2 Thematic analysis: changes in discourse over time ...... 200 5.2.3 Beginning: "Finding my own pathways"...... 200 5.2.4 Barriers: "I find it hard"...... 202 5.2.5 Communicating: "I'm very picky about how I say it" ...... 204 5.2.6 Acceptance: "So that's how it was with you" ...... 204 5.2.7 Comfort: "I feel a lot more comfortable"...... 205

5.2.8 Routine: "It's just part of the deal" ...... 206 5.3 PART 3: NEW CONVERSATIONS...... 207 5.3.1 Knowledge and attitudes quiz results ...... 207 5.3.2 Knowledge and attitudes quiz summary ...... 208 5.3.3 Thematic analysis: key themes...... 208 5.3.4 Resistance: "There's a stigma out there" ...... 209 5.3.5 New ideas: "Something I wouldn't have thought of before" ...... 212 5.3.6 Skilled questioning: "Using those questions" ...... 213 5.3.7 Motivational support: "You can lead a horse to water..." ...... 214 5.4 ABERRANT DISCOURSE ...... 215 5.4.1 Other barriers...... 215 5.4.2 A negative ABA experience ...... 216 5.5 INFLUENCE OF THE RESEARCHER ...... 216 5.5.1 Potential for bias ...... 216 5.5.2 Evidence of data validity ...... 218 5.5.3 Support through collaboration...... 218 5.6 SUMMARY OF CHAPTER 5 ...... 219 DISCUSSION AND RECOMMENDATIONS ...... 222 6.1 DISCUSSION OF MAIN FINDINGS ...... 222 6.1.1 Exclusivity ...... 222 Recommendations...... 224 6.1.2 Employment ...... 224 Recommendations...... 226 6.1.3 Duration ...... 226 Recommendations...... 228 6.1.4 Opposition to ongoing breastfeeding ...... 228 Recommendations...... 229 6.1.5 Motivational Interviewing theoretical framework ...... 230 Recommendations...... 231 6.1.6 Communication training ...... 231 Recommendation...... 234 6.2 DISCUSSION OF OTHER FINDINGS...... 234 6.2.1 Definitions ...... 234 Recommendation...... 235 6.2.2 Use of formula...... 235 Recommendations...... 236 6.2.3 Knowledge of breastfeeding management ...... 236

18 Recommendation ...... 237 6.2.4 New practice nursing behaviours ...... 237 Recommendation ...... 238 6.3 STRENGTHS ...... 238 6.3.1 Mixed methodology and triangulation...... 238 6.3.2 Clinical translational research...... 239 6.3.3 Generalisation from results...... 240 6.3.3 Involvement of primary care practitioners in research...... 241 Recommendation ...... 242 6.3.4 The Conversation Tool ...... 242 Recommendation ...... 242 6.4 LIMITATIONS...... 243 6.4.1 Potential for bias...... 243 Recommendation ...... 245 6.4.2 Community views ...... 245 Recommendation ...... 246 6.4.3 The nature of general practice...... 246 6.4.4 Mothers' busy lives ...... 249 6.4.5 The mothers' perspective ...... 250 Recommendations ...... 250 6.5 SUMMARY OF FINDINGS...... 251 6.6 SUMMARY OF RECOMMENDATIONS...... 252 REFERENCES...... 254 LIST OF APPENDICES ...... 275 References for the appendices...... 324

LIST OF TABLES

Table 1.1 Breastfeeding for health in developed countries (2) ...... 29

Table 1.2 Any formula feeding, Australia vs. Norway ...... 38

Table 1.3 Any breastfeeding, Australia vs. Norway ...... 38

Table 2.1 Number of support RCTs reviewed, by provider training and outcome ...... 79

Table 2.2 Results of RCT literature review, by training and outcomes ...... 93

Table 3.1 Checklist of items to include when reporting a cluster randomised trial...... 104

Table 3.2 MI studies of health behaviours in parents of children 0-5 years...... 117

Table 4.1 Results of RCT Randomisation...... 158

Table 4.2 RCT participants' baseline characteristics ...... 159

Table 4.3 Baseline maternal environment relating to breastfeeding...... 160

Table 4.4 Baseline planning by the mother for timing of weaning and work...... 161

Table 4.5 Birthing and hospital factors related to breastfeeding...... 162

Table 4.6 Percentage of study population breastfeeding for each outcome...... 163

Table 4.7 Any breastfeeding at 4 months ...... 166

Table 4.8 Full breastfeeding at 4 months (24-hour recall) ...... 167

Table 4.9 Full breastfeeding at 4 months (24-hour recall) adjusted...... 167

Table 4.10 Exclusive breastfeeding at 4 months (24-hour recall)...... 168

Table 4.11 Exclusive breastfeeding at 4 months (24-hour recall) adjusted ...... 168

Table 4.12 Exclusive breastfeeding to 4 months (since birth recall)...... 168

Table 4.13 Any breastfeeding at 6 months ...... 169

Table 4.14 Full breastfeeding at 6 months (24-hour recall) ...... 170

Table 4.15 Exclusive breastfeeding at 6 months (24-hour recall)...... 170

Table 4.16 Exclusive breastfeeding to 6 months (since birth recall)...... 171

Table 4.17 Breastfeeding rates of study population vs. Australian population ...... 174

Table 5.1 Discourse data classification and other data for process evaluation ...... 176

Table 5.2 Practice nurse presence and data collection points...... 178

Table 5.3 Feasibility of the Conversation Tool...... 193

20 Table 5.4 Practice nurse quiz paired samples statistics...... 207

Table 5.5 Practice nurse quiz paired samples test results ...... 208

Table 5.6 New conversations using the Conversation Tool ...... 209

LIST OF FIGURES

Figure 1.1 Definitions of breastfeeding...... 33

Figure 1.2 Estimated number of practice nurses in Australia...... 68

Figure 2.1 Ten Steps to Successful Breastfeeding (BFHI)...... 73

Figure 2.2 Seven Point Plan for Sustaining Breastfeeding in the Community...... 73

Figure 2.3 Example of systematic search strategy...... 75

Figure 2.4 Systematic literature review flow diagram...... 76

Figure 3.1 Chronological sequence of Support for Ongoing Breastfeeding study...... 103

Figure 3.2 The CONSORT methods flowchart...... 106

Figure 3.3 Location of Australian large rural population centres (286)...... 111

Figure 3.4 The Conversation Tool (Version 2) ...... 120

Figure 3.6 Topics in Breastfeeding Confidence...... 125

Figure 3.5 Training workshop sample program...... 128

Figure 3.8 Reflective practice exercise...... 137

Figure 3.9 Triangulation of data sources across thesis chapters ...... 147

Figure 3.10 Interpretive methods for three evaluations of the discourse...... 149

Figure 4.1 The CONSORT flowchart with results...... 157

Figure 4.2 Cohort infant feeding practices by age...... 165

Figure 4.3 RCT Outcomes for full breastfeeding (24-hour recall)...... 172

Figure 4.4 RCT Outcomes for exclusive breastfeeding (24-hour recall) ...... 172

Figure 5.1 Nursing context ...... 183

Figure 5.2 Mothering context...... 187

Figure 5.3 The community context ...... 189

Figure 5.4 General practice context...... 191

Figure 5.5 Research context...... 195

Figure 5.6 Learning context ...... 197

Figure 5.7 Distribution of practice nurse age ...... 198

Figure 5.8 Range of practice nurse age...... 199

Figure 5.9 Practice nurse children and breastfeeding experience...... 199

Figure 5.10 Beginning intervention delivery...... 201

Figure 5.11 Barriers to intervention delivery ...... 203

Figure 5.12 Communicating the intervention ...... 204

Figure 5.13 Accepting responses to the intervention...... 205

Figure 5.14 Comfortable intervention delivery ...... 206

Figure 5.15 Routine intervention delivery ...... 206

Figure 5.16 "There's a stigma out there"...... 210

Figure 5.17 Managing resistance...... 211

Figure 5.18 Avoiding resistance...... 212

Figure 5.19 Interpreting mothers' responses ...... 212

Figure 5.20 "Something I wouldn't have thought of before" ...... 213

Figure 5.21 "Using those questions"...... 214

Figure 5.22 "You can lead a horse to water..."...... 215

Figure 5.23 "OK to speak up and to make mistakes"...... 218

Figure 6.1 Giving the bottle...... 236

Figure 6.2 Notes on a practice recruitment attempt...... 248

Chapter 1

Introduction and Background literature

THESIS INTRODUCTION Breastfeeding is a complex, interdependent relationship between a breastfeeding mother and her breastfed baby or older child. It occurs as one of a mother's diverse societal roles, and these roles "affect her self perception and ability to feed and interact with her infant." (1)(p13) Breastfeeding is promoted in Australia because it is significant for the health of women and newborn infants. (2) "Medically speaking, there is no doubt breast is best." (3)(p53) The World Health Organisation recommends that breastfeeding be practised exclusively for the first 6 months of infant life, and not ceased before the child is two years old, (4) based on the principle of optimal nutrition as a human right.

Research question Does the introduction of motivational support from trained general practice nurses lead to an increase in exclusive breastfeeding rates and ongoing breastfeeding rates?

Thesis plan This thesis takes the approach of the "dominant Western mode of scientific, objective methodology" in the main, although as one anthropologist put it, "all researchers are influenced by their particular cultural context". (1)(p13) Additional perspectives have been gained using mixed methodology.

Chapter 1 Background literature review The importance of breastfeeding is generally agreed, and Chapter 1 opens by considering importance, measurement options, and the poor status of exclusive and ongoing breastfeeding in Australia.

The cultural context of breastfeeding is reviewed. Protection and promotion of sustained and exclusive breastfeeding are explored, with a focus on the role of health professionals. The need for more support for mothers to sustain breastfeeding is established. Breastfeeding rates have been addressed by many international research teams. Major systematic reviews are examined to identify a supportive role for primary care health professionals, to address modifiable barriers to ongoing breastfeeding.

The study described in later chapters only indirectly presents the views of mothers. Mothers' viewpoint is critical as breastfeeding outcomes depend on mothers' actions.

24 Considering this, Chapter 1 makes a point of hearing the voices and experiences of mothers, as reported in journal articles and in grey literature.

Chapter 2 Systematic literature review No major systematic review has specifically addressed the efficacy of various training options for health professionals who support breastfeeding mothers in community based or primary care settings after maternity discharge. Chapter 2 uses high-level evidence from randomised controlled trials to address this question, and points toward a promising local approach. This review also considers the communication framework of included interventions, as a new contribution to the field.

Chapter 3 Methodology In this chapter the design and conduct of a cluster randomised controlled trial, set in general practices in two small rural cities, is described. Training of practice nurses and the theoretical framework of motivational interviewing are explained. Development of instruments is described, including the Conversation Tool, a flowchart used by practice nurses to guide an intervention with mothers who attend for infant immunisations. Recruitment of mothers and participant flow are shown according to CONSORT guidelines. Breastfeeding outcome data was collected by telephone interview at 4 and 6 months of infant age. Methods to enable process evaluation are described.

Chapter 4 randomised controlled trial results The results of the randomised controlled trial are presented.

Chapter 5 Process evaluation The practice nurse experience and perspectives regarding this trial are explored in this chapter. The pre and post training quiz results, practice nurse demographics, and a topical literature form a background to the discourse analysis. This evaluation measures the feasibility and acceptability of the intervention in a routine practice setting and indicates the extent of compliance with intervention protocol. It also considers the perspective of the researcher and interactions with the subject.

Chapter 6 Discussion and recommendations The contribution of this research to the field of lactation support is discussed. Strengths of this study as clinical translational research are presented, with reference to issues for conducting trials in primary care. Limitations relating to generalisability and system-

level change are presented. Recommendations are made for future research and policies.

BACKGROUND LITERATURE REVIEW

1.1 THE IMPORTANCE OF BREASTFEEDING Health is the most common rationale for breastfeeding research and advocacy, (1)(31)(p243f). However, breastfeeding is important also for food security, naturalist, environmental, economic, and human rights reasons, as discussed in this section.

1.1.1 Importance as a biological imperative Breastfeeding is a dynamic mammalian connection between a woman and her growing offspring that is vulnerable to disruption, "a transitional stage from pregnancy to a period of relative autonomy for both the mother and the child." (5). Across the 20th century, rapid social and technological change has been associated with large fluctuations in breastfeeding rates, (4) with the greatest disruption occurring in the 1950s and 1960s. (6)

Breastfeeding into childhood was once essential for survival. This is still the case in many developing countries.

Oh sure... I would love to counsel every HIV-positive mother about her choices... But then I wouldn't be living in Uganda and I wouldn't be talking to my own people. I would be living in America and I would be talking to your people... Twenty-seven percent of babies born to infected mothers become infected from breast-feeding. In rural areas 85% of babies will die from dirty water used in formula... you don't need a medical degree to figure out which of those odds to take.a

This perspective reached communities in the developed world in December 2010, when Australians in southeast Queensland experienced a natural disaster with cyclone-related flooding. The online community was anxious about infant feeding and food security. (7)

a Dr Miro of the Makerere University Medical School of Uganda when he spoke to the New York Times in 1998, (7)(cited p159)

26 ...other places nearby us have been cut off for a while now. They are running out of formula and nappies. The health nurse told me this morning that there are people there trying desperately to re-establish breastfeeding.

Online responses in the discussion thread included

That's one of the saddest things I've heard this week. Dumb question, but what happens if they can't re-establish BF and the formula supplies run out?

Another participant replied

...they are either evacuated, or a food drop is organised. It is also really important that mothers know that they can keep breastfeeding even if they are stressed/displaced etc. They need our support.

It seems that generally in developed countries, it is no longer expected that an infant's needs will be met by breastfeeding, or that a mother will receive skilled, knowledgeable breastfeeding support. From the perspective of a Dutch woman who opposed breastfeeding, and told a researcher, "I am not a cow!" (8)(p17) it seems that substitution of breastfeeding with bovine milk formulas is routinely expected. Certainly, this was implied in the submission from Heinz to the Australian parliamentary inquiry into breastfeeding. (9)(p131) Heinz wanted to see "the use of formula milk be depoliticised and treated objectively as a routine aspect of baby care". When mothers consider formula normal and routine, they may consider breastfeeding to be an unfamiliar luxury and non-essential. Dr Mitchell Gruich, a New Orleans paediatrician, (10) wrote about his hospital discharge rounds on the day following Hurricane Katrina in 2005. There was no running water or electricity in the city.

Even after lengthy discussions of the benefits of breastfeeding, especially now, mothers of the first three newborns chose to use formula.

The "rhetoric of 'choice' in infant feeding" (11) is considered at the end of this chapter in the context of support. As anthropologist and feminist writer Van Esterik commented, (12)(pS45)

Choice, of course, only exists when options are fully available, including information regarding possible consequences of different methods of infant feeding.

Falling breastfeeding rates of the mid 19th century led to rising concern among individuals from a range of disciplines, about the status of infants, young children, and women. Collaborative calls for action included the WHO Innocenti Declaration on the protection, promotion, and support of breastfeeding (1990). (4) International responses led to recovery of breastfeeding initiation in many countries, however, sustained breastfeeding remains uncommon in most developed countries. (13, 14)

1.1.2 Importance as a health issue Human health and development outcomes merit the protection, promotion, and support of breastfeeding. (4) It is worth noting that this perspective often focuses on giving of human milk rather than on breastfeeding as an interpersonal action. The health importance of exclusive and sustained breastfeeding is the loudest voice in breastfeeding promotion. (4) There are calls for the health risks of formula feeding to be presented. (15) This thesis itself has a framework based on health priorities, largely exploring the improved use of health-focused knowledge and evidence-based health intervention strategies by health practitioners, within health facilities.

Breastfeeding contributes to healthy infant development, to infant protection and to healthy maternal reproductive and endocrine function. This contribution occurs over a period of time, and requires ongoing breastfeeding to achieve its full potential. The human "reproductive strategy" has been described as (5)(pS4)

...usually producing a single newborn who is very immature at birth and whose development into the adult is slow and requires not only many years but also sustained parental fostering.

Data from developed countries (2) revealed poorer maternal and infant health outcomes when breastfeeding was reduced or absent, with evidence now showing relationships between some outcomes and the extent of breastfeeding. Long and short- term health outcomes for the infant and mother are detailed in an evidence report prepared for the US Department of Health and Human Services, published in 2007. (2) Systematic review and meta-analysis of controlled trials and cohort studies revealed consistently better infant outcomes for breastfed infants compared to formula fed infants, for a variety of health outcomes. This report also identified consistently better maternal outcomes for women who had breastfed compared to those who had not, for a number of outcomes.

28 The authors (2) noted limitations in the breadth of their review, in that they did not address studies on nutrition issues such as micronutrient deficiency, found to be important by others. (16) Nor did they review studies in the field of intervention trials for breastfeeding promotion.

Table 1.1 Breastfeeding for health in developed countries (2) Outcomes for which breastfeeding is protective

Infant Maternal acute otitis media breast cancer non-specific gastroenteritis ovarian cancer severe lower respiratory tract infections type 2 diabetes atopic dermatitis post-partum depression asthma (young children) obesity type 1 diabetes type 2 diabetes childhood leukaemia sudden infant death syndrome (SIDS) necrotising enterocolitis

The Promotion of Breastfeeding Intervention Trial (PROBIT) was conducted in the Republic of Belarus and had 13,889 participants. PROBIT demonstrated increases in both exclusive and ongoing breastfeeding in the intervention group, compared to controls. (17) The six-year follow-up from PROBIT published recently provides strong evidence of differential outcomes for cognitive function: children in those clusters randomised to intervention had significantly higher verbal IQ and improved reading and writing according to independent standardised teachers' ratings, compared to controls. (18) They did not consider relative intensity of breastfeeding. Dose-dependent relationships have been identified for the risk of diarrhoea, febrile respiratory illness, and otitis media. (19) These were identified at two levels, firstly comparing fewer vs. more breastfeeds (degree of substitution) and secondly comparing duration of 3 months to duration of 6 months or more (duration of feeding before weaning completely onto substitutes). (19)

Breastfeeding is associated with attentive and responsive mothering (20) and increased active mothering time (21), which may relate to some of the developmental

outcomes as described and may also lead to greater physical safety from death by accidental injury. Chen et al. compared data for 1,204 representative infant deaths from accidental injury (aged 1-11 months) to data for 7,740 living one-year-old children in the USA, with analyses weighted for birth weight and race, and found children who had been breastfed had a lower risk than those who had never been breastfed, OR 0.59 (95%CI 0.38-0.94). (22) Morales in her Masters dissertation (23) found that greater maternal closeness was associated with increased frequency of feeding times each day, more a feature of breastfeeding than of formula feeding.

The authors of the major systematic review (2) did not address the impact of lactational amenorrhoea, which has a significant influence on women's health through increased child spacing. (24-26) New research is suggesting longer-term effects on women's metabolic health such as the risk of type 2 diabetes, particularly for women diagnosed with gestational diabetes (HR range: 0.14 to 0.56; p=0.03). (27)

Risks of formula use Other than the risks from loss of breastfeeding-related health influences, (28) reliance on formula involves additional risks to infant safety. Safe handling and preparation instructions must be followed when bottle-feeding. A large recent study (n= 1399) in USA found "Of 881 mothers who reconstituted powdered formula mixed with tap water for infants aged 1.5 to 4.5 months, 30% boiled the water first". Moreover, 55% did not always wash their hands with soap and water before preparation and 33% sometimes re-used bottle teats after only rinsing with water. (29) NHMRC advise health workers that powder must be added to cool boiled water; that quantities vary by manufacturer and that preparation instructions must be checked; that measuring scoops may be inaccurate; and that for safety, equipment should be sterile and each bottle feed should be reconstituted when needed. (30)

Formula feeding involves health risks from formula manufacturing errors, product contamination, untested formulations, and loss of species-specific nutritional and bioactive components. (19, 31)(pp 25-36)

1.1.3 Importance for environmental and economic outcomes Environmental concerns relate to the use of resources to produce and distribute formula and to manage by-products. (31)(pp 25-27) Decreased reliance by mothers on (mainly bovine) dairy based formula would reduce the energy consumption and waste

30 production associated with this industry (32) due to the lower environmental impact of breastfeeding compared to dairying.

The increased cost of health care, because of differential health outcomes from reduced breastfeeding, is considerable. A 1999 review of population-based studies in Tucson Arizona, USA and Dundee, Scotland (33) examined health costs for lower respiratory, otitis media and gastrointestinal infections. In infants aged up to 12 months who were no longer breastfeeding, the additional cost per child was $331 - $475 in managed care costs. In 2002 a costs analysis was performed using Australian Capital Territory data on breastfeeding rates, (34) hospitalisations, and the relative influence of breastfeeding on five specified health conditions. The authors (34) found that one to two million dollars of annual hospital expenditure was related to early weaning onto infant formula. These analyses did not include increased societal costs such as parental absence from employment to care for sick children. Attention has turned more recently to the health costs of managing obesity in the Australian population, (35) which is associated with reduced breastfeeding.

The economic contribution to society by women, in the time and breastmilk given to infants, has been raised by researchers. Economist Marilyn Waring noted that breastfeeding was work done by women, albeit unpaid and often unrecognised as such. (Cited, (36)(p19)) This unpaid work may be costly to families for whom maternity leave has minimal or no financial support. Without an accounting in Gross Domestic Product, the economic value of breastmilk provision is overlooked. (37)

1.1.4 Importance as a human right The importance of breastfeeding as a human right has been overshadowed by the health dialogue. Breastfeeding can be an empowering connection; for many women and infants this interdependent relationship is a source of pleasure, (38) and an engrossing personal journey. (20) Feminist writers point out that when women's right to breastfeed is undermined by barriers and inadequate support they are oppressed and exploited. (12)(pS41) It is noteworthy that breastfeeding rates in developed countries are lowest among disadvantaged women, relating to "their structural position in Australian society which places them in a position of vulnerability and subordination." (39)(p41) Women's rights include acknowledgement of their lived and embodied experiences and recognition of their work of nurturing as being valid, valuable and worth supporting. (5)(pS3)

...breastfeeding is women's work and should be valued as such. Families and societies need to accept their share of the task of supporting women to breastfeed.

Feminist discourse is about validating women's autonomy, liberation, experiences, and rights. (12)(pS42)

Breastfeeding is a paradigmatic feminist issue because it requires rethinking basic issues such as the sexual division of labor, the fit between women's productive and reproductive lives, and the role of physiological processes in defining gender ideology... Breastfeeding is a holistic act and is intimately connected to all domains of life - sexuality, eating, emotion, appearance, sleeping, parental relationships.

Australian women's studies author, Alison Bartlett, argues for recognition of the unique rhythms, passions, sensuality, creativity, perceptive attention, and endurance that are part of the lives of everyday breastfeeding women. (40)(p178) Feminist writers have argued against the medicalisation of events in women's reproductive lives, including breastfeeding, and have argued that mothers are persons whose autonomy must be recognised without denial of their experiences and rights. (41)

This thesis, while it is set in a health framework, does attempt to recognise the embodied experience of breastfeeding and to acknowledge the context of mothers' many responsibilities amongst which breastfeeding is performed.

1.1.5 Definition and measurement controversies Measures and definitions are used to describe breastfeeding practices and to evaluate health outcomes. These need to be precise and comparable. (42, 43) Reported research outcomes can influence public opinion through media and marketing, and can inform public policy makers. It is worth noting that infant health outcomes in breastfeeding research are generally assumed to relate equally to feeding of expressed breastmilk, (2) an untested assumption.

The behaviour of exclusive breastfeeding is related to duration as steps along a continuum of gradual weaning. Infants fully breastfed in the first six months are more likely to continue breastfeeding into the next six months. (44)

32 Measurement and analysis Outcomes studies commonly measure breastfeeding at specified time points with analysis for prevalence. (45) However, some studies maintain contact with each participant until breastfeeding has ceased completely, and use survival analysis. (45) Commonly used definitions, and those used in this thesis, are shown in Figure 1.1. An alternative approach is to divide the "Any breastfeeding" category further into intensity measures indicating the percentage of daily breastfeeds replaced by formula, (46, 47) or into partial and token, indicating the degree of nutritional contribution from breastmilk. (48)

Figure 1.1 Definitions of breastfeeding

Also may be fed: Definition:

Medicines Exclusive Full Any breastfeeding breastfeeding breastfeeding Water- based fluids

Formula or other milk

Spoon-fed or hand-held solids

Exposure to breastmilk substitutes Exposure data is used in the measurement of breastfeeding behaviours since birth. It may not be a marker of the mother’s breastfeeding intentions, which are strongly linked to duration outcomes. Breastfed newborns in Australian maternity hospitals risk being given water or formula without medical indication, at times without maternal consent, even where hospital policies specifically oppose this. (49)

Exposure to any breastmilk substitute prior to 4-6 months of infant age, commonly formula or solids,b relates to infant and childhood risks of allergy and atopy. For these health outcomes, measurement of any exposure since birth is required. (50, 51)

b See Abbreviations and glossary

The extent of formula or solids substitution for breastmilk is related to other health and development outcomes. Recently an inverse association was found between excess weight gain and more intense breastfeeding. (46) The intensity of breastfeeding merits examination because the effects are incremental or dose-dependent for many health outcomes.

Current use of breastmilk substitutes Current practice measures apply over short time periods such as one day, and are more sensitive to fluctuations and trends. This is in contrast to exposure, which is a retrospective measure, applying to consistent behaviour over long time periods such as four months. The potential for health promotion interventions to change breastfeeding behaviours or formula feeding behaviours, particularly in smaller studies, may be best tested using current practice measures. Consistency and comparability are limited when studies such as previous Australian National Health Surveys use non-specific terms such as "regularly" which are open to interpretation. (52)

In their “best practice” intervention study, Bonuck et al. found that while there was no significant difference between groups in exclusive breastfeeding, control subjects had lower breastfeeding intensity at thirteen weeks (odds ratio [OR]: 1.90; 95% confidence interval [CI]: 1.13-3.20) and 52 weeks (OR: 2.50; 95% CI: 1.48-4.21). (47) This means that in the intervention group, mothers replaced fewer breastfeeds with formula than occurred in the control group, which is a positive health behaviour change.

Intensity of breastfeeding is useful for observing the normalisation of regular or periodic formula feeding behaviours, which are implicated in early weaning. As some authors observed, “over the past hundred or so years artificial infant feeding has become so ubiquitous in many developed world settings as to be widely viewed as the standard way to feed infants." (53)(p 387) While exclusive breastfeeding since birth does the least harm, measurement of intermediate steps towards this goal may be valuable.

Recall time period Commonly in measurement of current practice, mothers are asked what their baby was fed "in the last 24 hours," and for clarity, results are described as 24-hour recall. Use of 24-hour recall enables international comparability of results. The World Health Organisation has developed indicators for countries to use in measurement of breastfeeding, and some of these use 24-hour recall. (54) Australian policy makers

34 refer to 24-hour recall. (42) Recall of exclusive breastfeeding, unless using a "since birth" measure, is most accurate over short periods.

The main disadvantage of the 24-hour recall survey method is that it reflects only one point in the week, which may not be a typical day. Authors acknowledge that this method tends to overestimate exclusive breastfeeding from birth (43, 54); there is opposition to the use of the 24-hour recall because it may lead to complacency by health policy-makers. (51) A Canadian study (55) discussed the transitions back and forth between partial and exclusive breastfeeding over time. The authors (55) concluded that the reference period of one week used in Canadian research was practical and evidence-based for infants up to 12 months of age, comparing reference periods of one to eleven days. Where 24-hour recall is the only data collected, conclusions about health outcomes may be flawed because the impact of inconsistent use of breastmilk substitutes is not recorded.

Recall of duration of breastfeeding retains accuracy over longer periods than does recall of exclusivity. One study found a one-month overestimation in duration of breastfeeding using a recall period of 1-3.5 years. (Gillespie et al., Cited in (56))

Describing the data Data collection and analysis techniques need to be matched to the goals of each study (57); accordingly, methods vary widely across studies. (45) Measurement of "breastfeeding/any breastfeeding" has involved either survival analysis, prevalence at a specified time point, or maternal recall over a specified time period.

Indicators that are used to measure breastfeeding vary widely and results are dependent on points of measurement. Hector makes the point that because solid foods are recommended from the day the infant turns 6 months of age, there is no place for measuring exclusive breastfeeding on that day i.e. "at 6 months." (56) A current debate in Australia is focussed on the appropriate time point at which breastfeeding should be measured and reported (including "up to", "under", "until", "for", or "at") any specified time point, for comparability over subsequent studies and internationally. (54, 56, 58, 59) These variations can make it difficult for government authorities to "measure changes in breastfeeding practices in the population and to plan and target programs and services to promote breastfeeding." (59)

Variant definitions of breastfeeding in published research are more the rule than the exception. This was detailed by the authors of a systematic review on breastfeeding and obesity. (60)(p1367)

"The feeding groups were defined as being mutually exclusive in 4 studies, the breastfed group included mixed feeders in 7 studies, and the formula-fed group included mixed feeders in 7 studies. In 2 studies in which infants were breastfed exclusively, the exclusiveness of formula feeding could not be gauged. The exclusiveness of initial feeding was unclear in 10 additional studies."

There is general agreement that breastfeeding is important, but currently no universal agreement about how breastfeeding is defined.

1.1.6 Recommended breastfeeding practices In response to research on health outcomes, Australia's National Health and Medical Research Council recommended breastfeeding exclusively for the first 6 months. They recommend continuing breastfeeding alongside suitable foods to 12 months "and beyond if both mother and infant wish". (30) These documents state that they support the WHO recommendations; however, those include continuing breastfeeding for two years or more. (54)

The Australasian Society of Clinical Immunology and Allergy now advises cessation of exclusive breastfeeding “from around 4-6 months” with introduction of solids. (61, 62) The Australian Dietary Guidelines of 2003 use both “about 6 months” and “at 6 months” interchangeably as the time to start solids (30) despite recommending exclusive breastfeeding "for" 6 months.

It is agreed that non-human milks for infants and young children are suitable without modification from 12 months; infant formulas are required to 12 months. (30, 61)

Recommendations depend on current scientific evidence and the views of a range of experts. However, like any public policy, they have political and funding consequences and are subject to debate, as noted in 2001. (42)(p11)

In Australia, the NHMRC has taken a cautious approach to duration, noting that some groups in the community react negatively to the suggestion of breastfeeding for two years.

36 The Australian Dietary Guidelines and Infant Feeding Guidelines are currently undergoing review, with the draft recommending exclusive breastfeeding "until around 6 months of age (22-26 weeks)."c

Authoritative bodies vary their recommendations over time, however new mothers rely on the advice and support of the older generation. In the past, women who are now grandmothers were advised to give additional foods from one month. (63) Guidelines for exclusive breastfeeding were updated in 2001 in Australia, following the World Health Organisation's lead. Previously exclusive breastfeeding was recommended for only 4 months. Although the change occurred more than 10 years ago, awareness of the change is poor, as evidenced from a recent Sydney study. The authors (64) interviewed 409 first-time mothers at 24 to 34 weeks of pregnancy about their breastfeeding plans and found only 61% were aware of current recommendations. Of these mothers, 42% intended to meet the recommendations. Awareness was significantly associated with maternal education, and mothers were 5.6 times more likely to plan to meet the recommendations if they were aware of them.

1.2 INFANT FEEDING IN AUSTRALIA Most Australian new mothers begin breastfeeding; Australia has a high initiation rate of almost 90%. (66) However, substitution of breastfeeds with formula begins early and increases quickly, particularly among disadvantaged mothers. Analysis of national infant nutrition data from 2005 (52, 65, 66) found that by 3 months over 40% were using substitutes, as shown in Table 1.2. (52, 65, 66). The least educated mothers were least likely to avoid formula substitution by 3 months. (65) By comparison, weaning was slower in Norway, where substitute use at 3 months was 30% and weaning by 6 months was 20% in population analyses. (67, 68) Preliminary results of the Australian National Infant Feeding Survey conducted in 2010-2011 (69) found no improvement since 2005. Solids were being fed to 35% of 4-month-old infants, and formula was being fed to 55% of 6-month-old infants. Similar to previous surveys, (65) by 6 months almost half of all infants were fully weaned, see Table 1.3..

Australian population rates of breastfeeding are far from the Australian recommendations. The National Health and Medical Research Council endorsed World Health Organisation calls to “promote, support, and protect breastfeeding”; these include continuation beyond two years. (30, 70-73) c http://consultations.nhmrc.gov.au/open_public_consultations/infant-feeding

Table 1.2 Any formula feeding, Australia vs. Norway Age at data collection Australia, 2005 (52, Australia, 2010 (69) Norway, 2002 (68) 65, 66) 3 months 40% 46%* 23% 6 months 63% 55%** 43% 12 months 75% 80%** **includes all non-human milks

Table 1.3 Any breastfeeding, Australia vs. Norway Age at data collection Australia, 2005 (52, Australia, 2010 (69) Norway, 2005 65, 66) (67) 3 months 64% 70% 90% 6 months 51% 60% 80% 12 months 23% 18%** **measured for infants aged 13-18 months

In 2007, an Australian parliamentary inquiry examined the health benefits and current status of breastfeeding and the challenges faced. (74) (75) As Hector explained to the committee, (74)(p39)

We have a lot of evidence already about why women self report that they do not breastfeed.... technical difficulties come up in the first few weeks generally and then issues of insufficient milk supply are another primary reason given for ceasing breastfeeding. It then moves on to return to work or study becoming more prominent. These are the reasons that mothers are giving, but mothers are often not aware of the actual underlying causes of why they are giving up or why they are not breastfeeding. A lot of the societal and environmental level determinants of why they are not breastfeeding are then impacting on them to perhaps create a situation in which they do not have enough milk and they are forced to give up anyway.

1.3 PROTECTION OF BREASTFEEDING

Measures exist to protect the breastfeeding relationship from disruption, once established, and include guidelines with variable legal powers. These offer a degree of encouragement for continued breastfeeding; however, they are far from sufficient, as this section will show.

38 1.3.1 Protection against unethical marketing of formula Skilful marketing may explain why, in the 2010 Australian National Infant Feeding Survey, 26% of women gave as their reason for not breastfeeding ‘infant formula as good as breastmilk’. (69) Similarly in a large and carefully representative US opinion study, a majority agreed that ‘breastfeeding is healthier for babies’ (55-75% of people depending on geographic region), however, only a minority agreed that ‘feeding a baby formula instead of breastmilk increases the chances the baby will get sick’ (13-38% of people depending on geographic region). (76)

Control of formula marketing is a measure to protect breastfeeding. The World Health Organisation's 1981 International Code of Marketing of Breast Milk Substitutes and subsequent World Health Assembly resolutions (The WHO Code) aimed (30)(p308)

...to protect the nutritional wellbeing of all infants in two separate but closely related ways: through protecting and promoting breastfeeding; and the appropriate use and marketing of breastmilk substitutes, bottles and teats when these are necessary.

Australia is a signatory to the Code, yet "the Code has no authority here and monitoring of compliance is limited." (77)(p39) Breastfeeding advocates have campaigned for implementation of The WHO Code in Australia, a call taken up by the House of Representatives standing committee on health and ageing. (9)(p142)

Breastmilk substitutes necessary for a very small minority of infants and are the preferred feeding option for some mothers. The NH&MRC guidelines state, "If an infant is not breastfed or is partially breastfed, the commercial infant formulas are the most acceptable alternative to breastmilk until 12 months of age." (30)(p289)d Formulas are the commercial products of corporate entities (78)(p199) who operate in a market economy. They are accountable to share-holders who expect them to increase their market share (formula use).

Historically in Queensland, "In 1969, a Nestlé baby book ... was given to mothers in all postnatal wards. From the late 1960s, free cans of artificial baby milks were also provided to mothers on hospital discharge." (79)(p69)

d Infant formula is "an industrially produced milk product designed for human consumption." (9) (p127) Milk proteins and electrolytes are modified, and vitamins, minerals, and vegetable oil constituents added, according to the National Health and Medical Research Council.

The 1992 Marketing in Australia of Infant Formula (MAIF) Agreement covers only parts of the WHO Code and, "unlike national legislation restricting the advertising of medications, provides no penalties for breaches." (80)(p233) Manufacturers may choose to become signatories to this voluntary agreement, which restricts infant formula marketing directly to mothers. However, brand recognition is freely used to promote product line claims such as through advertising of toddler milks. (81, 82)

MAIF signatories agree to "include clear information on... the benefits and superiority of breastfeeding" (p408-9); (30) the boxed notice on all formula tins states "Breast milk is best for babies" as the entire information on the "benefits and superiority of breastfeeding". The boxed notice also states, "Good maternal nutrition is important for breastfeeding" which is ambiguous. Nutrition is important for maternal health. Nutritional deficiency in breastfed infants is rare in comparison to the risk from infant formula. (83) While not literally inaccurate, the implication of the notice is that poor maternal nutrition may be detrimental to the mother's breastmilk and hence to her breastfed infant. It may be interpreted as an imperative for disadvantaged women to avoid breastfeeding, according to women in one USA study. (84)(p1761)

Breast milk is good for the baby if the mother keeps herself up. I’m not against it, but most mothers don’t take care of themselves. The milk could be poison.

Consumer information on one manufacturer's website entry page begins

Breast milk provides the optimal nutrition for babies and should be used if at all possible.e

However, subsequent pages use "optimal" to describe the effect of the company's standard product.

[Brand] Stage 1 formula gives babies 0-5 months a full range of vitamins and minerals, plus the right balance of proteins, carbohydrates and fats for optimal growth and development.

The website marketing implies a solution to common parenting concerns through using the company's specialised formulas.

e http://www.novalac.com.au/entry/entry-disclaimer.asp

40 There is also a [Brand] formula for infants who wake constantly due to hunger... Find out more about the health benefits of [Brand] and how the formulas work to help manage infant feeding problems.f

In Australia there is a high incidence of these infant feeding problems. [Brand] formulas may help reduce crying, leaving infants content and parents more relaxed.g

[Brand] Sweet Dreams formula provides a longer-lasting feeling of fullness in infants who wake often due to hunger.h

Auerbach reports on printed materials from five manufacturers, all similarly misleading in their information. The messages confuse the reader or leave a negative impression about breastfeeding. (85) They align formula feeding with contentment and independence of infants; night waking is associated with inadequate feeding, distress and dependence such that parents cannot relax. Marketing is designed to align formula feeding with contemporary values around parenting, (86)(p93) as discussed in Section 1.4.3.

Manufacturers also advertise infant feeding helplines; Dr Fielding of Wyeth (now Pfizer) Australia is recorded in Hansard from the Parliamentary Inquiry public hearing on 4 June 2007 (74) saying

We get 16,000 calls a month, and 25 per cent are healthcare professionals... Australia and New Zealand, but New Zealand represents just a small number of those calls.i

An alternative telephone helpline is provided by volunteers for the non-government organisation Australian Breastfeeding Association, with some government financial support.

More than 400 breastfeeding counsellors volunteer on the Breastfeeding Helpline each month, answering as many as 6,000 calls.j

f http://www.novalac.com.au/default.asp g http://www.novalac.com.au/about.asp h http://www.novalac.com.au/sweet-dreamsformula.asp i http://www.aph.gov.au/house/committee/haa/breastfeeding/hearings.htm; https://www.meandmychild.com.au/Pfizer_Nutrition_Careline/Contact_Us/ j https://www.breastfeeding.asn.au/breastfeeding-helpline

In a qualitative study with disadvantaged black women in the USA, the researcher observed a greater confidence in formula than in breastmilk. (87)(p80)

Helena told me that she was nervous about whether there were enough nutrients in her breastmilk. Febe had similar concerns. She said, 'I tasted my milk and it tasted sweet. And it didn't seem like iron would be in there anywhere.' They both decided to give some formula, which they believed would supplement their deficient milk.

Logically the health discourse on breastfeeding should be unquestioned; however, mothers are vulnerable to the sophisticated advertising techniques used by manufacturers of substitutes.

1.3.2 Protection in paid employment Returning to paid employment adds pressure to and risks separation of a breastfeeding relationship. Victorian State Member of Parliament, Kristy Marshall, was ejected from the parliamentary chamber in 2003 for attempting to breastfeed her 11-day-old baby in her workplace, leading to international publicity. (36) On this occasion, an archaic law restricted non-elected persons ("strangers") from the chamber. However, subsequent discussion in The Age newspaper online included numerous criticisms of her actions, from journalists, letter writers, and schoolchildren, many considering the roles of motherhood and employment as mutually exclusive. (40)(p77)

Mothers of infants older than one month frequently cite a return to paid employment as their reason for ceasing exclusive or all breastfeeding. Recent analysis of the Longitudinal Study of Australian Children found that weaning was more common for women who commenced fulltime employment within three months after the birth than for those not employed. (65) Other women who combined work and motherhood cited low milk supply as their reason.

Educated women with some workplace independence, like Bartlett, Giles, Hausman or Shaw who are all authors, are able to successfully negotiate the time and facilities needed to combine breastfeeding and work, and subsequently "use their personal breastfeeding experiences in their research as a means to stimulate discussion and debate about infant feeding in general." (36) This is not possible for many casual or low-paid women workers, and women of childbearing age make up 45-60% of the paid workforce in developed countries. (88) In 2007, the Cochrane collaboration was unable

42 to find any randomised controlled trial evidence on workplace interventions to support breastfeeding women. UK cohort data on 6,719 employed British/Irish white mothers of infants less than 10 months old found that women were less likely to sustain breastfeeding to at least 4 months (89)(p5)

the more hours they work, the earlier they return to work, or if they return for financial reasons.

Where family-friendly or flexible work arrangements were available, duration of breastfeeding was longer, and self-employed mothers or those working alone were the most likely to still breastfeed at 4 months in this UK study. Of interest is the context of UK statutory Maternity Leave payments to 4 months, paid at 90% of the woman's average weekly wage for the first six weeks. (89)

A comparison of maternity provisions and workforce participation published in 2003 noted that Swedish laws and policies, compared to the USA, gave mothers more opportunities (90)(p171)

...not only to exclusively breastfeed for longer periods but also to experience, along with their infants, the emotional intimacy potentially offered by the physical breastfeeding relationship.

This offer of 'equal opportunity' for women workers to breastfeed, particularly for those groups who are most vulnerable, considers the rights dialogue as well as the health dialogue. (90)

Duration of breastfeeding depends on reducing the separation of a mother and her baby. In a study of 810 employed breastfeeding mothers in the USA, 43% of mothers pumped/expressed milk at work. However, those mothers who physically breastfed during their working day achieved longer breastfeeding duration. (91) This study listed possible solutions, (91)(pS61)

...ways to enable direct feeding include on-site child care, telecommuting, keeping the infant at work, allowing the mother to leave work to go to the infant, and having the infant brought to the work site.

Canadian authors found that increased maternity leave provisions in Canada led to increased breastfeeding duration, according to national goals. (92)

The International Labour Organization Maternity Protection Convention 183 was designed to protect breastfeeding and included fourteen weeks of paid leave and

subsequent paid lactation breaks. (ILO, cited (89)) In Australia, commercial child-care costs are subsidised to encourage mothers to return to paid employment, and Australians had no government-supported paid maternity leave prior to 1 January 2011.k Paid Parental Leave is now paid at the minimum wage for eighteen weeks.

Australian employers may apply for Breastfeeding-Friendly Workplace Accreditation (BFWA) Program from the Australian Breastfeeding Association. This contributes to normalising sustained breastfeeding by providing support. (9)(p79)

Many Australian women in paid employment cannot rely on workplace conditions to protect breastfeeding. While useful strategies have been described, (91, 93) many mothers devise and advocate in isolation for their own strategies regarding ongoing breastfeeding.

1.3.3 Protection in public Australian mothers who wish to breastfeed in any public place have explicit legal protection through anti-discrimination legislation such as the Commonwealth Sexual Discrimination Act 1977 s5 (1A). (Cited (94)) In contrast, in one large study over 80% of Australian adults telephoned at random agreed that bottle-feeding was more acceptable in public places, and 70% agreed that there was not always a place to breastfeed when outside the home. (95) Public breastfeeding is seen as a scandal according to Bartlett (40) and seen as something too physical, secretory and uncontrolled, (96) too much a bodily function and hence seen as disgusting according to Nussbaum.l (Cited (94)(p23))

Public breastfeeding faces disapproval particularly when the infant becomes less passive and more interactive. The Best Start, an Australian parliamentary inquiry report, found mothers were reporting "that breastfeeding beyond 12 months elicits 'significant stigma and taboo' from the public" and cited one mother's submission:

The community views breastfeeding an older baby, let alone a toddler, as sick and 'child abuse'. I know of many women who are scared to breastfeed in public. I know of women who have been abused for doing so. k http://www.fahcsia.gov.au/sa/families/progserv/paid_parental/Pages/default.aspx l "Nussbaum asserts that disgust is a powerful socially constructed emotion that has been mobilised throughout time to exclude certain groups and persons from public life" (94)(p23)

44 The report concluded its comments on extended breastfeeding with the finding:

There seems to be a curious dichotomy in the community where an infant of 18 months is still considered to be totally dependent on their parents for everything including food but is perceived to be too old to breastfeed. (9)(p73)

The laws and the amendments specifically to prevent discrimination around breastfeeding are framed around infant health benefits and leave mothers open to demands to revert to expressed breastmilk. (94) Focus group research in Perth, Western Australia and literature of the Australian Breastfeeding Association both discussed discreet breastfeeding and breastfeeding etiquette. Bartlett comments: (40)(p77)

I suspect that this rhetorical strategy may well be enabling to individual women breastfeeding in a potentially hostile climate, even if that hostility now contravenes the law.

Mothers may be unaware of legal protections, and those mothers who have more education, resources or life experience are more likely to feel comfortable defending themselves against accusations of unacceptable social behaviour. (39) As one adolescent mum in the UK reported, (97)(p394)

You need a lot of courage if you are wanting to go out... you've just got

to hold your head up high and think I am doing this for my baby.

1.4 PROMOTION OF BREASTFEEDING Human rights were the basis of the WHO's call to "protect, promote and support breastfeeding," (4) however, breastfeeding promotion is generally equated with health education campaigns. The failure of simplistic messages has been noted for decades, (98) and additional perspectives need to be taken into account. Medical anthropologist Van Esterik explains that medicalisation of infant feeding portrays health as a commodity. This perspective means that health professionals set rules to be complied with by patients; relationships between health professionals and industry are strengthened as new treatments are sought for identified problems; and it "individualises human problems, removing them from their social and economic contexts". (99)(p114,115)

Women face contradictory messages and expectations about appropriate behaviour as breastfeeding mothers and of their breastfed infants. When breastfeeding is about milk

and health rather than about a relationship connecting a mother and her child, competing scientific claims are made for feeding alternatives. This influences public opinion about breastfeeding older babies. The comments of one mother, reducing the child to "it" and the reasoning to sketchy science, demonstrate impatience with ongoing breastfeeding. (39)(p40)

Well, I'd want to know why: why isn't it getting a cup instead.... When it's little, breastfeeding is good, it's an important oral sensation, I suppose, but you should introduce it to other stimuli.... There's got to be a cut-off point.

1.4.1 Promotion against cultural norms In developed countries and among the elite in developing countries, (100), ongoing breastfeeding is contrary to cultural expectations. (12) Haider commented that in developed countries, it is consistently the more advantaged mothers who have the resources to devote themselves to exclusive breastfeeding while the less advantaged mothers carry excessive burdens and do not have enough support to achieve this. In the developing world, the more advantaged mothers see no need to persevere with breastfeeding when they can afford to formula feed. (101)

According to the Australian parenting book The Mask of Motherhood, "All things being equal, breast milk is best for babies. Yet all things are not equal, not by a long shot." (102)(Cited p70) Cultural influences surround breastfeeding women and influence their understanding and actions regarding breastfeeding; anthropologists report that in societies where children are allowed to nurse "as long as they want" they usually self- wean without distress between 3 and 4 years of age. (103) However, women face challenges when they breastfeed outside the home, outside routines, outside the newborn period, or outside gender expectations. In such settings, as one Dutch father observed, "As a woman you must almost defend yourself against people." (8)(p18) Breastfeeding mothers of nine-month-old infants in the USA reported shortening their plans for breastfeeding duration as their babies grew older and they encountered increasing community opposition. (104) In a qualitative study, the researcher reported the experience of one such mother. (87)(p122)

She asserted that it was the only thing that got the baby to settle down at night. She claimed that everyone said to her, "Are you STILL breastfeeding?"

46 This mistaken disapproval relates to various aspects of breastfeeding including prioritising a baby's needs above the mother's needs. Such disapproval assumes that breastfeeding is unnecessary and formula feeding is a generally acceptable or equivalent alternative. (105)(p67)

'Why are you still doing this? You're sick... you're losing weight. Why are you trying to feed the child as well?'

Authors of a Western Australian study found that mothers were "vulnerable and sensitive to the comments and opinions of others" to varying degrees, which "were perceived by participants as a reflection on their mothering abilities". (105)(p66) Communication of this opposition, while at times subtle, was nonetheless clearly perceived by mothers. (106)(p425)

Nobody said specifically that it was time [to wean]. Although people were really thinking, in my family, that at 6 months it was time. They think a year is unbelievable. So people were wondering when I was going to stop. And I’d hear, “Oh, you’re still doing that?”

Breastfeeding is an activity of women and children that takes place among many other activities in their complex lives. In 2003 after six weeks of maternity leave, Kate Langbroek returned to the Australian television talk show The Panel. She happened to respond spontaneously to her nearby hungry baby while continuing her work. Her discreet breastfeeding became the subject of a media controversy around appropriate breastfeeding. As recorded by Bartlett in her book Breastwork, the same episode of The Panel had also shown "a clip of a woman washing the windscreen of a car with her own soapy breasts" which had generated no public comment. (40)(p189) These contrasting views on culturally acceptable uses of a woman's breasts shine a light on the dilemma for women of combining breastfeeding with participation as equal members of society.

Among one group of Western Australian mothers, there was a clear assumption that "only babies are breastfed." (105)(p69) Mothers affirmed this belief when they defended their ongoing d breastfeeding with an answer such as Gwen's (105)

Fourteen months just seemed kind of nice because he was still sort of babyish, he wasn't really a little boy yet...

In contrast, from the perspective of sustained breastfeeding as the norm, Mothering your nursing toddler described breastfeeding a child through the second year as an expression of relationship and nurturing, a gradual assistance through babyhood for

"babies on wheels." The author rejected the "prevailing belief that separation is maturity and connection is immaturity." (107)(p25) She noted that toddlers and preschoolers were discovering many unfamiliar situations. They may need a supportive connection when they "overextend themselves in their efforts to master new skills." (107)(p208-9)

Research with Australian mothers who practised extended breastfeeding revealed that "many had initially felt disgust for breastfeeding beyond infancy". Ultimately these women reported being influenced by their child's enjoyment, and "as their knowledge about breastfeeding increased and as they were exposed to long-term breastfeeding role models." (108)(p5) The author notes in earlier work that studies of the incidence of finger-sucking and dummy use in developed countries, which is associated with abnormal oro-facial development, have found prevalence as high as 50% in three to five year olds. In cultures where breastfeeding is unrestricted, finger sucking is rare. (109)

The breastfed children in this study believed that breasts were for breastfeeding. (109)(p8)

Thus, one mother described how her son ‘has commented on seeing a “Sports Illustrated” magazine that the lady had “beautiful breasts” and that she had them “all out ready for breastfeeding”’ (male, four years seven months).

Breastfed children said it calmed and relaxed them. Children reported that they breastfed because they felt close to their mother and because they liked the taste, "using phrases such as ‘better than ice cream’ or ‘as good as chocolate’." (109)(p10) However, as discussed below, health practices and community attitudes act to thwart ongoing breastfeeding, uninformed by evidence on the implications for infant development.

1.4.2 Promotion of the visible use of breasts Breastfeeding is rarely seen in print or audiovisual media. An examination of 334 breastfeeding articles over three years in Australian newspapers found only 1.3% were accompanied by breastfeeding images, 57% were neutral or negative about breastfeeding, and breastfeeding in public was the most common topic. (110) A similar study in the UK addressed television programs and newspaper articles; their content analysis found that "Bottle feeding was associated with 'ordinary' families whereas

48 breast feeding was associated with middle class or celebrity women" and they found "The media rarely present positive information on breast feeding." (111) Common symbols for motherhood and babies seen on congratulations cards are infant feeding bottles and dummies, and the use of either is associated with early weaning. (69)

It is likely that sexualisation of human breasts contributes to women's discomfort with the idea of continuing breastfeeding as their babies grow older, as Bartlett indicated regarding The Panel, above. (40) An earlier researcher into lactational amenorrhoea commented on the importance of continued breastfeeding, particularly for the health of disadvantaged women. (26)(p80)

It is ironical that Western cultures have chosen to worship the breast as a sex symbol but have accidentally abolished its role as nature's contraceptive.

The contrast between appropriate settings for breasts to be seen in public is addressed in The Politics of Breastfeeding. (78)(p120)

...in contemporary industrialised society where women's bodies and particularly breasts are used to sell newspapers, cars and peanuts, public breastfeeding provokes cries of protest from both men and women.

Palmer suggests that some of the protest from women may be grounded in unresolved anger or grief over personal 'failure' in breastfeeding. (78)(p121) Other women may protest out of embarrassment, relating to their own sense of modesty and to their own satisfaction with formula feeding. (102)(p76) Protests may relate to patriarchal concepts of ownership or masculine ideology; one woman's response to public breastfeeding is illuminating, “Yuck, those are for your husband!” (112)(p704)

If breasts are primarily sexual organs then it is culturally inappropriate for them to be witnessed or handled by children. It may be this factor, or perhaps concepts around child autonomy that influenced one Australian woman's views on weaning. (102)(p208)

I will be sad this time, because I don't expect to have any more children. Having said that, if my daughter doesn't wean herself around her first birthday, as her brother did, I probably won't go past eighteen months of age. I still feel awkward about the idea of feeding a walking, talking toddler.

1.4.3 Promotion and infant behaviour historically From the 1700s, mothers were discouraged from following the feeding cues of breastfed babies. Doctors advised four-hourly feeding with no feeds to be given at night. Dr William Cadogan was glad, in 1769, to see infant care now being supervised by "men of sense rather than foolish unlearned women". (Cited (78)(p 41)) These principles were advocated by Dr Frederic Truby King in New Zealand in his popular 1913 book Feeding and care of baby. (113) Together with emphasis on rapid weight gain and regimented birthing protocols, this focus on timed feeding led to more lactation failure and increased use of formulas. (114, 115) The Karicare formula was developed by Dr Truby King in response to such lactation failure.m (116) The Mothercraft Manual, (cited (117)( p7)) published in 1924 by the president of the Royal College of Midwives who had trained under Dr Truby King,n taught what was expected of babies and mothers in their first year together:

Self-control, obedience, the recognition of authority, and later, respect for elders are all the outcome of the first year's training... The baby who is picked up or fed whenever he cries soon becomes a veritable tyrant... while, on the other hand, the infant who is fed regularly, put to sleep, and played with at definite times soon finds that appeals bring no response, and so learns that most useful of all lessons, self-control... the conscientious mother has to be prepared to fight and win all along the line, in matters small and great.

In 1978, Sydney psychiatrist Peter Cook wrote, "Western medicine has not yet overcome a distrustful urge to regulate an infant's food intake in arbitrary and artificial ways." He noted that this urge "lacks biological precedent and scientific justification" (117)(p7) and perhaps owed more to the official church teachings of John Wesley (1703-1791 AD) and Saint Augustine (354-430 AD) who aligned infant behaviour with depravity. (117)

m "The Karicare brand is the legacy of Dr Truby King, the founder of the Plunket movement in New Zealand and the Karitane Product Society. Dr King's motto, "help the mothers and save the babies" inspired him to develop and manufacture hygienic and nutritious infant formula, and for almost 100 years the name Karitane or Karicare has been a familiar and trusted name in infant products." http://www.nutricia.com.au/infant00.asp accessed 4 May 2007 n http://www.britishlistedbuildings.co.uk/en-494491-elizabeth-house-hornsey

50 In the 1950s, Queensland's Department of Health and Home Affairs published Mother and Child, which "described the ingredients and preparation methods for home- modifying fresh cow's or goat's milk." (63)(p85) This practice was supported by popular USA paediatrician and author Dr Benjamin Spock, "By formula, Spock literally meant a recipe." (63)(p86) A study of 217 Sydney mothers in 1972 found that breastfeeding rates had fallen from 80% at birth to 20% at three months; a homemade modified cow's milk 'formula' was used by 40% of mothers. (98) "It was only at the end of the 1950s that competing brands came onto the Queensland market," notes a history of marketing from this era. (79)(p67)

A 1963 series of advertisements in the Australian Home Journal, assumed that the infant was not breastfed, heading it "Baby's First Course - Cereal or Bottle?" (63)(p87)

... there are some babies who are so fond of their bottle that they resent a more solid food till they've had a good drink of milk. Others happily lap up cereal first, then finish off with formula".

The perception of formula as normative, equally producing healthy vigorous infants like breastfeeding, and suitably enjoyed by infants persisted 40 years later in the book Motherguilt. (3)(p66)

Babies can thrive on breast milk; they can also thrive on formula. The 'breastfeeding Nazis' may think they know it all, but in fact, when it comes to how it wants to be fed, a hungry baby knows best!

Such backlash against moralising breastfeeding promotion acts to perpetuate a misunderstanding of infant behaviour and development. It alienates mothers from potential sources of support, and may undermine maternal confidence in breastfeeding.

Breastfeeding responsively or for comfort has been discouraged in an individualistic and hierarchical society. Societal change has also reduced the realm of influence for many women, as attitudes towards the value of mothering have changed. (5) One mother felt uncomfortable about her friend still breastfeeding a 12-month-old infant. (39)(p40)

I just get that feeling that her baby is just so dependent on her being there... it'd be good for the baby up to 6 months to have the best and you know, be closer to you and everything, but after that I sort of feel it would be nice to have a bit of independence, not being there all the time.

Many mothers faced persistent but uneducated advice to resist an urge to respond quickly or attentively to infant distress, (9)(pp69-73) and to avoid extensive carrying, as these may "spoil" a baby; mothers were told to allow the baby to "cry it out." (118) This advice demonstrates ignorance of the physiological interaction between mother and infant that governs milk supply. It demonstrates ignorance of the relationships between infant mental health, infant distress, infant cortisol levels, and long-term metabolic development. (119) Infant cries are at times attributed to an attempt to manipulate the parent, although this would require the cognitive sophistication of an older child's neurological development. (31)(p171) Dependence and vulnerability are poorly tolerated and independence is valued in contemporary developed countries, (96) contributing to lactation failure.

The role of mother-infant bonding in infant development may be poorly understood by parents. In the 2010 Australian national health survey, 29% of mothers gave as their reason for not breastfeeding "so my partner can share feeding." (69) The needs of the adult male partner for nocturnal intimacy have been argued as greater than and prohibiting the needs of infants for nocturnal food, assistance or comfort. (31)(p175) This argument may arise from a poor understanding of normal lactation management or of the links between ongoing breastfeeding and infant development.

1.4.4 Promotion with a research bias of "best" and "benefits" Language such as “breast is best” does not convey that breastfeeding is the physiological norm against which substitutes are to be considered for their adequacy or risk. (28) As this is the common parlance of promotion, (53) promotion is not commonly worded effectively.

It is noteworthy that research findings commonly posit the intervention as breastfeeding, compared to a control group in which formula feeding is the standard. Breastfeeding promotion that refers to evidence thus commonly describes "benefits" of breastfeeding, (15) although an examination of health professional communication reported that "loss framing" is more effective in influencing health behaviours than "gain framing". (120)

Titles and abstracts of research articles rarely describe formula feeding as risky. Moreover, their cautious tone means that associations are reported with misleading expressions such as "Breastfeeding and the sudden infant death syndrome". (15)

52 Contaminants have been found in formulas at various times (30)(p391ff) (31)o and crucial nutrients have been omitted unawares or through errors. (31, 83) These risks are rarely mentioned in breastfeeding promotion.

Health recommendations and policies are formed in response to research findings. However, research funding does not assume the primacy of breastfeeding over substitutes. High-level research on health outcomes is costly and funding priorities are influenced by medical priorities, academic benefits, public interest, and industry involvement. In the three-year period 1994-1996 in the USA, $40.4 million was awarded in funding for infant nutrition/breastfeeding/lactation research. Of these funds, 13.7% was "awarded to projects determined to have either a direct or indirect impact on ... increasing the incidence and duration of breastfeeding". (121)

1.4.5 Promotion methods of health professionals

Knowledge and attitudes Despite the health discourse and the deference to health professional opinions, health professionals are not skilled at breastfeeding promotion. Knowledge about the importance of breastfeeding and practical knowledge of lactation management have consistently been shown to have significant gaps, and lactation is not consistently covered in professional training. (122-127) Surgeons who offer cosmetic breast surgery do not consistently ensure that women are adequately informed of potential barriers to breastfeeding. (128, 129) One UK professional speaker recommended introduction of solids from 4 months of age, based on her experience with her first child rather than relying on her training as a paediatric dietician. (130)

Australian general practice registrars in a recent survey (131)(p284) were unlikely to promote breastfeeding.

...nor did the participants consider it necessary for a GP to ensure a mother had sufficient information to make an informed decision: "The role of the GP is not necessarily to try and change how they think, [but to] support what they want to do."

o The plastic, melamine, was unethically added to milk including formula in China, as discovered in 2008; known to have caused 6 child deaths and 300,000 cases of child kidney stones. http://www.time.com/time/world/article/0,8599,2010044,00.html accessed 17 February 2012

Personal experience of being the parent of a breastfed child was the main source of knowledge for many health professionals, whether positive or negative. (132-134) Training opportunities for health professionals are sporadic (135) and may be of lower quality than training programs in voluntary organisations.

This knowledge gap has an impact on the capacity of health professionals to promote breastfeeding. In one large Western Australian study (n = 2669), "unhelpful information from child health nurse" was significantly associated with breastfeeding cessation before 10 weeks for both multiparous and first time mothers. (136) Many mothers reported difficulties with the range of feeding advice presented by health professionals. The Australian Parliamentary Inquiry cited 34 submissions on this point and quoted one example. (9)(p100)

I left the hospital feeling very confused and rather alone in this brand new world of babies and breastfeeding.

Knowledge gaps may result in ill-advised advice to wean in the face of challenges, (137) or when a medication is prescribed that is safe to use with breastfeeding. (138)

Relegating breastfeeding to "what they [mothers] want to do" absolves the health professional of the need to seek professional education. In a USA study at paediatricians' offices, neither the office nurses nor the paediatricians raised the subject of breastfeeding with the mothers of their paediatric patients. Office nurses (of whom none had breastfed) considered breastfeeding a medical issue, while the (male) paediatricians assumed breastfeeding support from the (female) nurses would be most appropriate and skilled. Both declared that the practice was supportive of breastfeeding although their patients generally disagreed with this assessment. (106) Health professional ambivalence towards promoting breastfeeding, to avoid inducing guilt, was noted in both these Australian and USA studies, with the USA authors concluding:

They thought that offering more support to breastfeeding mothers would make those who weaned feel “guilty.” It would be “unfair” to those mothers who bottle-fed, according to many [healthcare professionals], because they would not receive as much praise for their mothering efforts as breastfeeding mothers.

The public health imperative is poorly understood. (31)(p45)

...there is a degree of professional ignorance which is historically quite understandable, but no longer tolerable. In a society where skill and

54 money can produce heart transplants and coronary bypasses, ignorance about preventative medicine such as breastfeeding is inexcusable.

Communication skills Mothers are vulnerable to cultural expectations, including the accepted place of breastfeeding promotion within structured advice given in medical settings. (40) However, as Graffy observed in his study of 720 British women, "They want to be listened to and encouraged without feeling pressurised." (139)(p185) Mothers had a higher recall of nutrition messages in a Brazilian study where 30 physicians demonstrated the counselling skills they were taught in the intervention. (140) When Swedish nurses had extensive additional training in how to communicate with mothers, the increase in intimacy between mother and baby was both perceivable by mothers and observable independently at three days and at nine months post-partum. (141) Effective health promotion relies on skilled communication at an interpersonal level.

Deterred by their minimal understanding of breastfeeding management and by the moral implications of common breastfeeding slogans, many health professionals have no language with which to promote breastfeeding.

1.5 SUPPORT FOR BREASTFEEDING Breastfeeding support is required at the structural level and at the personal level. A major barrier for increasing breastfeeding rates is the lack of structural or community support for breastfeeding. (92, 95) Support, like breastfeeding, involves a range of disciplines. The need for breastfeeding support is greater where protection is inadequate and where promotion fails to create the required society-level changes. Support was explored with UK adolescent mothers and the findings may be more widely applicable. (97)(p398)

...no single aspect of support was acceptable in isolation. The adolescents needed a synergistic combination of emotional, esteem, instrumental, informational and network support.

One attempt to overcome some of the "silo" effect of different disciplines involved in breastfeeding advocacy was the Baby Friendly Health Initiative. (142) The increased

duration of breastfeeding nationally in Switzerland was attributed to the BFHI in a large national study. (143) Accreditation with the BFHI required addressing health worker training, organisational structures and continuity with community women's support groups. BFHI for community-based services, places where mothers of babies older than six weeks seek information and support, has barely begun in Australia. Even in the hospitals, BFHI practices have been found to be incomplete. (144) In rural Australia there are fewer BFHI accredited hospital maternity services, and breastfeeding rates are lower than in urban areas. (Hector, cited (74)(p42)

Other contributions to inter-disciplinary breastfeeding support include restrictions on infant formula marketing, legislation for paid maternity leave, and support for breastfeeding-friendly workplaces. (142) However, as described above, these are only partially enacted in contemporary Australia. In view of the sparseness of these structural supports, women are acutely in need of personal support when they are breastfeeding.

1.5.1 Personal support needs to be offered Support is needed at the personal level in order for women to have "faith" in their milk as sufficient for their growing babies. (145) Mothers’ attitudes, confidence, and satisfaction play a significant role in their relationship and activities with their babies. (146) Self-efficacy and the confidence to declare a planned behaviour can be reliably measured and are related to feeding outcomes. (147) If a mother perceives herself inadequate in the face of her baby’s needs, her baby’s behaviour or in the opinion of others, she is likely to doubt the adequacy of her breastmilk. (148) Hence, the reason given for ceasing breastfeeding earlier than planned was “not enough milk” for 55% of Australian women in one cohort, (149) as is widely reported elsewhere. (150, 151)

Self-confidence and self-perception also influence help-seeking behaviours, (152) as does postnatal depression, which is associated with less help-seeking (153) and with earlier cessation of breastfeeding. (154, 155) The contribution to breastfeeding rates at 6 months of breastfeeding plans, expectations, and “faith in breastmilk” are greater than that of socio-demographic factors. (145, 156) While breastfeeding peer support services provide excellent care, their efficacy is limited by whether mothers access them. (157) Strategies to support maternal confidence and breastfeeding confidence are needed.

56 Trials of peer counsellor support are instructive. A Canadian study (158) provided peer telephone support and, after the initial contact by the peer counsellors, mothers were encouraged to make contact as desired. Although the majority of mothers received five or more contacts (mean 5.4 [SD 3.6]), only 9.3% were initiated by mothers. Likewise a large UK peer study (159), where intervention group mothers were encouraged to contact trained peer counsellors, found that only 53% of women attempted to do so, while 14% of the control group also did so. It seems that despite a need for support indicated by breastfeeding rates, and an appreciation of support when accessed (159), mothers did not commonly seek breastfeeding support outside their families. In a smaller Australian study, 25% of mothers with an identified need did not seek help. (160) An important factor to consider in provision of breastfeeding support is that mothers who may agree that they need help or support do not appear to seek it. (153)

Help-seeking behaviour may be limited by many factors. Knowledge limitations include: unawareness that a solution may be possible due to low knowledge (161); unfamiliarity with help-seeking among young women with no prior medical history (162); or previous unsatisfactory experiences with help-seeking (106) particularly if a "foolproof recipe for success" failed for them. (163)(p232) Interpersonal limitations include perceiving support services as targeted at others such as middle class white mothers, (164) or reluctance to seek help from an unfamiliar person. (97) Personal limitations include: a perceived need for stoicism and self-reliance in the face of adversity among rural women (165);fear of critical or negative responses (105); or shame at not achieving their own cultural expectations of ideal mothering (105, 152) regardless of competing pressures from societal change. (5)

1.5.2 Support for the embodied experience Women need support as they travel the "engrossing, personal journey" through the changing circumstances that motherhood comprises. (20) Breastfeeding is a physical and relational experience for women and for infants, integrally linked to mothering. Women differ widely in their responses to this new experience of spinning through of powerful emotion, intimacy, responsibility, sensuality, physiological reflexes and at times intense pain. Breastfeeding mothers encounter these changes in a setting of much attention from family, friends, and even strangers, which adds to the challenge. Author and poet Adrienne Rich writes about her embodied experience in Of Woman Born. (166)(p36)

Nothing could have prepared me for the realisation that I was a mother....That calm, sure, un-ambivalent woman who moved through the pages of the manuals I read seemed as unlike me as an astronaut. Nothing, to be sure, had prepared me for the intensity of relationship already existing between me and a creature I had carried in my body and now held in my arms and fed from my breasts.

As Rich suggests, mothering and breastfeeding are experiences fraught with ambivalence, extremes of emotion, and periods of great personal uncertainty, quite contrary to popular cultural views of motherhood.

A pleasurable and sensual experience of breastfeeding is described by many women and for Rich the intimacy of breastfeeding was acknowledged and accepted. (166)(p31)

I recall the times when, suckling each of my children, I saw his eyes open full to mine, and realised each of us was fastened to the other, not only by mouth and breast, but through our mutual gaze: the depth, calm, passion, of that dark blue, maturely focused look.

Others find the physical sensation unpleasant, unbearably painful or disturbing. (102)(p72)

It was more of a bitey feeling, like getting nibbled at. When it was happening, I'd be looking at her and she'd be looking at me, and I just felt uncomfortable.

Some women find it frightening to rely on their own bodies to meet their baby's needs; some are unable to adapt to a responsive rather than a measured life; some absorb and are guided by cultural values about the sexual breast; and some find their need for privacy or acceptance of cultural values prevent them from breastfeeding. (102)(p72)

I just thought, 'No way'. And I'd look at all the other mothers when they were breastfeeding and think, 'Oh, I'm glad I'm not doing that.'

A woman’s attitude to her own sexuality may inhibit her interest in breastfeeding. Body image or history of sexual assault or childhood trauma may cause an aversion to intimate physical contact as well as a loss of resilience and a loss of breastfeeding confidence. (148, 167) While there are cases of women who persevere with breastfeeding and overcome these barriers, (167) these factors may contribute to the sense of relief expressed by some women when they cease breastfeeding. A series of

58 interviews with Australian women from late pregnancy to 6 months after the birth revealed that for some women, this embodied experience was "connected, harmonious and pleasurable" while for others it was "disruptive, unpleasant and violent". (168)(p325)

1.5.3 Support from family The mother's partner is identified as the major source of personal breastfeeding support in many studies (169) (170-172) although there are differences with ethnicity. In an older USA study, black women identified a friend and Mexican-American women identified their mother as their primary support, (173) as did an adolescent mother in a UK study. (97)(p397)

I wanted my mum around because she's breastfed two children herself... I can remember her feeding my youngest brother.

Partner support is important to gay breastfeeding women,p while the encouragement of health professionals is particularly influential for breastfeeding women without partners. (174)

Partner support needs to be both practical and interpersonal. (170) Many Australian fathers are uncertain in their role, wary of negative community perceptions about men as carers or men being with children, according to NSW fathering researcher Richard Fletcher. He reports that men rely on role models to develop their approach to support, as they have few avenues for learning father-care. (175) A recent Western Australian study identified that both mothers and fathers feel "Dads do make a difference" in supporting breastfeeding. Mothers appreciated their presence, positive encouragement, and commitment; fathers reported difficulties finding their way, with a desire to advocate for their partner and child but with little targeted information or guidance. (176) Fathers presented an alternate perspective, but one that acknowledged the challenges faced by their partner. (176)

I guess there's still that well, it's like shame, and you don't want everyone looking at things [breasts] that have been private. And

p http://www.dailymail.co.uk/femail/article-1267923/Lily-Britains-baby-women-parents-birth- certificate--Mummies-tell-Dad.html

suddenly you've gone from being a sexual thing to a kitchen utensil. (Eric, age 38)

Public discussion and images of caring fathers may reference bottle feeding. "New Zealand was in the throes of a breast-feeding controversy this week, thanks to a rugby player shown doting on his daughter," reported the global online edition of The New York Times. An anti-smoking commercial filmed in Piri Weepu's home initially captured what the agency described as "a nice little poignant moment" of him bottle-feeding his 6-month-old daughter. However, in a routine government checking process "La Leche League expressed concern that the images were inconsistent with the Government's “breast is best” campaign and asked that it be removed." A local newspaper protested, “It was almost saying that what Weepu was doing was wrong” and online diatribe lauded the parenting of bottle-feeding fathers. A few voices protested that the bottle may contain expressed breastmilk, but common perception was reinforced by The New York Times, which noted the extensive use and acceptance of formula feeding by Maori and Pacific Islander mothers.q The needs for both regulatory protections and community-based promotion to support breastfeeding are confirmed in this example.

Family and friends need to be more aware of the importance of breastfeeding, and know more about how to help mothers to continue breastfeeding. (157, 177) Family may discourage women from ongoing breastfeeding based on the age of the infant or child. (105)(p67-70)

Mum is of the opinion that they should be off the breast by 4 months anyway and when we went to [another Australian state to visit] that's what she said, 'Are you still breastfeeding her?' (Mary)

I was also getting heaps of pressure from my husband by that stage [18 months] to stop. He, I think, had a bit of a hang-up basically with her breastfeeding when she was an older child, which I didn’t really have... And then towards the end he’d say...‘We’ve got to stop this... This has got to stop. When are you going to stop? She’s still going to do it when she’s five years old.’ (Helen)

Age-based comments may relate to a belief that breastfeeding is developmentally inappropriate as a child grows towards independence, ultimately defined as school- aged. q http://latitude.blogs.nytimes.com/2012/02/10/breast-feeding-controversy-hits-new-zealand- press/ accessed 17 February 2012

60 1.5.4 Support from peer organisations Some mothers access organised peer breastfeeding support. Attempts to improve support for women have come through community organisations with trained volunteers, internationally La Leche League, at the national level the National Childbirth Trust (UK) (159) or the Australian Breastfeeding Association, (6) and smaller regional groups such as Halton Breastfeeding Connections in Toronto, Canada. (158) Peer organisations may provide personal or structural support. (77)(p39)

Non-government organisations such as the Australian Breastfeeding Association, traditionally known for their provision of telephone counselling services and peer support for new mothers, are beginning to move into environmental support strategies such as breastfeeding- friendly facilities in public places and workplace provisions for breastfeeding mothers.

A variety of training was available to volunteers in these lay organisations. While many women found sufficient knowledge and acceptance of breastfeeding in the subculture of their own family or friends, (97, 178) some established trusting supportive relationships through such groups. (179)

1.5.7 Support for choice The discourse about health risks that informs "breast is best" campaigns carries a moral implication that a "good" choice complies with health recommendations (146, 180) and that mothers' decisions are made according to health priorities in the first 6 months. (86)(p92)

Yet mothers who feed with infant formula ... make their decisions based on other ideas about child care, nutrition, and the maternal body, as well as on life priorities and material constraints that may preclude breastfeeding.

The health imperative is supported by governments, who are particularly responsive to potential economic benefits; they may also be aware of the difficulties of achieving system-level change. (9, 73)

Breastfeeding promotion is ethically complex. Reinforcing the status quo rather than addressing the cause of the problem, through simplistic public health messages, "may easily slip into moralizing and blaming mothers for their infant feeding decisions." (99)(p122) "While it is 'known' that breastfeeding is better, our society is not structured

to facilitate that choice." (181) In this context, the "breast is best" message needs to be tempered by acknowledgement that breastfeeding women face many barriers, as discussed above. (53)

Mothers often fail to identify the causative systemic barriers and inadequate support for their lactation failure, resulting in a sense of personal failure or guilt. The physical capacity to lactate and nourish her infant is present for almost all women. (30) A more logical response to lactation failure may be grief or anger: without baby-friendly child- birthing practices or being shown how to attach their baby successfully, a mother's choice to continue breastfeeding may be an impossible choice. Yet self-blame is more common than anger among mothers who face this "choice". (8)(p17)

I found it awful and it hurt tremendously. One night, I decided to stop breastfeeding, even though I thought myself to be a bad mother, because that was the way it felt.

According to one review, “A major reason for the early use of formula is mothers’ perceived difficulty with breastfeeding rather than maternal choice." (158) Postnatal separation of mothers and infants is common following Caesarean Section births, and the resultant difficulty in establishing breastfeeding is associated with increased use of substitutes. Another review concludes “many mothers stop breastfeeding before they want to.” (45) This is a particular concern as caesarean section rates rise. (182)

The dominant discussion in developed countries is one of choice. (11) Infant feeding is presented as a personal matter of maternal choice and maternal control, with her baby and her breasts presented as potentially unwilling partners should she choose breastfeeding. (178) The language of choice has moved the discourse on breastfeeding from a "mutual physical dependence" and a "dynamic relationship" (183)(pp201-202) to a search for consumer satisfaction.

Both the relationship and the product used to nourish the infant are now commercially defined. Product constituents are described so women can "make their 'rational' choice accordingly". (99)(p122) In a society "where financial productivity is highly valued," (96)(p78) women may also choose to return to paid employment. Should a mother attempt to combine her paid work and mothering roles, she may face condemnation, such as the online feedback to Australian politician Kristie Marshall whose actions revealed her choice. (40)(p77) Because breastfeeding is not viewed in society as

62 "normal", a choice to maintain breastfeeding requires extraordinary effort for many women.

This perspective of choice fails to offer, to those women who desire it, any opportunity to sustain the "relational closeness in a unique way" (183)(p201) that defines the breastfeeding relationship for many women. Instead, the message of choice on the one hand devalues breastfeeding as a minimally important dietary option, while on the other hand it offers formula as a caring equivalent. (11)

[M]others make good choices when they follow their sensibility, or 'heart,' rather than their heads... current rhetoric from infant formula company websites articulates choice defensively... The response, then, is to argue that it is love that counts in infant feeding and care, not what goes into the baby.

Many women do intend to breastfeed and plan or have expectations about breastfeeding goals, (178), which are acknowledged by theoretically based interventions. (184-186) Women's breastfeeding goals may change as their understanding of breastfeeding changes or as their baby grows older. (108) Each day women face the choice of whether to continue breastfeeding, and are unable to rely on societal support unless they have the support of their partner.

The place of support is articulated by a World Breastfeeding Week publication from 2011. (187)(p142)

Communications all around should reassure mothers that they are not alone in their efforts to successfully breastfeed their infants. The feeling that they have the support of all people around them should encourage and empower mothers to successfully breastfeed their children and give them a good start to life.

1.5.5 Support from health professionals With the repeated demonstration of their poor understanding of lactation management, (124, 125, 133, 188) it appears that many health professionals are poorly equipped to support breastfeeding. As discussed above in Section 1.4.5, health professionals have come to be seen as experts on infant nutrition. Factors leading to this situation include historical advocacy for infant health and wellbeing, (63) (113) the medicalisation of

infant feeding, (99) and the ambiguous links between formula company marketing and doctors. (189-191) Formula company marketing continues to encourage women to seek health professional advice on how to feed their infants,r The views of health professionals have a significant influence on mothers' breastfeeding plans and outcomes. (155, 192)

A minority of mothers receive proactive breastfeeding support once they leave maternity services and continue beyond the neonatal period. This is despite the observation in a recent metasynthesis of qualitative studies that women "appreciate professionals who are proactive". (179)(p56) In her study of Australian general practice registrars, Brodribb noted that doctors felt an obligation to appear neutral about feeding methods and were "reluctant to share their professional opinion with mothers". They did not tend to address breastfeeding in any detail. (131)(p284)

'...like when I first see them after [birth] I ask how it's all going and if it's not a problem for them, that's sort of all I've left it at.'

Health professionals may describe themselves as supportive of breastfeeding. This attitude needs to be clearly communicated to mothers. In one study, (193) women whose physician advocated breastfeeding were more likely to plan for longer breastfeeding duration, than those whose physician expressed no preference. For the many women in this study who were uncertain about breastfeeding plans, a neutral attitude from their physician was clearly associated with weaning in the six weeks after the birth. Women gain negative impressions of support from an assumption that they are formula feeding, (97) or from the failure of professionals to give even the simplest encouragement. (106)(p426)

'[When breastfeeding a baby in the exam room, the pediatrician said] “What a lucky baby.” This was the first time anyone, any doctor or nurse anywhere . . . has said anything.' [Mother]

Assumptions of formula feeding are more common as the baby grows older, with only 22% of doctors in one study thinking to ask whether the mother of a toddler was breastfeeding before prescribing medication for the mother. (133) According to one literature review, (194)(p19) r http://www.infanurture.com.au/hungry.asp, https://www.wyethnutrition.com.au/ http://www.nestlebaby.com/au

64 Health care professionals may be a negative source of support if they provide women with inconsistent, inaccurate, or inadequate breastfeeding information.

Given the antithetical messages about ongoing breastfeeding from outside the health sector, support from health professionals is crucial. However, breastfeeding support from health professionals is limited by knowledge, attitudes and assumptions, time factors, (195-198) and communication skills. Women want their health professionals to communicate well and professionally. (199) Women want doctors and nurses to understand that they may have priorities and values other than distant health outcomes; to show them how their baby is happily thriving in their care, supporting their relationship with their baby; and to suggest options that may help them to continue breastfeeding. (139, 200)

1.6 OVERVIEW OF SYSTEMATIC REVIEWS Recent major systematic reviews indicate that supportive breastfeeding interventions can increase short-term and long-term breastfeeding outcomes. Mothers benefit from information, assistance with learning skills, and assistance with overcoming challenges. (45, 201) These and other systematic reviews (202, 203) call for increased support for breastfeeding mothers through new interventions, as well as effective translation of identified successful approaches into practice.

The tertiary setting has seen improvement in maternity practices through the Baby Friendly Hospital Initiative (BFHI); there is widespread support for interventions in primary care. (45, 201, 202) As de Oliveira et al. point out, "It is now time to build on the success of the Baby Friendly Hospital Initiative by extending it into primary care". (202) Britton et al. agree since they recommend "supplementary breastfeeding support as part of routine health service provision." (45) Chung et al. confirm that interventions based in primary care can increase breastfeeding rates. (201)

Although the focus of each review is slightly disparate, these four systematic reviews all included supportive breastfeeding interventions provided by health professionals, peer supporters/counsellors, or both. The authors noted that a combination of peer and professional support might be particularly effective for duration of breastfeeding.

Of the interventions considered in these reviews, those set solely in the antenatal period were less likely to have significant outcomes. For example, the large well-

designed study conducted in Melbourne, Australia (n=981) tested two antenatal interventions in a three-arm randomised controlled trial. The authors found no difference between the groups for duration and discussed whether this was related to a Hawthorne effect or perhaps also to a low intended duration, although intention data is not published. They commented on the challenge of demonstrating change in a setting of high baseline breastfeeding rates. (204) These reviews did note, however, that interventions that combined antenatal components with perinatal or postnatal components could be highly effective. Regarding postnatal support components, they found that health professional support was effective particularly for continuation of exclusive breastfeeding. However, professional support needed to be face-to-face. Postnatal telephone support seemed to be only effective when delivered by peer counsellor support.

In addition to breastfeeding outcomes, some studies also measured effects on acute health outcomes and in some trials; the specific health goal may have increased the motivation of mothers. The health importance particularly of exclusive breastfeeding has been previously established. (2)

All reviews noted the variability in interventions; the authors found direct comparisons between studies difficult due to different definitions, time points for data collection, periods of maternal recall, methods of analysis and outcomes reported.

1.7 DEVELOPING THE RESEARCH QUESTION

1.7.1 The need International research addressing the low duration of breastfeeding (including exclusively) is quite clear: mothers who begin breastfeeding will require support in order to continue. (45, 205) There is a need for innovative, evidence-based interventions to improve breastfeeding prevalence. While health professionals may be well-intentioned, support does not occur without knowledge, (206) hence interventions need to involve training that is well defined, reproducible, and effective. (157) In view of the breastfeeding recommendations, (30) interventions need to offer support to women who wish to breastfeed beyond 2 months. (45, 205)

66 1.7.2 Choosing the theoretical framework Motivational interviewing (MI) has been used successfully to achieve lifestyle change, a task that professionals frequently consider challenging and frustrating, (207) and more studies using MI to support breastfeeding mothers have been suggested. (186, 208, 209)

Conflicting values are barriers to communication; one study on the use of MI to reduce risk behaviours in HIV positive people noted the usefulness of an approach that avoided stigma and supported self-efficacy by identifying past successes. Strategies in this study included consistent MI training, the use of intervention providers who had extensive counselling skills, a protocol with freedom for the conversation to be directed by the client's areas of concern in order to build rapport, identifying which clients needed support with stress reduction, and planning for repeat visits.(210)

1.7.3 Choosing the setting and methods General practice, as the leading primary care discipline in NSW provides (211)(p7)

...a comprehensive approach to care, which includes disease prevention, community empowerment and multidisciplinary collaboration.

The general practice setting has the capacity to move away from simplistic messages and closer to the spirit of the Ottawa Charter for Health Promotion (WHO 1986), creating an enabling culture for change, as described in the Australian National Breastfeeding Strategy 2010-2015. (212) While primary health research is challenging, (213, 214) clustering by general practice is used increasingly to test interventions and to reduce potential for bias. (215, 216)

1.7.4 A potential role for practice nurses

Many rural Australian mothers have regular contact with their doctors' practice nurses, (211) and mothers with infants average several general practice visits before the infant reaches 6 months old. (Gunn cited p284 (131)) Australia has national funding incentives designed to support general practice care of familiess; where these are used, disadvantaged mothers, who have lower breastfeeding rates, may have access s The Medicare Australia "bulk billing" incentive, enables cost-free attendance by clients <16 years. http://www.medicareaustralia.gov.au/provider/incentives/medicare-initiatives.jsp

to a familiar care provider for regular child health visits. Mothers value a facilitative model of person-centred care in their relationships with health professionals, such as can be provided with continuity of caregiver. (179)

Practice nurses are accessible in rural areas, particularly since the introduction of targeted item numbers to fund practice nurse immunisation, in 2001-02. (217, 218) Numbers have risen rapidly in recent years due to incentives, as shown in Figure 1.2. In 2009, 87% of New South Wales general practices employed practice nurses. (211) Some patients view a practice nurse as more approachable than a general practitioner (219), which is important for the care of disadvantaged patients. (220)

Figure 1.2 Estimated number of practice nurses in Australiat

The nature of practice nursing Strongly held values of caring, both for patients' general health and for their emotions, guided practice nurses in their time use and their patient follow-up, according to a recent study. (221) The extent of nurses’ emotional work, and the value of this work to health outcomes, has not been fully explored. This "nurturing rationality" is hard work and requires skilful and delicate balancing of emotional connection with maintenance of professional role. (222)

In effective primary care, the focus is on teamwork so that “the work gets done”. (223) The changing role of Australian practice nurses is the subject of current discussion (218, 221, 224) and there are calls for more Australian research on current practice t http://www.phcris.org.au/fastfacts/fact.php?id=4824

68 and outcomes. (225) Practice nurses are “specialist generalists”, (223) whose roles cover clinical care, clinical organisation, practice administration, and integration.

High quality patient care is achieved across diverse practice styles. (223) General practices have diverse settings and business orientation. (226) More recently, one study explored the structural and values aspects of practice nursing. Within the "system" of daily nursing activities, employment conditions and their organisational role, these authors found that “nurses are creating a quality practice that is of their own devising”. (221)(p11)

Challenges to scope of practice for practice nurses There remains a need to address “funding of services, inter-professional issues, medico-legal concerns and a poorly defined scope of practice… and also the mentoring of these isolated professionals." (227)

Australian general practices are largely private businesses and heavily reliant on Medicare Australia item number funding u. (223) Most practice income still requires general practitioner attendance. (224) "Current funding systems often fail to recognise the importance of nurses’ contributions to quality and safety in primary care." (221)(p12)

The position of practice nurse has few established structures or traditions in Australia. Practice nursing has only recently achieved recognition as a discipline, with annual conferences by The Royal College of Nursing starting in 2003. The "fledgling practice nurse professional leadership" established the Australian Practice Nurses Association (APNA) in 2001, (227) conducting annual national conferences only since 2009.v

Enhanced nursing roles in preventive health care

University-based education since the 1980s supported more complex nursing roles and tasks. (228) Nursing interventions are effective for smoking cessation, (229) and a large Australian cluster RCT demonstrated effective nursing care in chronic obstructive pulmonary disease. (230) A greater role in health promotion would be supported by u Of services provided to Australian patients, 67.5 - 74.3% were bulk billed to Medicare Australia the universal national health insurance in the decade to 2009/2010, http://www.health.gov.au/internet/main/publishing.nsf/Content/medstat-dec10-tables-aa v http://www.apna.asn.au/

many general practitioners (219) and would interest many practice nurses, who currently are involved in health assessment, education and promotion, as well as research, (231) particularly with the support of their local Division of General Practice.w

This thesis describes a cluster randomised controlled trial, in a routine primary care setting, using practice nurses to deliver a motivational intervention. The intervention aims to provide appropriate support to breastfeeding mothers, using familiar professionals who are well informed about breastfeeding management, so that rates of exclusive and any breastfeeding may be increased.

1.8 CONCLUSION TO CHAPTER 1 Breastfeeding, although important to the health and wellbeing of mothers and their babies, requires more support. Current practices are less than optimal. Past attempts to promote breastfeeding have suffered from a failure to address cultural, structural, and commercial biases that favour infant formula use. Support must be tailored to the individual experience and motivations of each woman. Potentially, general practice nurses could offer such support during routine practice visits if they had an adequate knowledge base.

w reconfigured from 2011 as "Medicare Local"

SYSTEMATIC LITERATURE REVIEW OF RANDOMISED CONTROLLED TRIALS SYSTEMATIC LITERATURE REVIEW

Chapter 2

Systematic literature review

INTRODUCTION

The need for breastfeeding support As discussed in Chapter 1, mothers need support to continue breastfeeding (45, 73, 108) in a setting of cultural, structural, personal and health system challenges. (20, 95, 106, 151, 232) Commercial interests favour and promote substitution of breastfeeds with other foods, particularly bottles of manufactured dairy milks. (4, 79) This is detrimental to maternal and infant health (2) and contrary to women and children's rights (4, 11) however substitution is ubiquitous. (69) Effective approaches to address this situation need to be identified and instituted.

Intervention provider training The reviews by Britton et al. and Spiby et al. (45, 203) call for more research into appropriate training for intervention providers, while Chung et al. call for more research with trained health professionals. (201) In calling for a primary care extension of the BFHI, de Oliveira et al. (202) imply the need for training of primary health professionals; this is a mandatory part of the BFHI accreditation process. (233) There is scope for enhanced health professional support for breastfeeding mothers, in order to facilitate more ongoing breastfeeding. The capacity of the primary care workforce to meet this need has not been specifically addressed. The importance and nature of health professional education towards this goal has not been explored.

Two international and standardised accreditation processes exist regarding health professional support for breastfeeding mothers, and these are used in some of the studies in this review. Strong links exist between many trained personnel associated with these processes: firstly, certification of individuals by the International Board of Lactation Consultant Examiners (IBLCE) and secondly, accreditation of facilities by WHO and UNICEF jointly for the BFHI (see below).

In Breastfeeding Management (Appendix 2) Brodribb describes the emergence of international board certified lactation consultants (IBCLC), with certification exams held in Washington DC USA and Melbourne Australia in 1985, under the IBLCE. (6)(p487ff) While various training options are suitable, applicants are currently required to have at least 90 hours of lactation education and 1,000 hours of breastfeeding counselling experience over five years before applying for the certification examination.x x www.iblce.edu.au/

72 Figure 2.1 Ten Steps to Successful Breastfeeding (BFHI)y 1. Have a written breastfeeding policy that is routinely communicated to all healthcare staff. 2. Train all healthcare staff in the skills necessary to implement this policy. 3. Inform all pregnant women about the benefits and management of breastfeeding. 4. Place babies in skin-to-skin contact with their mothers immediately following birth for at least an hour and encourage mothers to recognise when their babies are ready to breastfeed, offering help if needed. 5. Show mothers how to breastfeed and how to maintain lactation even if they are separated from their babies. 6. Give newborn infants no food or drink other than breastmilk, unless medically indicated. 7. Practice rooming-in, allow mothers and infants to remain together 24 hours a day. 8. Encourage breastfeeding on demand. 9. Give no artificial teats or dummies to breastfeeding infants. 10. Foster the establishment of breastfeeding support and refer mothers on discharge from the facility.

Brodribb also gives a history from the joint WHO/UNICEF statement (1989) on Protecting, Promoting, and Supporting Breastfeeding: The Special Role of Maternity Services, which was synthesised into ten practical steps, (See Figure 2.1) "against which hospitals and maternity facilities can be assessed and designated 'Baby Friendly'". (6, 234, 235) In the Ten Steps and in the more recently developed Seven Point Plan for Sustaining Breastfeeding in the Community, (See Figure 2.2), Step Two advises training for staff "in the skills necessary to implement the policy." With the inclusion of a primary care focus, the structure is flagged in the UK as UK Baby Friendly Initiative (BFI) and in Australia as Baby Friendly Health Initiative (BFHI). A package of tools including suitable BFHI training program curricula (233) is available to assist health facilities in the UK to prepare for accreditation, while Australia's community program has barely begun.z The curricula include training in breastfeeding management skills, knowledge and attitude training as well as goal-directed but

y http://www.babyfriendly.org.au/ z www.unicef.org.uk/babyfriendly/ http://www.babyfriendly.org.au/

patient-centred counselling skills training.aa Efficacy of the Baby Friendly initiative was confirmed through the PROBIT study. (17) Capacity of the training program to effect improved provider knowledge and attitudes was confirmed in related studies. (236)

Figure 2.2 Seven Point Plan for Sustaining Breastfeeding in the Communitybb 1. Have a written breastfeeding policy that is routinely communicated to all healthcare staff. 2. Train all staff involved in the care of mothers and babies in the skills necessary to implement the policy. 3. Inform all pregnant women about the benefits and management of breastfeeding. 4. Support mothers to initiate and maintain breastfeeding. 5. Encourage exclusive and continued breastfeeding, with appropriately- timed introduction of complementary foods. 6. Provide a welcoming atmosphere for breastfeeding families. 7. Promote co-operation between healthcare staff, breastfeeding support groups and the local community.

2.1 AIM This review aimed to identify key education strategies and theoretical perspectives associated with improved breastfeeding outcomes that were used in the conduct of postnatal, primary care, health professional interventions. It aimed to produce findings to guide the design of an evidence-based intervention for this primary care study.

This systematic literature review of randomised controlled trials with a postnatal timeframe sought high-level evidence from interventions suited to routine primary care. It examined trial evidence on a range of preparatory education strategies for health professionals, tested in interventions to support ongoing breastfeeding in these settings.

aa www.who.int/nutrition/topics/bfhi/en/ www.unicef.org/programme/breastfeeding/baby.htm bb http://www.unicef.org.uk/BabyFriendly/

74 2.2 METHOD

2.2.1 Search strategy In 2007 and again in August 2011, searches were conducted from the start of the databases for English-language articles. (See Figure 2.3) Internationally published studies were sought, testing interventions delivered by a variety of primary health professionals, from MEDLINE, EMBASE, and CINAHL databases. Medical subject heading (MESH) terms and keywords were used for breastfeeding or lactation, time factors, health promotion, education, or personnel, and for clinical trial.

Figure 2.3 Example of systematic search strategy Database(s): Ovid MEDLINE(R) 1946 to Present with Daily Update Search Strategy:

# Searches Results 1 exp Breast Feeding/ 22926 2 exp Lactation Disorders/ or exp Lactation/ 33223 3 breastfeed$.mp. 10226 4 breast feed$.mp. 26549 5 nursing mother$.mp. 760 6 exp Time Factors/ or exp Time/ 1023864 7 cessation.mp. 50128 8 stop$.mp. 78778 9 duration.mp. 327255 10 continu$.mp. 617022 11 exp Health Occupations/ 1206570 12 exp Health Personnel/ 331872 13 exp Health Promotion/ 44452 14 exp health education/ 123241 15 1 or 2 or 3 or 4 or 5 59740 16 6 or 7 or 8 or 9 or 10 1925672 17 11 or 12 or 13 or 14 1562191 18 15 and 16 and 17 1317 19 clinical trial/ 465552 20 randomized controlled trial/ 317921 21 controlled clinical trial/ 83363 22 19 or 20 or 21 583165 23 18 and 22 135 24 limit 23 to (human and English language) 130

Titles, abstracts and then full text articles, as shown in Figure 2.4, were assessed for inclusion, and assessed against the exclusion criteria.

Figure 2.4 Systematic literature review flow diagram

Relevant articles identified from MEDLINE, EMBASE, and CINAHL (n = 305)

Abstracts excluded (n = 222) Duplicate article or study 38 Not RCT 128 No RCT breastfeeding outcome 26 No post-discharge component 30

Full-text articles assessed for inclusion (n = 83)

Articles excluded (n = 51) Duplicate article or study 3 Not RCT 9 No RCT BF outcome ≥6 weeks 11 No post-discharge component 13 No Health professional delivery 12 Not comparable to Australia 4

Articles reviewed for provider training and significant outcomes (n = 31)

2.2.2 Inclusion criteria Studies of interest had postnatal components delivered by a health professional in a primary care setting, namely by telephone, by home visiting or at a clinic. They compared breastfeeding rate or duration in the intervention group against a control group. Randomisation offered each mother equal opportunity to enter either group. Breastfeeding outcomes used a range of definitions, time frames, and analytic methods. Health, satisfaction, and process outcomes were noted but were not the focus of this review.

76 2.2.3 Exclusion criteria

 Duplicate  No randomised controlled trial  No post-discharge component  No breastfeeding outcomes > 6 weeks assessed from a randomised controlled trial  No health professional postnatal delivery  Not comparable to Australia (> 30% difference in initiation or duration, documented in the country, population group or the control group)

Breastfeeding mothers in primary care were considered those who had returned from hospital to their community with their babies. Britton et al. (45) found that only the initiation of breastfeeding but not sustained breastfeeding was affected in studies that did not include a postnatal component. In Australia, tertiary-based maternity units are funded to provide universal maternal and infant care until infants are aged six weeks; transition to community-based care occurs over the first eight weeks after the birth. (212)(p6) Women are advised to access primary care services after this, although there is no barrier to earlier access. Interventions demonstrating outcomes only prior to six weeks could not reliably be compared to the Australian primary care setting.

Interventions were excluded if their target population had breastfeeding rates that were markedly different to Australian population outcomes. Breastfeeding outcomes in the study control group were used for this criterion where relevant, as the target sub- populations of some studies may have comparable breastfeeding rates to an Australian population.

2.2.4 Analytic approach

Organisation of results Studies were grouped according to the preparatory training of the intervention providers. They were further grouped by the country in which the intervention was delivered, for ease of comparison with an Australian population, and to indicate background breastfeeding rates. The characteristics of included studies are shown in Table 2.2, compared by population characteristics and size; the timeframe, setting, and frequency of intervention contact with the mother; trial goals and any use of a counselling framework in the intervention. Breastfeeding outcomes are compared by significant breastfeeding outcomes, outcome measures and effect sizes, and study

quality. Effects were discussed in more depth in the text and compared descriptively. Meta-analysis was not performed due to study variability and variability in reporting. Studies with significant breastfeeding outcomes were described individually; key points were drawn from studies without significant breastfeeding outcomes.

Assessment of quality Each included study was assessed for methodological quality using a descriptive evaluation. This addresses the range of methodologies included, following the systematic literature review by Spiby et al., which states "Quantitative quality scores were not allocated to studies to avoid possible implication of spurious accuracy and inappropriate interpretation of findings; this appeared particularly important in the context of the range of different research methods used." (203) Methodology was assessed for power calculations and appropriate sample size, sample selection, avoidance of contamination between intervention and control groups, appropriateness of design, consistency of breastfeeding definitions, clarity and completeness of reported methods, analysis by intention to treat and adjustment for clustering when necessary. Potential biases in the design of each study, as reported in the published article, were assessed as minor or major risks for bias. Hence, quality classified as good where extensive measures were taken to avoid potential bias in high quality trials. Quality was classified as fair where occasional potentials for bias were noted. Quality was classified as poor if results may be unreliable due to widespread or multiple methodological concerns.

There is a general risk of bias across the included studies in that only published randomised controlled trials were eligible for inclusion. Much of the breastfeeding research literature is omitted, as it is not at this level. Others have considered small projects that respond to government initiatives or policies, with notable reviews completed in New South Wales (237) and in the United Kingdom. (238) While other studies may reveal useful and valid strategies, this review aims to form a clear evidence base for intervention development and to facilitate any recommendations arising from the results of this project.

2.3 RESULTS Interventions from 31 international studies remained after exclusions as shown in Figure 2.2. They are grouped by intervention provider training and by efficacy in Table 2.1. Full details of these interventions are shown in Table 2.2, immediately before

78 Section 2.4. Of these studies, seventeen reported statistically significant breastfeeding outcomes and fourteen did not, compared to their control groups.

Table 2.1 Number of support RCTs reviewed, by provider training and outcome

Training Not Significant EBF Duration EBF+ Duration Total BFHI 0 2 1 2 5 LC 3 1 2 1 7 Other 3 0 2 1 6 Not specified 8 3 1 1 13 Total 14 6 6 5 31 EBF=exclusive breastfeeding, BFHI=Baby Friendly Health Initiative, LC=lactation consultant

Successful interventions Five studies reported statistically significant increases in both duration and exclusive breastfeeding. (17, 239-242) Six studies reported increased duration alone. (47, 243- 247) Six studies reported increased exclusivity alone. (185, 248-251, 252 ) These seventeen studies were from Australia (n=2), Belarus (n=1), Brazil (n=3), Canada (n=1), Denmark (n=1), France (n=1), India (n=1), Italy (n=1), Thailand (n=1), The Netherlands (n=1), and the USA (n=4). Interventions in these studies provided postnatal support to breastfeeding mothers from a range of health professionals including physicians, midwives, neonatal nurses, child health nurses, community nurses, and nutrition educators. One was an older study published in 1987; six were relatively recent and published in 2000-2005, while eleven were more recent and published in 2006-2011.

Unsuccessful interventions From the perspective of this review, the studies without significant breastfeeding outcomes were considered unsuccessful. These thirteen studies were from Australia (n=3), (253-255) Canada (n=2),(256, 257) Nepal (n=1),(258) New Zealand (n=2),(259, 260) The Netherlands (n=1),(261) and the USA (n=5). (262-266) They were published from 1998 to 2011 inclusive.

DETAILS OF INTERVENTIONS AND OUTCOMES

Interventions are listed in this section according to the type of training given to intervention providers, as described in the publications. This listing is matched to Table 2.2, at the end of this section, which summarises the studies for ease of comparison.

2.3.1 BFHI provider training Five included studies with effective interventions reported using BFHI programs to train intervention providers. Two studies reported 40 hours of training, three reported sixteen to eighteen hours of training.

Two large, well-designed trials were conducted in Belarus (n = 16,491) and Denmark (n = 1,336). Each was set in a general population of well mothers and infants who intended to breastfeed, and each involved cluster analysis of postnatal care.

The PROBIT study in Belarus (17) demonstrated increased exclusivity and duration of breastfeeding at every measurement point with large effect sizes. Exclusive breastfeeding was significantly maintained through the first 6 months. Full breastfeeding percentages at 3 months in the intervention group were 51.9% vs. 28.3% in the control group (adjusted OR 0.28; 95%CI 0.16-0.49), while the odds of having been weaned were halved at infant ages of 3, 6, 9 and 12 months (each adjusted odds ratio 0.5; 95%CI 0.32-0.73) in the intervention group compared to controls. Also significantly reduced were risks of gastroenteritis by 40% and risks of atopic eczema and other rashes by 45% compared to controls. Support was provided at routine health checks held at postnatal clinics by a team of physicians and nurses who had all received BFHI training. The consistency of training and the comprehensive protocols were strengths of this study, which was designed to test the BFHI across 30 hospitals and associated infant health clinics. Mothers had six contacts to 12 months of infant age, including four in the first 6 months.

The study in Denmark (252) demonstrated a reduced cessation of exclusive breastfeeding. The cessation rate was 14% lower in the intervention group than in the control group at 6 months (HR=0.86; 95%CI 0.75–0.99). Breastfeeding support, including a booklet, was offered at home visits by "health visitors", registered nurses who had all received BFHI training. Mothers in the intervention group received one to three contacts when their infants were aged 1 - 5 weeks. Particular attention and additional visits were provided to multiparous mothers with only brief breastfeeding histories. On average, each intervention family received an additional 0.4 visits compared to controls. Mothers in both intervention and usual care reported high levels of satisfaction with care, perhaps indicating that mothers are not aware of the level of efficacy of the support they receive.

80 Three smaller studies that reported on the use of BFHI training to equip intervention providers were each focused on an additional specific outcome, namely dental caries, respiratory morbidity, or the volume of breastmilk substitutes used.

Two Brazilian studies (239, 248) examined dental and respiratory outcomes respectively from subsets of the authors' earlier large trial published in Portuguese. Participating mothers had given birth to well, full term newborns at an urban hospital in Brazil that cared mainly for a lower socio-economic population. Intervention mothers received nutrition advice included exclusive breastfeeding to 6 months, introduction of healthy foods, food hygiene, and avoidance of unhealthy feeding practices. They were provided with ten contacts up to 12 months of infant age, including seven contacts in the first 6 months, through home visits by trained outreach nutrition educators. The intervention protocol also involved weekly mentoring of the intervention providers by a skilled practitioner, probably assisting compliance with intervention protocols.

The Brazilian dental outcomes study (248) reported longer maintenance of exclusive breastfeeding in the intervention group (2, p=0.000), lower odds of dental caries (OR=0.52; 95%CI 0.27-0.97) and significantly fewer mean decayed tooth surfaces (p=0.03) compared to controls. This study did not measure breastfeeding duration. The respiratory outcomes study (239) reported increased exclusive breastfeeding at 4 months in the intervention group (44.8% vs. 28.2%; RR= 1.59; 95%CI 1.21-2.07; p=0.001) and increased maintenance of breastfeeding to 12 months (52.8% vs. 41.9%; RR=1.25; 95%CI 1.02-1.55; p=0.032) compared to the control group. This study reported protection from respiratory morbidity in the past month with fewer affected children in the intervention compared to control group 23.3% vs. 39.7% (RR=0.59; 95%CI 0.43-0.81); the number of families requiring intervention for one child to avoid morbidity (number needed to treat) was 6.1.

The third Brazilian study was set in a less disadvantaged Brazilian population than the studies described above. In this smaller study, (n = 157) (243), the authors provided a frequent home visits program to 4 months of infant age, provision of audiovisual and printed support material and telephone support on request. Ongoing breastfeeding was increased at 4 months in the intervention group compared to control group, with 84% vs. 71% breastfeeding (p=0.04); however, changes in exclusive breastfeeding were not significant. The authors also measured the volumes of breastmilk and of water received daily by some of the infants at 4 months, using radioisotopes (intervention n = 37; control n = 31). There was a trend to less volume of substitution in the intervention

group: mean water intake (ml) (alone or mixed with formula) was 107 vs. 195 daily (p=0.16) while mean breastmilk intake (ml) was 761 vs. 723 daily (p=0.48) compared to the controls. The range of outcomes was wide and the study was underpowered to detect a difference <100ml hence the results for intake volumes were not significant.

These support interventions were provided for both disadvantaged (239, 248) and more advantaged (243) mothers in Brazil and also for a broader sample of mothers in Belarus (17) and in Denmark. (252) Interventions ranged from high intensity prolonged home visits (239, 243, 248) to only a few home visits (252) or use of routine clinic attendances. (17) These studies clearly support the efficacy of the complete BFHI intervention, demonstrating improved breastfeeding outcomes. They support the use of BFHI training to equip health to improve breastfeeding support across a variety of populations and settings.

BFHI training was not used in any of the thirteen included studies that were unable to report significant breastfeeding outcomes.

2.3.2 Trained lactation consultants Four included studies with significant breastfeeding outcomes involved health professionals trained and working in lactation. This thesis ascribes the description of lactation consultant according to the publications (267); however, only one of these is identified as International Board of Lactation Consultant Examiners Certified. (244) The detail and standard of training in breastfeeding management and support was unknown but presumed to be greater than general nursing training. All four interventions were set in the USA, and the three larger studies involved disadvantaged women.

The 1987 intervention by Frank et al. (240) compared three trial arms to the usual care at hospital discharge at that time. Usual care meant giving mothers a discharge pack that contained equipment for formula feeding and marketing materials from formula manufacturers. In this trial, mothers were randomised to four arms. Group 1 received an alternative "research" discharge pack. Group 2 received eight or more supportive telephone calls to 3 months in addition to the usual discharge pack. Group 3 received both the research pack and the counselling. Group 4 received only the usual discharge pack, being the usual care control group. The authors noted some contamination of groups in hospital and at discharge. Despite this, using survival analysis, the authors demonstrated a breastfeeding duration of significantly more days in Group 3 compared

82 to control (60 vs. 42; 2 =3.13; p=0.038). Exclusive breastfeeding at 2, 3, and 4 months was increased in Group 3, compared to control (at 4 months 2 =3.28; p=0.035 one tailed, log linear analyses). Exclusive breastfeeding rates were initially increased at 2 months in Group 2 (35% vs. 23%; 2 =5.09; p=0.005 one tailed) compared to controls in Group 4. However, this increase was not sustained. The authors suggested this environmental factor was relevant, and that marketing in the usual manner to postnatal mothers at that time was effective. There were no significant outcomes from Group 1, which is consistent with other research reporting the need for face-to-face support from health professionals in order to increase breastfeeding rates. (45)

A small study with predominantly white, educated, first time mothers provided intervention group mothers with one maternity hospital contact using an additional video. (185) Three to six weeks of breastfeeding log were then required from mothers. The log recorded daily self-monitoring of the length and number of feeds, use of substitutes, expression, and replies to open questions about the mother's feelings. Weekly telephone contacts for the first three weeks encouraged submission of completed logs. Their intervention followed Bandura's social cognitive learning theory and it was designed to promote self-efficacy. At their monthly preventive primary care visits for the first six months, the primary care providers for both the intervention and control mothers completed a new questionnaire about feeding and weight patterns, which for the intervention mothers may have functioned synergistically with the earlier intervention components as a face-to-face avenue of support for the mothers' motivation to continue breastfeeding. Exclusive breastfeeding at 6 months, measured as full breastfeeding, was increased among intervention mothers, compared to controls (10/41 vs. 3/43; 2 =39.12; p=0.000). The theoretical basis is sound; use of the log and the involvement of routine providers are promising; however, this is a small convenience sample with incomplete analysis, which weakens the study due to potential for bias.

Interventions described by Bonuck et al. (47) and Wambach et al. (244) supported clients of the Supplemental Nutrition Program for Women, Infants, and Children (WIC). These low-income minority women were mainly black or Hispanic, and in the latter study, also single teenaged first-time mothers. The relevance of these studies to Australian practice is reduced in the context of modest baseline breastfeeding rates (69% initiation among the teenagers, 65% any breastfeeding in the first study). These were of course complex interventions to address the needs of disadvantaged

populations, and they provided several antenatal as well as postnatal contacts. The magnitude of skilled supportive contact and, in the adolescent study, peer delivery of some of the postnatal phone contacts may explain their success in maintaining duration.

Bonuck et al. (47) provided extensive antenatal contact but minimal postnatal contact, relying mainly on the use of a weekly self-report diary. These authors reported increased breastfeeding duration measured at 20 weeks in the intervention group (53.0% vs. 39.3%; p<0.028) and increased breastfeeding intensity up to week nine, compared to the rate in the control group. Exclusive breastfeeding was unchanged.

Wambach et al. (244) reported increased duration of breastfeeding among intervention group teen mothers compared to control teen mothers (2 =16.26; p<0.01), but also found no significant difference in exclusive breastfeeding. Despite a fair initiation rate, by three weeks 73% of all breastfeeding mothers were using formula supplements. They also found that none of the predictors for initiation was predictors for exclusive breastfeeding. These two studies demonstrate an increase in sustained breastfeeding alongside early and continued use of formula by mothers, factors that may relate to population behaviours, intervention components, or social context.

These four successful lactation consultant studies between them address environmental factors, theories of self-efficacy, the value of peer support, and the varied life circumstances of individual women. It is possible that two essential components of their success relates to the presence of trained providers and to their deeper understanding and targeting of societal and personal barriers to be overcome by mothers in order to continue breastfeeding.

Three small studies did not report significant breastfeeding outcomes despite the use of lactation consultants in a community setting.

A Canadian study (256) with mothers of very low birth weight infants (<1500g) used lactation consultant support in hospital and then postnatal through clinic attendances. This was a motivated, white, relatively advantaged population, and the study had problems with contamination as the control group identified community lactation consultants as their most used resource for breastfeeding advice. The median duration of breastfeeding in both groups was 17.4 weeks by survival analysis, with no difference identifiable.

Another small study in the USA with low income WIC participants (265) provided lactation consultant contact, face to face antenatally and in hospital. However, the postnatal contact was by telephone apart from two women for whom a home visit was arranged due to complaints of pain. As reported, the intervention approach was directive and skills-based rather than client-centred. Although exclusive breastfeeding at one week in the intervention group was twice that of the control group, none of the outcomes was significant when measured at three months. The lack of face-to-face postnatal support is a methodological weakness; there were also the small sample size, high loss to follow-up and high failure of the intervention group to receive key intervention components (half did not receive hospital contacts) which affect the quality of this study.

An earlier, small USA study with a general population (262) offered two postnatal home visits in the first two weeks. The intervention also focused on preventing excess fatigue. However, it was described as a pilot project; duration of breastfeeding at 6 months was almost twice the control group but was not statistically significant in this small sample.

2.3.3 Other specified provider training Three studies with significant breastfeeding outcomes reported some details of the training they developed for intervention providers. Two were quite similar, recent Australian studies set in lower income urban areas, and both increased duration of breastfeeding.

Kemp et al. (245, 268) described sustained home visiting provided to "at risk" disadvantaged mothers by child health nurses trained for this study. The theoretical approach was described as "strengths-based". It involved helping mothers to care for their own needs, helping parents to interact with children in supportive ways, and improving socially supportive networks. Training aimed for a 'respectful parent-nurse partnership' in a whole family approach. Ongoing breastfeeding was only one part of a complex intervention known as Miller Early Childhood Sustained Home-visiting (MECSH), with the study protocol published earlier by the same multi-centre team of authors. (268) MECSH provided many, frequent, prolonged antenatal and postnatal home visits. Breastfeeding duration was increased by the intervention compared to controls (16 weeks vs. 8 weeks; effect size (d) =0.49; mean difference 8 weeks; 95%CI 3-13; p=0.002) in the context of low background breastfeeding rates, although

confidence intervals are wide. The authors attributed much of this difference to behaviour change among overseas-born mothers across the whole cohort, whose breastfeeding duration in weeks was greatly increased (26 vs. 11; mean difference 15; 95%CI 7-22; p<0.001) compared to the lesser impact on duration in weeks among Australian-born mothers (10.3 vs. 5.5). The program was very resource-intensive; hence, despite the wide-ranging focus a significant change in breastfeeding outcomes was achievable in a needy population.

The second Australian study (246) also provided breastfeeding support at home visits. However, it was a larger and less intensive study, it recruited pregnant first time mothers, and it focused on family risk factors for childhood obesity as the Healthy Beginnings Trial. Intervention providers were community nurses who were trained by health promotion practitioners in a guidelines-based approach, which aimed to be family-friendly. Nurses addressed infant feeding, nutrition and play as well as family physical activity, nutrition and social support. They used a checklist of discussion points and provided a range of resources during home visits lasting between one and two hours. No effect was found on exclusive breastfeeding outcomes; the "many inconsistent messages" these mothers faced regarding when they should introduce solids was reported elsewhere by the authors. (64) The intervention significantly delayed regular use of solids at less than 6 months and significantly increased duration of breastfeeding in weeks compared to the control group (17 [95%CI 13.9-20.4] vs. 13 [95%CI 10.1-15.0]; p=0.05) with a hazard ration for breastfeeding cessation of 0.82 (95%CI 0.68-0.99).

Both these Australian studies increased breastfeeding duration but breastfeeding rates at 6 months remained lower than the national mean (42% and 32% respectively, compared to the national mean of 49%), (52) which is not unexpected in such disadvantaged populations, as discussed in Chapter 1.

The third effective study to describe a tailored training program for intervention providers was set in France in a general population. (241) Primary care physicians and paediatricians prepared for their intervention with a series of two 2.5-hour training sessions, held in the month prior to the start of the trial. The curriculum was developed in response to guidelines and aimed to improve breastfeeding knowledge as well as counselling skills. Adult learning techniques were used including discussions, role-play, and printed materials. Intervention breastfeeding mothers (in a setting of 70% breastfeeding initiation) were offered one additional postnatal contact at two weeks, at

86 the clinic of a trained provider, as an additional preventive health visit. There was an 80% uptake by intervention mothers, and the study was well conducted and analysed. Despite some contamination, as 7% of control group mothers also accessed the intervention, exclusive breastfeeding was maintained at 4 weeks using 24-hour recall (83.9% vs. 71.9%; HR=1.17; 95%CI 1.01-1.34) compared to controls. Median duration in weeks was longer in the control than the intervention group (18 vs. 13; HR=1.4; 95%CI 1.03-1.92). It appears this clinic visit functioned similarly to those in the PROBIT study, (17) indicating the value of also training doctors in lactation management and counselling.

Three large studies that did not report significant breastfeeding outcomes also specified the tailored training programs they used to equip providers. Common to these studies was training of various quality focused on counselling, without preparatory training in breastfeeding management.

A New Zealand (260) study with smokers recruited to a multi-arm study found no difference in duration measured at 4 months. Women had contact with their home- birthing midwife for breastfeeding education and support within routine care at multiple antenatal and postnatal contacts, until their infants were six weeks old. Control group midwife training is not reported. Intervention group providers received information most likely about health importance, printed materials, and informal discussion with a lactation consultant. They had occasional discussion sessions through the fifteen months of the trial with a lactation consultant or midwife. One study arm also provided a quit smoking motivational intervention, for which midwives received four hours training in motivational interviewing, and outcomes for the dual approach were promising. However, there were difficulties with recruitment and the breastfeeding intervention focused more on immediate problem-avoidance than on maintaining ongoing breastfeeding.

Concluding their report on a large (n = 698), well conducted study in a general population of Dutch mothers which found no significant breastfeeding outcomes, Kools et al. (261) declared every part of their intervention to be ineffective. However, components of this study merit more evaluation. The intervention was delivered during routine care contacts: by maternity nurses providing antenatal home visits, by child health nurses providing postnatal home visits and clinic care, and by physicians providing child health clinic care. Additional components were the funding of lactation consultant referrals for intervention group mothers, and a mothers' communication

booklet for integrated provider records where mothers were asked to log their breastfeeding problems, barriers, and motivation before each provider contact. The comprehensive involvement of all primary care providers and the care coordination are worth considering. This study may have been compromised by the problem-focused log that was used, compared to the achievement focus of logs used by others. (185, 251) Another possible weakness is the rather directive "health counselling" used, compared to the client-centred model of the BFHI; the prompt sheet for providers gave matched solutions for identified problems, without interpersonal exploration of the woman's experience. Training was focused on the "health counselling" model; adequate breastfeeding knowledge was assumed and only a skills update was used. Kools et al. noted that 19% of nurses scored less than 5.5/10 in a pre-trial breastfeeding knowledge assessment although mean outcomes were considered acceptable. The providers reported that they found their task challenging, which is consistent with inadequate preparation or mentoring. The contamination of clusters by independent rotations of trained staff was also a weakness.

Minkovitz et al. describe outcomes of the Healthy Steps for Young Children Program (263), another large study (n = 1987) in the USA focusing on child safety, development and health with specific training for this goal. The early childhood nurses who provided the five or more postnatal contacts were assumed to have adequate breastfeeding skills for the small component that comprised breastfeeding support. Postnatal contact with the trained provider occurred at physicians' clinics, where the study experienced some contamination, since both intervention and control families attended each clinic. Breastfeeding outcomes were not significant, perhaps related to the minimal focus and lack of adequate training.

2.3.4 Provider training not specified Training for health professionals delivering the interventions was either not specified or not provided in the remaining 13/31 studies included in this systematic review. Five of these had significant breastfeeding outcomes: they considered asthma prevention (n=2); skin contact for low birth weight neonates; acupuncture for low breastmilk supply; and empowerment-focused ongoing education.

A large multifactorial asthma prevention intervention in Canada (n = 493) (247) aimed to reduce inhaled and ingested allergen exposures for infants at high risk of asthma. "Trained research nurses" encouraged breastfeeding for 4 to 12 months with solids introduced from 6 months at one antenatal home visit and at home visits at 2 weeks, 4,

88 8 and 12 months after birth. Breastfeeding initiation was high (92%) yet this intervention still significantly increased breastfeeding duration measured at eight months (61% vs. 50%; p=0.02) compared to controls. Intervention group mothers were also less likely than control group mothers to introduce solids before 4 months (19.5% vs. 49.5%; p<0.001). Continued breastfeeding was only one of five major behaviour changes addressed with these mostly educated and likely motivated families. However, they also faced surprisingly mixed messages: while the authors noted evidence of improved child asthma outcomes following exclusive breastfeeding, (269) their methods included provision of free partially hydrolysed formula (donated to the study by manufacturers) to all intervention families "when necessary and after weaning." Supply of free formula is known to be unsupportive of breastfeeding, as discussed. (240) This study did not measure exclusive breastfeeding.

A smaller study set in The Netherlands (n=89) also focused on primary prevention of asthma. In this subset of a larger study, the intervention described by Gijsbers et al. (251, 270) addressed mainly breastfeeding behaviours. A "trained research assistant", provided two antenatal and one postnatal home visits, to mothers with diagnosed asthma in their families. The intervention considered social influence, attitude and self- efficacy theories. Contacts provided mothers with basic breastfeeding information, advice on allergies and asthma and a booklet (developed from prior focus groups) identifying barriers and giving tips and personal stories. Mothers also completed a brief daily feeding diary, which may have reinforced self-efficacy as proposed by others. (185) The theoretical foundation, use of the diary and evidence-based supportive print materials are strong points in this study's methodology, although the higher education and motivation of these mothers most likely contribute significantly to the outcomes. Breastfeeding initiation was 88%, with significant maintenance of exclusive breastfeeding measured at 6 months (48% vs. 27%; p<0.05) in the control group compared to the intervention group, apparently using a since birth definition based on the diaries.

A small study in India (n =100) also used minimal postnatal resources for breastfeeding support, in a specific health context. The authors report an intervention with mothers of low birth weight babies (<1800g) who were ready for oral feeding. (249) Neonatal nurses provided support for Kangaroo Mother Care (KMC) daily for at least one month in hospital with follow-up-up contacts weekly for another two months at a hospital clinic. The KMC intervention significantly supported exclusive breastfeeding by intervention group mothers than by control group mothers, measured at 3 months (14/50 vs. 6/50, a

16% difference; p<0.05). It was also shown to support maternal-infant attachment. Relevance of this small, specialised study with health-based maternal motivation to the general Australian population is limited; however, the use of clinic-based support and the utility of increased mother-infant contact for breastfeeding outcomes are noteworthy.

A small study (n=84) set in Italy explored the effect of acupuncture, which was provided for the mothers of approximately one month old babies who were referred to a breastfeeding clinic with "insufficient milk supply". Neri and her colleague who performed the acupuncture were both medical practitioners. (250) They described their experience, theory, and careful maintenance of consistency with acupuncture techniques. They were pleased with the significantly increased exclusive breastfeeding rate in the intervention group immediately after the three weeks of treatment and again when the infants were 3 months old using 24-hour recall (98%vs. 60%; p<0.03 and 35% vs.15%; p<0.03), compared to the control group. In both groups, 50% had been exclusively breastfeeding at baseline. The authors pointed out in the discussion that mothers may have been anxious about an unsettled baby and mistakenly attributed this to an insufficient milk supply; hence, perhaps the acupuncture treatment program supported the mother by relieving anxiety. The intervention required mothers to lie still for 30 to 40 minutes six times in three weeks, while control group mothers had three contacts at the breastfeeding clinic in this same period.

A small study in Thailand (n=71) was guided by a strong theoretical and personalised approach. (242) Intervention mothers received a three-hour antenatal class and eight postnatal telephone calls up to 6 months, supported by home visits when "problems with exclusive breastfeeding" were identified. The nurse educator (training not specified) provided antenatal education and postnatal support using a "knowledge sharing practices with empowerment strategies (KSPES) program", designed to motivate behaviour change, facilitate open conversation and to support reflection, autonomy and perseverance. This approach was adapted from a diabetes self- management education program, and proved significant for maintaining exclusive breastfeeding at 6 months (20% vs. zero; p=0.005) and full breastfeeding at 6 months (60% vs. 5.3%; p=0.0002) compared to controls. The background initiation rate was 90%; by three months there was a significant difference from the control group in any breastfeeding (100% vs. 66%; p<0.0001) which was maintained to 6 months.

90 There were eight studies without successful outcomes that did not report details of intervention provider training, four larger studies (n>340) and four smaller studies (n=41 to n=200). Large studies were set in Australia, Nepal and the USA while smaller studies were set in Australia, Canada, New Zealand and the USA.

A recent large Australian study (255) used midwives with unspecified breastfeeding training to advise mothers about problems and to identify their support needs during 12 telephone calls up to six weeks of infant age. Weekly home visits were offered but no data was provided about uptake. Perhaps a strengths-based approach and some regular home visits from a provider with specific breastfeeding management training may have helped.

Likewise, an earlier large Australian study (253) found that 6-month breastfeeding outcomes were unchanged by an early postnatal routine visit to the usual primary health doctor (general practitioner), at two weeks rather than six. There was no breastfeeding training undertaken by the doctors, whose breastfeeding management knowledge was assumed sufficient.

Health workers in a large study in Nepal (271) were trained (not specified) and mentored to provide a single contact for health education to mothers (n = 393) at three months postnatal, however breastfeeding support was one of five significant education tasks and not surprisingly there was no change in exclusive breastfeeding measured three months later. There was a large loss to follow-up after recruitment.

A large USA study (266) with disadvantaged mothers enrolled in WICcc (n = 341), using support from a registered nurse (training not specified), serves as a reminder that even with a topical message and very frequent contact in the first two weeks, telephone support from a health professional tends to be ineffective. Again in this study a wide range of topics were addressed in each phone call, possibly weakening the effect.

A smaller Australian study with illicit drug-using mothers (254) provided extensive postnatal home visiting using a non-judgemental style and flexible content in a multifocal intervention by research midwives (training not specified). Insufficient breastfeeding management training may have limited their effect, and the study was very broadly focused and underpowered. cc WIC=Women Infants and Children supplementary nutrition program, available until three months post partum, providing medical care including supply of free formula.

In a general Canadian population, McQueen et al. (257) conducted a pilot study using an experienced postnatal care registered nurse to provide one hospital contact and follow-up telephone contact at one week. The study aimed to support self-efficacy; assessed provider views and experience; used cluster randomisation; and took care to avoid biases; but was not powered for significance.

A New Zealand study focused on supporting mothers of preterm infants (259) with an extensive program of home visiting support from a registered nurse (training not specified) and the availability of daily telephone support. Breastfeeding was maintained at 6 months by 36% of mothers, with no significant difference between groups. Hospital breastfeeding support was extensive and mothers were probably highly motivated.

An earlier small study with USA WIC mothers (264) used a team to support mothers and described increased duration of any breastfeeding measured at 4 months (16/21 vs. 8/20 i.e. 75% vs. 40%) but the authors only reported calculations on exclusive breastfeeding outcomes, which were not significant. The authors assessed health costs of daily support in hospital by a lactation consultant, postnatal contacts by a community nurse (unspecified training) in the first month at home, and telephone calls by a peer counsellor, sixteen in the first eight weeks then weekly to 6 months. The cost analysis was positive, particularly if peer-training costs were shared across a larger group of mothers.

Table 2.2 Results of RCT literature review, by training and outcomes Author, Country (n) Mothers Prior Use of a Postnatal contacts Sig Sig Significant breastfeeding End Quality Year BF% (or included in contacts Postnatal Counselling EBF Dur results as reported (mo) CG%) the study AN 3o Intervention Theory HV C Ph Baby Friendly Health Initiative (BFHI) training Kramer et Belarus 16,491 General D BF support at  6   EBF at 3,6 mo 12 Good al., CG BF at routine visits to FBF at 3mo adjusted 2001 (17) 3 mo 12 mo OR=0.28; 95%CI 0.16-0.49 60% BF at 3,6,9,12 mo adjusted OR=0.5; 95%CI 0.32-0.73 Albernaz et Brazil 157 Urban, middle 1 BF support, video,  7  EBF NS 4 Fair al., 92% income telephone hotline BF at 4 mo 2003 (243) (272) 84% vs. 71%; p=0.04 HR=2.06; 95%CI 1.04-4.10 Feldens et Brazil 378 Urban, low Support for  10  EBF at 4 mo 12 Fair al., 92% income healthy nutrition 35% vs. 16%; 2 p=0.000 2- 2007 (248) (272) tailed Vitolo et al., Brazil 395 Urban, low Support for  10   EBF at 4 mo 12 Fair 2008 (239) 92% income healthy nutrition 44.8% vs. 28.2%; RR= 1.59; (272) 95%CI 1.21-2.07; p=0.001 BF at 12 mo 52.8% vs. 41.9%; RR=1.25; 95%CI 1.02-1.55; p=0.032 Kronborg et Denmark 1,336 General More support if  1-3  Cessation rate at 6 mo 6 Fair al., 99% (129) poor BF history; 7.7% vs. 4.9% 2007 (252) booklet; to 5 w HR=0.86; 95%CI 0.75–0.99 n number analysed; BF breastfeeding; BF% national BF initiation rate; CG% BF rate in control group, specified; n number analysed; Prior contacts AN 3o number of intervention contacts antenatal or in tertiary maternity unit; Postnatal contacts HV C Ph number of intervention contacts by home visit, at a clinic or by phone; Sig significant outcome; EBF exclusive breastfeeding; Dur duration of breastfeeding; End age of infant at end of data collection; Quality see text, section 2.2.4; FBF full/predominant breastfeeding; OR odds ratio; HR hazard ration; CI confidence interval; RR relative risk; NS not significant; D daily; d day/s; w week/s; mo month/s

Author, Country (n) Mothers Prior Use of a Postnatal contacts Sig Sig Significant breastfeeding End Quality Year BF% (or included in contacts Postnatal Counselling EBF Dur results as reported (mo) CG%) the study AN 3o Intervention Theory HV C Ph Lactation consultant involvement postnatal Wambach USA 289 WIC: low 5 1 Peer and  5 EBF NS 6 Fair et al., 69% income teen, professional team  Median duration 2010 (244) single, Black, support to 4w 177 vs. 61 days 2 =16.26; first time p<0.01 mother Pollard et USA 84 First time 1 Daily self-monitor  3  FBF at 6 mo 6 Poor al., CG BF at mother log; calls to 3w 10/41 vs. 3/43; 2 =39.12; 2010 (185) 3 w 65% p=0.000 Bonuck et USA 304 WIC: low 0-14 Weekly self-report  1  EBF NS 12 Fair al., CG BF at income, Black diary Duration at 4 mo 2005 (47) 2 w 65% /Hispanic 53.0% vs. 39.3%; p<0.028 Frank et al., USA 343 Fairly low Alternate 8   EBF at 4 mo 4 Fair 1987 (240) CG BF at income discharge pack; 2 =3.28; p=0.035 one tailed 4 mo counselling calls Median duration 54% to 3 mo 60 vs. 42 days; 2 =3.13; p=0.038 Pinelli et Canada 115 General; 10 Support at routine 4 12 Fair al., BF 40w VLBW infants visits; test weigh 2001 (256) 63% <1500g Petrova et USA 82 WIC: low 2 1 Supportive calls to 4 3 Poor al., CG BF at income 4 mo 2009 (265) 1 w 90% Pugh et al., USA 60 General To reduce fatigue  2 5 Poor 1998 (262) 71% (263, 272) n number analysed; BF breastfeeding; BF% national BF initiation rate; CG% BF rate in control group, specified; n number analysed; Prior contacts AN 3o number of intervention contacts antenatal or in tertiary maternity unit; Postnatal contacts HV C Ph number of intervention contacts by home visit, at a clinic or by phone; Sig significant outcome; EBF exclusive breastfeeding; Dur duration of breastfeeding; End age of infant at end of data collection; Quality see text, section 2.2.4; FBF full/predominant breastfeeding; NS not significant; ABF any breastfeeding; D daily; d day/s; w week/s; mo month/s; WIC Women Infants and Children Supplementary Nutrition Program; VLBW very low birth weight

Author, Country (n) Mothers Prior Use of a Postnatal contacts Sig Sig Significant breastfeeding End Quality Year BF% (or included in contacts Postnatal Counselling EBF Dur results as reported (mo) CG%) the study AN 3o Intervention Theory HV C Ph Specified training Kemp et al., Australia 208 Urban, low 10 For skilled  33  Median duration 24 Fair 2011(245) 80% income, multi- parenting 16 vs. 8 w; mean difference 8 (246) cultural, at w; 95%CI 3-13 w ; p=0.002 risk Wen et al., Australia 527 Urban, low 1 To prevent  5  Median duration 12 Fair 2011 (246) 80% income, multi- childhood obesity 17 vs. 13 weeks; cessation cultural HR=0.82; 95%CI 0.68-0.99; p=0.05 Labarere et France 226 General D Support at routine  1   EBF at 4w 6 Good al., 71% physician visit at 83.9% vs. 71.9%; HR=1.17; 2005 (241) <2w 95%CI 1.01-1.34 Median duration 18 vs. 13 weeks HR=1.4; 95%CI 1.03-1.92 McLeod et New 399 Smokers 8 [1] Education to 6 w; 5 4 Poor al., Zealand MI in one arm for 2004 (260) CG BF at smoking 6 w 71% cessation Kools et al., The 698 General 1 Health counselling  3 6 Good 2005 (261) Nether- booklet; log lands barriers; free 80% IBCLC referral Minkovitz et USA 1 987 General Parenting support  1 4 4 Poor al., 71% for child health 2001 (263) and development n number analysed; BF breastfeeding; BF% national BF initiation rate; CG% BF rate in control group, specified; n number analysed; Prior contacts AN 3o number of intervention contacts antenatal or in tertiary maternity unit; Postnatal contacts HV C Ph number of intervention contacts by home visit, at a clinic or by phone; Sig significant outcome; EBF exclusive breastfeeding; Dur duration of breastfeeding; End age of infant at end of data collection; Quality see text, section 2.2.4; HR hazard ration; CI confidence interval; D daily; d day/s; w week/s; mo month/s; [1] homebirth contact;

Author, Country (n) Mothers Prior Use of a Postnatal contacts Sig Sig Significant breastfeeding End Quality Year BF% (or included in contacts Postnatal Counselling EBF Dur results as reported (mo) CG%) the study AN 3o Intervention Theory HV C Ph Training not specified Chan- Canada 493 Family risk of 1 Reduce allergens  4  BF at 8 mo 12 Good Yeung et 92% child-hood of smoking, dust 61% vs. 50%; p=0.02 al., asthma mites pets, foods; Solids at 4 mo 2000 (247, leaflet 19.5% vs. 49.5%; p<0.001 273) Gathwala India 100 With LBW D Encourage 8  EBF at 3 mo 3 Fair et al., 95% infants on oral kangaroo mother 88% vs. 72%; p<0.05 2010 (249) (272) feeds, mean care for six hours BW 1800g per day, to 3 mo Neri et al., Italy 84 With 'low milk Acupuncture from  6  EBF at 2 mo 3 Fair 2011 (250) at 1 mo supply', 1 mo, for 2 w 98%vs. 60%; p<0.03 70% infants mean 1 mo of age EBF at 3 mo 35% vs.15%; p<0.03 Kupratakul Thailand 71 General Calls to 6 mo;  nc 8   EBF at 6 mo 6 Fair et al., 90% visits if needed to 20% vs. 0%; p=0.005 2010 (242) support EBF FBF at 5 mo 60% vs. 5.3%; p=0.0002 BF at 6 mo 100% vs. 66%; p<0.0001 Gijsbers et The 89 Family risk of Feeding diary; info  1  EBF at 6 mo 6 Good al., Nether- child-hood on allergy and 48% vs. 27%; p<0.05 2008 (251) lands asthma asthma; support HR=0.50; 95%CI 0.26-0.95 88% booklet n number analysed; BF breastfeeding; BF% national BF initiation rate; CG% BF rate in control group, specified; n number analysed; Prior contacts AN 3o number of intervention contacts antenatal or in tertiary maternity unit; Postnatal contacts HV C Ph number of intervention contacts by home visit, at a clinic or by phone; Sig significant outcome; EBF exclusive breastfeeding; Dur duration of breastfeeding; End age of infant at end of data collection; Quality see text, section 2.2.4; FBF full/predominant breastfeeding; HR hazard ration; CI confidence interval; D daily; d day; w week; mo month; nc likely but not clear;

Author, Country (n) Mothers Prior Use of a Postnatal contacts Sig Sig Significant breastfeeding End Quality Year BF% (or included in contacts Postnatal Counselling EBF Dur results as reported (mo) CG%) the study AN 3o Intervention Theory HV C Ph Training not specified, continued McDonald Australia 582 General 1 Weekly diary; nc 12 6 Fair et al., 91% calls to 6 w; visits 2010 (255) offered Gunn J et Australia 395 General Early routine 1 6 Fair al., 84% medical (GP) 1988 (253) check at 2 w Bartu et al., Australia 136 Illicit drug Support for child  8 6 Poor 2006 (254) 86% users development and parenting McQueen Canada 150 General 2 Supportive call at  1 2 Fair et al., CG BF at 1w 2011 (257) 4 weeks 80% Bolam et Nepal 393 General Education on 1 6 Fair al., 98% food, family 1998 (271) (272) planning, health; at 3 mo Gunn T et New 200 With preterm D Early discharge 10 6 Fair al., Zealand infants born with home 2000 (259) 82% <37 weeks support gestation Bunik et al., USA 341 WIC: low Calls to 2 w 10 6 Fair 2010 (266) BF at 1 income mo 74% Pugh et al., USA 41 WIC: low Visits to 4 w; peer 3 24 6 Poor 2002 (264) CG BF at income calls to 6 mo 1 week 90% n number analysed; BF breastfeeding; BF% national BF initiation rate; CG% BF rate in control group, specified; n number analysed; Prior contacts AN 3o number of intervention contacts antenatal or in tertiary maternity unit; Postnatal contacts HV C Ph number of intervention contacts by home visit, at a clinic or by phone; Sig significant outcome; EBF exclusive breastfeeding; Dur duration of breastfeeding; End age of infant at end of data collection; Quality see text, section 2.2.4; D daily; d day; w week; mo month; nc likely but not clear; WIC Women Infants and Children Supplementary Nutrition Program

2.4 DISCUSSION OF SYSTEMATIC LITERATURE REVIEW In this review 18/31 studies gave details on how they trained the health professionals who were providing the interventions to support breastfeeding. This included twelve that demonstrated significant breastfeeding outcomes and five that did not. Training appeared to be important for effective interventions, so it is worth considering why six interventions that used trained providers were unable to demonstrate efficacy.

Methodological issues weakened the three lactation consultant studies: a Canadian study was confounded by contamination of the control group (256); a well-designed USA study had too small a sample size for statistical significance as was described as "exploratory" (262); another small USA WIC study relied on health professional telephone follow-up and non-client-centred support. Methodological issues also weakened the three studies reporting tailored training programs. A New Zealand smoking cessation study (260) used very brief breastfeeding education, unlikely to differentiate intervention from control group providers; while one arm showed promise from combination with motivational interviewing training, the authors described recruitment difficulties as reducing study power. The large Dutch study relied on provider counselling training without provider training in breastfeeding management. (261) Access to lactation consultants was only by referral. Results may have been weakened by protocol non-adherence and contamination of control clusters that was identified by the authors due to rotation of staff between clusters. A large USA study used providers who had broad training in child health and safety for a study that addressed many healthy behaviours, (263) however breastfeeding knowledge of providers was assumed, and breastfeeding outcomes were unchanged.

On the other hand, of the studies without any specified preparatory training (13/31), five were effective. These included three studies with mothers motivated by specific relevant health concerns, the small theory-based acupuncture study (250), and one study with extensive theory-based personalised support. (242) Reporting the details of how intervention providers are trained is an area in which others have called for improvement. (45) In these five studies, provider knowledge of breastfeeding management and provider counselling training are unknown. However, common features of these studies are individual tailoring to the mother's situation, repeated contacts in a community setting, and face-to-face contact, although the extent of this is not recorded in the Thai study. (242)

The bulk of the included studies that had no significant breastfeeding outcomes are from those studies where provider training was not specified (8/31). Half of these were large trials with more than 340 participants.

As mentioned earlier, definitions and data collection points varied across studies. One advantage of the 24-hour recall question used by some studies was sensitivity to changes in mothers' behaviours following intervention contacts. Transitions in practices on a day-to-day basis are fairly common, (55) and current practice measurement may detect mothers who cease offering formula or solids after an intervention contact. While these changes may not prevent atopic sensitisation from earlier substitute use, (51, 269) there are many advantages for mothers and infants from a return to exclusive breastfeeding as measured by 24-hour recall, including secondary benefits from the implied increase in maternal confidence and motivation. (274)

The need for routine, knowledgeable support of breastfeeding mothers is clearly identified by Britton et al. (45) and theory-based studies have been called for. (203, 275) This systematic review provides new evidence that health professionals need training in breastfeeding management as well as in the counselling skills required to support confidence and motivation among breastfeeding mothers. Effective breastfeeding outcomes, this evidence suggests, derive not only from knowledge about health imperatives, but also from a deeper understanding of breastfeeding norms and barriers, and from a theoretical framework for client-centred support.

Mentoring of health professionals may be needed to build capacity for this role, as provided in one successful Brazilian intervention (239, 248) and perhaps needed in a Dutch study unable to demonstrate improved outcomes. (261) This systematic review shows with new clarity, in focusing on health professionals and postnatal care provision, that support may be effectively linked to routine health visit. (17, 185, 241, 252) It suggests that studies with a stronger base in primary care may offer benefits in the form of repeated contacts (17, 243, 246, 247) as the infant grows older. A consistent finding from this systematic review and others in the field is that effective health professional support needs to be provided face-to-face.

2.5 CONCLUSION TO CHAPTER 2 This systematic literature review supports postnatal intervention by health professionals in primary care in order to increase breastfeeding rates. This review indicates that to provide such support, the health professionals needed training both in breastfeeding management and in counselling skills. Successful studies tended to use a counselling theoretical framework with an individualised and proactive approach to mothers.

Fallon et al. assessed and concluded that there was an “urgent need for research” on breastfeeding support in Australia in the area of breastfeeding duration. They suggested “enhancing the knowledge of health professionals in the community, such as general practitioners, who are likely to provide breastfeeding support to mothers once they leave hospital.” (13) Observational Australian studies show that general practice-based support could be effectively provided by a practice-based lactation consultant (276) but that training is needed due to the low existing knowledge base. (131)(p79) The role of practice nurses is expanding and their numbers are increasing. (217, 225, 277) Partnerships between practice nurses and general practitioners in evidence-based care preventive and chronic care have been successful. (230) Evidence-based routine support for breastfeeding women merits consideration.

This thesis identifies, in Australia, an intervention opportunity linked to preventive routine child health care. This is attendance for infant immunisation. Immunisation occurs at local councils and community early childhood centres, but primarily in private general practices, where nurses are increasingly employed with the support of federal funding. Funding enables access by disadvantaged mothers, who have the most to gain from increased breastfeeding, and who are least likely to access care from early childhood nurse clinics. Personalised and patient-centred support to improve breastfeeding rates, particularly from a familiar provider, is supported by this review.

Based on the evidence presented, this thesis describes the development and conduct of a randomised controlled trial to evaluate an intervention to support ongoing breastfeeding. This study was designed to include the key elements of successful trials. Set in primary care, it uses practice nurses, at routine immunisation appointments. Practice nurses received training in breastfeeding management and counselling skills. The MI theoretical framework, enabled proactive support. Improving mothers’ access to trained peer counsellors was also attempted.

100

Chapter 3

Methodology

101

INTRODUCTION TO CHAPTER 3 METHODOLOGY

Introduce motivational support from trained general practice nurses to increase both exclusive breastfeeding and ongoing breastfeeding rates

Maternal and infant health outcomes are directly related to breastfeeding rates. More mothers may decide to continue breastfeeding if they have increased support available at their local general practice. This research primarily aimed to increase breastfeeding exclusivity and duration.

This chapter describes a cluster randomised controlled trial, clustered by and set in general practices. The chronological sequencing of trial processes is shown in Figure 3.1. An intervention and a preparatory training course were developed for this trial, outcome measures were defined, and general practices in rural Australia were recruited then randomised. Practice nurses from practices allocated to the intervention group received training to prepare for administering the intervention. Trained general practice nurses then administered a motivational intervention at routine appointments, using the Conversation Tool.

The trial participants were breastfeeding mothers, recruited by mail survey during pregnancy, who identified a recruited general practice as their usual place to see a doctor. When breastfeeding mothers attended an intervention general practice for routine infant immunisations, it was intended that they receive the treatment at 2, 4, and 6 months of infant age. Independent telephone interviews at 4 and 6 months collected data from mothers on breastfeeding practices for primary outcome measures.

During the period of the study, repeated interviews were held with intervention practice nurses. Data was collected for secondary outcome measures relating to the feasibility and experience of this practice nurse role in lifestyle change counselling around breastfeeding.

102

Figure 3.1 Chronological sequence of Support for Ongoing Breastfeeding study

Starting in 2007 Strand 1 Strand 2 Motivational Mothers Nurses interviewing Phase 1: preparation literature RCT systematic review literature PN literature review review Develop & pilot Chapter 2 Chapter 5 Conversation Tool Develop questionnaire Engage with Engage with for mothers' maternity APNA and Develop & pilot baseline data hospitals Divisions hospitals Divisions training program Develop Recruit, Recruit then breastfeeding baseline data randomise outcomes CATI Develop & pilot Babies born Intervention PN Pre & post and breastfed training training quiz

Phase 2: intervention Attend 2 mo Use of immunisation Conversation Tool Phase 3: data collection Repeated Attend 4 mo Use of interviews with immunisation Conversation practice nurses Telephone and updates interviews with Tool mothers at 4 and 6 months Attend 6 mo Use of immunisation Conversation Tool

Chapter 4 Chapter 5 Phase 4: analysis Results of Process RCT analysis

Finishing in 2011

APNA Australian Practice Nurses Association; CATI computer assisted telephone interview; mo months; PN practice nurse; RCT randomised controlled trial

This chapter conforms to the CONSORT statement for trial reporting as extended to cluster randomised trials (278), with Table 3.1 showing the checklist of items that should be reported in the methods section, as recommended.

103

Table 3.1 Checklist of items to include when reporting a cluster randomised trial

Methods topics Descriptor Section Participants Eligibility criteria for participants and clusters and the 3.1.3- 3.1.6 settings and locations where the data were collected Interventions Precise details of the interventions intended for each group, 3.2 whether they pertain to the individual level, the cluster level, or both, and how and when they were actually administered Objectives Specific objectives and hypotheses and whether they pertain 3.1.1 to the individual level, the cluster level, or both Outcomes Report clearly defined primary and secondary outcome 3.1.1 measures, whether they pertain to the individual level, the cluster level, or both, and, when applicable, any methods used to enhance the quality of measurements (e.g. multiple observations, training of assessors) Sample size How total sample size was determined (including method of 3.1.2 calculation, number of clusters, cluster size, a coefficient of intracluster correlation (ICC or k), and an indication of its uncertainty) and, when applicable, explanation of any interim analyses and stopping rules Randomisation: 3.1.5 Sequence Method used to generate the random allocation sequence, generation including details of any restriction (e.g. blocking, stratification, matching) Allocation Method used to implement the random allocation sequence, concealment specifying that allocation was based on clusters rather than individuals and clarifying whether the sequence was concealed until interventions were assigned Implementation Who generated the allocation sequence, who enrolled participants, and who assigned participants to their groups Blinding Whether participants, those administering the interventions, and those assessing the outcomes were blinded to group assignment. If done, how the success of blinding was evaluated Statistical methods Statistical methods used to compare groups for primary 3.4 outcome(s) indicating how clustering was taken into account; methods for additional analyses, such as subgroup analyses and adjusted analyses

3.1 RANDOMISED CONTROLLED TRIAL PROCESSES

3.1.1 Objectives and outcome measures Primary objectives were to demonstrate, compared to controls, a 15% increase in the number of women: . continuing to breastfeed at 4 months . exclusively breastfeeding at 4 months . continuing to breastfeed at 6 months . exclusively breastfeeding up to 6 months

104

Secondary objectives were to . consider the baseline demographics and factors that may influence breastfeeding decisions, for the women, in order to adjust for potential confounders . evaluate three currently used and alternate definitions of exclusive breastfeeding . explore the practice nurses' experiences in the research . assess the feasibility of and adherence to research protocols in general practice

3.1.2 Sample size calculations Winepiscope computer software was used to calculate the required sample size, allowing 10% additional participants for used of a clustered design.

There were 278 participants required to complete the study in order to demonstrate a 15% difference (2 tailed distribution with 80% power and 95% confidence) in breastfeeding rate between intervention and control groups.

3.1.3 Participant eligibility Eligibility was determined in stages as shown in the CONSORT flowchart, Figure 3.2. Pregnant women registered on the antenatal database at one of three local hospitals were invited to participate if they were due to give birth during the study period.

Exclusion criteria As shown in Figure 3.2, exclusion criteria were:  not planning to attend a participating general practice  not initiating breastfeeding in a local maternity hospital  not breastfeeding at 8 weeks after the birth

Parity and past infant feeding experience were not exclusion criteria, and women with multiple births were interviewed to collect data on one randomly selected infant. All participants were women sixteen years or older, with sufficient English to complete the mail survey and consent. A driving force behind the trial design was an aim to produce trial results that were applicable to the average rural general practice population.

The trial setting excluded women who gave birth before 34 weeks, as they are transferred out to larger metropolitan hospitals, and women giving birth at other hospitals had inaccessible hospital data including delivery date. Women who did not plan to attend a participating general practice were excluded, as they did not have equal potential to receive the intervention.

105

Figure 3.2 The CONSORT methods flowchart

106

3.1.4 Participant recruitment process during pregnancy The trial aimed to make three points of contact with mothers, following a modified Dillman Protocol (279, 280) for maximal response rate. Three components were developed for this purpose, a poster, survey package, and a postcard.

Poster Dillman identified that advance notice increases response rates. A laminated A4 poster using individual hospital letterhead was developed and supplied to each hospital. It was displayed in the booking-in room or the antenatal waiting room. (Appendix 3.1)

Mail questionnaire Identified pregnant women were sent a mailing containing a questionnaire, a cover letter on local hospital letterhead signed by their maternity care unit's manager, a study information page and a consent form together with a reply paid envelope. The development of the questionnaire is described later in this chapter. At each hospital, a midwife research assistant reviewed the bookings registered for each calendar month, gathered names and addresses, and sent a mailout each month to pregnant women at 28 - 38 weeks gestation. Those who wished to participate returned their completed questionnaire and consent form in the reply paid envelope and provided their telephone contact details for interview at 4 and 6 months after the birth.

Mail survey provided a low cost method of reaching participants over a wide geographic area. The emotive topic of breastfeeding also influenced the method selected for recruitment, as pregnant women who have not yet established breastfeeding may be threatened by a more confronting approach such as in-person interview, (281) with the potential for bias towards providing desirable answers. (282) Mail survey gave women opportunity to participate regardless of parity, birth type, or briefness of hospital admissions, compared to in-person maternity ward recruitment during business hours. This approach required no additional work by on-duty midwives, a critical issue in rural hospitals were there were staff shortages.

Survey administration during pregnancy enabled data collection on infant feeding intentions. In this study, it also enabled blinded allocation of mothers to intervention or control group based on their named choice of general practice site for maternal and infant care. Baseline data collection included variables identified in the literature as

107

relevant to breastfeeding duration, and maternal demographics, as described later in the results. The survey is shown in the appendices

Postcard reminder As Dillman points out, reminders, particularly in a different format from previous points of contact, increase response rates. A professional black and white photograph of a mother and baby was selected from the Australian Breastfeeding Association photo library. The postcard gave researcher contact details, a message of thanks to those who had already responded and encouragement to “reply today” to those who had not. (Appendix 3.3)

3.1.5 Participant local birthing hospitals Mothers from the city of Wagga Wagga and across Riverina towns and farming districts gave birth at the aging Wagga Wagga Base Hospital at no cost, or if they had private health insurance, usually gave birth at the newer private hospital, Calvary Health Care Riverina. Mothers from the twin city of Albury-Wodonga and across Border district towns and farming districts gave birth at the new Wodonga Regional Health Service, either at no cost or with private health insurance.

Hospital protocols differed regarding the gestation or stage of pregnancy they expected women to be registered on their antenatal database, or to ‘book in’. All hospitals preferred women to book in prior to labour. Bookings were expected from twelve weeks gestation at Wagga Wagga Base Hospital, from 20 weeks at Calvary Hospital, and from 28 weeks at Wodonga Regional Health Service.

Maternity care differed between hospitals. The majority of pregnant women at Wagga Wagga Base Hospital attended midwives at a hospital clinic, and gave birth with hospital midwifery care, supported by public obstetric specialists and trainees. Pregnant women booked at Calvary Hospital attended private obstetricians’ rooms and gave birth under the care of a private obstetrician and hospital midwives. At Wodonga Regional Health Service, a minority of pregnant women had private obstetric care; the majority of women with low risk pregnancies attended the rooms of a general practice obstetrician, and gave birth under the care of hospital midwives and the daily on-call general practice obstetrician.

Support for initiation of lactation was similar at the three hospitals, despite the differences in their accreditation status for the Baby Friendly Health Initiative (BFHI), as

108

discussed at the start of Chapter 2. (283, 284) Wodonga Regional Health Service gained BFHI accreditation 19 June 2008. Wagga Wagga Base Hospital began the application process in 2008. Calvary Hospital had not yet made application for BFHI accreditation. Each hospital had a lead International Board Certified Lactation Consultant, who in each hospital was employed on a routine midwifery roster with a full midwifery caseload matching that of other midwives on the same shift. None of these hospitals had a specified position to employ midwifery lactation support or specific funding to provide lactation training to other midwives.

Engagement with maternity unit managers and lactation consultants was required for ethics approval and for recruitment of midwifery research assistance.

3.1.6 Midwifery research assistance Confidentiality of hospital pregnancy and birthing data was maintained through employment of a midwifery research assistant at each hospital to manage recruitment of mothers. Murphy et al. discuss ethical concerns around confidentiality and the importance of separating recruitment from intervention to avoid bias in cluster randomised trials. (285)

Duties Organisational structures, computer software and ethics requirements differed at each hospital, influencing specific tasks. The role is detailed in Appendices 3.4-3.6. The midwife research assistant arranged display of the poster, accessed hospital records, addressed recruitment mailings and reminders each month and collected birth data for consenting women.

Data management At Wodonga Regional Health Service, data was available on the Birthing Outcomes System software. Reports were run and data entered in Excel for the researcher.

At Calvary Hospital, data was stored electronically, in paper registers and in paper files. Birth factors data was manually entered into blank fields on a supplied report for consenting women and posted this to the researcher for data entry.

At Wagga Wagga Base Hospital, the midwife research assistant used OBSTETRIX software to run reports and supply birth factors data in Excel format to the researcher. Uniquely, women book into Wagga Wagga Base Hospital from twelve weeks gestation.

109

Additional crosschecking was required as described in the ethics amendment to avoid sending a mailing to any woman who may have experienced pregnancy loss after booking in (Appendix A3.6).

3.1.7 Setting

Rurality The trial was designed to involve a broad sample of rural women, covering a wide geographic area to achieve an adequate sample sizedd. The trial's geographic setting was pragmatic; two rural Australian cities with population size >50,000 were geographically separated by less than two hours of road travel. Each region included public and private maternity services, public maternal and child health "clinics", immunisation clinics offered by local government, a community controlled Aboriginal health service, and private general practices where 92% of child immunisations occurred.

Wagga Wagga in the Riverina region of New South Wales is this state’s largest inland city, population 60,000. The twin city of Albury-Wodonga bridges the Murray River border between New South Wales and Victoria, population 80,000. Melbourne and Sydney are more than four road hours away, to the southwest and northeast. The study population represented those living in or close to these rural cities.

3.1.8 General practice recruitment Clustering reduces contamination of results at the staff or patient level of the intervention. Within each general practice, there are teamwork approaches and nursing routines that structure the care of all patient attendances in a particular category such as infant immunisation. There is also contact between patients in waiting rooms. Bias is best avoided when allocation of clusters to each group is truly random, and when the identification and recruitment of participants is undertaken in a process that is independent of the group allocation. (285) Clustered trials are useful for testing behavioural interventions to improve primary care practice, but they need to adjust for clustering statistically, and attend to potential adherence and recruitment issues. (216) dd According to the 2008 census and the Australian Standard Geographical Classification, 68% of Australians live in capital cities, 20% live in inner regional centres, and 12% live in outer regional, remote, or very remote locations. Remoteness is defined by road distance and population centre size. Outside metropolitan areas, there are eighteen identified large rural centres, each with a population of less than 100,000. Eight are inland in New South Wales and Victoria, including the two in this trial. (286)(286) Australian Institute of Health and Welfare. Rural, regional and remote health: A guide to remoteness classifications AIHW cat. no. PHE 53. Canberra: AIHW; 2004 March.

110

Figure 3.3 Location of Australian large rural population centres (286)

Inclusion criteria for clusters were  general practice located in Wagga Wagga or Albury Wodonga  employment of at least one practice nurse  immunisation of infants conducted independently by a practice nurse

Geographic regions in Australia are served by local general practice organisations, referred to in this thesis as Divisions. This trial involved practices within the boundaries of two adjacent Divisions: Riverina Division of General Practice and Primary Health (Riverina), and Albury Wodonga General Practice Network trading as the Border Division of General Practice (Border). Engagement with the Divisions and with the Australian Practice Nurses Association (APNA) were important aspects of preparation for the trial, as shown in Figure 3.1 and detailed at the end of this chapter.

Like many other Divisions nationally, at the time of this study Riverina and Border Divisions each employed a nurse whose role was to support the region’s practice

111

nurses. This role, funded by the national health program Nursing in General Practice (NiGP) was to arrange and promote local educational meetings, and to co-ordinate up- skilling and certification. Provision of immunisation and wound care by practice nurses drew specific national funding into each practice hence these were common education topics.

The process of recruiting practices for research is not to be underestimated according to New Zealand experience, (213) and this study concurred. General practice recruitment used a doctor-to-doctor approach, and team meetings were held during lunchtime meetings at each practice, lasting up to 30 minutes. Data was entered in SPSS 18 for initial analysis. Additional workplace visits were arranged to practice nurses in each of the fifteen recruited practices for initial knowledge and attitude quiz. Randomisation was not conducted until the completion of cluster recruitment.

3.1.9 Randomisation The recruited general practices were randomised as clusters as shown in the CONSORT flowchart, Figure 3.2. In this trial, the fifteen recruited general practices were initially stratified by region (Border or Riverina) to account for differing birth hospital experiences in each region. Secondly, they were stratified by the size of their infant population into two categories. Infant population was determined by a report from software in each practice on the number of patients aged 0 to 1 year of age registered at that practice. In each practice the practice manager ran the report and provided the information verbally.

Stratified general practice clusters were randomised by computer-generated numbers, using codes to represent the practices. In the Border region, three general practices were randomised to the control arm and four to the intervention arm. In the Riverina region, four general practices were randomised to the control arm and four to the intervention arm. A total of seven control and eight intervention sites were identified.

Participants were blinded to randomisation, being unaware of the group allocation of their chosen general practice when they entered the trial; however blinding was not possible for intervention practice nurses. Collection of outcome data was blinded, conducted by Computer Assisted Telephone Interview from Hunter Valley Research Foundation, without knowledge of allocation until the completion of the trial.

112

3.1.10 Statistical methods Participant data was collected by mail survey in pregnancy and by telephone interview at sixteen weeks and 26 weeks after the birth. Data was entered initially in SPSS, with every tenth entry double-checked for accuracy of data entry.

Exclusion criteria were applied by selecting cases in SPSS and deleting non-selected cases. Exclusions were described in section 3.1.3 and shown in Figure 3.2. Briefly, they were determined by the general practice attended by the woman, breastfeeding in the maternity hospital, and whether she was breastfeeding her infant at 8 weeks of age.

Analysis for the primary outcomes was on an intention-to-treat basis using STATA. Descriptive statistics were used to summarise the background characteristics of participants. Crosstabs were used to confirm effectiveness of the randomisation using Pearson chi square.

Adjustment for clustering, with fifteen clusters, was performed in STATA, although considerable variation was expected in the breastfeeding patterns of individual women. Binary logistic regression was used to analyse breastfeeding prevalence and exclusivity at 4 months and at 6 months. Where baseline characteristics were significantly unevenly distributed across the groups (p<0.05), outcomes were adjusted individually for each of these in the regression analysis.

3.2 INTERVENTION PROCEDURES This trial aimed to test an intervention, the setting and aspects of which are described in this section. Figure 3.1 shows the progress of the trial through the sequential phases of preparatory literature review, and initial engagement, then into the conduct of trial processes, and finally to outcome evaluation. Two strands are illustrated in this figure. Mothers, the first strand, experience progress through pregnancy, birth, and breastfeeding, with data collection at each stage as described. Nurses in the general practice clusters, the second strand, experience progress through training, delivery of the intervention, and cycles of interviews and updates. Alongside these two strands, instruments and tools are developed and used.

The intervention setting was the workroom of the general practice nurse during an immunisation appointment. Use of an established setting was intended to assist

113

integration of the intervention into routine nursing practice, routines being important to establishing a change. (287) Locating the intervention at the time of immunisationee was used similarly in a trial that delayed the introduction of weaning foods by West African mothers. (288)

The intervention method is a form of client-centred counselling known as motivational interviewing. How this approach could be effective in promoting healthy lifestyle choices is explored in a literature review below.

The intervention itself is an opportunistic conversation with breastfeeding mothers, to explore their experiences and responses to breastfeeding recommendations. The conversation follows steps on a desktop Conversation Tool. Use of a flowchart has been shown to prompt nurses successfully to talk with parents as a matter of routine, even with a daunting subject such as potential child abuse. (289) The flowchart can be adapted by the practice nurse within her routine appointment schedule.

The particular skills, knowledge, and attitudes required to deliver the intervention are developed in a training program, described below. Training is supported by resources that are described later in this chapter. Each of these components is intrinsic to the Conversation Tool intervention.

3.2.1 Literature review: motivational interviewing for health promotion

Background MI literature The intervention process is based on motivational interviewing, a goal-directed, client- centred counselling approach, used to assist behaviour change and first described by William Miller. (290) This method relies in part on Prochaska and Di Clemente's trans- theoretical model of change, (184) and Carl Rogers' counselling methods. (291) The seminal text by William Miller and Stephen Rollnick described motivational interviewing in detail particularly as used in the addictions field. (292) More recent reviews of the application of MI in the context of parents with young children have provided summaries of the method (293, 294), noting that paediatrics is a new area with few ee “A priority for the study was to test an intervention model that could be used at local health centre level by local doctors and nurses in continuation with everyday clinical practice. The intervention was therefore linked with routine immunizations at the health centre in order to obtain appropriate and realistic conditions. In order to assess the impact of the intervention, the study was designed as a randomized trial.” (288)(p741) (288) Jakobsen M, Sodemann M, Molbak K, Alvarenga I, Aaby P. Promoting breastfeeding through health education at the time of immunizations: a randomized trial from Guinea Bissau. Acta Paediatr. 1999 Jul;88(7):741-47.

114

reported trials. Key points for consideration in designing such an intervention were explored by Hecht et al., offering lessons from the experience of their community- based research team. They emphasised negotiating the client's agenda, exploring decision-making, and personalising feedback with the client. With regard to training the MI providers, they emphasised maintaining the "spirit" of MI, ongoing case discussion over time to build skills and the value of recorded sessions as a supervision tool. (295) In another exploratory study, rural healthcare providers found the idea of community- based brief MI intervention daunting, without any prior topical training (regarding physical activity) or any modelling of how to incorporate techniques into practice. (296)

Efficacy of MI has been extensively examined across a range of fields. A systematic review of 72 randomised controlled trials (published 2005) found MI more effective than traditional advice. This applied to behaviours including alcohol, weight, and asthma management, using MI in both brief and longer formats, delivered by nurses, psychologists, or physicians. Repeat consultations were particularly useful. (297) The critical elements of successful interventions were identified by Moyers et al., who found that interpersonal style was crucial. (298) These authors concluded that interpersonal skills must be built in order to enable an intervention that is consistent with the defined manner of conversational interaction: collaboration, eliciting talk about change, and emphasis on client autonomy. Miller and Rollnick's work was foundational to the Center on Alcoholism, Substance Abuse, and Addictions at the University of New Mexico USA, whose website bibliography lists 999 journal articles and 51 theses on outcome studies with motivational interviewing.ff

Systematic MI literature search with structured review A systematic Medline search was conducted using the term 'motivational interviewing' and MESH terms for nutrition, breastfeeding, infant, and child, which identified 58 journal articles. Studies were included where they addressed outcomes from MI interventions with parents of children aged less than five years in a community setting. Titles and abstracts were reviewed: exclusions were made for unrelated setting (2), unrelated population group (17), lack of trial outcomes reported (31) and duplicate study (4). Fifty-four papers were excluded.

ff http://casaa.unm.edi/mi.html

115

Results Four outcome studies (186, 299-301) were identified from this literature review and an additional four studies were identified from references and reviews. (302-305) Initial findings are promising although evaluation of motivational interviewing to improve the health of young children is an emerging field. (See Table 3.2) Each of these eight studies reports careful attempts to achieve consistency with the training described by Miller. (292) However, as others have noted, (293) the extent of training and assessment of proficiency vary widely across studies and it is important that fidelity checking be incorporated in future studies in this field.

These eight studies with parents of 1,078 children and infants found improvements in breastfeeding duration, dental caries, passive smoking exposure and in parental involvement in child social, oral health and treatment adherence behaviours. The smaller studies have less power to detect changed outcomes. The four larger studies had greater effect. They were associated with an increased number of contacts between the care providers and the participating family members. The study by Gross et al. led to improved duration of breastfeeding. The motivational breastfeeding intervention involved clinic environment change and acknowledgement of ambivalence, with posters, a video of mothers' stories about importance and barriers, and mothers were offered conversations with MI trained staff. (299) With five of these studies relating to aspects of nutrition (breastfeeding, oral health, and obesity), further research in this field is indicated.

116

Table 3.2 MI studies of health behaviours in parents of children 0-5 years

Sample Study Fidelity Reference Child Age Health Field Intervention Outcomes n Design Check

Gross Prenatal Breastfeeding 115 Cluster TG1:motivational videotape Significant: BF at 8 and 16 wks, Yes 1998 to 6 mo RCT, TG2: MI peer counselling at least all TG. (299) 4 weekly for 16 wks clusters TG3: both interventions CG: usual WIC support

Wilhelm 2-4 days Breastfeeding 73 RCT TG: MI and IBCLC session at 2 days, NS: BF at 6 mo No 2006 2 wks, 6wks. (186) CG: IBCLC assessments only

Shaw 17-27 mo Conduct disorder 120 RCT TG: family check-up, feedback MI Significant: More maternal Yes 2006 session plus ≤ 6 follow-up sessions involvement, at 4 yrs; less child (305) (mean 3.26) destructiveness at 2-3 yrs; child CG: no treatment aggression NS

Weinstein 6-18 mo Dental caries 240 RCT TG: pamphlet, video, 45 min MI Significant: reduced dental caries Not 2006 session, postcard x2 phone call x6 at 1 and 2 yrs, more effect than stated (301) over 20 wks education alone CG: pamphlet, video only

Freudenthal 6-24 mo Dental caries 72 RCT TG: MI session and follow-up phone Significant: increased brushing, Yes 2010 call at 1 and 2 weeks later reduced utensil sharing; attitude (300) CG: No formal oral health program change NS n sample size; RCT randomised controlled trial; CT controlled trial; TG treatment group; CG control group; BF breastfeeding; IBCLC International Board Certified Lactation Consultant; NS not significant; MI motivational interviewing; WIC women infants and children's supplementary nutrition program

Page 117 Page 118

Sample Study Fidelity Reference Child Age Health Field Intervention Outcomes n Design Check

Emmons < 3 yrs Parental smoking 291 RCT TG: MI home visit x1 30-45 mins Significant difference in home Yes 2001 plus phone call x4 over 6 mo nicotine levels; cessation NS (302) CG: usual care

Nock & 2-12 yrs Treatment 76 RCT TG: MI session 5-45 min x 3-8 (mean Greater adherence to treatment; Yes Kazdin attendance (for 6.4), additional child sessions if age > moderate to large effect with 8 2005 behaviour) 7 yrs sessions (304) CG: usual care

Schwartz 3-7 yrs Obesity 91 CT TG1: one paediatrician MI visit NS: Trend of decrease in body Yes 2007 TG2: paediatrician & dietician MI visit mass index (303) plus repeat at 3 months n sample size; RCT randomised controlled trial; CT controlled trial; TG treatment group; CG control group; BF breastfeeding; IBCLC International Board Certified Lactation Consultant; NS not significant; MI motivational interviewing

3.2.2 The Conversation Tool to guide the intervention Use of decision aids may help providers to be supportive, communicate information, and negotiate patient decisions. These may direct brief counselling, prompt identification of barriers, and guide appropriate referral. (306) However, these are mainly designed for use by patients. (307)

Development of the Tool A systematic literature review was conducted in Medline to explore the use of desktop office tools. Titles, selected abstracts, and selected papers were assessed. The search strategy used the subject heading ‘reminder systems’ and the terms tool, desktop, guide and flowchart, excluding papers focusing on computer-based systems, for applicability to non-computerised community practices. This review was supplemented with additional papers recommended by colleagues.

Three published tools were considered, developed to prompt generalist health workers as they discussed various behaviours with clients. A flowchart addressing child abuse risks was developed by an emergency department in Bristol, England, to prompt nurses to consistently ask all patients, despite potential discomfort (289) and this reduced gaps in patient records. A list of prompts addressing illicit substance use, including sample responses and questions, was developed for the international brief intervention trial ASSIST III to direct health workers and this trial had successful outcomes. (308) Thirdly, a colour-coded visual guide to identifying client attitudes to smoking was developed for the Australian Indigenous SmokeCheck program to prompt Aboriginal health workers according to client responses. (309) As with these examples, the Conversation Tool was designed for non-specialist counselling. An adaptation of motivational interviewing by the Royal Australian College of General Practitioners is available to guide opportunistic conversations to promote healthy adult nutrition, based on the Five A’s approach to tobacco cessation counselling. (310) These are ask, advise, assess, assist and arrange. However, no such resource targeting infant feeding was identified.

The choice to continue breastfeeding is consistent with the Relapse Prevention stage of decision-making described by Prochaska and Di Clemente. (184) They describe how the client's level of commitment may vary over time. The significance of suggesting a planned duration for breastfeeding outcomes is consistent with the theory of planned behaviour. (145, 156)

119

Figure 3.4 The Conversation Tool (Version 2)

120

Format of the Tool The colour A4 laminated flowchart (Figure 3.4, previous page) moves downwards through three horizontal panels. Panels are organised as sequential stages each with its own specific task following the motivational interviewing (MI) approach: beginning with Engagement (Panel 1), moving down to Response and action (Panel 2), and closing with Moving on out (Panel 3). Researcher contact details are included. The intended setting of the conversation appears in the title bar.

Engagement panel This first panel has simple questions about the mother’s situation and her response to the recommendations. Information giving is indirect, to avoid provoking resistance. Reflection and affirmation by the practice nurse are prompted. Mothers not currently breastfeeding are screened out.

The initial step is to broach the subject. Open questions allow the mother to tell her own story and enable the practice nurse to tailor specifically any information she then gives the mother. Drawing out the mother’s experience and views maintains consistency with the spirit of MI. Reflecting, expressing empathy, and giving specific goal-related affirmation are also consistent with the spirit of MI. (292) The conversation is guided towards a possible continuation of breastfeeding.

Contrary to the quantisation that is part of “Ask” in the model by Fiore et al. (310) this step does not involve numbers or measurement. Discussion around the role of weights, times and measurements in breastfeeding assessment suggests that these may be destructive to mothers' confidence and mothers' responsiveness to their babies' cues. (311) Breastfeeding recommendations are offered for discussion in a non-directive manner, in contrast with the “Advise” of the Five A’s in which the desired behaviour is strongly urged. In the spirit of MI, genuine acknowledgement of client autonomy is important. (292)

Response and action panel

The second panel guides the practice nurse to assess the mother’s response to the proposed healthy behaviour, in terms of motivation or readiness to continue breastfeeding. The practice nurse follows her assessment with the action of assistance with motivation. This panel has options, each with explicit questions and prompts so that the practice nurse takes the appropriate action for each mother, dependent on her readiness.

121

Three coloured boxes sit alongside each other visually representing alternative responses that could be made by each mother. The red, amber, and green boxes visually prompt the practice nurse to assess the mother’s motivation, respectively as stopping, unsure, or continuing breastfeeding. Possibility of movement across the colours is indicated by a horizontal arrow.

For the mother who plans to stop breastfeeding, the red first box prompts the practice nurse to give a fact or two on the importance of breastfeeding while accepting the mother’s autonomy. Indirect information giving again aims to avoid provoking the mother to defend and reinforce her position. This action supports the mother to consider ongoing breastfeeding. Following piloting this action was also discussed as "planting a seed".

For the mother who is unsure or ambivalent about continuing, the orange box prompts the practice nurse to ask the mother about firstly, the barriers to progress and secondly, about motivating factors for continuation. This action develops discrepancy and creates a supportive space for the mother to realise her ambivalence or doubts. While acknowledging the challenges, the practice nurse supports the mother to consider ongoing breastfeeding with a summary of her answers, which may tip the balance for the mother towards a choice to continue. Following piloting this was also discussed as "weighing it up".

For the mother who plans to continue breastfeeding, the green box prompts the practice nurse to give anticipatory guidance, (312, 313) suggesting potential barriers ahead and eliciting possible solutions. This action supports the mother’s self-efficacy. The practice nurse supports the mother to plan for challenges such as separation from her baby, breastfeeding in public or negative feedback from other people. Following piloting this was also discussed as "looking ahead".

Moving out panel

The third panel focuses on offering follow-up, helping the mother to access ongoing support and information when she walks out the door, which is a process associated with maintenance of breastfeeding. Resources are offered rather than arranged, unlike Fiore's model. (310)

122

Mothers may lack the knowledge or confidence required to contact support people or services, despite the demonstrated efficacy of skilled peer support. (158) The Conversation Tool guides the practice nurse to offer local specialised breastfeeding assessment services, to normalise help-seeking by mothers, and to offer the booklet Breastfeeding Confidence. (See section 3.2.3.2) Again, in this panel, the practice nurse provides basic information, elicits doubts, and helps to build the mother’s confidence.

Testing of the Tool The initial draft of our Conversation Tool was reviewed in three cycles by a Motivational Interviewing Network Trainer. The final draft was reviewed by eleven experts across fields of breastfeeding support, general practice, motivational interviewing, and public health research.

Appropriateness of the tool for use by practice nurses was checked by literature review, by consultation with three practice nurses, two senior academics in the field and two local general practice organisations and by review with the policy and research committee of The Australian Practice Nurse Association. (314)

Practice nurses in the Border region piloted use of the tool before the intervention began leading to the slight change in descriptions in the second panel, described above.

3.2.3 Training workshop aims In brief, as detailed in this section, training had seven main aims.  Improve practice nurse breastfeeding knowledge.  Improve practice nurse breastfeeding attitudes.  Develop practice nurse communication skills.  Introduce practice nurses to motivational interviewing.  Stimulate reflective practice.  Orient practice nurses to materials.  Motivate practice nurses to deliver the Conversation Tool intervention.

The materials and process of practice nurse training are described in this section: the resources and the curriculum, with descriptions of their role in the delivery of the intervention, which aimed to increase the support available to breastfeeding mothers.

123

One-day training sessions for nurses have been shown to produce appropriate changes in practice, where the training sessions focused on using behaviour change counselling, as with this research. (315) A systematic review of seventeen trials (316) indicated that patient-centred consultation style could be taught to health professionals working in primary care. It was unclear from these studies whether this improved health outcomes.

3.2.4 Resources provided at the training workshop Local and established resources were used for practice nurse training wherever possible. A resource folder was given to each practice nurse who attended the training, with training materials and for future reference. Practice nurses were also given two posters and a quantity of literature for handout to mothers.

Contents of the resource folder (detailed in Appendices) included  Conversation Tool (colour, laminated).  One Breastfeeding Confidence booklet  Other handout literature.  Local maternity services directory.  Training session handouts and worksheets.

A Breastfeeding Confidence booklet was offered to each breastfeeding mother, according to the Conversation Tool, at the 2-month immunisation visit. Other material was offered to mothers at the practice nurse’s discretion.

Breastfeeding Confidence booklet This booklet, published by the Australian Breastfeeding Association, (317) addressed twelve common questions raised by breastfeeding mothers of babies aged 0 – 3 months as shown in Figure 3.6. It was a 34 page, DL sized booklet, in glossy full colour. Information was presented in point form with large print, including quotes from mothers. Key points were highlighted and pages were illustrated with photographs. As handout literature, this booklet was selected as an attractive, accessible, and reliable source of breastfeeding information.

During breastfeeding management training, attention was drawn to pages in the booklet that illustrated and summarised topics, and use of the booklet with mothers was discussed. Written material has supplemented verbal information in successful

124

Figure 3.6 Topics in Breastfeeding Confidence

1. What should I expect? 2. What breastfeeding provides 3. How often should I feed my baby? 4. How to position your baby for feeding 5. How do I know if I have enough milk? 6. Looking after yourself 7. What’s the let-down reflex? 8. Some tips if you have sore breasts or nipples 9. Expressing and storing 10. Getting out & about 11. Your baby’s weight 12. Where to find out more about breastfeeding About the Australian Breastfeeding Association Who to call for assistance with breastfeedinggg

interventions (243, 252, 273); isolated use of written materials, as Webb et al. point out, is consistently shown to be ineffective. (318) Face-to-face contact with breastfeeding mothers is a feature of successful interventions by health professionals, as identified in Chapter 2.

At the time of this intervention, this booklet was new to most mothers unless they had access to it through the Breastfeeding Helpline or from Calvary Hospital in Wagga Wagga. Pregnant women who booked in to Wagga Wagga Base Hospital were given a NSW Health booklet about breastfeeding; compared to Breastfeeding Confidence it was more text-based, without photographs and had an instructive rather than problem solving approach.

Starting family foods leaflet

This leaflet explained why and how to introduce “solids” to infants from 6 months of age, and also addressed weight gain and sleep concerns. It was a double-sided A3 page folded to DL size with a similar format to Breastfeeding Confidence, in glossy full colour. The approach was breastfeeding-friendly and it provided various options to seek additional information. Misconceptions about these topics can prematurely end

gg The most common questions raised by mothers of babies aged 0-3 months, ABA (formerly Nursing Mothers’ Association of Australia) data from regular surveys of breastfeeding counsellor volunteers on the 24-hour telephone helpline. (317)

125

the period of exclusive breastfeeding. (6) Starting Family Foods, published by NSW Health, was developed by a team of nutritionists and ABA volunteers on the NSW Central Coast.

Practice nurses were given the leaflet to hand out from their second contact with a mother. As a convenience sample of NSW general practices, the intervention practices in this trial were not familiar with the leaflet prior to the trial. At the time of this intervention, mothers had access to this leaflet if their general practice was aware of NSW Health publications and had ordered this leaflet, however it was not widely promoted.

Balancing breastfeeding and work leaflet This leaflet covers considerations for breastfeeding mothers returning to employment, briefly listing options for preparation, negotiation, facilities, childcare, and maintenance of breastfeeding. The 2004 version is available on the website of the national Department of Health and Ageing. It was produced as part of the National Breastfeeding Strategy at the time, in association with the University of Adelaide and the South Australian Employers’ Chamber of Commerce & Industry Inc. (319) It prints as a double-sided A4 leaflet folded to DL with line illustrations in single colour, formatted in question-and-answer points. For this study, the contact details were updated for printing.

The need for this resource was identified during the pilot phase of the trial. In the process of the conversation, if practice nurses identified that mothers were returning to work they offered this leaflet.

Local services directory This compilation of contact details was developed for the study from information provided by local Divisions, ABA groups and state health services, either online or through personal approach. Services included community clinics with early childhood nurses, day-stay parentcraft centres run by the public hospitals, ABA mothers’ groups, helpline telephone numbers, and websites. These contact sheets could be photocopied as handouts.

Ready access to local referral options may result in more referrals; with specific encouragement and recommendation, more mothers may seek skilled help. Awareness of suitable referral options for mothers in difficulty may increase practice nurses'

126

comfort levels when broaching the subject of ongoing breastfeeding. Clifford et al. found that primary care practitioners working with behaviours around alcohol use were less likely to undertake a motivational brief intervention when they perceived a lack of appropriate referral options. (320)

Posters

Two glossy laminated posters featured infants over 4 months of age. These were developed by the Australian Breastfeeding Association to increase community awareness and acceptance of breastfeeding as a routine method of infant feeding. An A3 colour poster showing positive photographs of breastfeeding mothers and an A4 black and white poster of a baby's face showing information about exclusive breastfeeding were provided to each practice nurse.

For pragmatic reasons, implementation of poster display in the general practice, either in the patient waiting area or in the room used by the practice nurse, was determined by each practice nurse. Poster display was intended to build a welcoming environment for breastfeeding mothers.

3.2.5 Delivery of the training sessions The curriculum was adapted from three established courses as described below which covered communication skills, motivational interviewing, and breastfeeding management. Curriculum was also developed to cover community support for breastfeeding mothers and reflective practice. A sample program is shown in Figure 3.5 and Powerpoint slides are shown in Appendix 17.

Training was delivered as two sessions of five hours each, held one week apart, as described below. The importance of educating the health professionals who deliver breastfeeding interventions is highlighted in the systematic literature review in Chapter 2. The majority of training time incorporated an eight-hour professional course in breastfeeding management. Photographs of mothers who were breastfeeding older babies were added to strengthen the focus of the professional course material on increased duration of breastfeeding. The style of training delivery was guided by the theoretical framework of motivational interviewing.

Before starting the intervention practice nurse training, a compressed training module was piloted with practice nurses in a non-participating practice, as described below. For practice nurses who were employed late in the fourteen months' trial duration, a

127

compressed training module was used, with additional home study and telephone follow-up, also described below.

Figure 3.5 Training workshop sample program

Session 1: 5 hours 12.30 pm Lunch and quiz 1:00 pm Welcome and Introduction 1:20 pm Communication with mothers 2:30 pm Motivational interviewing Intro to Conversation Tool 3:40 pm Afternoon tea 3:55 pm Importance of breastfeeding 4:15pm Hazards of formula feeding 4:25pm DVD 4:45pm Positioning and attachment 5:00pm Anatomy and physiology of the breast 5:20pm Questions Summary, intro to reflective journal and close

Session 2, one week later: 5 hours 12.30 pm Lunch and meet with ABA counsellors 1:00 pm Welcome, discuss reflective practice exercise Review Conversation Tool 1:20 pm Communication with mothers – revision 1:45 pm Breastmilk composition Frequency and duration of breastfeeds Expressing 2:10 pm BFHI: Ten steps to successful breastfeeding BFHI: Seven point plan for community health centres 2:25 pm Management of common problems 3:35 pm Afternoon Tea 3:50 pm Issues impacting on breastfeeding Case study 9 5:00 pm Community resources 5:05 pm Review Conversation Tool 5:20 pm Questions, summary and close

128

Training was delivered locally in each regional centre to enable minimal travel time away from the workplace. Repeat sessions ensured that all intervention nurses had consistent training. Team delivery involved a registered nurse and a general practitioner to deliver peer-based training in an interactive adult learning format. The trainers' professional and lactation expertise was shared from the perspectives of midwife and general practice obstetrician and perspectives of International Board Certified Lactation Consultant and ABA breastfeeding counsellor.

Adult learning techniques were used, with purposive use of reflection and story telling. Adult learning principles assume that adults are self-directed, self-motivated, and come to learning with past experience. (Brookfield, Knowles, Cited p618 (321)) In accordance with these principles, brief didactic presentations were interwoven with small group work; extensive individual question and answer; and illustrations. Normal breastfeeding behaviours of older babies and toddlers were presented through discussion of photographs of older babies and toddlers breastfeeding, and through story-telling using brief case histories and examples from the personal or counselling experience of the presenters. Presentation style was open to the nurses raising their expectations and personal experiences for discussion. They were introduced to trained volunteers who visited, with their children, with opportunity to discuss their experience and approach.

Training addressed the process of change on two levels; training aimed to facilitate change in the breastfeeding behaviours of mothers and aimed to facilitate change in the health promotion behaviours of practice nurses. One-day training sessions for nurses have been shown to produce appropriate changes in practice, where the training sessions focused on using behaviour change counselling consistent with this approach. (315) The method of training delivery was designed to model and be consistent with the spirit of motivational interviewing, described by Miller (292) as involving autonomy, evocation and collaboration. Trainers recognised practice nurse autonomy in affirming ideas for flexible intervention delivery in individual workplaces. Trainers evoked practice nurse experiences, reflections and potential solutions to challenges and barriers encountered. Trainers collaborated with practice nurses in offering new knowledge and workplace resources, helping to develop their nursing practice over time.

129

3.2.6 Curriculum for the training sessions

Breastfeeding management No basic breastfeeding knowledge was assumed. The foundational curriculum “Breastfeeding Management in a Baby Friendly Health Service, an 8 hour structured course for health professionals working with mothers and babies in maternity settings” as described earlier by authors Byrne and McIntyre, (322) was generously made available at no cost by Australasian Lactation Courses (ALC). It is suitable for flexible delivery and the 2008 edition was used with minimal adaptation. Course materials, delivery recommendations, and topics are described in the course introduction (Appendix A5.4). PowerPoint slides for this study were developed from course transparencies (Appendix A5.5, A5.8).

From a scientific perspective, breastfeeding is a reference behaviour against which the use of infant formula is an intervention with associated health risks. This approach was used wherever possible through the training sessions. To establish the importance of breastfeeding and of the breastfeeding recommendations (30), health risks to formula- fed babies were discussed according to the varying levels of evidence (323) (Appendix 5.7.3). Historically formula feeding has been over-recommended by Australian health professionals. (63)

The curriculum was divided across the two training sessions. In the first session, diagrams, photographs, and the DVD ‘Follow me mum’ were used to illustrate the anatomy of the maternal breast, the infant mouth, and correct positioning and attachment or latch. Physiology of breast milk production was discussed with reference to Lactfacts 2:10 Increasing the Milk Supply and course diagrams about relevant hormonal responses. The importance and mechanisms of breastfeeding formed the knowledge foundation from which problem solving could emerge.

In the second session (Appendix A5.7 Day 2), a written exercise on the components of breastmilk reinforced the importance of breastfeeding recommendations. Discussion on frequency and duration of feeding was illustrated with case stories known by the trainers or by participating practice nurses who had breastfeeding experience. Common behaviours of older breastfeeding babies and their mothers were discussed. For example, time use differences have been identified from the Longitudinal Study of Australian Children. (324)(p11)

130

By age 11–12 months or older [<15 months], if still breastfeeding, infants were spending about one and a half hours a day less on breastfeeding than the 3–4 month old infants.

Options for expressing breastmilk were covered with reference to the breastmilk storage table in Breastfeeding Confidence.

Cultural context and the level of support in the family's environment were discussed across both sessions using PowerPoint slides about the Baby Friendly Health Initiative, community campaigns and the community support activities of the Australian Breastfeeding Association.

In the second session, potential barriers to continuation of breastfeeding were discussed in pairs or small groups using problem-solving exercises; practice nurses used the participant material, Lactfacts and Breastfeeding Confidence to seek solutions. Common challenges considered were nipple pain or trauma, thrush (Candida species) infection, blocked milk ducts and mastitis, and actual or perceived low milk supply. Breast refusal was mentioned briefly using a case story, due to time constraints. Potential situational barriers were discussed using relevant photographs of women managing in these situations. Discussion addressed strategies around breastfeeding in public; when employed, fatigued, or getting out of the house; and in the context of relationships with older children, the woman’s partner, and the grandparents. Training closed with discussion of several case presentations that addressed substance use, infant weight gains, maternal life choices, and maternal distress.

Communication skills

Section 3.2 of the WHO/UNICEF BFHI 20-hour course for maternity staff, Session 2: Communication skills (233) was used with permission from WHO. This module was expanded to 75 minutes across the two sessions, and included role-plays and worksheets. In the supplied seven role-playing demonstrations, a child’s doll was used as a mock baby according to module instructions. The final role play addressed the context of a potential HIV diagnosis but this was edited to potential alcohol use, for greater relevance to rural Australia. Worksheets were used early in the second session, to reinforce learning from the first session.

Communication Skills was the first module delivered, in order to undergird training sessions. The Conversation Tool developed for the intervention has several

131

communication prompts, and these were pointed out as the required skill was addressed in this module. As discussed earlier in this chapter, these skills are foundational for successful motivational interviewing. (298) This module of the curriculum was interactive and emotive. Understanding and learning can be increased when the topic is connected to personal experiences. (325)

Demonstration role plays were a means of developing insight for the participants, using a “how not to” version prior to the “how to” version. In this way myths were challenged and assumptions revealed, such as recommending formula for perceived low milk supply. Common pitfalls were revealed, such as relying on personal experience. Demonstrations also allowed the group to bond with humour. Skills covered included non-verbal communication; use of open questions and reflection; empathic and non- judgemental responses; acknowledging the mother’s feelings, achievements and practical needs; and providing information as brief options in accessible language. This module concluded by discussing follow-up, help-seeking and referral, and the process of summarising the conversation.

Graffy found in his study of 720 British breastfeeding women that "They want to be listened to and encouraged without feeling pressurised." (139)(p185) However, in Australia breastfeeding information has become part of the biomedical instructive literature. Mothers are vulnerable to cultural influences and this biomedical context may be counterproductive. (40) We concluded that the method of communication was as important as the message content, particularly because early weaning relates to various combinations of low knowledge, low confidence, and low support.

Communication training for health professionals working with mothers has been a component of both successful interventions to increase breastfeeding rates, and in studies assessing related outcomes; Chapter 2 of this thesis notes the success of the Baby Friendly Health Initiative (BFHI) training courses, jointly developed and regularly revised by the WHO and UNICEF, in prolonging both exclusive and any breastfeeding. BFHI training emphasises communication skills as foundational, and this approach is supported by outcomes research into maternal knowledge and behaviours. A study in Brazil found an association between higher maternal recall of nutrition messages and greater use by physicians of the counselling skills learned in specific training. (140) In Sweden, researchers found that when Swedish nurses received extensive communication skills training, there were persistent positive changes in intimacy

132

between the mother and the baby. These changes were both perceived by the mother and independently observed at three days and at nine months post-partum. (141)

Motivational interviewing This component was based on an MI training program developed and delivered by Joel Porter, an experienced clinician and presenter in this field, and kindly made available by him for this study. It was informed by work for conference presentations made in 2007. (326) One hour on this topic in the first session covered the concepts and their use in the Conversation Tool.

Practice nurses were introduced to motivational interviewing as “a person-centred, goal-oriented approach for facilitating change through exploring and resolving ambivalence.” (327)(p138) The concept of ambivalence and different stages of readiness in decision-making around breastfeeding was also addressed with reference to Australian women's stories of their baby-feeding decisions in pregnancy. This article by Sheehan et al. (178) was provided in the resources folder with permission.

Questioning and role play exercises led the practice nurses to consider  the sources and outcomes of ambivalence towards breastfeeding,  the destructive potential of an authoritarian approach to an ambivalent person,  and the limitations of common non-person-centred helping roles.

Evidence demonstrating successful outcomes with a motivational interviewing approach as well as reports of provider satisfaction were presented.

The stages of change were presented with reference to the Conversation Tool, with particular attention to pre-contemplation, contemplation and action. These stages were described as “traffic light” messages from the mother that showed her potential responses to the idea of ongoing breastfeeding: stop, caution, or go ahead. The purpose of the three matched nursing actions was explained, namely to support autonomy rather than to provoke confrontation or resistance, to assist weighing up, and to discuss strategies.

This approach aligned with nursing priorities in patient care, (See the practice nursing literature review, section 1.7.3) anticipating that practice nurses would want to avoid potentially distressing conversations with mothers. Motivational interviewing in its essence is a caring approach that offers a communication solution for caring health

133

professionals. Hauck et al. observed that mothers have found others’ expectations of them burdensome, describing feelings of distress, anxiety, and guilt. (105) Breastfeeding decisions are part of the complex and challenging environment that is motherhood, according to the many Australian mothers whose "true feelings about motherhood" are documented by Ita Buttrose and Penny Adams in their book Motherguilt. (3) In this context, MI was emphasised as an appropriately supportive approach with non-directive provision of information, an improved health promotion method.

The interactive elements of motivational interviewing were explored in several steps using a model, with the major focus being on the spirit of motivational interviewing: autonomy, evocation, and collaboration. (292) Quotes from mothers were interspersed to illustrate the need for this approach, “it’s almost like you as a person are gone and we need your body and we need those breasts and we need those nipples out." (178) It was emphasised that the mother would make her own choices (autonomy), as the practice nurse worked alongside her (collaboration), drawing out the mother’s own perceptions, experiences, and potential solutions (evocation). This support could lead the mother’s behaviour choices towards the breastfeeding recommendations. There are behavioural similarities between the choice to continue breastfeeding and a person’s choice to comply with other healthy lifestyle recommendations. (186, 328) It was an aim of training that practice nurses be able to recognise these similarities, to see how motivational interviewing describes these settings, and to use the motivational interviewing approach with breastfeeding mothers.

Micro skills had been covered in the Communication Skills module; change talk and commitment talk were mentioned only briefly due to time constraints and due to the earlier elements having more evidence for efficacy.

Mother-to-mother support

Volunteer trained breastfeeding counsellors led informal discussion during a refreshment break in either session, depending on their availability. They were accompanied by their own babies and young children, simulating aspects of the support group experience. Volunteers explained their code of ethics, education qualifications, and approach to supporting mothers.

Each practice nurse had opportunity to engage informally with at least one volunteer trained breastfeeding counsellor during the sessions. Practice nurses spontaneously

134

raised case examples and experienced receiving management suggestions from the counsellors. The potential impact of support for mothers from experienced breastfeeding women was also addressed at various points through the training sessions.

This study aimed to influence the referral patterns of the practice nurses in order to increase mothers' access to ABA and other supportive resources. Evidence from literature and systematic reviews shows that peer support can prolong breastfeeding. (45, 194) This study aimed to increase mothers’ access to the Australian Breastfeeding Association whose trained volunteer counsellors are experienced breastfeeding mothers. Potential avenues for access have been discussed above.

A friendly introduction to some trained ABA volunteers, demonstrating their educational and parenting roles in a meeting setting with the nurses, was intended to influence the referral patterns of the nurses. Beliefs, behaviours, attitudes, and intentions toward professional and self-help groups were assessed in a study of 975 US mental health professionals across 109 agencies. (329) Of these, 35% had never made a referral, although community-based professionals were more likely to refer. They noted that of those professionals experienced in referring to support groups, 92% would refer again in future. However, contrary to evidence about relative efficacy, most professionals preferred referring to professional-led groups, and expected them to be more effective. The authors suggested, to increase collaboration and referral, strategies such as increasing professional contact and experience with these groups and their meeting processes, and simulating a group meeting during professional training. The invitation of counsellors and their children to a workshop session was according to this process.

Putting referral pathways into practice can be challenging, and creating a sense of familiarity was intended to support this process, for the benefit of mothers. One author conducted an extensive background literature review of referral patterns to consumer- run groups, noting that "in published studies, referral rates to self-help groups are generally 50% or less." (330) He found that "Familiarity with existing resources appears to be a key factor." However, he noted, "Awareness is a necessary but not always sufficient condition for referrals". His own study recruited 301 USA mental health professionals across 46 states. He found significant associations between referral rate and workplace collaboration as well as between referral rate and contact with consumer-run groups and programs. His work focused on the needs of people with psychiatric disabilities; however, his observations of professional-lay organisational

135

interaction have some relevance. Given the increasingly evidence-based methods of these groups, he concluded, "The potential for increased referral to and collaboration with consumer-run programs is striking". (330)

Reflective Practice At the end of the first session, the Reflective Journal was introduced, with reference to a diagram of Gibbs’ reflective cycle. (331) (Figure 3.8) An informal personal handwritten record was to be brought for discussion in the second session. Nurses were asked to raise the subject of breastfeeding with one patient in the intervening week, to use patient-centred communication, to communicate one message about the importance of breastfeeding, and afterwards to reflect on this experience. In the second session, trainers guided each nurse in turn to discuss experiences and reflections following the steps in the Gibbs reflective cycle. Trainers used and demonstrated a motivational interviewing approach to the nurses’ experiences of undertaking a new behaviour.

Stories created and discussed in this exercise were used to address practical and personal barriers. Discussion took place in an atmosphere of trust and sharing, acknowledging the challenges, and using the reflective process to support the Practice nurses in “survival and development”, to use Reid's description, as they incorporated the intervention into routine practice. (332)(p308)

Intervention efficacy depends on the actions and attitudes of those tasked with intervention delivery; while the training addressed attitude change more broadly as well, this component of training focused also on the practical implications of introducing a new nursing procedure into daily work. Attitude change is a key factor in changing health service practices to better support breastfeeding mothers (333); it is also an aim of the WHO/UNICEF 40 hour course for the Baby Friendly Health Initiative which was used to develop the ALC course used in our Communication Skills training. Reflection on one’s own actions is an important part of acting more effectively and possibly transforming the workplace environment. According to Argyle and Schön's theory of action science, a model that uses internal commitment, free and informed choice, and valid information will produce more effective behavioural strategies than a traditionally more defensive model will produce. (334)(p330)

136

Figure 3.8 Reflective practice exercise

------REFLECTIVE-----JOURNAL------

Completing a reflective journal helps nurses to evaluate their clinical experience. Reflecting on practice is a way nurses can learn from experience.

For this exercise, we want you to have a conversation with a pregnant or breastfeeding mother who comes to see you: - practice an open question or two about her views on breastfeeding. - reflect on what the mother tells you - ask if she would like to hear some facts about breastfeeding

Following this interview, you are to complete a reflective journal. Please complete your journal as handwritten notes as soon as possible following your client interview. For your journal please also record any facts about your day that you feel were relevant to your experience.

Completing this journal is for your personal learning. Did you feel you were able to put into practice what you learnt about communication skills? Try to identify factors that made practising these skills easier or more difficult. Please refer to Gibbs’ model of reflection to assist you with this exercise.

Your journal will not be submitted however, we will discuss the journals at the second session.

Gibbs’ model of reflection (1988)

Description What happened? Action Plan. Feelings If it arose What were you again what thinking and would you do? feeling?

Conclusion Evaluation What else What was good and could you bad about the have done? situation? Analysis What sense can you make of the situation?

137

The “situations of practice” where professionals such as nurses conduct their daily work are characterised by “complexity, uncertainty, instability, uniqueness, and value conflict” according to Schön. (335) In such situations, self-awareness is important for the professional who needs to learn a new practice, to “deal with negative self feelings” during the learning process, as Reid puts it, (332)(p307) and to respond effectively to each individual patient's situation. Reid described reflective practice as “the process of using reflection to influence how practitioners approach and respond to varying situations.” (332)(p305) Hence, teaching reflective practice was seen as a key component of preparation for the intervention.

3.2.7 Piloting the training delivery and curriculum The compressed four-hour module used to pilot the training process was evaluated by the trainers (Appendix A4.10) and the evaluation was discussed with three experienced educators.

Content and delivery

The pilot module introduced the research methods, Conversation Tool, and workshop program. The communication skills module was delivered as a one-hour interactive component and the full 40-minute motivational interviewing presentation was delivered. Breastfeeding management was summarised, covering the importance of breastfeeding and risks of artificial formula; breast anatomy; physiology of breast milk production; feeding patterns; and expressing. The 20-minute DVD Follow Me Mum by Rebecca Glover, about positioning and attachment, was used. The session concluded with 30 minutes of small group work on Case Study 5, the unsettled baby, practising open questions and offering management suggestions, followed by group discussion.

Summary of evaluation Practice nurses needed easier access to evidence for the importance of breastfeeding. They were unfamiliar with the idea of opportunistic health promotion around ongoing breastfeeding, and in their nursing practice they were possibly reassuring mothers about weaning in inappropriate contexts. They may be reluctant to discuss breastfeeding in some contexts: lack of personal experience; personal unease with the concept of sustained breastfeeding beyond 6 to 12 months; or when a mother found breastfeeding challenging. Structural aspects of the delivery session and the introduction of resource materials needed to be clarified, and time management improved.

138

3.2.8 Alternative brief format training Nursing staff turnover was anticipated within the intervention practices during the two years of trial preparation and conduct. There is little Australian published data on this issue, although in Ohio USA one large study identified 53% staff turnover in primary health care practices during a two year period. (336) Due to funding constraints and pragmatic reasons, a brief training session of three hours duration was developed for any nursing staff employed after the trial started.

Sessions were held in each region locally, at rooms of the Divisions. One trainer delivered the program working with a minimum of two nurses in each session. One session was required in each region during the study.

Complete training resources were provided. The initial communication skills module with demonstration role plays was retained entirely. The curriculum was abbreviated in delivery. Case studies and most problem-management exercises were transferred to home study options or to update meetings held later. The motivational interviewing module was abbreviated to the rationale and one role play activity with the Conversation Tool being used as the model for addressing stages of change. Mother- to-mother support was addressed only with reference to the resources. The reflective practice exercise was explained and individual follow-up was planned by telephone interview.

3.2.9 Supporting practice nurse to hold new conversations Practice nurses who were delivering the intervention were considered both agents and subjects of change processes. In the interests of maintaining the integrity of the intervention, ongoing mentoring was provided for the Practice nurses, consistent with processes identified in Chapter 2, the systematic review of RCTs in this field. Researcher contact details were given to the Practice nurses who were encouraged to make ready contact however mentoring was provided mainly as a pro-active support process.

Development of a new nursing behaviour can be slow. In an action research project described in the compiled report on projects funded by the United Kingdom's Infant Feeding Initiative, nursing researcher Mary Price introduced and evaluated a new nursing procedure. While the context differs, the procedure was similar, being a deliberate conversation between midwives and pregnant women about the option of skin-to-skin contact after the birth, to promote breastfeeding. Price found that memory

139

prompts for the nurses were required, to enable full incorporation of the new nursing procedure (238) hh.

Workplace visits Each practice nurse was visited in her workplace once or twice during the cluster recruitment process, and once at two to four weeks after the training sessions. Research appointments were held in the practice tearoom or in the practice treatment room, individually at six practices and in pairs at one practice in each region, over fifteen to 40 minutes.

During workplace visits, the practice nurses’ physical workspace was noted, with informal assessment of practice layout and of waiting room resources for patients. Visits established rapport with the practice nurse and explored individual practice facilities and dynamics. These factors facilitated communication and co-ordination with the practice nurses.

Post-training visits were used to administer the post-training questionnaire (Appendix A4.6), to conduct a semi-structured interview (Appendix A4.8) and to ensure nurses had sufficient supply of handout literature.

Workplace telephone calls Each nurse was telephoned at least twice at the practice over the trial duration following the workplace visit, at a suitable time in her appointment schedule. These calls aimed in part to support each practice nurse in her motivation to deliver the intervention, as a new nursing process. (See Appendices for details) Beyond qualitative data collection, interviews enabled individual mentoring of each nurse as she instituted a new nursing process, providing encouragement, information, and support for specific challenging cases.

Evening education update meetings

In each region, practice nurses attended at least one update, which was structured as case-study-based continuing education. These were held at the local Division of

hh “Finally, remembering was facilitated, recognising that new practice may be hard to incorporate into a busy day’s work... A focus group identified a problem of remembering to discuss skin-to-skin contact... A comment was made about the difficulty in remembering everything about the benefits of skin-to-skin contact, so a bullet point list was produced... Together, these facilitated the giving of quality information about the value of skin-to-skin contact rather than just ‘do you want it?’” (M Price, cited) (238)(p164-165)

140

General Practice, one evening after work, with light refreshments and handouts provided. Four update sessions were held across the duration of the trial, beginning six weeks after the start of the intervention. Handouts were mailed to those who did not attend. Cases were selected from Royal Australian College of General Practitioners' case-based self-assessment program issues on breastfeeding. (337-339) Update meetings aimed to  provide breastfeeding management education in more depth on selected topics  revise and reinforce communication skills and the motivational interviewing approach  enhance practice nurse fluency with the Conversation Tool  discuss hurdles and successes in delivery of the intervention  qualitatively explore the extent of intervention delivery  support practice nurses in their motivation to continue to deliver the intervention  identify any need for additional telephone support or breastfeeding topic education

Meetings were scheduled in the format of continuing professional education run by Divisions on other topics, to facilitate ease of attendance and consistency of delivery. Case-based discussion allowed peer group learning and addressed practice nurse interest in further education, as identified in the literature review in this section.

3.3 DATA COLLECTION This section describes the selection, adaptation, and development of instruments used in data collection.

3.3.1 Primary outcome measures The primary outcome measures in this study are eight maternal breastfeeding behaviours. Breastfeeding outcomes were assessed by data collected at sequential Computer Assisted Telephone Interview (CATI). (See section 3.3.4.2 below) Four outcomes were developed from data collected at 4 months. The remaining four outcomes were developed from data collected at 6 months when the questions were repeated for any mothers who had not weaned before the 4-month interview. The outcomes are:  any breastfeeding (any additional food or fluids also permitted)  full breastfeeding in the past 24 hours (any water-based fluids also permitted)  exclusive breastfeeding in the past 24 hours (no other foods or fluids)  exclusive breastfeeding since birth (no other foods or fluids)

141

3.3.2 Secondary outcome measures Breastfeeding outcome data from the CATI enabled evaluation of the three currently used and alternate definitions of exclusive breastfeeding. Maternal questionnaire data enabled consideration of the baseline demographic profiles of the groups and of factors that may influence breastfeeding decisions in order to adjust for potential confounders. The instrument is described below in section 3.3.4.1 and attached in appendices.

Practice nurse breastfeeding knowledge and attitudes quiz, conducted prior to and again approximately two weeks after training gave evidence of a training effect. The instrument is described below in section 3.3.4.3 and attached in appendices. Analysis is presented in Chapter 5.

Practice nurse interviews explored the practice nurses' experiences in the research and assessed the feasibility of and adherence to research protocols in general practice, acting as a qualitative approach to process evaluation and analysed in Chapter 5.

3.3.3 Potential confounders Potential confounding variables were identified from the literature as described in Chapters 1 and 2 of this thesis. Variables considered in this study are  baseline demographics: age, education, income, parity, marital status, language  breastfeeding culture (markers for the woman's and family's breastfeeding knowledge and attitudes): limited past breastfeeding experience, breastfeeding experience of the infant's maternal grandmother, strength of the partner's preference for breastfeeding  breastfeeding intentions: any plans for duration, specific planned duration, planned timing for introduction of solids (marker for planned duration of exclusive breastfeeding), any plan for return to paid work or study, specific planned return to work (marker for breastfeeding management plans during periodic separation from the infant)  maternity factors: type of delivery, regional anaesthesia, birth weight, gestation, initiation of breastfeeding in hospital, use of infant formula

142

3.3.4 Development of data collection instruments

3.3.4.1 Mothers’ baseline questionnaire This survey (Appendix A4.4) was constructed for pregnant women and administered during pregnancy. It was mailed with the recruitment consent form, an information page, and an introductory letter from the local Nursing Unit Manager on hospital letterhead. It enabled data collection on demographics, variables known to impact on breastfeeding outcomes and infant feeding intentions, and identification of usual general practice.

The 29-question survey was developed from and included sixteen questions from Forster et al., an Australian survey recruiting pregnant women interested in breastfeeding (204) and three questions constructed by simple extensions of these questions. The remaining ten questions comprised three novel questions, three for data management and randomisation and four for standard variables derived in part from the survey developed by Lawlor-Smith et al. (276) and reviewed for validity by Arbon et al. (340) Sources are detailed in Appendix A4.5.

The novel questions were numbers 3, 7, and 27. Question three, planning for introduction of solid foods, was developed to allow comparison of goals with exclusive breastfeeding duration outcomes. Wording allowed for historical variations in the recommended duration of exclusive breastfeeding. Question 7 was developed to assess practical preparations for infant feeding choice, hoping to explore the relationships between preparation, breastfeeding outcomes, and motivation. Question 27 was developed to assess a known potential barrier to continued breastfeeding, namely separation from the baby for work or study(92, 341), to explore the relationship between this known potential barrier and planned breastfeeding duration in this study population.

The complete questionnaire was assessed by four experienced researchers for face validity and by four mothers of young children for readability.

3.3.4.2 4 month and 6 month breastfeeding outcomes (CATI) The Computer Assisted Telephone Interview (CATI) detailed in the Appendices was used to collect data on current, and retrospective breastfeeding practices, using the 24- hour and since birth recall periods. Mothers were asked up to fifteen questions, with some skipped when answers indicated they were unnecessary.

143

Eight of the survey questions were sourced directly from the 24-hour recall questions of the World Health Organisation (42) which correspond to the recommended breastfeeding indicators. Medicine use was not asked, as it is permitted within the definition of exclusive breastfeeding. (54) The finalised CATI survey was reviewed by three breastfeeding researchers by email and three breastfeeding mothers by telephone.

The CATI was contracted to the Hunter Valley Research Foundation (HVRF) following tender, and funded through Australian government grants to the Rural Clinical School UNSW. Sequencing was adjusted with guidance from HVRF. Repeat calling for up to two weeks was used for those not answering. Interviews lasted about five minutes. Assessment of mothers' satisfaction was not included in this brief CATI, as accurate and informative assessment of satisfaction required instruments that were more extensive. (342)

There is currently no consensus on the measurement of breastfeeding intensity. The Australian National Health survey overlooks occasional substitution; it uses the question “At what age was [child’s name] first given [substitute] regularly?” to define exclusive breastfeeding, and the question “Including times of weaning, what is the total time that [child’s name] was breastfed?” to define duration of breastfeeding. (42) Discussion with four experienced researchers in the field led to the creation of Questions 11 and 12, about use of infant formula or solid foods over the previous week, which are similar to Canadian data collection. (55) Future analyses of these options may contribute to discussions on the definition and on the breastfeeding behaviours of non-exclusively breastfeeding mothers.

Interviews at 6 months are common; an interview at 4 months was added to take account of the changes in Australian recommendations for the introduction of solid foods; previously "4 to 6 months", currently "from 6 months". (30)

3.3.4.3 Practice nurse knowledge and attitudes quiz This 36 question survey (Appendix A4.6) assessed breastfeeding knowledge and attitudes using check box options for answering “agree”, “disagree” or “don’t know”. In addition, practice nurse demographics, previous training in motivational interviewing or breastfeeding management, personal parenting experience, and breastfeeding experience were collected.

144

This survey was completed by each practice nurse in the researcher’s presence, with brief introduction and no use of reference materials. Intervention and control practice nurses completed the survey before the training sessions began, and intervention nurses repeated the survey approximately four weeks later.

Topics covered by the 22 knowledge questions were maternal and health, milk supply, attachment, exclusivity and duration recommendations, challenges and safety. Topics covered by the fourteen attitude questions were psychosocial factors for the mother and the role of the nurse in breastfeeding promotion. Demographic questions included age, nursing qualification, parity, and personal breastfeeding experience.

Survey construction is detailed in Appendix A4.7. Knowledge questions were selected or adapted from Brodribb’s survey of Australian General practice registrars, (132) Byrne’s assessment tool incorporated in the ALC Breastfeeding Management training package, (322) and Martens’ survey for rural Canadian hospital nurses. (343) Attitude questions were developed from Martens' survey of nurse breastfeeding attitudes (343) and from Coleman and Wilson’s survey of practice nurse attitudes to assisting health behaviour change in the field of tobacco cessation. (344) For face validity, the constructed survey was assessed by four breastfeeding researchers, four practice nurses and one lactation consultant. It was piloted with Practice nurses from the Southern Highlands Division of General Practice, by email and fax.

3.3.4.4 Practice nurse semi-structured interviews Repeated short telephone interviews (described in Appendix A4.8) were held over the 14-month period of the trial with each nurse, documenting her reflections. In each interview, the practice nurse was encouraged to reflect on her training and learnings, on her experience of implementing the intervention, and on any challenges she described. Interviews used open questions around implementation, use of the components of the Conversation Tool and the practice nurse's experiences interacting with mothers of young infants. Answers were explored where necessary. Expected interview duration was 10 to 30 minutes. Interviews were recorded in writing, contemporaneously on paper or by touch-typing. The process of contacting practice nurses for each workplace interview is described in Appendix A4.9.

145

Use of an iterative approach for qualitative data collection (345) meant that later interviews were more likely to include open questions about the individual sections of the Conversation Tool. Time factors limited how many components of the Conversation Tool were discussed in each interview. Additional questions could include "What about the recommendations?" or "How are you handling mothers who are in the green box?" Further exploration also addressed the nature of interpersonal factors, "How are people responding?" and feasibility "How are you going with the time issues?"

Interviews documented the experience of the nurses as they incorporated the intervention process in their daily work. They enabled reinforcement of key training and intervention messages to consolidate learning. They also enabled process assessment, exploring the extent of intervention delivery per protocol using the Conversation Tool and any barriers to delivery. Finally, they helped to co-ordinate nurse attendance at training update events.

3.5 MIXED METHODS This thesis used the framework of the RATS guidelines (relevance, appropriateness, transparency, soundness) for presenting qualitative research, described by Clark. (346, 347) Qualitative data and other practice nurse data is grouped in Chapter 5 to explore the feasibility, the adherence to protocol and the experience of change described by the practice nurses.

3.5.1 Use of mixed methods in this thesis Figure 3.9 illustrates how the chapters of this thesis work together, providing alternate views of the same events. Firstly, the systematic literature review of randomised controlled trials, in Chapter 2, looked at "what can work." (348) Secondly, the quantitative methods of the randomised controlled trial in Chapters 3 and 4 looked at "what works". Thirdly, Chapter 5 provides a mixed method process evaluation, through the lens of practice nurses who were delivering the trial intervention, looking at "how it works". These three perspectives together may provide a more insight into the impact of breastfeeding support for mothers, offered by trained general practice nurses, on breastfeeding rates, than each provides individually. The appropriateness of using triangulated methods in this way is debated by Silverman (349)(p278) as will be addressed in the discussion in Chapter 6. However, each element of a project may contribute to the overall picture, according to Bazeley, particularly where a study aims

146

to understand both process and outcome, using multiple or mixed research methods. (348)

3.5.2 Use of mixed methods in practice nurse data Figure 3.9 shows the three sources of practice nurse data that are used in the process evaluation in Chapter 5. Sequencing of practice nurse data collection was determined by design aspects of the trial, as shown at the start of this chapter in Figure 3.1.

Figure 3.9 Triangulation of data sources across thesis chapters

Practice nurse Practice literature nurse quiz review knowledge & attitudes

Chapter 2 Chapter 5 Literature Process Review evaluation

Chapters 3 and 4 Practice RCT nurse discourse

THESIS Support for ongoing breastfeeding

RCT randomised controlled trial

The main practice nursing literature review was conducted early in the trial preparation phase and is located in Chapter 1. The pre- and post-training quiz and the demographic data were collected at practice nurse training, later in the trial preparation phase. Discourse data collection for qualitative analysis began at practice nurse training and continued through the trial intervention and data collection phases. These sources were drawn together in Chapter 5.

3.5.3 Collection of discourse data The major discourse data set was the telephone interview data. Discourse data was collected from practice nurses in four data sets, which were combined for analysis: . training workshop reflective practice exercise data . post-training practice visit data

147

. post-training individual semi-structured telephone interview data . post-training evening update meeting data

Discourse data collection occurred before and during the randomised controlled trial, over 21 months, July 2008 - March 2010. Anonymity was maintained by using a coded identifier for each practice nurse, (numbered in order of initial research contact) and by removing identifying details about practices and patients.

3.5.4 Analysis of discourse data The discourse data were prioritised over the other data sources analysed in Chapter 5. The process evaluation was divided into three parts; each had individual objectives, an individual meaning drawn from the practice nurse discourse data and each drew additional meaning from other practice nursing data. This analytic approach relied extensively on the framework of Sandelowski, USA academic and nursing researcher, as cited by Creswell et al. (350) in their review of the role of qualitative data in mixed methods research.

In analysis of the discourse data, the research question was prioritised over other questions arising from the discourse examination, as it was in the thesis as a whole, enabling the evaluation of the trial processes.

Practice nurse interview data was read through, and coding notes were made in the margin (see Figure 5.3). On the second reading, coding was repeated. Sub-themes were identified inductively in cycles of data review and then entered in a column on a data grid according to the model in one study, (351) in Excel software, for context analysis.

Initial data analysis evaluated the feasibility of this study in this context, and results are presented in Part 1 of Chapter 5. All sub-themes on the data grid were categorised into topical contexts relating to the setting and the tasks of the study. These contexts were each added to the data grid as an individual column, and marked against relevant sub- themes. Data was sorted for each context in turn so the themes in each context could be identified, and the degree of linkage with other contexts (where a sub-theme was relevant to more than one context) was examined. It was possible then to draw out common views, particular insights and dissenting views expressed in the practice nurse discourse, against each context. Discourse data in Part 1 related also to the background practice nursing literature in Chapter 1.

148

Returning to the data, a second analysis evaluated treatment fidelity over time by the practice nurses, and results are presented in Part 2 of Chapter 5. Interview data were read again chronologically to consider the extent to which each practice nurse reported using the Conversation Tool, influenced by distance in time from the training workshop, gradual increase in skills and ease with more familiarity, and repeated researcher contact by telephone and at update meetings. Following the protocol was a new behaviour for the practice nurses; their responses to undertaking this new behaviour are drawn into sequential themes. Discourse data in Part 2 related also to the personal background of the practice nurses and the results of the descriptive statistics are shown in Part 2.

Returning to the data once more, a third analysis evaluated the process of change as perceived by the practice nurses in their reflections on intervention delivery, and results are presented in Part 3 of Chapter 5. Practice nurse reflections are drawn into key themes to give insight into the challenges for, and possible solutions to, practice nurse support for ongoing breastfeeding. The reported experiential learning of these practice nurses is related to their change in knowledge and attitudes after the training workshops, compared to prior to training, and the paired samples T Test results are shown in Table 5.5.

Figure 3.10 Interpretive methods for three evaluations of the discourse Sort discourse chronologically for each practice nurse

Read the discourse Coding (note sub-themes in margins) and creation of data grid

Re-read the discourse Part1: Feasibility

Context theme analysis

Re-read the discourse Part 2: Fidelity

Chronological theme analysis

Re-read the discourse Part 3: New behaviours

Key theme analysis

Aberrant discourse

Reflection on the role of the researcher

149

3.6 CONCLUSION TO CHAPTER 3 In this chapter, the processes and components required for the conduct of a robust clinical trial have been described. The trial was developed in response to the systematic literature review, and from focused literature reviews around each element of the intervention. Methods used in this trial were designed for the particular context of rural Australian general practice. They took into consideration the realities of conducting a clinical trial in a primary care setting.

The trial was designed, prepared, and conducted with careful consideration of the theory of motivational interviewing. Within this theoretical perspective, an initiative in health promotion could be conducted while simultaneously acknowledging the individuality of both interviewers and interviewees. The goal of improved health outcomes for breastfeeding mothers and infants was approached systematically yet with flexible and pragmatic program delivery. This approach harnessed the power of both health professionals and mothers to take on board and continue new behaviours, when provided with evidence-based, brief, and appropriate support.

150

INNOVATION The literature describes a range of effective interventions to support breastfeeding exclusivity and duration, conducted in the setting of primary care, as identified in the systematic literature review.

This study adds to the field an examination of . breastfeeding practices in a large group of rural NSW and Victorian women . intervention delivery by practice nurses employed by general practices . intervention setting, in Australia, of a routine preventive health care visit . intervention purposive linkages, created between primary health care providers and an evidence-based consumer support group . intervention theoretical framework of motivational interviewing, used beyond the first six weeks of breastfeeding . evaluation of a practice nurse preventive health care intervention delivery, using qualitative methods . intervention delivery involving development of short-course training for general practice nurses to support continuation of breastfeeding . exploration of the definition of exclusive breastfeeding, comparing outcomes of three different definitions at the same time points with the same population

151

ETHICS HREC 07246 Ethics approval was gained from The University of New South Wales Human Research Ethics Committee. This approval covered the general practice clusters. It was accepted by Calvary Healthcare Riverina to cover use of names and addresses from the antenatal database of Calvary Hospital Wagga Wagga for recruitment and to cover additional data collection from this site. Progress was reported annually as required.

HREC 08/GSAHS/19 Greater Southern Area Health Service Human Research Ethics Committee gave their approval, approved the required Site Specific Application, and approved six amendments. This covered recruitment of participants using names and addresses from the antenatal database of Wagga Wagga Base Hospital, and additional data collection from this site.

JHEC 306/08/13 The Joint Hospitals Ethics Committee approved recruitment of participants using names and addresses from the antenatal database of Wodonga Regional Health Service, and they approved one amendment that covered additional data collection from this site.

ACTRN 12608000361303 Before recruitment began, the trial “Support for Ongoing Breastfeeding” was registered with the Australian New Zealand Clinical Trials Register.

152

ENGAGEMENT The Riverina Division of General Practice and Primary Health endorsed the introduction of this research project to practices in the rural Riverina city of Wagga Wagga NSW.

The Albury Wodonga Regional General Practice Network trading as Border Division of General Practice endorsed the introduction of this research project to practices in the twin town known as Albury-Wodonga. On the border, Albury NSW and Wodonga Victoria form bridges across the Murray River.

The Policy and Research Committee of the Australian Practice Nurses Association endorsed the involvement of practice nurses in this research project and the dissemination of trial progress and outcomes to members of the association through their journal. An introduction to the trial with background literature review was published in Primary Times in 2008. (314)

Ethics approval for recruitment and hospital data collection required liaison with the Maternity Nursing Unit Managers of Wagga Wagga Base Hospital, Calvary Hospital, and Wodonga Regional Health Service. Each of these nurse managers discussed and supported the involvement of their units in this research project and the potential for midwife research assistant employment.

Through membership of the Australian Breastfeeding Association's Lactation Resource Centre, resources were made available, in some circumstances at no charge. Discussion and presentation at conferences including those organised by ABA enabled linkage and exchange. (326, 352) Personal peer contact with local volunteer ABA breastfeeding counsellors maintained local links.

153

PREPARATION FOR THE TRIAL

Above: 2008 delivering survey envelopes to Albury-Wodonga midwife research assistant

Above and below: 2008 Practice nurse training sessions

154

RESULTS OF THE RANDOMISED CONTROLLED TRIAL

Chapter 4

Results of the randomised controlled trial

155

4.1 INTRODUCTION TO CHAPTER 4 This chapter contains the results of the cluster randomised controlled trial and identifies breastfeeding outcomes for participating mothers according to group allocation in order to determine the efficacy of the intervention. The checklist of items to include when reporting results of a cluster randomised trial, from the extension to the CONSORT statement published by Campbell et al., was used as a guide for this chapter. (278)

Pregnant women due to give birth in two rural Australian cities from August 2009 to October 2010 were recruited by mail survey. The RCT included 330 mothers attending fifteen general practices who were breastfeeding at infant age 2 months. The intervention was administered at 2, 4, and 6 months. Breastfeeding outcomes were measured from data collected at 4 months and at 6 months. Participant flow is shown in Figure 4.1.

4.2 RESPONSE RATE Of the 18 eligible general practices 15 were recruited, see Appendix 1.

Mail surveys were sent to all 3,127 pregnant women booked to give birth over fourteen months at the three local hospitals. In the calendar year prior to the start of this trial there had been 2,959 births at the three local hospitals, according to verbal information from maternity staff who consulted their Maternity Unit birthing records. The distribution was Wagga Wagga Base Hospital 777, Calvary Health Care Riverina 564, and Wodonga Regional Health Service 1618. Due to a lower birth rate in the study year compared to the previous year, observed during ongoing monitoring, recruitment was continued for fourteen months. We allowed for variations between predicted and actual dropout rate and aimed to achieve the sample size of 278 women determined in the sample size calculation in Chapter 3, Methodology.

Expected losses had been calculated. Standard approach to population mail survey is to anticipate a 20% response rate. We also anticipated 75% attendance at recruited general practices, 75% breastfeeding rate at 2 months at the commencement of the intervention, and 80% participant retention from recruitment to completion of 6 month CATI survey.

As seen in Figure 4.1, there was a 34.4% response rate to the mail survey including those women who declined to participate. This meant that 991 women were assessed

156

for eligibility. Following exclusion of 500 women who were attending a non-participating general practice (50%) and other exclusions specified in Figure 4.1, 330 women were enrolled.

Figure 4.1 The CONSORT flowchart with results

157

The randomised controlled trial excluded mothers who ceased breastfeeding before eight weeks. These mothers could have no access to the intervention, which was delivered to breastfeeding mothers of infants aged 2, 4 and 6 months.

4.3 ASSIGNMENT General practices were randomised using the methods described in Chapter 3, resulting in assignment of 176 mothers to the control group and 154 mothers to the intervention group across the fifteen general practices. Participant flow is shown in Figure 4.1. Assignment is shown in Table 4.1.

Table 4.1 Results of RCT Randomisation

Frequency Percent

Control 176 53.3

Intervention 154 46.7

Total 330 100.0

4.4 PARTICIPANT FLOW At 4 months, contact was made with 308 (93%) of 330 participants.

At 6 months, either breastfeeding cessation by 4 months was documented or contact was made with 322 (97%) of 330 participants. At 6 months 8/330 mothers were unable to be contacted. Distribution was equal between groups. The 6-month outcomes included data collected from 22 participants who could not be contacted at 4 months. Distribution of these 22 participants was disproportionate, 16/176 in the control group and 6/154 in the intervention group however due to the initially larger size of the control group this discrepancy was not considered likely to bias results.

Data for these 22 participants was inferred for "Any breastfeeding at 4 months" hence n=330. Data was inferred for "Exclusive breastfeeding until 4 months (since birth recall)" in the two cases where participants continued exclusive breastfeeding until 6 months hence n=310.

Data for these 22 participants could not be inferred for "Exclusive breastfeeding until 4 months (since birth recall)" for participants who had ceased exclusive breastfeeding at an unspecified time prior to 6 months (20/22). Data could not be inferred for "Full breastfeeding at 4 months (24-hour recall)" or for "Exclusive breastfeeding at 4 months

158

(24-hour recall)" because the timing of data collection was critical for these definitions, hence n= 308 for these outcomes.

4.5 DATA

4.5.1 Baseline data In baseline data, the intervention group is compared to the control group to confirm equal distribution of potential confounding variables. Variables that may impact on breastfeeding outcomes are classed as demographic, social environment, maternal intention, or perinatal characteristics. See Appendix 5 for the specific survey questions. Baseline data also allows some comparison with other Australian population surveys.

4.5.2 Demography Table 4.2 presents summary demographic information from the 330 mothers, all of whom were breastfeeding their infants at 8 weeks of age. In Australia, early breastfeeding cessation is associated with younger age, lower education or income (171), less availability of partner support (176), and lower English language skills. (194) Pearson Chi-Square test confirmed that these characteristics were distributed well. No statistically significant difference was found between the two groups in this study.

Table 4.2 RCT participants' baseline characteristics

Control Intervention Total Siga Variable % (n) % (n) % (n) Age group (years) n=325 (172) (153) 0.167 below 20 2.3 (4) 1.3 (2) 1.8 (6) 20-29 32 (55) 43.8 (67) 37.5 (122) 30-39 62.2 (107) 51.6 (79) 57.2 (186) above 39 3.5 (6) 3.3 (5) 3.4 (11) Education (highest qualification) n=324 (173) (151) 0.359 Secondary or less 16.2 (28) 11.9 (18) 14.2 (46) Industry qualification 34.1 (59) 31.1 (47) 32.7 (106) University degree 49.7 (86) 57.0 (86) 53.1 (172) Income (ABS quintiles) n=294ii (154) (140) 0.486 Lowest (1) 10.4 (16) 13.6 (19) 11.9 (35) Middle (2 - 4) 62.3 (96) 64.3 (90) 63.3 (186) Highest (5) 27.3 (42) 22.1 (31) 24.8 (73) Partnered 96.0 (169) 94.8 (146) 95.5 (315) 0.643 First language English 96.0 (167) 96.1 (146) 96.0 (313) 0.972 a Pearson Chi-Square ii Australian Bureau of Statistics gross household incomes, 2008, catalogue 6523.0

159

4.5.3 Maternal environment Table 4.3 describes characteristics of the mother's experience and social environment that have been shown to impact on breastfeeding outcomes. Early breastfeeding cessation is associated with parity (353) and with previous breastfeeding experience. (146) Not having been breastfed oneself as a baby is used as a proxy for having a maternal grandmother of the newborn who has no capacity to support breastfeeding based on personal experience. (154, 354) A partner who does not encourage breastfeeding is a strong predictor for early cessation. (161) There was no statistically significant difference between groups using Pearson Chi-Square test.

Table 4.3 Baseline maternal environment relating to breastfeeding

Control Intervention Total Siga Variable % (n) % (n) % (n) Previous children (number) n=326 0.713 0 33.0 (58) 39.0 (60) 35.8 (118) 1 34.1 (60) 28.6 (44) 31.5 (104) 2 21.6 (38) 22.7 (35) 22.1 (73) 3 or more 10.3 (18) 8.4 (13) 9.4 (31) Past experience breastfeeding beyond 6 months n=320 46.8 (80) 40.9 (61) 44.1 (141) 0.294 Was breastfed herself as a baby n=316 0.760 81.7 (138) 81.0 (119) 81.3 (257) Partner's preference for infant feeding n=314 0.763 Bottle feeding 0 (0) 0.7 (1) 0.3 (1) No preference 4.2 (7) 4.1 (6) 4.1 (13) To be supportive 50.0 (84) 50.0 (73) 50.0 (157) Breastfeeding 45.8 (77) 45.2 (66) 45.5 (143) a Pearson Chi-Square

4.5.4 Planning

Table 4.4 describes for each group at baseline the mothers' feeding intentions, prior to the birth of their babies. The period of exclusive breastfeeding ends at the introduction of water, other milks, other fluids, or solids. In this study mothers were asked when they planned to "introduce solids" as a proxy for intended duration of exclusive breastfeeding as this terminology is commonly used and most predictable. (355) Recommendations for exclusivity vary across different authorities and over time, with many Australian women unaware of current national health recommendations. (64) Breastfeeding outcomes are more strongly associated with maternal intentions than with maternal demographics. (156)

160

Actual or planned separation of mothers and infants for maternal employment or study, particularly in the first 6 months, is a risk factor for reduced breastfeeding rates. (89, 341, 356) As shown in Table 4.4 randomisation did not create equal distribution here, indicating the need for adjustment in analyses.

Of those who had made plans, more mothers in the intervention group (70%) compared to the control group (55%) planned to return to paid work/study before this baby turned one. Many of these women planned to return to paid work/study by the time this baby was 6 months old, a characteristic that described more of the intervention group (30%) than the control group (21%). This difference between the groups in this baseline characteristic was statistically significant (p<0.05) using Pearson Chi-Square.

Table 4.4 Baseline planning by the mother for timing of weaning and work

Control Intervention Total Siga Variable % (n) % (n) % (n) Any plan for breastfeeding duration? n=326 0.106 Unsure 23.0 (40) 30.9 (47) 26.7 (87) Yes 77.0 (134) 69.1 (105) 73.3 (239) Duration of breastfeeding (planned, months) n=326 0.451 < 4 4.0 (7) 3.2 (5) 3.6 (12) 4 - 6 12.6 (22) 11.0 (17) 11.8 (39) > 6 60.3 (105) 53.9 (83) 57.0 (188) unsure 23.0 (40) 30.9 (47) 26.9 (87) Solids starting time (planned age, months) n=312 0.917 4 - 6 36.5 (61) 33.3 (49) 35.0 (110) Around 6 or later 63.5 (106) 66.7 (98) 65.0 (204) Any paid work/study in the next 12 months (planned) n=326jj 0.040 No 36.8 (64) 25.0 (38) 31.3 (102) Unsure 17.2 (30) 15.8 (24) 16.6 (54) Yes 55.6 (80) 70.3 (90) 62.5 (170) Paid work or study in the next 12 months (planned, months) n=272 0.040 (144) (128) ≤ 6 months 20.7 (36) 29.7 (38) 27.2 (74) > 6 months 30.6 (44) 40.6 (52) 35.3 (96) None 44.4 (64) 29.7 (38) 31.3 (102) a Pearson Chi-Square

jj "Any paid work... (planned)" includes mothers who were unsure of their plans. "Paid work...(planned, months)" is a category limited to those with a specified time period relating to their employment/study plans.

161

4.5.5 Perinatal factors Table 4.5 presents descriptive information about the context in which the mothers began breastfeeding, with all births occurring in hospitals. Reduced exclusivity and duration of breastfeeding are associated with maternity hospital practices such as medical birthing interventions with more complex types of birth (143) and the use of infant formula. (154) Reduced duration of breastfeeding has been associated with use of epidural analgesia, (357), low infant birth weight (358) and gestation under 40 weeks. (359) There was no statistically significant difference between groups using Pearson Chi-Square.

Table 4.5 Birthing and hospital factors related to breastfeeding

Control Intervention Total Siga Variable % (n) % (n) % (n) Type of delivery 0.446 Caesarean, emergency 9.1 (16) 13.0 (20) 10.9 (36) Caesarean, planned 20.5 (36) 15.6 (24) 18.2 (60) Vaginal, assisted 15.3 (27) 18.2 (28) 16.7 (55) Vaginal, unassisted 55.1 (97) 53.2 (82) 54.2 (179) Regional anaesthesia 41.5 (73) 42.9 (66) 42.1 (139) 0.800 (epidural/spinal) Use of infant formula in 5.1 (9) 8.4 (13) 6.7 (22) 0.227 hospital Birth weight <2500g 3.4 (6) 1.3 (2) 2.4 (8) 0.422 Gestation <40 weeks 43.8 (85) 46.1 (71) 47.3 (156) 0.691 a Pearson Chi-Square

4.6 ANALYSIS

4.6.1 Intention to treat Analyses for outcomes of the trial are based on the intention to treat convention. That is, where the pregnancy questionnaire identifies that the participant will attend an intervention practice, she is assigned to the intervention group and analysed as receiving the intervention. Most women in these regions attend their local general practice for infant immunisation. For the purpose of analysis, intervention group participants are assumed to have received the intervention according to trial protocols. Implications of this analytic method are addressed in Chapter 6. The two study groups analysed are control and intervention; outcome measures are reported for intervention group outcomes compared to control group outcomes.

162

No quantitative per protocol analysis was conducted; the extent of intervention delivery per protocol is considered using mixed methods in Chapter 5.

4.6.2 Adjustment for clustering Allocation to study group was by practice, so outcome analyses were conducted with adjustment for clustering by practice. These analyses were conducted using STATA.

4.6.3 Outcome measures The overall rate for "Any breastfeeding at 4 months" was 89% and "Any breastfeeding at 6 months" was 79% as shown in Table 4.6. These outcome measures were used as duration of breastfeeding, where other studies have recorded the inverse as cessation rates.

Outcomes for exclusive breastfeeding were analysed using three different outcome measures. The primary aim was to assess the efficacy of the intervention. A secondary aim, in view of the differing scientific opinions regarding definitions of exclusivity as discussed in Chapter 1 (51, 56), was to compare outcomes for each definition in the same population. Definitions used were "Full breastfeeding (24-hour recall)" for comparison with national data (42), "Exclusive breastfeeding (24-hour recall)" for comparison with international data (54) and "Exclusive breastfeeding (since birth recall) " to examine more recent expert recommendation. (51)

Results are shown for these four outcomes, at 4 months and again at 6 months. The eight breastfeeding outcomes are shown in Table 4.6.

Table 4.6 Percentage of study population breastfeeding for each outcome

Outcome variable Valid Missing Percentage (n) (n) breastfeeding Any BF at 4 months 330 0 89 Full BF at 4 months (24-hour recall) 308 22 67 Exclusive BF at 4 months (24-hour recall) 308 22 60 Exclusive BF to 4 months (since birth recall) 310 20 40 Any BF at 6 months 322 8 79 Full BF at 6 months (24-hour recall) 322 8 17 Exclusive BF at 6 months (24-hour recall) 322 8 14 Exclusive BF to 6 months (since birth recall) 322 8 8 BF=breastfeeding; n=number; any=breastmilk plus other intake; full=breastmilk plus only water-based fluids; exclusive=breastmilk only

163

4.6.4 Points of measurement Breastfeeding outcome data was collected at two time points. The protocol was telephone contact at 4 months and again at 6 months for those who were continuing to breastfeed. Numbers analysed at each time point are greater than the required sample for power calculation (n = 278) as identified in Chapter 3.

As discussed above, 4 months 24-hour recall data was collected from 308 mothers 86.4% were contacted at 16-18 weeks, and 10% at 19-21 weeks. Mean contact date was at 17.08 weeks SD 1.35 weeks.

Data was collected from 322 mothers for 24-hour recall at 6 months; 82% were contacted at 24-25 weeks, 12.4% at 26 weeks, and 5.6% after 26 weeks. Mean contact date was at 24.92 weeks, SD 1.23 weeks.

For clarity of comparison between the three definitions of exclusivity in this study, noting that most participants were contacted prior to 6 calendar months of age (26 weeks), outcomes were described for simplicity as "at 6 months (24-hour recall)" or "to 6 months (since birth recall)".

Outcomes for all study participants at these two time points are shown in Figure 4.2. This cohort perspective shows four alternate measures that have been used to define breastfeeding in a range of studies. The importance of clear definitions is illustrated.

164

Figure 4.2 Cohort infant feeding practices by age

100 90 80 70 60 Percentage 50 40 30 Any BF 20 Full BF (24-hour recall) 10 EBF (24-hour recall) 0 0 EBF until (since birth recall) 2 4 6 Age group in months

BF=breastfeeding; any=breastmilk plus other intake; full=breastmilk plus only water-based fluids; exclusive=breastmilk only. Note that 2-month values for Any BF are measured. Other 2-month values are a calculated average.

4.6.5 Definitions for reporting Definitions used are consistent with national and international monitoring recommendations. (54, 56) In this study, current practice was defined as the previous 24 hours.

Any breastfeeding Infant are fed with breastmilk and also potentially any additional liquids or solids. This includes infant feeding practices ranging from complete reliance on breastfeeding to meet all fluid and nutritional requirements to giving only one breastfeed per day; may be defined retrospectively or for current practice.

Full breastfeeding Infants are fed with breastmilk and any additional water-based liquids. This measure is defined by avoidance of all non-human milks and all solids. Full breastfeeding must be defined for current practice because maternal recall of the timing of commencement of solids or non-human milks has been found to vary in accuracy. (56)

165

Exclusive breastfeeding Infants are solely fed breastmilk. Exclusive breastfeeding in this study, in accordance with WHO definitions (54), is defined by the non-use of water-based liquids, non- human milks, or any solids to feed the infant. Administration of medicines is permitted in this definition. It may be defined for current practice or by since birth recall for up to three years. (51)

4.7 OUTCOMES

4.7.1 Any breastfeeding at 4 months The measure "any breastfeeding" allows any other liquid or solids as well as breastmilk, and results were available for all 330 participants. (See Table 4.7) There was no significant association between group allocation and any breastfeeding at 4 months.

Table 4.7 Any breastfeeding at 4 months

Group No Yes OR 95%CI p % (n) % (n) (OR) Control 11.4 (20) 88.6(156) Intervention 11.0 (17) 89.0 (137) 1.03 0.50 - 2.12 0.93 Total n=330 11.2 (37) 88.8(293) Adjusted for fifteen clusters by practice

As shown in Table 4.4 there was a significant difference in distribution across groups in two variables. "Any paid work/study in the next 12 months (planned)" p=0.04 was described for all women, including those who were unsure of their plans. "Paid work or study in the next 12 months (planned, months)" p=0.04 was described for women who planned to return to work in the first 6 months, after 6 months, or not at all in the first 12 months. Data were available for 326 and 272 women respectively. Adjustment was performed but there was no significant difference due to the intervention in any breastfeeding at 4 months adjusted for either of these variables, OR 1.05 (95%CI 0.50- 2.22) p=0.89 and OR 1.03 (95%CI 0.46-2.29) p=0.94 respectively.

166

4.7.2 Full breastfeeding at 4 months (24-hour recall) The measure "full breastfeeding" allows water-based fluids as well as breastmilk. Results at 4 months are shown in table 4.8. There was no significant association between group allocation and full breastfeeding at 4 months.

Table 4.8 Full breastfeeding at 4 months (24-hour recall)

Group No Yes OR 95%CI p % (n) % (n) (OR) Control n=160 36.3 (58) 63.8 (102) Intervention n=148 29.7 (44) 70.3 (104) 1.34 0.84-2.15 0.22 Total n=308 33.1 (102) 66.9 (206) Adjusted for fifteen clusters by practice

Adjustment was performed for "Any paid work/study in the next 12 months (planned)" with results available for 304 women but there was no significant difference due to the intervention in full breastfeeding at 4 months, OR 1.53 (95%CI 0.93-2.50) p=0.09.

Adjustment was also performed for "Paid work or study in the next 12 months (planned, months)" with results available for 254 women which did show a significant difference due to the intervention, as shown in Table 4.9.

Table 4.9 Full breastfeeding at 4 months (24-hour recall) adjusted

Group No Yes OR 95%CI p (planned work/study) % (n) % (n) (OR) Control n=131 39.7 (52) 60.3 (79) Intervention n=123 28.5 (35) 71.5 (88) 1.95 1.03-3.69 0.04 Total n=254 34.3 (87) 65.7 (167) Adjusted for fifteen clusters by practice

4.7.3 Exclusive breastfeeding at 4 months (24-hour recall) The measure "exclusive breastfeeding" allows no food or fluids other than breastmilk in the past 24 hours. Results at 4 months are shown in table 4.10. There was no significant association between group allocation and this outcome at 4 months.

167

Table 4.10 Exclusive breastfeeding at 4 months (24-hour recall)

Group No Yes OR 95%CI p % (n) % (n) (OR) Control n=160 43.8 (70) 56.3 (90) Intervention n=148 35.1 (52) 64.9 (96) 1.44 0.90 - 2.29 0.13 Total n=308 39.6 (122) 60.4 (186) Adjusted for fifteen clusters by practice

Adjustment was performed for "Any paid work/study in the next 12 months (planned)" with results available for 304 women but there was still no significant difference due to the intervention in exclusive breastfeeding at 4 months, OR 1.58 (95%CI 0.94-2.68) p=0.09

Table 4.11 Exclusive breastfeeding at 4 months (24-hour recall) adjusted

Group No Yes OR 95%CI p (planned work/study) % (n) % (n) (OR) Control n=131 46.6 (61) 53.4 (70) Intervention n=123 35.0 (43) 65.0 (80) 1.88 1.01-3.50 0.047 Total n=254 40.9 (104) 59.1 (150)

Adjusted for fifteen clusters by practice

4.7.4 Exclusive breastfeeding to 4 months (since birth recall) This measure allows no food or fluids other than breastmilk, since the birth of the infant, according to maternal recall over the previous four months. There was no significant association between group allocation and exclusive breastfeeding to 4 months as shown in table 4.12.

Table 4.12 Exclusive breastfeeding to 4 months (since birth recall)

Group No Yes OR 95%CI p % (n) % (n) (OR) Control 62.1 (100) 37.9(61) Intervention 57.4 (85) 42.6 (63) 1.27 0.85 - 1.89 0.24 Total n=310 60.0 (186) 40.0(124) Adjusted for fifteen clusters by practice

168

Results were available for 304 and 254 women respectively. Adjustment was performed for the variables "Any paid work/study in the next 12 months (planned)" and "Paid work or study in the next 12 months (planned, months)." There was no significant difference due to the intervention with either adjustment, OR 1.37 (95%CI 0.87-2.17) p=0.18 and OR 1.37 (95%CI 0.05-2.23) p=0.20 respectively.

4.7.5 Any breastfeeding at 6 months The measure "any breastfeeding" allows any other liquid or solids as well as breastmilk, and results are available for 322 mothers, as shown in table 4.13. There was no significant association between group allocation and any breastfeeding at 6 months.

Table 4.13 Any breastfeeding at 6 months

Group No Yes OR 95%CI p % (n) % (n) (OR) Control 21.5 (37) 78.5(135) Intervention 21.3 (32) 78.7(118) 1.01 0.62 - 1.63 0.97 Total n=322 21.4 (69) 78.6(253) Adjusted for fifteen clusters by practice

Results were available for n=318 and n=266 women respectively. Adjustment was performed for the variables "Any paid work/study in the next 12 months (planned)" and "Paid work or study in the next 12 months (planned, months)." There was no significant difference with either adjustment, OR 1.06 (95%CI 0.50-2.22) p=0.89 and OR 1.15 (95%CI 0.68-1.97) p=0.59 respectively.

4.7.6 Full breastfeeding at 6 months (24-hour recall) The measure "full breastfeeding" allows water-based fluids as well as breastmilk. Results at 6 months are shown in table 4.14. There was no significant association between group allocation and full breastfeeding at 6 months.

Overall, in the study population 16.8% were fully breastfeeding, compared to 18.4% of infants at 25 weeks in the 2001 Australian National Health Survey (ANHS). (52) Direct comparison cannot be made however, as the ANHS excluded infants who were fed solids or substitute milks "regularly", whereas exclusion in this study applied to infants who were fed these in the previous 24 hours.

169

Table 4.14 Full breastfeeding at 6 months (24-hour recall)

Group No Yes OR 95%CI p % (n) % (n) (OR) Control 82.6 (142) 17.4(30) Intervention 84.0 (126) 16.0 (24) 0.90 0.53 - 1.55 0.71 Total n=322 83.2 (268) 16.8(54) Adjusted for fifteen clusters by practice

Adjustment was performed for the variables "Any paid work/study in the next 12 months (planned)" and "Paid work or study in the next 12 months (planned, months)." There was no significant difference due to the intervention with either adjustment, OR 0.95 (95%CI 0.52-1.76) p=0.88 and OR 0.86 (95%CI 0.49-1.52) p=0.60 respectively.

4.7.7 Exclusive breastfeeding at 6 months (24-hour recall) This measure of "exclusive breastfeeding" allows no food or fluids other than breastmilk in the past 24 hours. Results at 6 months are shown in table 4.15. There was no significant association between group allocation and this outcome at 6 months.

Table 4.15 Exclusive breastfeeding at 6 months (24-hour recall)

Group No Yes OR 95%CI p % (n) % (n) (OR) Control 86.0 (148) 14.0(24) Intervention 85.3 (128) 14.7 (22) 1.06 0.57 - 1.94 0.85 Total n=322 85.7 (276) 14.3(46) Adjusted for fifteen clusters by practice

Adjustment was performed for the same two variables but there was no significant difference either adjustment, OR 1.16 (95%CI 0.57-2.36) p=0.68 and OR1.01 (95%CI 0.56-1.86) p=0.96 respectively.

4.7.8 Exclusive breastfeeding to 6 months (since birth recall) This definition measures the avoidance of all breastmilk substitutes since birth, according to maternal recall over the previous 6 months, with the majority of data collected prior to 26 weeks as described in section 4.6.4 above. Again there was no significant association between group allocation and exclusive breastfeeding to 6 months as shown in table 4.16.

170

Table 4.16 Exclusive breastfeeding to 6 months (since birth recall)

Group No Yes OR 95%CI p % (n) % (n) (OR) Control 92.4 (159) 7.6(13) Intervention 91.3 (137) 8.7(13) 1.16 0.39 - 3.45 0.79 Total n=322 91.9 (296) 8.1(26) Adjusted for fifteen clusters by practice

Adjustment was performed for the same two variables but there was no significant difference due to the intervention with either adjustment in exclusive breastfeeding to 6 months, OR 1.31 (95%CI 0.37-4.63) p=0.68 and OR 0.91 (95%CI 0.26-3.15) p=0.88 respectively.

4.8 SUMMARY OF RCT RESULTS

4.8.1 Primary outcome measures

4.8.1.1Duration No significant difference in duration of breastfeeding ("Any breastfeeding") was measured at 4 months or 6 months, analysing for any breastfeeding by intervention group participants compared to control group participants, adjusted for clustering by practice.

4.8.1.2 Exclusivity: 4 month data collection A significant difference in exclusive breastfeeding (24-hour recall) due to the intervention was measured at 4 months. The intervention successfully supported women to maintain full breastfeeding (24-hour recall) at 4 months, OR 1.95 (95%CI 1.03-3.69) p=0.041. (Figure 4.3) The intervention also successfully supported women to maintain exclusive breastfeeding (24-hour recall) at 4 months, OR 1.88 (95%CI 1.01- 3.50) p=0.047. (Figure 4.4) These outcomes are for women attending the intervention general practices, compared to women attending the control general practices. They are adjusted for defined plans for return to paid work or study as there was a significant difference in the distribution of these plans between the groups and for clustering by practice. They define current breastfeeding practice (24-hour recall) for 254 infants whose mean age was 17.08 (SD 1.35) weeks. These outcomes are reported using established measures of breastfeeding intensity. (42) No significant difference was found between groups in exclusive breastfeeding to 4 months (since birth recall).

171

4.8.1.3 Exclusivity: 6 month data collection No significant difference was found in exclusive breastfeeding under 6 months, analysing for full breastfeeding or exclusive breastfeeding using current practice (24- hour recall) data collection. Again, there was no significant difference found in exclusive breastfeeding to 6 months using or retrospective (since birth recall) data collection.

Figure 4.3 RCT Outcomes for full breastfeeding (24-hour recall)

120

100

60 FBF Intervention FBF Control 40 Percentage FBF 20

0 0246 Infant age in months

Note: 0, 4, and 6-month outcomes are measured. 2-month values are a calculated average.

Figure 4.4 RCT Outcomes for exclusive breastfeeding (24-hour recall)

120

100

60 EBF Intervention

40 EBF Control Percentage EBF Percentage

0 0246 Infant age in months

Note: 0, 4, and 6-month outcomes are measured. 2-month values are a calculated average.

172

4.8.2 Comparison of outcomes for three alternate exclusivity definitions Outcomes for exclusive breastfeeding clearly differ from each other for each of the three definitions at 4 months, as seen in Table 4.6 in this chapter. The 24-hour recall questions, for full breastfeeding and exclusive breastfeeding, showed that the majority of women were exclusively breastfeeding according to these definitions, 67% and 60% respectively at 4 months. However, use of the since birth recall question showed that a minority, less than half of all women, 40%, were exclusively breastfeeding according to this definition. The 24-hour recall question was able to distinguish differences of the outcomes in this intervention study, which was not possible using the since birth recall. These findings suggest that where attempts are made to increase breastfeeding intensity, which impact on short-term health outcomes and hence health and personal costs (2, 201), evaluations should use current practice outcome measures.

At 6 months, outcomes for exclusive breastfeeding were all quite low, although again the 24-hour recall questions showed more women breastfeeding exclusively using these definitions, 17% and 14% respectively, compared to the since birth recall, 8%. The implications of these differences for selection of national indicators are addressed in the Chapter 6.

4.8.3 Use of infant formula There was no difference between groups in the introduction of formula. At 4 months 11% of all infants (39/330) were not breastfeeding (4% had planned this). However, 93/310 were fed with formula, according to CATI data, and hence at least 18% (54/291) of breastfeeding infants were mixed feeding. At 6 months 21% (69/322) of all infants were not breastfeeding (15% had planned this), 43.5% (140/322) were being fed with infant formula and hence 28% (71/253) of breastfed infants were fed both.

4.8.4 Outcome of background analyses In the study population there was a high proportion of breastfeeding. At 2 months, all women were breastfeeding and at 6 months, most were breastfeeding. Most women participating in this study (89%) had post-secondary qualifications and their household income was more than twice as likely to be in the highest quintile as in the lowest quintile. The significance of these sample characteristics for generalisation to a wider population is addressed in Chapter 6.

173

4.8.5. Comparison to national data In the context of a study population who were all breastfeeding at 2 months, 11% had weaned by 4 months of age, and 21% of study mothers had weaned by 6 months of age, with 79% continuing to breastfeed at least once per day at 6 months.

National comparison of outcomes for "Any breastfeeding" can be made, see Table 4.17, using the evaluation by Donath and Amir of data from 1,883 Australian children aged less than three years of age for the 2001 Australian National Health Survey. They analysed breastfeeding data using life table methods and noted that 49% of children were still breastfed at 25 weeks of age. (52) Comparison of full breastfeeding rates in the study and in the Australian populations must be considered cautiously, however, because of the different time periods used to create the definition. The ANHS excluded infants from the category of full breastfeeding if they were fed solids or substitute milks "regularly", whereas in this study infants were excluded if they were fed these substitutes in the previous 24 hours.

Table 4.17 Breastfeeding rates of study population vs. Australian population

Outcome variable Study Australian population population breastfeeding breastfeeding Any BF at hospital discharge 100 83 Any BF at 6 months 79 49 Full BF at 6 months (24-hour recall) 17 18

This outcome suggests that the enrolled population has characteristics that differ from the population sampled for the 2001 Australian National Health Survey (NHS), which will be addressed in Chapter 6.

174

Chapter 5

Process evaluation

175

INTRODUCTION TO CHAPTER 5 PROCESS EVALUATION In this thesis, the randomised controlled trial demonstrated a positive effect on breastfeeding at four months in mothers who specified their employment plans for the first year after the birth. This may have clinical significance in relation to maternal employment. In order for this research to make a truly effective contribution to policy and practice in this field, the process of conducting this research must be clearly understood.

This chapter presents an analysis of data collected from practice nurses throughout the randomised controlled trial. The different data sources are shown in Table 5.1, indicating in which parts of the chapter they are analysed; mixed methods are used to evaluate practice nurse data. Process evaluation provides understanding of how the trial worked. Reasoning for the use of mixed or multiple methods, sequencing, and the methods of practice nurse data collection and analysis are described in Chapter 3.

Table 5.1 Discourse data classification and other data for process evaluation

Purpose Discourse data set, classification Other data sets

Part 1 Six contexts Literature review Feasibility analysis Nursing Mothers The community General practice Research Learning

Part 2 Six chronological stages Demographics Fidelity analysis Beginning (Adherence to Protocol) Barriers Communicating Acceptance Comfort Routine

Part 3 Four key themes Knowledge and Experience analysis There's a stigma out there attitudes quiz before and after (New Conversations) Something I wouldn't have thought of before training Using those questions You can lead a horse to water

176

After practice nurses were trained, they delivered the intervention in the course of their daily work. Their reports on their experiences over this time make up the body of discourse data, which is analysed in three different ways, as shown in Table 5.1, for three separate purposes. For additional depth in the analysis description of practice nurse demographics and past experience, assessment of practice nurse knowledge and attitudes with pre- and post-training quiz results, and a practice nursing literature review (located in Chapter 1 and summarised below in 5.1.1) are also used.

5.0.1 Results of discourse data collection Discourse data was collected from nineteen practice nurses as outlined in Table 5.2. Of these, 15/19 received training in July-August 2008, before any participants received the intervention. The remaining four practice nurses began employment in an intervention practice during the trial period and received the brief training module, two prior to the last eight months of the intervention and two prior to the last seven months of the intervention. There were other staff changes during the period of the trial. Two practice nurses were lost from the trial when they ceased employment at a recruited general practice and two took maternity leave without returning. One practice nurse took maternity leave and returned within the trial period. One practice nurse changed employment from one intervention practice to a different intervention practice, which is not considered a loss.

Table 5.2 shows the duration of involvement in the trial for each practice nurse, compared to the total trial duration, the extent of each trained practice nurse's presence in her general practice, and an indication of the extent to which her experience is represented in the discourse data. Perspectives of those practice nurses with a greater number of data collection points may also be present more in the discourse data, depending on the length of each interview. These practice nurses received more mentoring support, as the contact points had a dual function. There are implications for adherence to protocol from the number of data collection points.

Staff turnover and staff learning curves meant that some mothers who attended these general practices had reduced access to a trained practice nurse. Skill development is evaluated qualitatively later in this chapter in Part 2 Fidelity; there was no formal per protocol analysis of the fact or extent of intervention delivery with each mother. No maternal data adjustment was performed in the context of arrivals or departures of practice nurses, following the intention to treat protocol. This is of benefit for translational research, which is addressed in Chapter 6.

177

Table 5.2 Practice nurse presence and data collection points Practice Employed Employed Reflective Practice Telephone Update Nurse during per week at practice visit data interviews meetings ID code intervention intervention recorded recorded recorded attended (months) practice (occasions) (occasions) (number) (number) (days) PN1 16 3 1 1 4 2 PN2 4 3 1 1 0 0 PN3 16 4 1 2 3 2 PN4 16 5 1 1 4 2 PN5 16 5 1 1 4 3 PN6 16 5 1 1 3 3 PN7 16 5 1 1 3 2 PN8 9 3 1 1 3 1 PN9 16 4 1 1 1 3 PN10 16 1 1 1 2 0 PN11 10 3 1 1 3 1 PN12 12 2 1 1 1 1 PN13 8 3 1 1 3 1 PN14 16 2 1 1 2 1 PN15 16 1 1 1 2 0 PN16 9 3 0 0 2 0 PN17 8 2 0 0 1 1 PN18 8 2 0 0 1 2 PN19 8 3 0 0 1 1 Total PN data collection points 15 16 43 26 ID identity; PN Practice Nurse.

Practice nurses were initially asked, "Have you seen any mums for baby immunisations?" to elicit their opportunities and experiences with using the Conversation Tool. The first question generally led to case descriptions of mothers' breastfeeding stories, descriptions of the practice nurse's reflexive emotions and thoughts, and description of the practice nurse's subsequent actions with the mother. The second question was "How are you going with using the Conversation Tool?" Interviews explored experience with the different components of the Conversation Tool, resulting in discourse around stigma, work practices, skills, knowledge, and attitudes.

178

Analysis and results of the discourse data are presented through Parts 1-3, below.

5.0.2 Results of descriptive data collection Practice nurses were asked to complete background information prior to training. This survey was completed at the time of the knowledge and attitudes quiz, by fifteen nurses at the start of the intervention. Analysis and results are presented below in Part 2 of this chapter.

5.0.3 Results of training effects data collection The pre-training quiz was administered to fifteen practice nurses when each nurse attended her first training session. This was the entire nursing staff of the eight general practices randomised to the intervention at the start of the trial. The post-training quiz was administered in person four weeks after completion of training, to these fifteen practice nurses in their workplaces. Practice nurse training and the development of the quiz are described in Chapter 3. Analysis and results are presented in Part 3 of this chapter.

5.0.4 Structure of the discourse data analysis The discourse data is the major dataset in the process evaluation, and results of thematic analysis are presented. Analysis of discourse data is based on single researcher coding, using a coding framework developed in consultation with supervisors. Validation is based on research presentations and discussions of stories and theme analysis with a range of professionals in the field.

Evaluation of the discourse data is presented in three parts, and each part views the same discourse data independently. Throughout the analysis, boxed figures contain all relevant quotes, to show how broadly each topic was raised, and to illustrate the diversity or similarity of experiences and perceptions.

Context analysis of the discourse data in Part 1 Feasibility, as shown in Table 5.1, created contextual themes that included four settings (nursing, mothers, the community, and general practice). Two tasks (research and learning) also became apparent as important. Within these contextual themes was a range of subthemes. Some sub-themes were specific to a particular context while others linked several contexts.

179

Chronological analysis of the discourse data in Part 2 Fidelity created sequential themes, steps the practice nurses undertook as they adapted to incorporating the trial tasks into their working day. These adaptations included changes in attitude, development of skills, and altered procedures. Six chronological themes were identified, (beginning, barriers, communicating, acceptance, comfort, and routine). These themes describe a process of change, which is the framework of this thesis, this study, and the Conversation Tool intervention with mothers. The process of change is approached from the perspective of motivational interviewing.

Analysis of the key themes in the discourse data in Part 3 New Conversations revealed how the practice nurses came to understand their role in supporting ongoing breastfeeding. Four key themes arose: "There's a stigma out there"; "Something I wouldn't have thought of before"; "Using those questions"; and "You can lead a horse to water".

5.1 PART 1 FEASIBILITY Part 1 Feasibility explores the opportunities and barriers for practice nurse involvement in this type of study. It draws on a review of Australian practice nursing literature and a context analysis of the discourse data.

The context in which research is conducted has an impact on the research outcomes. A mixed methods approach enables a more complete understanding of outcomes in a complex context like primary health. Key constructs may impact on the conduct and the outcomes of the trial. The primary outcome measures, improved breastfeeding rates, may make a lesser scientific contribution if the trial itself is not feasible to replicate. Assessment of feasibility is a secondary objective in this study (section 3.1.1)

Tasks, interpersonal factors and structural factors all affect research outcomes. General practice is task-oriented by nature, yet successful outcomes in general practice rely on social interactions. Creswell et al. (350), in their mixed methodology review, cite the use of mixed methods "to explain the process that task-oriented people use in participation in social situations." (p5)

5.1.1 Summary of practice nursing literature analysis The Australian practice nursing literature describes a growing body of caring professionals, working often in isolation and with little mentoring. (211, 217, 218, 227)

180

Hence, they value networking and support. Their skill and experience levels vary widely, however their work as "specialist generalists" is defined by a complex interaction of clinical, administrative, organisational and co-ordination roles. (221, 223) Nurses appreciate adult-learning style education, and the opportunity to integrate new activities into their work if these will improve patient care. (218, 221, 224) More research into how such new activities may impact on outcomes is needed. (225)

Practice nurses are masters at individualising care for patients across a wide range of general practice styles (222, 226) and patients find them accessible. (219) The major barriers to a greater role in primary care interventions are their need for specified training, (360) support from their general practitioner employers, (219, 230) and a funding structure that supports more independent practice. (221) Recent funding changes that prioritised nurse immunisation (225, 277) may create opportunities for concomitant maternal and infant health promotion.

5.1.2 Thematic analysis: context Each context area was explored for insights into the feasibility of the research project, taking into account four setting contexts and two task contexts. Practice nursing is a complex professional activity where key skills include providing links and co-ordinating care. Therefore, it is not surprising that sub-themes often linked across several contexts. The settings included Nursing, Mothers, Community, and General practice and the tasks included Research and Learning.

5.1.3 Nursing context: providing care

Patient care It was clear from the data that "Caring for the mother" was a definitive nursing role, consistent with the literature, and was raised in the majority of interviews as shown by the comments in Figure 5.1.Caring is central to high quality nursing practice, directed in support of those patients specifically booked to see them. Contemporaneous research notes of one interview record: "As if these are MY patients, the importance of being able to better meet their needs is greater than other demands on my time about other people's patients" (PN5). Some practice nurses clearly identified how they gave priority to caring, wherever possible. "Caring for the mother" may cause hesitancy in using research tools, according to other contemporaneous research notes, "very deliberate about not wanting to offend HER patients." (PN2) Several practice nurses reported their disapproval of breastfeeding promotion by specialist nurses when insufficient care

181

was shown towards the mother's emotional needs. Tasks needed to be perceived as caring before they were acceptable to the practice nurses.

"Connecting with the patient" linked nursing and mothering contexts. Practice nurses described this sense of connection as resulting from quality nursing care. "You do feel you've got a bit of a bond with them because you've gotten a bit personal with the questions." (PN10) The patient's whole family, various members of which may attend along with the mother, fell into this sub-theme, making the tasks of the intervention a natural fit in her work. This connection then enabled further intervention or, if they were unable to achieve it, limited any intervention, "It depends on what kind of connection you can make with them" (PN7). Connection was enhanced by the continuity of care that is a feature of general practice, enabling repeated contacts with mothers. Connection featured the use of intuitive skills "I knew right away there was a problem, your gut feeling tells you that" (PN1).

Working "Getting the job done" was a sub-theme that linked the nursing context with the general practice context, and is discussed further under the general practice context. Practice nurses reflected on the limitations imposed by business structures and tasks. "What practice nurses do" linked nursing with both general practice and learning contexts. Reflections in this theme involved practice nurses expressing sorrow that practice managers, receptionists or doctors seemed to overlook their capacities and contribution to the workplace. One nurse who was leaving was disappointed that she wasn't being replaced, "It does make you feel a bit undervalued, considering our lists are pretty busy." (PN11) "Peer-to-peer learning" linked nursing and learning contexts. Education was valued by the practice nurses, and valuable insights were gained from other practice nurses who understood what was useful and relevant to them. "I looked over at [another PN] and I could see she felt the same"(PN15). The iterative process of reporting challenges or strategies from other practice nurses in subsequent telephone interviews became an important learning tool. Consistent with the practice nurse literature, collegial support was appreciated, "good to hear I'm not alone!" (PN3). "Local connections" linked the nursing and the community contexts, describing resources that these rural practice nurses used to benefit their patients. One spoke of the local maternity and child health day-stay unit. Another emphasised the range of local options that she offered mothers.

182

Research The processes of the trial linked a number of contexts together, as discussed below. For example "Having the time available" linked the General Practice, Nursing, and Research contexts in terms of trying to fit all the tasks into a busy day.

Figure 5.1 Nursing context

'Have you ever heard of the mother and baby unit?' She said yes, but she didn’t know how to get to it, 'Well I think that if you’re prepared to go, that would be a good place. They will observe you feed the baby and work out what’s going on, and they’ll also give you some good points.' She’d actually put him onto formula a week before and in that week she’d actually changed her formula twice! She was worried the baby was constipated... 'Would you be interested in me contacting them?' By this time she’s crying, admitting that she’s just not managing. So I just rang, and I was lucky I got someone that I know, and they fitted her in the following work day. 'You make sure you pop in and tell me how you get on.' I don't think this outcome would have happened with this girl if she had booked in a pap with a doctor. She was booked for the pap and the immunisation so she had an hour. I went over that, and had to catch up. I knew that I had an hour to work with this situation.... In [closer appointment scheduling] I still would have addressed it, but I wouldn't have the same outcome. PN1 A lot of midwives they still seem to harass mums on the feeding. Rather than 'OK, mum's made their decision, we give her the information and walk away', they still feel the need to push, [get] someone's back up, give them the heavy duty. It is still a thread in the tone of the whole experience which is actually kind of sad. I guess that's probably why, too, I've kind of gone through with that [approach to mothers of] 'happy, healthy mum and a happy, healthy baby.' PN 3 I do their vaccination and to settle the baby I often ask if they’d like to breastfeed the baby here, to settle the baby in the privacy. That’s supporting them, isn’t it? PN19 You sort of know whether she's coping, or if she's just at the limit, if you're watching body language and that. PN3 These discussions aren't very long because the immunisation's been done and I'm trying to do paperwork and I know there's someone else [waiting]. PN9 I just think, with the parents, and the mother, we've got more of a rapport with them. PN6 I'm still getting them back at five years, I see my name repeatedly in the baby book. PN9 I think [practice staffs] don't really understand what I do, it's not the same as what the doctor does, don't you think? PN1 Since that Monday, it's been a heap better, because I've picked up lots of tips from the other girls.... That last get-together really helped me. To get other perspectives, it helps you to look outside the box. PN13 I said who are you talking to, have you got anyone to talk to? So that's why I gave it to her, the ABA and the early childhood nurses, I said just in case. PN3 Sometimes you just get run over by so many kids and the doctor keeps you waiting and you just don't have time. PN 107

183

5.1.4 Mothering context: work

Becoming a mother An awareness of the magnitude of entering motherhood entered the discourse of most practice nurses as seen in Figure 5.2. An Australian qualitative study had a similar finding. (361) The feasibility of influencing the behaviours of mothers with this intervention depended on how well practice nurses understood and were able to work with mothers. In this thesis, the only data collected directly from mothers was their breastfeeding outcomes. The context of mothering is filtered through the lens of the practice nurses' perspective, which is naturally influenced by their own experiences. These are addressed in Part 3, and described by one practice nurse as "our own baggage" (PN16). However, the practice nurses' capacity to listen and learn, following training, proved to be more relevant than their personal experience. Practice nurses often saw women's weaning decisions in the light of enormous changes in their lives. When "there's so much on your plate after a new baby"(PN3), weaning could bring relief to some mothers. "You give the baby a bottle and off you go"(PN4). Yet as another practice nurse reflected, it was worthwhile raising the subject of continuing breastfeeding, because some mothers were still willing to consider the idea.

"Mother enjoying/proud of breastfeeding" was a theme describing relaxed and satisfying conversations. Most nurses saw enjoyment of breastfeeding as a clear facilitator for a mother's openness to the idea of ongoing breastfeeding, particularly where older children were of school age or where the mother appeared confident. Some practice nurses found this enjoyment unexpected. "Mother's response to baby's behaviours" was related to mothers' weaning decisions, sometimes made on the basis of insufficient knowledge.

Workload "Mother's busy workload" was raised often in the discourse. In families with numerous small children, there was a sense that mothers often reported weaning early because of an overwhelming workload. "Life is too busy" (PN10). The workload seemed to be accepted as one of the harsh realities of motherhood. One practice nurse described her patient "finding it hard to cope with being a new mother and having three under five and all those nice things that people have to cope with." (PN5) Family size may not be a barrier to breastfeeding, to the surprise of some practice nurses. In these families, the mothers were more likely to be more educated and have greater personal resources. It was a common perception among the practice nurses that older children

184

added more weight to a mother's thoughts about early weaning, particularly in a setting of social disadvantage.

The challenges of adding paid employment to their workload affected some mothers' confidence in breastfeeding according to several practice nurses. "A few working mums a bit anxious"(PN16). This is interesting in view of the literature about the importance of confidence for ongoing breastfeeding. This "Social context" linked the mothering and the community contexts. Return to paid employment was raised often as a deterrent to ongoing breastfeeding, due to structural barriers, "And I can understand, especially... if you don't have a private place at work." (PN11) However, a few practice nurses were impressed by the way a mother handled this challenge with confidence. In the context of anticipating paid employment, as seen in the results of the RCT, practice nurses were able to support many mothers.

Accessing support Practice nurses perceived social support to be important to the continuation of breastfeeding. "Social context" included a mother's friends, who may or may not be able to help the mother to continue breastfeeding. One practice nurse encouraged pregnant women to contact the local Australian Breastfeeding Association counsellors before the birth to incorporate these women into their friendship circle. She explained, "You will more access a friend than you will access an organisation"(PN3). This view was supported by another practice nurse who was expecting her own baby and who made a personal connection with a breastfeeding counsellor. Isolation was seen as detrimental when a woman found breastfeeding and mothering difficult. Farming families and transient defence forces families were seen as isolated. It was clear to the practice nurses that living near supportive people contributed to a mother's ability to happily connect with her baby and continue breastfeeding. They also noted that when family or support people had minimal breastfeeding knowledge or experience, their support for the mother could easily result in early weaning. Practice nurses assessed their own success at being supportive to a mother in distress from the mother's conversation, mood, and body language, "She was a lot more relaxed when she went out"(PN1).

Practice nurses did not speak of any role for themselves in addressing the mother's social context directly. They functioned as kind observers who willingly harnessed health support services but did not report harnessing social support services for

185

mothers. Some found it hard to see how the intervention could help mothers in difficulty. Others found it hard to see any role for the intervention or for broader community supportive networks if breastfeeding was going well and "mum's face lights up when you ask" (PN3). These perceptions are contrary to evidence from motivational interviewing trials. Evidence supports the deliberate offer of support, which can achieve health and lifestyle gains, and which assists with maintaining these gains. (312, 362)

The theme of "Explaining what went wrong" linked the mothering and learning contexts. Practice nurses reported the stories of mothers' experiences and reflections of weaning earlier than desired or recommended. According to the practice nurses, mothers' decisions were based on their own learning. Their knowledge about breastfeeding management and parenting came from their own experiences, and in many cases, from what they learned from family and friends. Occasionally practice nurses observed that a mother had learned about continuing breastfeeding by deliberately seeking information, which tended to be from a more educated mother. Practice nurses found it satisfying when they helped mothers to learn more about their breastfeeding options.

Research Finally, in the context of questions asked within the research intervention, "Mothers not wanting to talk about it" linked the mothering context and the research context. Practice nurses presented the mothers' resistance to questioning as independent of general practice or community contexts and generally unrelated to their own role. Rather they attributed it to each mother's response to the research-generated conversational activity. Some mothers simply did not engage in the conversation that the practice nurse attempted to initiate. Not all mothers were willing to engage on the subject of ongoing breastfeeding. The power of such interactions to deter further attempts at intervention delivery depended on the extent of the practice nurse's confidence, skill development and access to mentoring as discussed in Part 2.

More broadly, there was no reflection in the discourse data on any mother's interaction with a particular general practice, other than practice nurses expressing a desire to have patients continue to attend their own practice. There is no data on the mothers' perceptions of the general practice setting, and little insight on how the mothers perceived the practice nurses, although the practice nurses noted changes in expressed opinions about breastfeeding possibilities.

186

Figure 5.2 Mothering context

Thankfully, she has two friends who've been able to support her. PN8 She was out on the farm; that might have been part of the problem. PN1 We all bring a bit of our own baggage! [But it's about] addressing the patient and their problems and leaving your own baggage behind. PN16 She's thrown in to the ultimate of women's roles but doesn't know how to adapt to it yet. Her comfort zone is back in the life she had before. PN4 I get a few who just want to have more freedom, get their bodies back; some of them seem to be open to suggestion. PN11 I'm surprised at the number who are happy to sit there and say 'No, I'm having a good time with it.' They look like there's time to be relaxed about it. PN3 Some have even said how much they enjoy it and they want to keep going. PN11 [S]he can have all the time in the world with this one, and she's open to this, 'I'll have as long as I can.' She's not under a load to do anything else... you could see her face, really excited when she was talking about it. PN3 She wasn't aware of the recommendations but she was quite happy with that. She was one of those arty types but then they don't always breastfeed, do they? PN10 Lately I have heaps of mums at 6 months who have quit because the babies didn't seem to get enough... They say she cried a bit more often... Sometimes it’s really funny they bring in this cuddly baby with lots of fat and they say, “I don’t think they’re getting enough”. You hear often that it’s something they’ve been told and they don’t know who to listen to and that’s only fair I wouldn’t know either. PN7 Third child, two and a half hourly demanding, no time for the other kids, she decided that was it. PN8 Not so surprised with the mums who are having a hard time with it. Would you say low socio- economic? There's generally more kids, there's less income. PN3 She had all four with her, they'd been to the park, the library, she was so calm! The two week old was feeding well. The other kids were a bit distracting; I'd got the distraction box down for them. PN10 Sometimes mums will bring it up about returning to work, that seems to be a fairly common conversation that I have. I actually have one lady, I felt really good about it actually, I gave her the return to work pamphlet, she was thinking of giving up, she didn't think the baby would tolerate going from bottle to breast, she said 'I've heard that a lot of babies get confused', and I said well a lot of babies don't'. PN11 They're circulating in a younger group and those women haven't even had babies. PN4 My sister-in-law sent me a gift subscription to ABA, her cousin is like a group leader or something. I met her at my nephew's first birthday party on Saturday... She's really lovely. It's good to put a face to a name because I had the details and that but I just hadn't gotten around to getting in touch. PN13 [T]he number of service [Defence Force] people that come through that don't have support, they're the ones that appreciate what you give them. PN5 I'm using most of what was given to us to be able to appropriately work with all patients but mainly our mums that need just some help with everyday stuff. PN5 That baby's one and she's still taking that baby to work PN3 I think she needed one day at a time, I didn't think I should talk about the recommendations. PN10

187

They really do make an earnest effort to tell you why they went off [the breast], all of the sagas and what went wrong...It's venting. PN3 Some of them are at mothers' group [and they are getting advice:] 'swap to a different formula' [One woman who I saw was using] Phenergan - 'My mother says it will put him to sleep'. She hasn't actually gone to a doctor, to check the baby. PN1 She was feeding and she wanted to feed, so I sent her over to the mother and baby unit. But she's done away with the wrapping that they gave her, because the grandmother told her. PN1 They talk quite a lot about it; they talk about their girlfriends and what they’ve done, if they were told to put the baby onto the bottle by grandparents. They’re not hunting down or trying to find any information anywhere else, they just listen to their family. And the mothers may be visiting them every day or every second day, they’re always there. PN6 The baby was only 4 month old and they told her it shouldn’t be feeding overnight and the sister’s babies 'All slept through the night and they were bottle fed, and you need to look at the feeding...' PN11 She was going to give up at 10 months, she was going back to work so we had a talk about continuing and she was quite open. She was quite surprised you could, and quite pleased. PN5 It looks like she'd done some pre-research, maybe she is a nurse but I didn't ask her things. She certainly had her head around the breastfeeding stuff... It was nice to see one that has that info, she's the first one I've come across who's really on top of it as a first time mum, it sort of gave me a shock, I run into a lot of second time mums who've got it together, but not the first time ones. PN3 Had a client with a 2 month old, a defence couple. She said she was going to wean at 6 months, going back to work. It was a 'don't go there' type of presentation. PN11 Sometimes I think 'I may be talking to the wall here' PN15 A couple of times I've talked about WHO [recommendations] but: silence sometimes PN8 It varies between mums, some don’t really seem to want to talk at all about anything; they’re the hardest ones. Then there are the ones you can have fun with, they’re happy to talk about it. It depends on what kind of connection you can make with them. Sometimes it could be me, there’s so much going on here... I guess that’s just part of this job. PN7

5.1.5 The community context: out there The broader community setting in which general practices exist and function was barely addressed in the discourse. "Community attitude to breastfeeding" identified by practice nurses in terms of "out there"(PN1), is discussed in Part 3 in the key themes. The lack of discourse specifically on wider community issues such as media images, retail marketing of substitutes or paid maternity leave may be due to a low level of awareness of structural barriers to breastfeeding, or perhaps due to the practice nurses' more individualised approach to health. There was no explicit reflection on broader regional or national breastfeeding knowledge or support. Only one practice nurse commented on public broadcasting and she immediately reflected on the research context. She commented "The minute I heard I wondered how that would go with your study"(PN13). Practice nurses did reveal an awareness of repeated messages in the wider community; however, the semi-structured interview method deliberately focused on workplace conversation and interaction with mothers.

188

The community and learning contexts were linked by discourse about the impact on mothers of low levels of breastfeeding management knowledge. (See Figure 5.3 for quotes) Sometimes inaccurate advice from community members closest to the mother could be given quite forcefully. Practice nurses found it satisfying to support these mothers when possible.

Figure 5.3 The community context

A lot of them don't know, they think there's no reason to continue beyond 12 months. PN1 The grandmothers and the mothers - the children haven't been sleeping - 'Oh, put 'em on the bottle'. They're mostly younger mothers. PN6 I also find that women are pressured to stop at 2 months, 3 months; they hear 'If they're crying you've run out of milk.' PN7 One mum said she loved the breastfeeding but her family, sister and mother-in-law or mother who didn’t fare well breastfeeding... [Their] other comment undermining her ... she was criticised for feeding it too much or too often. It wasn't that often. I said 'How do you feel about it, do you feel it's too much?' and she said 'No.' And I said 'Breastfed babies aren't like bottle fed, they can't be fed too much, they'll only take what they want' ... the baby was only 4 months old and they told her it shouldn't be feeding overnight. PN11 In the media ... they were saying that solids should be started at 4 months. PN13

5.1.6 General practice context: part of a team

Quality nursing Several practice nurses spoke about specific details of the physical and financial structures of their workplace, as expressed in the comments in Figure 5.4.These structures could be barriers to "Quality nursing care", impacting on both patient care and practice nurse job satisfaction. They were mostly seen as beyond the control of the practice nurse. In some practices, these barriers impacted on the feasibility of any nurse-led intervention. Immunisation-specific funding gave more independence and control over appointments for some nurses than for others.

To a varying degree, funding structures and workforce issues have an influence on which patients will see the practice nurse, and for what care. "I'm actually wondering if [other practice] takes a few because they're bulk billing... And it's been easier to get in [to other practice] because they've got lots of doctors" (PN1). Staffing levels affected the "Busy nursing workload" and impacted on the service that could be offered. As primary care organisations, general practices provided emergency care. Unpredictably busy days could impact on "Having the time" and restrict the service to the basics of care.

189

Supporting the nurse According to several practice nurses, some general practices failed to value higher- level professional nursing practice. Rapid immunisations were a priority in the appointment schedule at more than one practice leading to practice nurse dissatisfaction. Practice floor plans and business structures had the potential to be barriers to effective practice nursing, while unbalanced management structures could lead to friction and dissatisfaction. In these settings, practice nurses perceived their employers as unsupportive.

”Practice support for the practice nurse" came from good teamwork and internal respect in other workplaces. Staff supported each other efficiently to improve the care of patients. These practices were considered open to practice nurse initiatives for structural change. A common staff vision for the workplace was appreciated, and where structures remained restrictive, teamwork could overcome structural barriers and provide effective personal support. Support on a personal level was valued as indicated by the comment "when I've been sick they've been really helpful here, really supportive" (PN1). In some cases, supportive employment conditions could balance out a discrepancy in pay per hour compared to hospital work. Some practices supported educational needs of new staff, using the local Division of General Practice, as noted in the literature. One practice also provided support with workplace tasks that commonly fell to the practice nurses in other practices, such as ordering supplies or sterilising equipment.

From the variance in the discourse, the diverse range of general practice is clear, even in this small sample. In any practice, the level and style of care provided is vulnerable to changes in funding structures and workforce movements over time.

190

Figure 5.4 General practice context

I could fill my day up helping people valuably but I wouldn't make any money... It has a valuable place but Medicare hasn't recognised that, and there's no way I can get around that. PN1 I'm down in the back room all day, I don't speak to anyone except the patients, I don't know what's going on, and it's not very satisfying. They do have team meetings but there's no room for that kind of comment. PN1 The receptionists have a huge amount of power and they're the ones who make all the bookings and if you complain, they get cross. PN1 Work won’t pay study leave; I’m really pissed off about that... I’m rather disappointed that we are not getting the support that is so much needed... I said I’m not asking you to pay for it, it’s just my time. PN5 When you're doing 40 patients a day, it's not what I wanted... I had 30 fluvax, they were booked five minutely. PN5 They're down two doctors here so we've been taking on a lot of the injections; they've been flat out, poor things. PN10 Occasionally there's those days when you just have to get them in, get the body on the seat, get them out again. PN3 I don't have to do drug orders or sterilising, I have clinical support so I'm just here to see patients. PN9 [H]ere it was only about a dollar [less than hospital nursing per hour]. And I enjoy the flexibility and the hours here so that makes up for it for me a bit. You have to feel like you're part of a team, not dispensable. PN11 We're looking at the 2, 4, and 6 months as double appointments... It's generally a squeeze with the 2 months anyway... There's been maybe 15 minute to 20-minute appointments, that's to change from October, it's been needing to happen for a long time. PN11 I had one woman who came in with some mastitis and this was her second baby so she knew about it, and she'd been doing all the right things. The girls [receptionists] said 'Can you talk to this lady?' And I was fully booked. Then when I found out I told her to come straight in, and I got [the doctor] to see her so she could start on antibiotics straight away. PN4 Like my mum used to say, what you miss out on the roundabouts you make up on the swings. And I've got support here, everybody works like that, they don't watch the clock either. PN4 We need to extend [the time] on the 2 month. I can extend it. They’re giving me 15 minutes [grimacing tone of voice]. But if there’s a space before or after, and the girls are pretty good, they tell me if it’s an 8-week immunisation. PN6 [New PN] is planning to do the course. She's talking to [staff member] at the division about that. She will be immunising. PN13

5.1.7 Research context: to use and to adapt

Practice nurses reflected on the research tasks they were undertaking, and their comments on this process are grouped at the end of Section 5.1.7 in Figure 5.5. Their discourse provided insight into their experience of "Following the Conversation Tool" within their work context. Practice nurses confirmed the usefulness of the Conversation Tool format. Consistent with the literature, these practice nurses wanted education. They also wanted to provide targeted quality care. Practice nurses reflected on their motivational interviewing task from the perspectives of how to know what to do, how to be able to do it themselves, and how to weave it into nursing practice in the context of

191

practice structures. Evaluation of practice nurse interest in the process of intervention delivery provided a means to evaluate the feasibility of the Conversation Tool intervention.

Each component of in the Conversation Tool intervention was explored across sequential practice nurse interviews. Practice nurses were specifically asked to reflect on their experiences in attempting to deliver the various components. Contextual themes arising are identified in Table 5.3, which compares the tasks of the practice nurses with the context themes. Discourse analysis showed that brief training in motivational interviewing was feasible for some practice nurses. However, others needed a greater mentoring process to develop their skills, as demonstrated in Part 2 of this chapter, Section 5.2.

Engaging, opening the conversation Affirmation, one practice nurse explained, "Comes fairly natural for me anyway." (PN5) Some practice nurses were able to affirm mothers in their regular responsibilities, such as comforting a child, or providing healthy meals, "Your milk's better than anything you can give them off the supermarket shelf"(PN9).

Giving the recommendations was challenging for most practice nurses, as discussed in Part 3 in the key theme "There's a stigma out there". Generally, this process only became feasible with mentoring, ongoing reflection, and education over time, discussed below in Part 3. Successful strategies included specifically eliciting knowledge of the recommendations first, or a permission-giving approach. The process of facing this stigma with gentle action, according to the proverb discussed in Part 3 Key themes, "You can lead a horse to water but you can't make it drink," made sense to the nurses. They needed tools to defuse any confrontation, and communication skills were discussed in training and mentoring. In describing their actions, practice nurses revealed their consistency with the spirit of motivational interviewing: drawing out experiences and ideas from the mother (evocation), collaboration, and acknowledgement of autonomy.

192

Table 5.3 Feasibility of the Conversation Tool

Panel Task Themes arising from discourse Engage Identify breastfeeding mothers "Affirm the mother" (Research context) Ask about experiences

Affirm, reflect Offer recommendations "Offering information" (Research Elicit response context) Assess response

Motivate Action for pre-contemplation "Planting a seed" (Research context) (Coloured boxes) - the red box, not ready

Action for contemplation "Weighing it up" (Research context) - the orange box, ambivalence

Action for committed "Looking ahead" (Research context) - the green box, ready

Move out Offer resources "Offering information" (Research (Resources) - booklet, local resources, context) referral "Affirm the mother" (Research context)

Encourage return visits "Connecting with the patient" (Nursing context)

Coloured boxes for motivation Practice nurses were variably able to understand and enact the motivational tasks that matched the mothers' readiness to continue breastfeeding. In the Conversation Tool, practice nurse actions following the recommendations were guided by the "coloured boxes".

The red box was for mothers who were planning to stop breastfeeding. Offering unsolicited information on an emotive topic was very uncomfortable for the practice nurses, particularly if they risked offending their patients. "You don't want people to feel that you're telling them they're doing the wrong thing." (PN19) Describing the red box action as "Planting a seed" was helpful imagery for most practice nurses who were then more able to continue with this task to some extent. (See Part 3 Key themes, and Version 3 of the Conversation Tool in Appendices)

193

"Weighing it up" involved practice nurses using the questions in the orange box to prompt decision-making, when a mother was unsure about continuing. A practice nurse who had previous motivational interviewing training reflected that this meant "being able to recognise ambivalence and lead them across to the positive side of breastfeeding"(PN4). Another with no prior MI training was daunted by this task, and worried "when I get one who's a bit unsure that's really going to test me"(PN14). However, there was a sense that most practice nurses had moved away from authoritative advice giving since the MI training. This demonstrated their capacity to accept a level of ambivalence, an essential aspect of the motivational interviewing approach.

Some practice nurses were skilled at "Looking ahead", the "green box" action for mothers who were open to continuing. Many discussed strategies for mothers to continue breastfeeding despite plans to return to work. Several discussed expressing of breastmilk. For a few practice nurses, however, performing the action of anticipatory guidance required them to make greater personal change. In a busy workplace, they were accustomed to moving on without discussion when a mother had no breastfeeding concerns at the time.

Resources, moving out from the conversation Offering the information booklet "Breastfeeding Confidence" was a more familiar task and not described as problematic by any of the practice nurses. Practice nurses also readily referred to the available breastfeeding support services provided by government or non-government organisations, in a non-directive manner "have you heard of ABA?" (PN6) They also encouraged mothers to seek breastfeeding support from their own network.

"Offering information" was discussed at various steps through the Conversation Tool. This sub-theme particularly applied to the recommendations and to the avenues available for ongoing support and follow-up. One description of the motivational interviewing approach to non-judgemental information giving is "elicit, provide, elicit". (363, 364) Using an MI framework, the Conversation Tool guided the practice nurse to explore experiences, offer recommendations, and elicit response. Practice nurses frequently demonstrated flexibility and skill in following the spirit of motivational interviewing, although this depended on their level of training and their personal capacity for client-centred consultations.

194

Figure 5.5 Research context

I say 'Children grow really well on breastmilk and it's a comfort'. PN6 I'm asking the questions while I'm drawing up the vaccinations, it's a Conversation Tool that I can use and I can adapt it, using the time, you're not just sitting there doing nothing. PN4 If they ask me for something, I tend to put it back to them, for them to come to me with the answers themselves, and I say that sounds like a really good idea. Most of the time they do know, but they're waiting to be told they're doing the right thing. PN5 I've had a bit of the eyes raised, 'Oh, that long?' ... I don't want them to get the ire up, 'Don't want to see that bird again'. PN9 You're at least suggesting it, that you don't have to give up breastfeeding... you've put the idea in their head anyway. PN11 One mum had breastfed the last for seven months. I said 'You know the recommendations?' PN5 [I had] one woman who wanted to quit and she was told she wasn't allowed to. The MACH [nurse] was pushy. I had to explain to this woman that it's her choice. PN7 It just guides me a bit because I can see my boxes, you know the three, the stop, go and maybe. PN1 I get the negative and the positive instead of just saying. PN6 The husband wants to give a bottle at night, so I said 'what about expressing your breastmilk?' and she said 'Oh, I hadn't thought of that' PN6 If you can offer support and resources, that's the main thing. PN14 That little booklet [Breastfeeding Confidence] is a great thing; I personally wish I'd had it. Even if things are going well, I can say 'have a flip through when you have time' PN12 'Have you ever heard of the mother and baby unit?' She said yes, but she didn't know how to get to it... 'Would you be interested in me contacting them?' PN1 I told her if she needed any assistance we were here, told her about the breastfeeding association, and that’s all. PN12

195

5.1.8 Learning context: trained in whatever we do

New knowledge, skills, resources Practice nurses reflected on the process of learning and change they were undergoing within the research project, and their thinking processes can be clearly seen from their comments in Figure 5.6. They discussed learning with respect to several other contexts. Linked with the general practice and nursing contexts, they discussed what they had learned about breastfeeding and through the research tasks, and how this affected their nursing work. Linked to the community and mothering contexts, they discussed their increased awareness and skill in supporting breastfeeding mothers. Learning was also related to the practice nurse's own confidence and fluency in the tasks, which is discussed in Part 2 and Part 3 of this chapter.

One source of learning was the mothers themselves. One practice nurse explained a learning experience that she had encountered. A mother requested that she vaccinate the baby during a breastfeed, telling her about evidence that breastfeeding reduces pain sensation. (365) After her experience with this mother, the practice nurse became keen to give immunisations into both legs of any baby during a breastfeed without moving the baby. This practice nurse had subsequently shared this learning with other nurses who provided local council-funded infant immunisation clinics. Learning from mothers could also be about community perceptions or practices such as the prevalence of formula substitution or of brief breastfeeding durations.

Practice nurse training sessions were an opportunity to learn about breastfeeding management. This was an uncommon education topic for health professional education and yet often very relevant. "Let's make us as professional and trained in whatever we do." (PN5) The training was an opportunity for learning to use community volunteer resources. Learning about local government resources such as the mother and baby support unit was also valued, as there were no formal or co-ordinated links between general practices and public health services.

Poor health professional knowledge or skills Practice nurses found it worrying and disturbing to encounter low levels of breastfeeding management knowledge among other health professionals. They recognised a poor approach or misinformation from other health professionals as unsupportive and as contributing to poor outcomes. Poor medical communication was reported by mothers and witnessed by practice nurses, "There was even one who was

196

giving solids at three months. I was talking with her and the doctor came in and said 'Don't do that!' He just said that and than he left" (PN6). On this occasion, the practice nurse used her knowledge and communication skills from the training workshop to support the mother.

Figure 5.6 Learning context

It's something new that one of the mums asked to do here and I showed [other clinic nurses] ... how to do [the immunisation during a breastfeed] without moving the baby over and they said 'Ooh, can you do that?' so I'm out there teaching as well! PN5 I've handed out the cottage number too, I don't know HOW many times. PN5 I’ve had one lady that was told by the early childhood [nurse] to supplement the night time feed for the baby, that witching hour, ‘I haven’t got enough milk'. No, maybe we need to look at you, you might need some support. PN5 You know what I’ve discovered, more than half of them aren’t breastfeeding! They started breastfeeding in hospital but the majority stopped at about 4 – 6 weeks, some of them even just after they left hospital. I never noticed it until we started doing this study, you don’t realise. PN6 I've been asking the mums who come in for the 6-month immunisation and you know they're all giving bottles or solids! I was amazed. PN6 One mum that really wanted to give up, I gave her your pamphlet [Breastfeeding Confidence], she was really pleased. You know, I'm surprised how many haven't heard of the Breastfeeding Association. Anyway, I said 'They're really good; it's really worth getting in touch with them.' PN6 But I was really interested in why she was doing this; I asked 'why are you giving him solids?' And she said it was because her mother told her to. She didn't know and she asked, and that's what her mother said.... I said to her your baby doesn't really need that, he only needs breastmilk, it's about his development, and she said yes he wasn't really eating it, he was pushing it about with his tongue and I said 'that's right.' PN6 I came out of that training and I really felt cheated, I wish I'd known that when my babies were born. PN15 [They] actually started solids at 4 months, and one of them said one of the paediatricians felt that the reasons our kids have so many allergies is because they start solids too late, and they were recommended to start at 4 months. I was gobsmacked. PN14 They said to her 'You're doing it all wrong'... well that's what she said, she's a young mum and she felt the baby wasn't getting enough milk so she put him straight on the bottle after that. PN4 It's made me realise the attitudes of health professionals can really affect how the mother feels. PN13 One was asking about re-establishing feeding because she had been told she couldn’t breastfeed while she was having some medication. ... She’d been told she couldn’t get it going again. I said 'That’s not true, you ARE able to'. So I gave her all the information to get in touch with the Breastfeeding Association. I said 'I'm no good to you'. She said 'I'll look at that.' PN5 I have one lady that has re-established, she was told she had to give up because of a medicine. PN5

197

5.2 PART 2: FIDELITY Part 2 Fidelity explores the process of attitude change and skill development, which relates to compliance by intervention providers. It draws on a descriptive analysis of practice nurse demographics and a chronological analysis of the discourse data to determine the extent of treatment fidelity in the process of conducting the trial, one of the secondary objectives of this study (section 3.1.1).

Demographic analysis of the practice nurses' personal background provides depth to the discourse analyses. It illustrates a little of where the practice nurses started out, in terms of familiarity with breastfeeding and motivational interviewing, as a background to the evaluation of where they arrived. For many health professionals who discuss breastfeeding with mothers, their own experience is their main source of knowledge. (125, 132)

5.2.1 Personal background descriptive analysis Intervention practice nurses were aged from 32 to 61 years (mean 43.9 years) as seen in Figure 5.7. Half of these nurses aged 36 to 50 years old as seen from the interquartile range in Figure 5.8.

Figure 5.7 Distribution of practice nurse age

198

Figure 5.8 Range of practice nurse age

Most intervention practice nurses (n=12) had personal experience as mothers, as shown in Figure 5.9. One third (n=5) had over two years of total breastfeeding experience with two or three children.

These descriptive results indicate that for only a minority of practice nurses, personal experience was consistent with the Australian recommendations for breastfeeding duration. (30) In contrast, three had no children; another four had breastfeeding experience that clearly differed from the recommendations. The results for the remaining three are unclear.

Figure 5.9 Practice nurse children and breastfeeding experience Count of practice nurses of practice nurses Count Number of own children Total own breastfeeding experience (months)

199

Few practice nurses (3/15) had previous training in breastfeeding management, and these nurses were all older than the mean age of 43.9 years. Few practice nurses (3/15) had previous training in motivational interviewing, and these nurses were older than the mean age. With one nurse having received training in both areas, only one third of practice nurses (n=5) had prior exposure to educational components used in this research project.

5.2.2 Thematic analysis: changes in discourse over time In forming conclusions about the effects or size of effects of any treatment, a Type II error may arise where fidelity is low. In this case a false negative result may occur, where the capacity of the intervention to influence the outcome has not been detected. (366) There is a risk of this type of error with real-world trials in complex settings. (367) Chronological discourse analysis together with the personal background data indicate that in some practices treatment fidelity was achieved earlier than in other practices. Nurses who had prior training in relevant components were able to incorporate the intervention more quickly into their regular practice.

Personal breastfeeding experience could be a positive or negative influence on adherence to trial protocols. For some nurses, personal experience led easily to confident delivery of the intervention, "I was trying to explain about the eczema. We talked a little bit about it, [how] it makes a difference... And comfort, that's a lot of why I fed mine.'" (PN6). Other practice nurses initially withheld essential components of the intervention due to assumptions based on their personal experience.

Certain chronological themes were common to most nurses, although the individual emphases for each nurse differed. These themes provided insight to the process for each practice nurse of changing her routines to incorporate the intervention, a key factor in organisation change and required adherence to protocol. (287) Mechanisms of the trial were considered, with reference to the theoretical framework of motivational interviewing.

5.2.3 Beginning: "Finding my own pathways" Much of the practice nurse discourse focused on their learning process, as the boxed quotes reveal in Figure 5.10. Remembering the Conversation Tool and the need for open questions were initially challenging for many practice nurses and at times, they forgot to use various components. Targeting the right patients and remembering the components of the intervention took time to learn. These processes were presented

200

during training, but each nurse needed to adapt these processes to her own workplace setting and this did not occur immediately. Training and mentoring started two months before the RCT participants reached the first intervention point, in anticipation of this need. However, the speed of adaptation was also a feature of individual practice nurse personal characteristics and some needed more time to practice and more mentoring than others.

Practice nurses needed to learn new techniques to use the motivational interviewing approach as illustrated in the Conversation Tool. This approach required offering information, rather than simple advice giving. Finding their own way to use this approach took more time for some practice nurses than for others.

Early discourse shows tentative attempts. For some practice nurses, contextual factors around mothering, their personal background, or the structures of their workplace influenced how quickly they could find their own pathway. One of the challenges to implementing the intervention was practice nurse perceptions that there may be no need for any intervention. Early in the project, mothers who were breastfeeding may not have been offered the recommendations or the green box action of anticipatory guidance.

Figure 5.10 Beginning intervention delivery

Every now and then I keep going back, have I asked the questions? PN3 I'm just getting it down, finding my own pathways. PN4 I need to check the wording sometimes. PN4 I do actually enjoy sitting down and just discussing breastfeeding with women. PN4 I kept thinking 'I know there's other questions, what were they?' PN5 Don't think I mentioned the recommendations, I got lost, that's an issue for me in the big picture. PN8 It would help me if you had a CD of some of these questions, I could listen to it while I was going to sleep. It's as though I'm changing my mindset. PN8 I'm still not doing immunisation, trying to get accredited. PN15 I'm trying to remember [to deliver the intervention] and I've done maybe one out of ten. It's because of the Boosterix, the mum's there and the dad [both needing immunisation due to new NSW Health policies]. All of a sudden they're gone, out the door and you think O God I forgot to ask about the breastfeeding. And when you do, it turns into a three-way conversation. PN9 To be truthful, the minute they say I'm only breastfeeding, I just leave it alone, they're doing it, they're fine. Do they plan to continue? Yes, as long as we can. PN9 I've also got a background of I wasn't a successful breastfeeder, so I don't want to be making them feel guilty, so it was sort of also having a look at where I come from and my history and then having a look at finding ways so they don't feel uncomfortable. PN16 I feel a bit funny telling people to feed to two, when I've never done it and I don't know how hard it is and I don't really have the words, I don't know what to say. PN19

201

5.2.4 Barriers: "I find it hard" Several early challenges were described by the practice nurses in developing fluency with the process of delivering the research intervention. (See Figure 5.11 for their reflections) None of the practice nurses found the process instinctively straightforward.

Those practice nurses for whom their personal experience differed from the recommendations described the additional challenge of attitudinal change in order to find a genuine response. "We all bring a bit of our own baggage! [But it's about] addressing the patient and their problems and leaving your own baggage behind" (PN16). Some who had no immediate personal experience were conscious of not wanting to add burdens to the busy workload of mothers, which for many also included paid employment. However, increased reflection was a tool to overcome these barriers.

Informing patients of the breastfeeding recommendations was a task the practice nurses found daunting. Simply broaching the subject of breastfeeding experiences was a challenge for many practice nurses. In the personal one-to-one setting of the consulting room, and wanting to maintain good relationships with their patients, they feared possible negative responses. For all the practice nurses, the open questioning was a new strategy that required skill development. These barriers were overcome with time, practice and mentoring.

In the early days, when women answered the initial questions in a way that indicated they were acting contrary to the recommendations, practice nurses were uncertain how to respond. Unfamiliarity with the motivational interviewing approach due to only receiving brief training, affected their confidence. There was anxiety about entering further discussion with mothers who said they planned to cease breastfeeding, the red box: "I'm a bit worried about that one." (PN103). They needed time to come to understand how to focus on accepting patient autonomy, a helpful solution described in Part 2 in the key themes.

For one practice nurse, encountering a mother whose breastfeeding management was being guided, perhaps erroneously, by another health professional, was a challenging situation and she was uncertain how best to support the mother.

The motivational interviewing tasks were conceptually challenging because, by definition, patient responses were unpredictable. Other barriers were contextual as discussed in Part 1; one example is the busy general practice workplace. Barriers relating to skill development were a threat to treatment fidelity in the early days of the

202

trial. Some barriers relating to context such as funding priorities were persistent while others such as a busier day were random. It was not clear whether these barriers became easier to overcome with time.

Identifying and acknowledging aspects of delivery that were difficult was a key part of the reflective practice component of the training workshops. Practice nurses attempted the intervention and were then able gradually to develop their own strategies to manage the challenges they faced.

Figure 5.11 Barriers to intervention delivery

I'm still a bit mucked up on the age group. PN1 Probably haven't done it as much as I wanted to, I haven't had a lot of bubs that have been breastfeeding. I'm hoping to get back into it; I've forgotten some of what we went through with the course. PN2 I'd been thinking there would be all these women and they'd all be in the program but I'm not seeing very many. It's a bit off-putting! PN3 Sometimes for me the vocabulary, the specific English, it is hard to find the right words. PN7 Sometimes when you ask an open-ended question two hours later they're still there so I find that a big challenge, time management. PN8 I felt a bit at sea with some of her responses. PN8 In something like this, I know the importance of it, but it's actually delivering it. PN8 They start [exclusive breastfeeding], and then they just have one formula [bottle] at night. I still find that difficult to... what's the word? It takes more time and effort to enter into that conversation. PN8 I had one the other day who was breastfed and she was comp feeding because she didn’t think she was making enough milk, I think that was through the early childhood nurse that she was seeing, they were suggesting that's what she do. I just asked... what was concerning her? That the baby wasn’t putting on a lot of weight; but otherwise was quite a happy little baby... Because it was through them and not a well-meaning relative I can’t really oppose what they are suggesting, and I can’t really judge what the early childhood nurse had seen. Despite what you know about weight gains and milk supply. She knew a bit, to make more milk you need to feed more and all that. It is tricky when you’ve got a qualified early childhood nurse handling that, I didn’t think it was in her best interest to put doubt in the mother’s mind. She was happy to feed as long as the baby was interested. Still offering breast, I think it was more a night comp from memory, and or if she’d fed and he was still unsettled. Must have been a 2 month. She was happy, and the baby was more settled and she was feeling more settled, because she was worried... I felt kind of like; it would have been difficult to put your own advice in especially if it was different from what she was getting... The thing that irritates them most of all is conflicting advice, particularly within a system. That’s hard because everyone has different ideas and beliefs. It stuck in my mind, probably because I had another one yesterday. PN11 Some days I've come on and it's been double booked everywhere so it's hard to think, just getting through everything. PN11 It's hard though for someone who hasn't had any babies and as a young woman, I find it hard... I sort of feel like that if I was to bring it up with them, they'd be looking at me, thinking 'how would you know?' PN19 Work - that's a barrier in me, I'm in awe of it; I just think, how are you going to manage? I couldn't do it. I haven't had children but I've been around people [raising children]. PN8

203

5.2.5 Communicating: "I'm very picky about how I say it" All the practice nurses found it challenging to develop their communication skills, although for some the changes were greater than for others. (See Figure 5.12) Several commented on how they were expanding these skills into other areas of their work, as discussed in Part 3. As discussed in Chapter 3, interpersonal skills are the foundation of effective motivational interviewing. (298) Adherence to the Conversation Tool prompts increased as the practice nurses increased their communication skills development. They were more able to act in the spirit of motivational interviewing, listening to the mother, collaborating with her, and recognising her autonomy. The gradual change over time towards a more consistent motivational interviewing approach was associated with improved treatment fidelity, according to the practice nurse discourse, even when mothers may have been receiving contradictory information.

Figure 5.12 Communicating the intervention

I think I've got enough skills to give them the questions in a way that they are going to feel comfortable with; it's all about the wording. PN5 Probably a little bit different, instead of more giving advice, asking them what they have achieved FIRST, I say can you tell me firstly, what they have tried and then give advice, and listen to them. PN6 Perhaps I've changed, yes I think I have, I've changed my way of asking questions PN10 I was seeing an older woman, taking her blood pressure four hourly and I was telling her not to... I was disagreeing with her. She got all defensive. Then I backed right away, I said 'I'm not coming across how I want to, why do YOU feel you need to...' It went really well then. We had a bit of a laugh. PN11 All I said was the current recommendations say... unless it’s on the advice of a paediatrician. Future recommendations may come in. Our doctors here still recommend 6 months. PN11 I'm very picky about how I say it, I try hard not to be pushy about it PN12 With the open questions you can just sort of weave it into the conversation PN13

5.2.6 Acceptance: "So that's how it was with you"

While it was not the focus of this research, practice nurses frequently encountered women who had ceased breastfeeding their babies, or who had introduced solids earlier than recommended. "I suppose my biggest issue is the way to speak to mums when they are not breastfeeding... or mixed..."(PN8). As noted in Part 1 in the literature review and the discourse analysis of the Nursing context, "Caring for the Mother" is driven by key nursing values. Practice nurses wanted to avoid the tone of moral judgement implicit in simple health messages, as discussed in Chapter 1. They described a range of caring strategies to respond to mothers who were not

204

breastfeeding according to recommendations, which helped them to overcome some of the barriers described above. In Figure 5.13, their strategies are shown in their own words. Once a practice nurse had built an accepting approach into her pathways, she was more likely to deliver the intervention according to protocol.

Figure 5.13 Accepting responses to the intervention

'Is that something you feel comfortable with?' I have a talk with them about what they want to do and whatever's going to work for them, that's fine with me. PN3

I say I don't really care so long as there's a happy healthy baby and a happy healthy mother. PN3 I try to make them understand I don't care whether they do or they don't, it's up to them. PN3 I said to her 'You had to do what was right for you, but it's not always like that.' PN4 I find when I come across a mother who's not breastfeeding, I have to quickly readjust myself PN8 It's like when you ask them if they breastfed, and if they hadn't you say 'so that's how it was with you', well that opens up things doesn't it? PN10 I would always ask why, to give her a chance and see if there could be something simple that could be fixed for next time. PN14 Some mums who are half way in between breastfeeding and bottle feeding, they start off defensive, you can see that, but you soon put them at ease and then they get really chatty. PN16

5.2.7 Comfort: "I feel a lot more comfortable" For the practice nurses, "Feeling comfortable" was a critical facilitator for following the Conversation Tool. This theme was seen when they discussed their general level of comfort with a task as shown in Figure 5.14. On other occasions, it also referred to a practice nurse's level of comfort in a specific conversation with a particular mother, "otherwise I would have been more than happy with giving her more information"(PN11) or to her assessment of the level of comfort felt by the mother in the conversation, "finding ways so they don't feel uncomfortable" (PN16).

Over time, most practice nurses developed a high level of comfort with the task. This occurred as they developed strategies to deal with the stigma associated with their task and became more fluent in using their communication skills and in structuring the pathways of their immunisation appointments. With increased breastfeeding knowledge and with problem solving around motivational tasks that took place during mentoring, they became comfortable with broaching the subject and with supporting mothers wherever the conversation led.

205

Figure 5.14 Comfortable intervention delivery

Now I would have the confidence to talk with them. PN4 I feel comfortable talking with women about their breasts. PN4 It's really positive that they're happy to ask your advice and being able to answer truly or if you don't know, you can give them a tool to find out. PN5 I feel more comfortable about it now. It's the knowledge. I haven't had babies so I've never breastfed; knowing what's normal and what can be fixed and how it can be fixed. PN7 I'm feeling more comfortable, and it is just practice, and I usually have a squiz [at the Conversation Tool] before they come in, just as a refresh, to prepare what I'm going to say to them, but it's coming more naturally now that I've had more practice. PN11 I've definitely learned heaps more than I knew before. Before I wouldn't have been comfortable raising the topic because I didn't know very much, but now I feel a lot more comfortable. PN13 It's been actually really easy to bring up the breastfeeding. Getting very comfortable with the Conversation Tool. PN16 I think it's getting easier compared to when I first started. I didn't really have much confidence in the whole thing, especially since I hadn't had any of my own. I feel a bit more comfortable asking now than I did earlier in the peace. PN17

5.2.8 Routine: "It's just part of the deal" Finally, treatment fidelity was dependent on routines. Over time, most practice nurses became very comfortable using the Conversation Tool. Once they developed a level of comfort with the Conversation Tool tasks, they were confident in their routine use of the tasks. Later practice nurse discourse presented how they had incorporated the intervention processes into the routine of their daily work. Their conversation as seen in the boxed quotes in Figure 5.15 no longer had a sense of strain and they were more consistently adhering to the intervention protocols.

Figure 5.15 Routine intervention delivery

It's all part of the deal, that's just part of what I do with the vaccinations PN3 I've just got a routine that I work through with everybody PN4 What I've started doing is I print out my list of patients for the day and I highlight the ones that are in this category and before I call them in I have a quick look at the Tool first PN8 I've developed that pattern in my head now - whereas before you might have just said 'How's the breastfeeding going?' and not taken it any further PN10 It sort of seems a bit more natural now, you just go with it PN11 I'm starting to feel now that it's just something that I always talk about PN13 I always remember because I have the [Tool] with the vaccine side effects sheet PN14 It might actually be a bit of a habit - that's ok, even if one person out of five gets some use out of it PN14 It's now just part of the consultation when they come in PN16

206

5.3 PART 3: NEW CONVERSATIONS Part 3 explores the experience and effect of the study from the practice nurse perspective, one of the secondary objectives of this mixed method study (section 3.1.1). It draws on evidence of change from pre- and post-training paired quiz t-test analysis and presents key themes analysis of the discourse data.

Practice nursing is overwhelmingly a female profession. This chapter aims to hear these women's voices as recommended in feminist mixed methodology study. (350) In this study, the influence of motivational interviewing as a theoretical framework is seen in the practice nurses' experience of beginning and continuing new conversations. The use of an interpretive and theoretical framework can support the research outcomes of a mixed methods research project. (350)

In this process evaluation, test score analyses from a quiz conducted pre-training, and repeated four weeks after training, are presented regarding practice nurse knowledge and attitudes around sustained breastfeeding. The quiz is detailed in Chapter 3. These analyses are used to enrich the examination of the practice nurse new conversations, similar to the use of scores in other mixed methods research. (350)

5.3.1 Knowledge and attitudes quiz results The Breastfeeding Knowledge Quiz (BFKQ) and the Breastfeeding Attitude Quiz (BFAQ) were assessed, and the percentage correct provided a final score for each survey. Pre-training and post-training scores for each practice nurse were entered as repeated measures in SPSS (368) and analysed using paired t-tests to identify the significance of any changes. Table 5.3 shows the mean scores for the paired samples, and table 5.4 shows the t-test results and level of significance.

Table 5.4 Practice nurse quiz paired samples statistics

Mean N Std. Deviation Std. Error Mean Pair 1 Baseline knowledge .742 15 .084 .02156 Post-training knowledge .896 15 .066 .01710 Pair 2 Baseline attitude .798 15 .075 .01950 Post-training attitude .903 15 .067 .01726

207

Table 5.5 Practice nurse quiz paired samples test results Paired Differences 95% Confidence Std. Interval of the Sig. Std. Error Difference (2- Mean Deviation Mean Lower Upper t df tailed) Pair 1 Baseline knowledge - -.154 .082 .021 -.199 -.109 -7.27 14 .000 Post-training knowledge Pair 2 Baseline attitude - -.105 .065 .017 -.141 -.069 -6.25 14 .000 Post-training attitude

5.3.2 Knowledge and attitudes quiz summary Paired samples t-tests were conducted to evaluate the impact of the training intervention on practice nurses’ scores on the BFKQ and on the BFAQ.

There was a statistically significant increase in BFKQ scores from pre-training (M = 74.2, SD = 8.4) to post-training (M = 89.6, SD = 6.6), t 7.27(14), p<.001 (2-tailed). The mean increase in BFKQ scores was 15.4, with a 95% confidence interval ranging from 10.8 to 19.9. The eta-squared statistic (.79) indicated a large effect size. Practice nurses experienced a large increase in their knowledge of breastfeeding management, during the initial training and early practice with reflection.

There was a statistically significant increase in BFAQ scores from pre-training (M = 79.7, SD = 7.5) to post-training (M = 90.3, SD = 6.7), t 6.25(14), p<.001 (2-tailed). The mean increase in BFKQ scores was 10.5 with a 95% confidence interval ranging from 6.9 to 14.1. The eta-squared statistic (.74) indicated a large effect size. Practice nurses also experienced a large positive change in their attitudes towards breastfeeding during the initial training and early practice with reflection.

5.3.3 Thematic analysis: key themes Key themes arose from the discourse data set in the cycles of reading and re-reading, as described in Chapter 3. They related to practice nurse use of various components of the Conversation Tool and associated tasks, as shown in Table 5.6.

208

These themes provided insight to changes in practice nurse understanding and subsequent behaviours. Boxed quotes show the prevalence and characteristics of each theme across the group of practice nurses and over time with individual practice nurses.

Practice nurses underwent training that was delivered in manner consistent with the spirit of motivational interviewing, with extensive interaction, an understanding of the general practice setting and non-judgemental exploration of group views. They increased their knowledge, changed their attitudes, and changed their usual practice to incorporate the intervention processes. Training included reflective practice, breastfeeding management, communication skills, and motivational interviewing components, as discussed in Chapter 3.

Table 5.6 New conversations using the Conversation Tool

Panel Task Key themes arising from the discourse Engage Identify breastfeeding mothers "Something I wouldn't have thought of Ask about experiences before" Affirm, reflect "Using those questions"

Offer recommendations "There's a stigma out there" Elicit response "You can lead a horse to water" Assess response

Motivate Action for pre-contemplation "There's a stigma out there" (Coloured - the red box, not ready "You can lead a horse to water" boxes) Action for contemplation "You can lead a horse to water" - the orange box, ambivalence Action for committed "Something I wouldn't have thought of - the green box, ready before"

"Using those questions"

Move out Offer resources (Resources) - booklet, local resources, referral Encourage return visits

5.3.4 Resistance: "There's a stigma out there" The discourse examination in Part 1 and Part 2 mentioned the context of public opinion about breastfeeding beyond babyhood, and noted that this was a barrier to intervention delivery. In Chapter 1, the issue of public ignorance of health recommendations was

209

raised. (64) Some practice nurses were initially surprised by the power of opposition to the recommendations. The prevalence and intensity of maternal distaste challenged and shocked most of the practice nurses, whose reported experiences are shown as boxed quotes in Figure 5.16.They also struggled to understand the marked contrast between this attitude and their evidence-based health message.

The Conversation Tool instructed practice nurses to inform mothers about the breastfeeding recommendations. The text reads, "I don't know if you've heard, the recommendations are breastfeeding for at least 1 - 2 years, and not to start solids before 6 months (how would that work for you?)". Practice nurses reflected on the negative community attitudes toward breastfeeding duration that they encountered when performing this task. The theme of stigma could be seen in descriptions of how mothers responded when offered the recommendations. This theme was also seen in practice nurse comments as they interpreted how mothers were responding, or when practice nurses reflected on how they expected mothers to respond. Stigma related to expectations of breastfeeding, breasts, mothering and appropriate behaviour, as discussed in Chapter 1.

For some practice nurses, this theme reflected their own initial attitudes prior to training. In these cases, the practice nurse experience of change during training was remarkable.

Figure 5.16 "There's a stigma out there"

There's a certain stigma about that out there. PN1 ... a taboo area... PN4 I did get 'There's a stigma out there, feeding beyond 12 months.' PN8 They're very negative. PN11 You still get a little bit of shock, horror. PN11 I have had the odd raised eyebrow about that... [One mother said her family asked] 'Oh, when are you stopping that breastfeeding caper?' She was more or less saying to me 'Is it normal to want to feed this long?' PN11 I put in the one to two years and you still get a lot of laughing at that stage. PN11 The more I think about it I think 'Oh Yeah it's not that bad'. At that meeting when you said two years I thought 'Do people really DO that?!' PN12 My sister-in-law said 'That's disgusting!' I said it's just that we're not seeing it. PN12 They look at you like you're a leper when you tell them that. 'I'm not going to do THAT!' PN14 Because some of them think, 'I wouldn't be feeding my baby when she's 18 months old' because they think people will look down on them. PN14

210

One practice nurse who received only brief training (PN18) described her initial reluctance to follow the intervention processes when asked, "How do you feel about using the recommendations, telling people?" Stigma, low knowledge, or insufficient practice could be responsible for individual reluctance.

In order to persevere with the intervention tasks and to avoid a mutually distressing interaction with a mother, practice nurses used a range of strategies. Examples of these are presented in their own words in Figure 5.17. Avoiding emotional turmoil was a topic discussed repeatedly in interviews and update meetings. Discussion included the need to provide the recommendations before the mothers began early substitution, in view of the evidence of falling exclusive breastfeeding rates between age 2 and 4 months. Mentoring support involved acknowledging that continuing to breastfeed older babies is currently minority behaviour in Australia. The suggestion to one practice nurse to "Try it as permission-giving" was met with "Yeah, that might be easier, that's a good idea" (PN11).

Figure 5.17 Managing resistance

I'm a little bit standoffish because I've never done it before and I don't know anything about it apart from what you're telling me. PN18 'The recommendations are, but whatever suits you', yes, I guess I could say that and it's still getting the information out there, isn't it? PN8 I've said 'You need to do what you're comfortable with.' PN9 Usually I soften it by saying 'Certainly not like it is now, five or six times a day,' Then they kind of go 'Oh, I suppose so.' PN11 I say 'By then it's probably one or two times per day' so they're 'That's not as bad.' PN11 I try hard not to be pushy; I say 'This is what they're saying to us research-wise.' PN12

Three practice nurses did not report facing expressions of distaste. They described strategies they had developed for presenting the recommendations and their approach is seen in Figure 5.18. Perhaps one reason why PN6 received less negative feedback is because she modified the Conversation Tool recommendations, avoiding any mention of "two years". The other two practice nurses declared very clearly their non- judgemental message that they respected mothers' autonomy.

211

Figure 5.18 Avoiding resistance

I ask them do they know what the recommendations are? They'll say yes or no or they'll come up with a number, and I'll say nationally this, internationally they do this, so it's up to you, unless there is an issue and then we're going to work through it. PN5 I ask them how long they want to continue, a lot of women have the idea that you only breastfeed for 6 months. I say 'Have you ever thought of 12 months?' and they say 'Well if it works.' PN6 I actually say to them 'I haven't had a child, I'm just here to be a resource', maybe that's why they don't get defensive. PN17 Some practice nurses discussed possible alternative reasons for mothers' negative reactions to hearing the recommendations; their interpretations of the mothers' perspective are shown in Figure 5.19. These included the inability to look beyond the present or possibly unfamiliarity with normal breastfeeding management.

Not all mothers responded negatively, with particularly the mothers of 6 month old babies taking the idea on board as something novel but worthy of consideration. Some mothers with previous experience of ongoing breastfeeding were not at all troubled by the recommendations. When they had a positive reception, practice nurses said they gained satisfaction from supporting ongoing breastfeeding. One practice nurse found the message was very well received by a pregnant woman as anticipatory guidance.

Figure 5.19 Interpreting mothers' responses

These 2 monthers are just trying to get through the next few months. PN11 I think they're just a bit surprised, they think 'Will I still have milk then?' PN9 She wasn't aware of the recommendations but she was quite happy with that. PN10 Most have said 'I'm happy to continue as long as baby wants to.' PN11 She said 'It's just like getting back on a bike, really.' PN13 Talking with a pregnant mum... I linked her in with ABA, gave her the information. She was really excited and I was excited too, I enjoyed having that information to give her. She said 'the specialist gave me no guidance on anything.' I felt good after it, so I've been searching for my folder. PN4

5.3.5 New ideas: "Something I wouldn't have thought of before" Much of the work of the intervention brought with it new experiences for the practice nurses. This was strongly expressed in their reflections during the interviews, as presented in the boxed quotes in Figure 5.20. They were learning new ideas and knowledge. They continued to be open to learning across the curriculum areas covered by the training, through mentoring and update meeting learning opportunities. This new knowledge brought with it new awareness of the breastfeeding practices of mothers,

212

and new understanding of mothers' experiences and options. As the practice nurses gained new understanding, their attitudes changed, and this was an ongoing process beginning with the training workshops (demonstrated at the start of this section with the quiz results) and supported by telephone mentoring.

New skills in motivational support were developed and the discourse gives insight to this learning process. Practice nurses articulated that they were stepping into areas previously unknown to them. They saw the training and mentoring as essential facilitators for the intervention process, both in knowledge and in skills. This is consistent with the systematic literature review in Chapter 2 and with the views noted in the practice nursing literature review.

Figure 5.20 "Something I wouldn't have thought of before"

Yeah, I'll give that a go. That would work with me, easy. PN3 It's something I wouldn't have thought of before. PN4 ...and I don't like to give the incorrect information. PN5 More than half of them aren't breastfeeding... I never noticed it until we started doing this study. PN6 Yeah that's helpful, I'm glad I remembered to ask about that. PN11 I can't believe how much I didn't know. PN12 Prior to my education, I would have gone 'Oh crikey, got no idea!' PN16 That's a valid point actually, I've never thought about saying it around that way. Saying what the recommendations are but saying 'These are the options it's all up to you.' PN19

5.3.6 Skilled questioning: "Using those questions" As discussed in Chapter 3 in the motivational interviewing literature review, effective communication skills are essential for success of interventions that use a motivational interviewing framework. Workshops and ongoing contact included formal teaching of communication skills and techniques, and prompts were printed on the Conversation Tool. For most of the practice nurses, open questions and suggestions rather than advice were new approaches to interacting with their patients. In their descriptions of their learning experience, shown in the boxed quotes in Figure 5.21, the level of conscious effort this required can be seen. Again, the training and mentoring led to skills development, as an ongoing learning process. However, they were very satisfied with the results of this effort with few exceptions. Practice nurses' communication skills facilitated their conversations and connection with mothers. Finding this satisfying, they

213

began to build the questioning and the motivational interviewing approach into their work more broadly.

Figure 5.21 "Using those questions"

I'm using it with all the people that I come in contact with, I tend to use it, I use our open-ended questions and get them to talk to me, that seems to work exceptionally well. PN5 I need to train my mind into those gentle, planting seeds, open questions. PN8 My priority is, open up the feeding discussion, how are they feeding, how's it going. PN8 You know to ask an open question, you can get a little bit more. It's all about how you say things, isn't it? PN10 [She was monitoring her blood pressure at home every four hours and I was telling her this was unnecessary, and she was] initially quite defensive, looking down. [Suddenly I] remembered the motivational interviewing and I stopped and said 'I’m sorry, that came across all wrong, what I want to know is, why do you feel you need to take your blood pressure this often?’ And she just sort of changed instantly... I’ll definitely try to use it more. PN11 I'm thinking about it on a daily basis, the open questions. PN11 Like the other girls, I've been using the questioning in lots of other things I do. PN13 Using the Conversation Tool on other patients - When they give you one-word answers, you can get them to elaborate a bit more. PN15 It just took me reading back over how to ask the questions and how to explore other things, more just explore it and me having a look at how I ask, and rephrasing. PN16

5.3.7 Motivational support: "You can lead a horse to water..." The motivational interviewing approach was acceptable to practice nurses, as it fitted well within their own values of caring. The discourse analysis revealed a disapproval of insensitive or forceful actions towards mothers by health professionals, as discussed in Part 1 in the Learning context. This is consistent with the nursing value of caring, as seen in the practice nursing literature. Practice nurses wanted to help mothers to feel successful in their parenting and did not want to be confrontational about breastfeeding. In Australia, as discussed in Chapter 1, few mothers achieve the recommended breastfeeding goals and many do not achieve their own goals. Practice nurses did not want to burden struggling mothers, as discussed in Part 1. Use of the Conversation Tool was a means to offer information and support to increase breastfeeding rates, provided in a non-threatening and non-judgemental manner.

One practice nurse, who lived on a small farming property herself, described the intervention process by referring to proverb 'you can lead a horse to water but you can't make it drink'. This proverb encapsulates the concepts of the Conversation Tool. The need for support is identified and support is deliberately offered in a collaborative manner. However, there is a pragmatic acceptance that the response is then up to the individual. Attempting to force them to swallow it is pointless.

214

Practice nurses described how they had changed their approach to mothers; increasingly they supported autonomy. A dramatic change in their attitude to motivational interviewing with mothers at any of the three 'coloured boxes' stages can be seen in the boxed quotes in Figure 5.22. They now also provided breastfeeding information and motivational support to continue breastfeeding and avoid substitutes. Practice nurses were sensitive to maternal individual responses to breastfeeding information, and they supported mothers in making individualised, informed, choices.

Figure 5.22 "You can lead a horse to water..."

Often the suggested idea becomes their own. PN3 Being able to... lead them across to the positive side of breastfeeding. PN4 They take it all in; they still have got their own ideas. PN4 So it's open to them to work out what they want to do. PN5 [I had] one woman who wanted to quit and she was told she wasn't allowed to. The Maternal and Child Health [nurse] was pushy. I had to explain to this woman that it's her choice. PN7 Sometimes I think the less you speak to them the more they take in... That way it doesn't sound like a lecture, let them walk out the door thinking about it. PN9 Some people you feel like you've really go to back away, I've had a couple going back to work and they're very adamant ... You get the vibe to back off, they've made up their mind, you get the impression they're kind of sensitive about it, 'Oh yeah, I've given them the best'. You're at least suggesting it, that you don't have to give up breastfeeding... You've put the idea in their head anyway. PN11 Sometimes I think 'I may be talking to the wall here'. I look at getting the information, and then get the information to them, you know the old saying, 'You can lead a horse to water'? PN15 The first one I felt I didn’t want to be too pushy, shoving it down the throats of mothers that couldn’t, that sort of thing, and if they say they’re going to stop. So after that it just took me reading back over how to ask the questions and how to explore other things... having a look at finding ways so they don’t feel uncomfortable. I don’t want to be giving them a guilt trip. PN16

5.4 ABERRANT DISCOURSE

5.4.1 Other barriers Some later discourse featured reflections on failure of intervention delivery not due to skills or confidence but due to distance from training, " So I sort of lost my momentum in that time"(PN19). Circumstances interfered at times, "The vaccine fridge failed"(PN9), "Our vaccine fridge failed; the compressor"(PN14). Personal issues may have impacted on work performance, "My husband's... out of work, no pay. My mind has been a bit distracted"(PN8); "I was a bit tired yesterday and my brain wasn't really working at all." (PN18)

215

5.4.2 A negative ABA experience Most of the practice nurses spoke positively about the Australian Breastfeeding Association (ABA) as a resource for mothers, although reports of mothers who had established contact with ABA were rare. However, one practice nurse reported her conversation with a mother of a 4-month-old who had recently introduced a night-time bottle of formula. "'Have you heard about ABA?' She came along to antenatal classes, and it was sad to hear this: There was an older person [from ABA who spoke to the class] and within fifteen minutes, her husband squeezed her knee and said 'Let's leave' because it didn't turn out to be an encouraging talk. That didn't endear her to ABA. [After the birth, she wanted to phone ABA] Tried a few times, and it was difficult to get through and the husband said 'Why are you ringing them after the antenatal classes?'" (PN8) In this setting, the ABA resources were clearly less acceptable to the husband than to the woman herself. This mother overcame initial barriers and undertook help- seeking from ABA, but faced additional barriers. Clearly, her husband's views are significant, as he may be her primary support person, as discussed in Chapter 1. For community-based resources to provide support, they must be both accessed and accessible.

5.5 INFLUENCE OF THE RESEARCHER

5.5.1 Potential for bias This process evaluation is of course influenced by the perspective and involvement of the researcher. Any researcher is not independent from the research process. There can be no doubt that because I was the interviewer in this study collecting discourse data, there would be some influence on the data collected from my additional roles as designer of the intervention, PhD candidate, trainer, quiz assessor, and mentor.

Practice nurses were accountable to me to some extent in the tasks they were undertaking on my behalf and under my instruction. Potentially this may influence practice nurses towards wanting to give the right answers, leading to an error of falsely positive conclusions from the discourse analysis (Type I error). (366) The discourse reveals that they perceived that the trial processes and outcomes were linked to me as the researcher. On the topic of early solids recommended by a paediatrician, "And I wondered how that would affect your research, you're telling them one thing and someone else is telling them something else" (PN14). Conversely, practice nurses may work harder with patients because of their personal commitment to the researcher and

216

the research project, leading to true positive outcomes, which may be difficult to replicate. Asked if an incoming new practice nurse would be supported to have training to deliver the intervention, the reply was "Oh yes and we don't want to muck up your study" (PN13). No monitoring was conducted for these possible effects.

Practice nurse awareness of their research responsibilities increased with time, revealing a gap in the original training, regarding development of research skills. Personal commitment to supporting the researcher was revealed in a later interview, "I do ask if they're in the study because I don't want to miss them. I feel really bad because I haven't had any in the study for ages and I don't think I'm helping you with your research "(PN3). This comment also reveals that while the measured breastfeeding outcomes relate to enrolled trial participants, there was an unmeasured community effect from broader use of the Conversation Tool by practice nurses with other patients. It is possible that trial effects were greater than measured, but no monitoring was conducted of women not participating in the trial.

Consciously using the theoretical framework of motivational interviewing, I acted as a collaborator and mentor to assist the practice nurses to change their usual nursing practice and consultation style, "thanks Megan for your email, I appreciate your encouragement and support. It is just at times when we have heaps to do it is a bit hard to balance it all so calmly"(PN8). I assisted them to change their attitudes, to varying degrees. I supported the practice nurses by using motivational interviewing techniques, reflecting on their experiences, affirming their intention, effort, and achievements, offering new information and options, helping them to weigh up the steps towards using the entire Conversation Tool, and addressing barriers they encountered. Barriers were situational and not always predictable, such as new international concerns about a pandemic influenza and questions about whether mothers needed to choose between breastfeeding and immunisation. "I got your email, the one about the advice on breastfeeding and swine flu." (PN8) Such issues were only covered very briefly in the training. In my researcher role, I supported them by facilitating meetings and by sharing their stories and case examples with each other. These actions are recommended in the training of health workers to deliver a motivational interviewing intervention. (292) In this role collaboration was essential, and collaboration has the potential to create a perceived obligation for the practice nurses to report their successes more than their struggles. However, commitment of professionals to working together is associated with effective breastfeeding promotion. (369)

217

5.5.2 Evidence of data validity Evidence from the discourse, however, suggests that the record is reliable and valid, and the boxed quotes in Figure 5.23 reveal that practice nurses were describing their experiences quite freely. The discourse contains many revelations of non-adherent behaviours, at times with spontaneous declarations of truthfulness, suggesting that practice nurses did not feel compelled to talk about only positive experiences or compliant behaviours. My aim within the spirit of motivational interviewing was to emphasise practice nurse autonomy, to support them to find their own autonomous pathways. I attempted a motivational interviewing approach of drawing out the practice nurses' experiences with non-judgemental language and open questions. This approach is consistent with truthful discourse that could then be examined phenomenologically with validity. As one practice nurse commented, "I really enjoyed the sessions as well... I found that you were always relaxed and positive and it was OK to speak up and to make mistakes" (PN7).

Figure 5.23 "OK to speak up and to make mistakes"

I have forgotten some of what we went through with the course. PN1 Yes, I'll be honest, I do... PN5 Some of them I've discussed the work situation with them; I'll have to start that with the younger bubs. PN6 ...but I didn't have the Tool, I didn't even think of the Tool. PN8 To be truthful, the minute they say I'm only breastfeeding, I just leave it alone, they're doing it, they're fine...[Giving the recommendations?] Maybe not, I don't know. PN9 I probably haven't asked that too much... in the orange, I probably haven't framed it like that... PN11 Sometimes a bit forgetful. PN11 I just thought 'I'm not going there, I don't know about that.' PN14 I don't really show them particular bits [in Breastfeeding Confidence]. PN 17 I keep forgetting to use it. PN 18 I will be very honest, not that I have forgotten about it, but I had a month off... so I sort of lost momentum in that time. I'm hopeless... I feel I need more training, in all of it I guess! I know after I went [to the training], for the first few weeks after that I was quite good at it and was able to talk with people, but then... PN19

5.5.3 Support through collaboration

An illuminating event occurred at one of the first evening update meetings. I had invited a local hospital-based midwife who was also an International Board Certified Lactation Consultant (IBCLC) to build practice nurse support networks. I made a note on the dynamics I perceived at that meeting, and I subsequently raised this with one of the practice nurses (PN4) who had attended.

218

"PN4 found the IBCLC very knowledgeable and interesting, while I had been concerned about the IBCLC's tension and anxiety that the practice nurses may be doing the wrong thing and taking on too much. In the view of PN4, the practice nurses had willingly explained their role with mothers and they were quite confident with this. They saw that the IBCLC did not understand a practice nursing scope of practice. The IBCLC did not understand the PN role as 'first port of call' for many mothers. And they were not threatened by the IBCLC." I had explained to PN4 that I was concerned about the adequacy of my capacity to support the practice nurses. She replied, "You are such a good mentor for us and we learn so much out of what you say and how you answer our questions. You are so approachable." (PN4) Not fully satisfied, I questioned her again, explaining the context that "I am medically trained rather than nursing trained." To this she replied simply "You're one of us"(PN4), perhaps placing me as a fellow member of a general practice team and viewing the training as peer support rather than creating a hierarchical accountability.

This conversation illustrates the complexity of the researcher's roles in this project. Perhaps it is also reassuring that my use of the motivational interviewing approach may have created a space where the practice nurses could speak honestly of their joys and struggles in learning the processes required for the intervention. They became able to deliver a consistent intervention and in some cases to undertake behaviour changes in their usual nursing practice. "When you're taking histories you have to learn to open up your communication that way. So you were effective in lots of ways Megan, not just the breastfeeding!" (PN10)

5.6 SUMMARY OF CHAPTER 5 Practice nurses in this study described how they were generally able to change their conversation style with their patients. The trans-theoretical model of behaviour change, which undergirds the motivational interviewing approach, anticipates that behaviour change is challenging but possible with support. (292) With training and mentoring, most practice nurses achieved a level of comfort with and a routine use of the Conversation Tool. In this, they were able to deliver the intervention according to protocol, more often than not, and increasingly over the sixteen months of the trial.

Barriers to delivery of the intervention described by practice nurses included the opposition of the broader community to ongoing breastfeeding, a relatively slow

219

process of skill development, the context of mothers' busy workloads and at times, frustrations related to the structures of their own general practice workplace. In a Dutch cluster randomised controlled trial of early childhood nurses delivering breastfeeding support in a health-counselling model, 56% found it "(fairly) difficult to carry out" and 44% reported that they had longer consultations. (261)

As seen in Chapter 2, training in both breastfeeding management and communication skills are components of major successful breastfeeding interventions. Some practice nurses already had personal breastfeeding experience. However, as a group they were able to demonstrate increased knowledge, shown by the paired t test results, t 7.27(14), p=.000 (2-tailed) and more positive attitudes t 6.25(14), p=.000 (2-tailed). This finding concurs with a UK study that compared the breastfeeding knowledge of practising midwives before and after completing the 20-hour WHO/UNICEF Breastfeeding Management Course, using the pre-validated Breastfeeding Support Skills Tool. (370)

The role and functionality of the Australian Breastfeeding Association mother-to-mother support merits further study, as seen from the aberrant discourse. However ABA printed resources were welcomed by both practice nurses and mothers.

From the perspective of these nineteen practice nurses, particularly the initial 15, delivery of the intervention was seen as acceptable once they became fluent in its use and adapted it to their own setting. They gained satisfaction from the use of their knowledge and skills in helping their patients. They reported that their patients responded occasionally with disinterest but often with pleasure in the conversations. They did not report any patients who were angered or distressed during their conversations. Where they perceived a 'red box' reaction, practice nurses had the skills to move on to other topics and sufficient respect for their patients to avoid being pushy. A significant number of patients received support through the work of these nurses, and this process evaluation supports the positive outcome analysis from the randomised controlled trial in Chapter 4. Evaluation of the process behind the conduct of the research indicates that overall it was a feasible, safe, and satisfactorily conducted trial, from the perspective of the practice nurses.

220

Chapter 6

Discussion and Recommendations

221

INTRODUCTION TO CHAPTER 6 DISCUSSION AND RECOMMENDATIONS This thesis contributes to the field of breastfeeding research by demonstrating the role of primary care health professionals in Australia in increasing breastfeeding rates. It is based on international research showing the need for breastfeeding and counselling knowledge and skills in order to achieve these goals. In this thesis a barely tested concept, that of using motivational interviewing to support breastfeeding mothers to maintain breastfeeding, has been supported with evidence from a randomised controlled trial.

The study described in this thesis achieved one of the primary objectives listed in Chapter 3, (See 3.1.1) and all of the secondary objectives. As a scientific endeavour using randomised controlled trial methodology, this study was able to demonstrate a positive outcome. There was a statistically significant increase (>10%) in exclusive breastfeeding at 4 months, compared to controls.

There was no significant difference in exclusive breastfeeding at 6 months, or any breastfeeding at 4 months or 6 months, between intervention and controls.

This thesis also describes the achievement of the secondary objectives also listed in Section 3.1.1, namely identifying any relevant confounder (the need to adjust for pre- existing work or study plans) and comparison of results according to three alternate definitions of exclusive breastfeeding(exclusive breastfeeding 24-hour recall, full breastfeeding 24-hour recall and exclusive breastfeeding since birth recall). In process evaluation, which used mixed methodology, the intervention was found to be sufficiently feasible to build into usual routines. Practice nurses demonstrated increasing intervention fidelity through the period of the trial, and openly described their own experience of undertaking a new behaviour.

6.1 DISCUSSION OF MAIN FINDINGS

6.1.1 Exclusivity

Consistent with other health professional interventions, this study demonstrated improved outcomes for exclusive breastfeeding. For convenience, Figure 1.1 is shown again on the next page. By definition, mothers who were influenced by the intervention were more likely breastfeed every time they felt their baby needed food, drink or comfort, rather than to offer formula or solids or even water, compared to mothers in

222

the control group, at 4 months. It is likely that one mechanism for this success is increased maternal confidence and knowledge, as the most common reason that mothers give for introducing substitutes is a perceived low milk supply. (148, 371) Based on the evidence presented in chapter 1, likely sequelae of this change are fewer infant infections (2, 19) and for some women, delayed return of fertility, (24-26) although health outcomes were not assessed in this thesis.

Societal response to breastfeeding concerns is to use formula or to defer to health professionals; often these presumed experts have little understanding of breastfeeding as a dynamic and individualised relationship. (20) As detailed in Chapter 1, societal factors place many challenges before mothers who wish to continue breastfeeding. Any attempts to provide support must consider these challenges By contrast, in this study, increased breastfeeding support was available from trained health professionals. Practice nurse training aimed to increase awareness and consideration of challenges faced by mothers, and some of the practice nurse resources were specifically directed at helping mothers to deal with these challenges. (See 3.2.4)

Figure 1.1 Definitions of breastfeeding

May also be fed: Definition:

Medicines Exclusive Full Any breastfeeding breastfeeding breastfeeding Water- based fluids

Formula or other milk

Spoon-fed or hand-held solids

As Chapter 2 proposed, a successful increase in breastfeeding rates is more likely from interventions that provide breastfeeding and counselling training, particularly within a counselling theoretical framework. In this study practice nurses were thus equipped to support mothers to gain knowledge and confidence in breastfeeding, as the most suitable feeding method in the first 6 months, and as of continuing importance.

National targets have been proposed for full breastfeeding. In 1993, a 3-month target of 60% was set. (70)(p211) Available evidence indicates that population rates for full breastfeeding are close to this target, with 57% were fully breastfeeding. (52) In this

223

study, the 4-month full breastfeeding rate was 71% in the intervention group and 60% in the control group. (Adjusted, see Table 4.9) Hence, the fall in full breastfeeding rate over time was less steep for this study's control group than for the Australian population, and much less steep for the intervention group. Outcomes for exclusive breastfeeding showed a similar pattern. This study demonstrates that 4-month exclusive breastfeeding rates can be increased above the targets in similar populations to this study, with appropriate additional support. These findings need to be broadly communicated. (372) Targets for 3-month full breastfeeding are achievable.

National targets for 6-month breastfeeding rates, using various definitions of exclusivity, are more controversial. In 1993, a 6-month full breastfeeding target of 50% was set (to be achieved by the year 2000). (70) Ten years later, in 2003, this same target was reset (to be achieved by the year 2006) (30)(p307) with a new proposal for a 6-month full breastfeeding target of 80% (to be achieved by the year 2013). Twenty years later we are not even close to the original target.

Measurement points have been moved for clarity, because the introduction of solids "around 6 months" ends exclusive breastfeeding without necessarily indicating any substitution of breastfeeds with formula. In the 2010 ANHS, the population 5-month full breastfeeding rate was 36%. (52) This is not simply an Australian issue, as the 5-month full breastfeeding rate was 17% in Norway, as discussed in Chapter 1. (67, 373) More research is needed before any 5-month or 6-month full breastfeeding target can be identified, so that it can be of some value.

Recommendations 1. Disseminate trial findings to local and regional communities and to national bodies. 2. Introduce additional breastfeeding support for all mothers in order to increase exclusive and full breastfeeding at 5 months, in accordance with new indicators in the Australian National Breastfeeding Strategy 2010-2015. (212) 3. Explore reasons for the low rates of exclusive breastfeeding in developed countries at 6 months.

6.1.2 Employment The initial results in this study were not significant, until adjustment was made for mothers' plans for returning to work/study. There was initial confounding by this variable. This study makes a new contribution in finding that employment planning is

224

associated with breastfeeding outcomes. This finding demonstrates the importance of controlling or adjusting for maternal work/study intentions in the design of breastfeeding interventions.

Employment planning may influence the help-seeking behaviours of breastfeeding women, and the support they subsequently receive. Intention to return to paid employment before 6 months, compared to before 12 months, was significantly associated with attendance at breastfeeding support groups in one study (p=0.02) with predominantly white, middle income women in the USA. Attendance by these women at support groups compared to non-attendance was significantly associated with a mother meeting her own breastfeeding goals, 81% vs. 51% (p=0.04), although numbers were small (n=68). (374)

Employment plans that are described in the second or third trimester of pregnancy may change after the birth following the experience of mothering. This thesis did not re- examine mothers' employment planning after the birth. Wilhelm et al. found that return to work by 6 months was significantly lower than originally planned among intervention group mothers, 83% planned and 49% returned to work. However return to work by 6 months was the same as planned in the control group, 86% then 85%, which was significant despite the small sample (n=62),  2 (1) = 10.02; p = .002. (186) This difference may be influenced by pre-existing breastfeeding self-efficacy or pre-existing level of work commitment, which merited adjustment in this study. There may be an interaction between the intervention and employment planning by these women. However, women who have financial pressure to return to paid work may not have the flexibility to change their plans, as indicated by a large UK study relating maternal income to timing of return to work. (89)

Actual maternal employment was not assessed in this thesis, so it is unknown whether this finding reflects the impact of actual employment on breastfeeding outcomes. An Australian trial described in Chapter 2 measured intention to return to work before 6 months but this was equally distributed by randomisation. (255) Other studies have demonstrated an association between entering employment and complete or partial cessation of breastfeeding, (65, 91, 341) particularly if this employment is full-time and commences soon after the birth. The nature of employment is associated with breastfeeding duration, even in an advantaged population. In a large Australian rural study, women who gave birth at a private hospital were more likely to be exclusively

225

breastfeeding at one month 1.4 (95% CI = 1.0–2.0) if prior to the birth they worked in professional-type employment compared to other forms of employment. (375)

Recommendations 4. Increase community awareness of breastfeeding protections and strategies for mothers planning paid employment outside the home. 5. Provide additional support to pregnant and breastfeeding women who are planning to return to paid employment/study after the birth, in order to increase exclusive and full breastfeeding at 5 months.

6.1.3 Duration In this study there was no significant difference in any breastfeeding at 4 months or 6 months, between intervention and controls. Low rates of ongoing or continuation of breastfeeding in Australia, measured as breastfeeding duration or any breastfeeding, need to be addressed.

Due to the choice of methodology, in this study the 2-month any breastfeeding rate was 100%, clearly that of a population sub-group. In the Australian population, by comparison, the highest 2-month any breastfeeding rate was 85%, being for women with the education level of bachelor degree or higher. (69) These women comprised 53% of this study population.

Progressively between 2 months and 6 months, one fifth of the mothers in this study weaned their babies completely onto formula feeding. In this study the 6-month any breastfeeding rate was 79%. In the Australian population, by comparison, breastfeeding rates fell from 90% at birth so that the 6-month any breastfeeding rate was 60% in the 2010 ANHS overall (69) and 73% among mothers with a bachelor degree or higher. (See Table 4.2)

National targets have been set for breastfeeding duration. In 1993, 3-month and 6- month any breastfeeding targets of 80% were set (to be achieved by the year 2000). (70)(p211) However, breastfeeding declines steeply in the first month, with 15% of mothers not continuing. With a more gradual decline, another 15% have weaned completely onto formula by 6 months leaving the 60% 6-month any breastfeeding rate far below the 6-month target of 80%. (70)(p211) Evidence from earlier ANHS studies showed that by 12 months, one quarter of Australian mothers had already ceased both

226

breastfeeding and formula, relying instead on unmodified cow's milk which increases the risk of iron-deficiency. (376, 377)

Evidence for the importance of sustained breastfeeding beyond 12 months and sustained exclusive breastfeeding to 6 months is provided by systematic reviews (2, 45, 201, 205) and detailed in Chapter 1. Health outcomes are identified in some of the intervention trials described in Chapter 2 including gastroenteritis, eczema, (17) dental caries, (248) and respiratory morbidity. (239) The group of infants for whom breastfeeding ceases prematurely has a greater burden of disease and lifelong disadvantage. Examination of breastfeeding duration over the three sequential national surveys demonstrated a widening divide between the highest and lowest socio- economic groups over ten years. Progressively, breastfeeding duration measured at 3 months, 6 months, and 12 months has fallen in the lowest group from 1995 to 2005, that social determinants of chronic disease (378) may in part be moderated through infant nutrition. (66)

Possibly this study did not demonstrate change in any breastfeeding duration (any breastfeeding) at 4 or 6 months in the intervention compared to the control group because of the challenge of demonstrating increased duration in a setting of high initiation rates. (204) Breastfeeding duration was high for both groups and both groups already met the Australian targets for duration. (70)(p211)

In this study, comparison of breastfeeding duration to maternal goals revealed a discrepancy at 4 months. Of the 330 women in the study, thirteen planned to wean by about 4 months, but 39 had actually weaned at 4 months (Section 4.8.3). The reasons were not explored. Others have noted a strong association between intention and actual duration. (156, 379) It is possible that these women needed more support between the birth and 4 months, and it is unknown whether they received the intervention. Assessment and control of treatment fidelity were not conducted quantitatively; possibly these women did not receive the intervention. (294)

Qualitative analysis in Chapter 5 indicated that with adaptation the conversation could be less confronting about duration. The Conversation Tool was revised after the intervention and Version 3 (see Appendices) needs to be tested; it may influence breastfeeding duration.

227

Larger postnatal-only interventions examined in Chapter 2 that increased breastfeeding duration were more extensive than in this study, using a greater number of contacts or in some, multiple home visits. Of the twelve studies that increased duration, only three, set in Canada, Denmark and France reported improved duration with less than three intervention contacts in the first six months after the birth. (241, 247, 252) Other successful studies used a greater number of contacts with mothers, a comprehensively supportive health service, or used peer-counsellor involvement. (See Chapter 2)

Recommendations 6. Trial an adapted Conversation Tool intervention, with processes to improve and monitor receipt of the intervention, in a similar population, in order to increase duration of breastfeeding at 6 and 12 months (current practice). 7. Trial and evaluate the Conversation Tool intervention with relatively disadvantaged women. 8. Promote in Australia moves towards comprehensively supportive primary care health services, and evaluate primary care services in accordance with the Baby Friendly Seven Point Plan. 9. Explore novel approaches to increase the access of all breastfeeding women to established and skilled peer support.

6.1.4 Opposition to ongoing breastfeeding Aberrant but pervasive societal opposition met many attempts to provide support for ongoing breastfeeding. There was a tone of disapproval to the questioning of mothers about "still breastfeeding". One practice nurse reported that a mother was asked by her family, "Oh, when are you stopping that breastfeeding caper?" (Figure 5.16) Negative community reactions to breastfeeding based on the age of the child are influential even at the level of health policy, as discussed in Chapter 1. (42) Promotion of breastfeeding until one or two years of age, consistent with the recommendations, is hence contrary to cultural expectations and beliefs. This became one of the key themes in the practice nurse discourse, "There's a stigma out there" and as one practice nurse observed, "a lot of women have the idea that you only breastfeed for 6 months." (Figure 5.18)

Disapproval also relates to the length of an individual feed or to the frequency of breastfeeds, from those who are unaware of normal lactation management. As one practice nurse reported (Figure 5.3)

228

The other comment undermining her was ... she was criticised for feeding it too much or too often. It wasn't that often. I said 'How do you feel about it, do you feel it's too much?' and she said 'No.'

Again, this may be due to concern for the welfare of the mother who is considered to be working unnecessarily hard in her method of feeding, or may be a concern about whether the infant is adequately nourished, as another practice nurse reported. (Figure 5.3)

I also find that women are pressured to stop at 2 months, three months; they hear 'If they're crying you've run out of milk.'

Here the discourse quickly moves to a question about enough milk, and encourages the use of supplements "they were told to put the baby onto the bottle by grandparents" (Figure 5.2). Bottles of formula are suggested because the infant is thought to need more nourishment than the mother's perceived inadequate milk. Another practice nurse reported the opinions of grandmothers when the perceived needs of the mother for rest or independence are considered more important than breastfeeding, (Figure 5.3)

The children haven't been sleeping - 'Oh, put 'em on the bottle'. They're mostly younger mothers.

Possibly this stigma may be resisted when a mother is encouraged to articulate how she values her breastfeeding relationship and when she is supported with individualised strategies to overcome structural barriers. There is room to improve breastfeeding duration in the broader Australian population, particularly among disadvantaged women. (66)(p256)

Women from lower-income families are less likely to breastfeed for a number of reasons, including less family support for breastfeeding, less ability to seek help with breastfeeding problems, less flexibility with working arrangements, and concerns about breastfeeding in public.

Adequate support for ongoing breastfeeding, particularly for the underprivileged, (148) necessitates improved access to skilled support from health and peer organisations, improved breastfeeding protections for employed mothers, (89, 91-93, 319, 380) improved breastfeeding knowledge of families (157) and improved community attitudes to breastfeeding in public. (381, 382)

Recommendations 10. Improve Australian family and community knowledge of the importance and management of breastfeeding beyond 6 or 12 months.

229

11. Address comprehensive health promotion campaigns towards community attitudes on breastfeeding in public. 12. Improve measures to protect the breastfeeding relationship for women in out-of- home employment.

6.1.5 Motivational Interviewing theoretical framework Based on the literature, this appears to be the first randomised controlled trial to demonstrate improved breastfeeding outcomes using a motivational interviewing approach.

Motivation and success in sustaining breastfeeding are associated with whether the mother has a positive past breastfeeding history. (383) Randomisation accounted for this influence in this trial. Some psychosocial characteristics are modifiable though motivational interviewing, particularly intentions. These are more strongly associated with breastfeeding outcomes than demographic characteristics. (156) Goal-setting, active problem-solving and challenging unhelpful beliefs are strategies described by women as helping them to continue breastfeeding, in a Queensland mixed methods study. (384) These are strategies incorporated in the motivational interviewing approach, varying according to the mother's motivational response (the 'coloured boxes'). While confrontational or more standardised methods may make it harder to talk about breastfeeding, the MI approach enables conversations about behaviours.

As described in Chapter 3 (3.2.1), a breastfeeding motivational intervention with disadvantaged women in the USA showed a trend towards increased breastfeeding duration but this was not statistically significant. The study was conducted with a small convenience sample (n= 73).in a two-group design, and was not randomised. The intervention acknowledged ambivalence using a video of mothers' stories about importance and barriers, and mothers were offered conversations with MI trained staff. (299)

In this thesis, practice nurses found that working in the spirit of MI enabled them to support mothers even in initially challenging conversations (Figure 5.13).

Some mums who are half way in between breastfeeding and bottle feeding, they start off defensive, you can see that, but you soon put them at ease and then they get really chatty

230

Wilhelm et al. tested the use of MI in the USA to address the ambivalence and support needs of rural breastfeeding women. (186) Participants came from a similar socio- economic background to those in this thesis. Contacts were in the first week, at two weeks and at six weeks after the birth. All mothers received a conversation at each time point with a lactation consultant, who did not have MI training. Intervention mothers also received a motivational interview with a research nurse, whose MI training may not have been interactive and who did not receive breastfeeding training.

The intervention by Wilhelm et al. used a complex flowchart and focused on responses to ratings scales. (186) It may have provoked resistance in mothers who were planning to stop breastfeeding, and it did not involve anticipatory guidance for mothers who were confidently continuing breastfeeding. Protocol at contacts included routine test weights before and after breastfeeds, a technique prone to inaccuracy and to raising maternal anxiety. (6)(p77)

The authors concluded that MI strategies fit well with a nursing emphasis on client- focused decision-making and mutual goal setting and were worthy of further research. (186) In this thesis, the discourse analysis of practice nurses using the Conversation Tool, who received both MI and breastfeeding management training and who conducted responsive conversations with mothers, concurs with this conclusion.

Evidence is building for the routine incorporation of a motivational interviewing approach into breastfeeding support. The review of motivational interviewing studies conducted with parents in Chapter 3 (See 3.2.1) identified eight studies, five with successful nutrition-related outcomes.

Recommendations 13. Provide motivational interviewing training and a period of mentoring to all primary care health professionals who have regular contact with breastfeeding mothers. 14. Explore the use of an extension of the Conversation Tool intervention to pregnant women as a means of influencing breastfeeding intention and potentially duration.

6.1.6 Communication training The development of communication skills through training has an influence on the success of interventions. Chapter 2 revealed that improved breastfeeding outcomes

231

were associated with interventions provided by multi-skilled professionals, knowledgeable about breastfeeding management and using a person-centred approach. These components were common to successful interventions in countries as disparate as Thailand, (242) Italy, (250) The Netherlands (251) and Australia (246, 268) and with women from both disadvantaged (239) and well-resourced backgrounds. (243) Perhaps the key to effectiveness in these diverse settings was the common feature of an approach based on listening to the individual and individualising the message while maintaining the same goal orientation. (1) Health professionals may have breastfeeding knowledge, but a less collaborative working style, even with the same material, can reduce the intervention's effect on breastfeeding outcomes. (369)

The qualitative literature and grey literature explored in Chapter 1 presented breastfeeding as a complex, individualised, paired (with singleton births), and interdependent human behaviour. This is acknowledged by the BFHI training materials described in Chapter 3. There is a good fit between BFHI materials and the theoretical framework of motivational interviewing, as both of these are person-centred and goal- oriented. Both acknowledge the need for timely, empathic, and skilled support. Both are well founded on international evidence, suggesting that they are adaptable to the needs of diverse cultures.

Training is needed in order to increase breastfeeding support from primary care health professionals; a variety of professionals has regular contact with breastfeeding mothers. In Chapter 1, the practice nursing literature review revealed that practice nurses were interested in behaviour interventions for preventive health; however, they both wanted and needed additional training to conduct them. In Chapter 5, the influence of the researcher was examined. Consistent with the description in Chapter 3, training and mentoring of the practice nurses was conducted within a motivational interviewing approach, to enhance their adherence to the intervention processes. Researchers introducing a behavioural intervention need to be able to mentor professionals where they find new processes frustrating, or when they find it puts them into a place where they feel unqualified to deal with the patient. (214)

More support is needed so that staff can have access to such training. (135) Training enables systems change, "one of the easiest elements of successful health promotion to conceive, but the most difficult to implement." (385) Relatively few nurse communication-training programs are evaluated. The 20-hour Breastfeeding Management Course, organised by BFHI teams, was shown to be successful in

232

teaching midwives to support breastfeeding mothers, using a validated instrument. (370) A review of fourteen studies found that, overall, training effects on nurse skill and behaviour were limited, and patient outcomes were mostly unchanged. (386) The authors noted the importance of process variables including the characteristics of the relationship with the teacher and among participants. (386)(p130)

Nurses often prefer warm, enthusiastic, friendly teachers with a great deal of expertise. Reciprocal sympathy and support, respect and feedback among participants may contribute to the success of the learning process.

Two studies found positive effects on nurses' behavioural change, and teachers in both of these studies had a nursing background. One of these stressed the importance of giving participant feedback in a "safe environment", with feedback about their good points first to reduce the risk of undermining confidence. (Kihlgren et al., Faulkner and Maguire, cited (386)(p141-142)) It is relevant that of these largely ineffective training programs, none considered non-verbal communication. Studies that used pre- and post-test mostly did so immediately before and after training, however, as identified from the practice nurse discourse data in this study, (386)(p143)

...it may take some time before the newly acquired or modified behaviours are integrated into the daily routine.

Promotion of breastfeeding that uses information about breastfeeding management rather than using imperatives is likely to increase the confidence of women. This was observed by one practice nurse in a conversation about early solids that were started on a grandmother's advice (Section 5.1.8; Figure 5.6).

The doctor came in and said 'Don't do that!' He just said that and than he left....

I said to her 'your baby doesn't really need that, he only needs breastmilk, it's about his development,' and she said yes he wasn't really eating it, he was pushing it about with his tongue, and I said 'that's right'

Information that acknowledges women's embodied life with their baby and their current achievements is likely to increase the perceived importance of breastfeeding for women. (299)

233

Recommendation 15. Fund and encourage uptake of communication skills training for all primary health care professionals who have regular contact with breastfeeding mothers, consistent with the component of the BFHI 20-hour course used in this study.

6.2 DISCUSSION OF OTHER FINDINGS

6.2.1 Definitions This study explored the role of three definitions used for "exclusive" breastfeeding: 1) No other food or drink in the past 24 hours (current practice). 2) No other food or milk in the past 24 hours (current practice). 3) No other food or drink in the past 4 / 6 months (retrospective recall since birth). Significant outcomes were detected for the first two definitions, at 4 months, compared to the control group. They provided evidence that the intervention could significantly influence the behaviours of breastfeeding women. This is one goal of breastfeeding interventions and implies improvement in a range of dose-dependent health outcomes compared to control groups.

No change was detected in the third definition at 4 or 6 months, compared to the control group. The implication of this finding is that levels of atopy would be similar in both groups, all other influences being equivalent.

This third definition was not sensitive enough to detect a significant outcome from the influence of the intervention on mothers' behaviours. Mothers were not asked the timing of any distant past substitute use. This definition may be affected by use of formula in maternity wards, which in some hospitals is quite prevalent and is specifically addressed by the BFHI.

The 6-month exclusive breastfeeding rates were quite low using all three definitions, as discussed in Chapter 4 and there was no difference between intervention and control groups. The first two definitions at 6 months were similar to each other and to Australian population data. Use of the third definition would halve the outcome.

Comparisons cannot accurately be made between results from this study and results from the Australian National Health Surveys due to heterogeneity in definitions. The ANHS excluded infants from the category of full breastfeeding if they were fed solids or substitute milks "regularly", whereas in this study infants were excluded if they were fed

234

substitutes in the previous 24 hours. Unlike the clarity of Canadian data collection, which specifies a seven-day recall, (55) the term "regular" was open to varying interpretation by respondents. (42)

Definitions used for breastfeeding outcome measures are the subject of current debate at state and federal levels in Australia, and national leadership is needed. Monitoring of breastfeeding rates is part of the National Breastfeeding Strategy. (212) Proposed Australian national indicators include the third definition, (measured at each month of age, 0-6 months). Use of this indicator in population surveys will either lead to less accuracy (more distant recall) or more cost (greater numbers of infants under 6 months to be surveyed). (387) Collection of other dietary intake is proposed using 24-hour recall for breastmilk, solids and formula. Some areas of Australia have more culturally or linguistically diversity, and their use of teas or syrups would be undetected with these measures. However, an outcome for "full breastfeeding" (24-hour recall) could be derived which would enable comparisons with results from this study. Exclusive breastfeeding in the past Australian National Health Surveys, South Wales and Queensland state surveys used the 24-hour recall definition, as in this study and as recommended by the World Health Organisation. (56)

Recommendation 16. Create a central body to provide cross-disciplinary breastfeeding research leadership and to give oversight to the National Breastfeeding Strategy.

6.2.2 Use of formula Use of formula by 28% of breastfeeding mothers at 6 months in this high socio- economic group, as described in Chapter 4 (See 4.8.3) implies a greater reliance on formula in the broader Australian population as breastfeeding rates are lower among disadvantaged women. (66) It is interesting that in this highly motivated and educated group, successfully continuing to breastfeed at 6 months, use of formula is so common. This merits further research into reasons for and intensity of formula use, particularly in view of the dose-dependent risks of a range of poor health outcomes with reduced breastfeeding.

Practice nurse discourse reflected broad community acceptance of formula substitution, with reports of opinions that "the bottle" was seen as a better choice for mothers than breastfeeding as shown in the comments in Figure 6.1.

235

Figure 6.1 Giving the bottle

She felt the baby wasn’t getting enough milk so she put him straight on the bottle after that. (Figure 5.6) You give the baby a bottle and off you go. (Section 5.1.4) I asked why the baby was having bottles, and the husband wants to give a bottle at night. (Figure 5.5) They told her it 'shouldn’t be feeding overnight' and the sister’s babies 'all slept through the night and they were bottle fed.' (Figure 5.2) And then they just have one formula [bottle] at night. (Figure 5.11) Conversation with a mother of a 4 month old who had recently introduced a night-time bottle of formula... (Section 5.4.2)

A recent USA risk-focused media campaign to promote breastfeeding was criticised as exaggerated, moralising and as failing to address human rights issues. (388- 390)(p376)

Risk-based campaigns increase our responsibility to ensure that ... our words and actions do not reflect insensitivity to the very real constraints [women] face.

Media campaigns cannot stand alone, as this breastfeeding advocate pointed out. In view of the need in Australia for broader support of ongoing breastfeeding, effective restriction of formula marketing is needed, according to the WHO Code. (81, 391, 392)

Recommendations 17. Explore in depth the factors relating to reliance on formula by breastfeeding mothers in relatively privileged groups. 18. Educate health professionals about the risks of reliance on formula over breastfeeding. 19. Legislate to restrict approaches to the marketing of formula that minimise the health risks of weaning earlier than recommended.

6.2.3 Knowledge of breastfeeding management Practice nurses demonstrated positive changes in knowledge, confidence, and attitudes following training. One study with dieticians on communication with diverse clients identified that MI training and skill development assisted with nutrition interventions and was associated with more positive self-concept on the part of the dieticians and more confidence in communicating with diverse clients. (393) Skill development in this study happened over time and with mentoring support, as demonstrated in the practice nurse discourse. (See Chapter 5 Part 2)

236

Practice nurses frequently encountered ignorance about the processes of breastfeeding in the broader community. At times mothers followed the non-evidence- based medical advice of previous generations, making decisions based on hearsay, and conceptualising breastmilk as a time-limited commodity. These mothers were aware of health messages about the importance of breastfeeding, but unaware of methods or reasons to sustain breastfeeding. As one practice nurse observed, "A lot of them don't know, they think there's no reason to continue beyond 12 months." (Figure 5.3)

While individual components of the intervention cannot be separately assessed for their contribution to the trial outcomes, findings of this study concur with reviews that increased exclusive and full breastfeeding at 4 months may be achieved by providing one-to-one, informed support environment to mothers. (45, 318)

Recommendation 20. Fund and encourage uptake of breastfeeding management training for all primary health care professionals who have regular contact with breastfeeding mothers, consistent with the BFHI 8-hour training package.

6.2.4 New practice nursing behaviours The conversation of practice nurses underwent change over time, suggesting that increased awareness and ongoing attitude change could overcome cultural stigma. This demonstrated adaptability is an asset to delivery of primary care, in the context of complex professional work. The contribution of practice nurses to primary care is insufficiently appreciated. (394)

The process of personal change is illustrated by one practice nurse. Her initial reaction to ongoing breastfeeding was shock and ignorance, "Do people really DO that!?!" (PN12) Following the training workshops and her own reflection, she found herself challenging the stigma, "The more I think about it I think 'Oh yeah, it's not that bad.'" (PN12) Subsequently she defended sustained breastfeeding when speaking with a family member. "My sister-in-law said 'That's disgusting!' I said it's just that we're not seeing it". (PN12) The use of awareness training and repeated cycles of mentoring and reflection helped this practice nurse to achieve her own change, a key imperative for using reflective practice particularly when carefully supported. (332) Use of MI interventions, particularly with the opportunity for repeat contacts, sustained this change, consistent with the MI literature review in Chapter 3.

237

Most practice nurses encountered one particular barrier to changing their conversation to follow the Conversation Tool. Those who followed the instructions and presented the recommendations as a stand alone concept for response, tended to face immediate, instinctive revulsion "They look at you like you're a leper". (PN14) However, those practice nurses who allowed for initial personal reflection by mothers on the concept of ongoing breastfeeding "I ask them do they know what the recommendations are?" (PN5) did not report facing this reflex distaste when they gave the recommendations. Their spontaneous adaptation of the Conversation Tool was a positive indicator of practice nurses making the intervention their own, a marker of intervention feasibility and of their own effective change in behaviour (See section 6.3.2 below). For the other practice nurses, an iterative approach was used to overcome this barrier, with a shift from an advice-giving tone to a permission-giving tone. This shift enabled practice nurses to become more comfortable and to build the intervention into their routines. Mentoring and building collaborative networks acts to support new behaviours. (352)

More recent breastfeeding studies have identified the importance of psychosocial characteristics, above that of demographic characteristics. (156) Earlier breastfeeding studies found demographic characteristics to be associated with breastfeeding outcomes.

Recommendation 21. Build reflective practice activities, with opportunity for supportive feedback, into professional training programs to support ongoing breastfeeding. 22. Recognise the key role of practice nurses within the team, in promoting positive patient outcomes.

6.3 STRENGTHS

6.3.1 Mixed methodology and triangulation

This study used mixed method process evaluation, with a form of triangulation, where different views of the trial together may provided a more valid answer to the research question than each provided individually. (395) This study aimed to create change, not only in a research setting but also in reality for breastfeeding mothers walking off the street and into their general practitioner's rooms. Use of multiple research methods made a valuable contribution to understanding both the outcome and the process of the trial.

238

As mentioned in Chapter 3 (Section 3.4.1), some question the validity of this approach, (349)(p278) while others consider this approach pragmatic and appropriate. (348) The qualitative documentation of gradual practice nurse skill development and of eventual routine use of the intervention make a valuable contribution to research findings, confirming the need for training and mentoring to accompany use of the Conversation Tool for successful results. According to an analysis of ten health promotion interventions in primary care, research design needs to foster evaluation, evaluate well, and evaluate the process of achieving the outcomes. (396)

6.3.2 Clinical translational research The trial provides strong support for the efficacy of the Conversation Tool intervention. The increase in exclusive breastfeeding and full breastfeeding at 4 months compared to control was demonstrated using an intention to treat analysis. Results were achieved in a real world setting, taking into account the intention of intervention practice nurses to provide breastfeeding support to each trial participant. There was no per protocol analysis of the effect of the intervention itself on each participant.

Locating this study in existing primary health care services is a particular strength, as it makes this a trial of a program that is directly replicable at the service level. (268) Practice nurses had significant freedom to determine how to deliver the intervention. (397) Interventions delivered by autonomous professionals within existing infrastructure and offered to all women, when not adding too much time to the patient's visit or too much burden to the staff, (396) are more likely to be translated into daily practice. (214) For effective translation, interventions need capacity to be individualised by different general practices. (396) In translational research, there needs to be opportunity for feedback and fine-tuning (214); modifications may help the integration of the intervention into practice, (287, 396) as demonstrated by the practice nurse discourse around how to present the breastfeeding recommendations (See section 6.1.4, above).

Consistent with findings from others, the intervention in this study is nurse-initiated. Other attempts to build an intervention into routine practice have identified that patients do not take the initiative. (158, 159, 396) Moreover, patient priorities and perspectives on valuable uses of appointment time periods have been found to differ from those of evidence-based professionals or researchers. (396)

Chapter 5 revealed a moderate level of staff turnover among practice nurses, as others have observed (336, 398) which is likely to be associated with some loss of treatment

239

provision to the intervention group. This is a reality in primary care trials using existing staff, and an important point when considering translation of research outcomes into practice settings. Chapter 5 also revealed that adherence was incomplete in the early stages of the trial, as the practice nurses experienced a learning curve, which has been reported in another primary care based breastfeeding intervention trial. (214) Other trials have addressed this issue using greater pre-trial training. (17, 243) Despite the likelihood that some mothers received minimal or no intervention, the overall outcome was significant, which supports the use of this intervention in other general practices. For those mothers who did receive the intervention, the intervention may be more effective than indicated from reported results.

Translational research is challenging but important and needs to be funded in order to achieve the results promised by research insights. Results may be small but very important at the population level. (399) Processes used in trials need to have effective outcomes, be translated into practice, and be sustained for ongoing population benefit. (400)

6.3.3 Generalisation from results Results were collected from women who gave birth at three hospitals, across two geographic regions, including both public and private patients. These women came from a range of family sizes, with only one third of them breastfeeding their first baby, and hence had a wide range of breastfeeding experiences and motivation. (383) Other than the requirement that women attending the recruited practices be breastfeeding at 8 weeks (the trial was targeted at breastfeeding mothers of infants being immunised) there were few exclusions. These factors increase the likelihood that participants may be representative of rural women in other regions, and that findings from practice nurse delivery of the Conversation Tool intervention may be applicable to a general population.

This study was set in small rural cities, and applicability to an urban or remote setting cannot be assumed. Study results are relevant to approximately 20% of Australians, who live outside major metropolitan centres, in areas where access to specialised services is moderately restricted. (286) Perhaps the support offered these rural breastfeeding women may be relevant to some urban women, if their barriers to continuation of breastfeeding (66, 401) are related to restricted access to services and greater need for primary care health professional support. (268, 378)

240

Additionally, practice nurse discourse data describes the use of the Conversation Tool intervention with a broad spectrum of breastfeeding women of all ages and socio- economic class and provides some insight to the whole population. Practice nurse discourse indicates that other breastfeeding mothers were willing to engage one-to- one, and seemed responsive to the motivational conversation.

I do ask if they're in the study because I don't want to miss them...I haven't had any in the study for ages and I don't think I'm helping you with your research. They're doing it [breastfeeding] and we're talking about it but they're not in the study. (PN3)

Trial outcomes were collected only from those women who agreed to participate in the study. However, because the intervention needed to become routine practice, the practice nurses were encouraged to use the Conversation Tool with all breastfeeding mothers, time permitting. Practice nurses' improved knowledge, attitudes, and skills were available for the benefit of all breastfeeding mothers with whom they interacted.

6.3.3 Involvement of primary care practitioners in research This study was primary care professional-led, -supervised, and -delivered in collaboration. The MI theoretical framework described the researcher's goal-oriented leadership of the study, under supervision, and the individualised intervention delivery by each practice nurse.

Interventions that are established by a research team who understand and appreciate the challenges of clinical care are more likely to become part of daily practice. Strong interpersonal skills, building rapport with the staff and motivating the staff are critical parts of this process. (214) In this context, involvement of primary care professionals in the research team (352) was one factor among others known to contribute to effective primary care research. This study also linked federal financial resources (PhD scholarship), (402) university academic and administrative resources, (403) Division of General Practice human resources and networks, (403) and the experiences of nurses working across a number of general practices. (404)

Conversely, involvement of primary care professionals can benefit their own research capacity. Understanding the processes of research requires a supportive environment, time and a level of capacity building. (196, 198, 352, 405-407) The professionals who are doing the intervention delivery need to be involved in management of the research processes. (399)

241

As a setting for conducting reliable research, primary care is challenging; heterogeneity of reporting makes it difficult to perform meta-analyses from findings. However, according to a recent evidence review for the US Preventive Services Task Force, health organisations in primary care are likely to increase breastfeeding rates more effectively than usual care. Further integration in primary care, broadly with community or lay support services and temporally with interventions in pregnancy, are recommended. (201)

Recommendation 23. Fund and encourage more involvement of primary care professionals in health research through the provision of well-supported research capacity building programs.

6.3.4 The Conversation Tool This study involved development and testing of a clinical tool and provides evidence for its efficacy. Prompts and reminder systems are particularly needed in primary care. (408) Providing a tool or instrument helps with protocol enforcement. (396) Providing reminders, providing point-of-care decision support tools, and strengthening the patient-clinician relationship, are all important components of successful translational research. (214) The intervention flowchart, the Conversation Tool, was developed with a strong evidence base and functioned as an effective reminder according to the practice nurse discourse. Other tools for nutrition counselling have followed the "Five A's" also used for tobacco cessation. (409) The Conversation Tool was designed to be more responsive to mothers' individual situations and less confrontational than the 5 As. Nurses were provided with relatively brief training to use this tool, and referral to more skilled or specialised services was built in to the intervention, as recommended. (410)

The Conversation Tool is an attractive, accessible guide which can be used easily by trained staff, and which has the potential to be adapted for use in other fields where broaching the subject is personally challenging.

Recommendation 24. Disseminate the Conversation Tool among other trainers so that they may consider adaptation or use of this flowchart by professionals they are training.

242

6.4 LIMITATIONS

6.4.1 Potential for bias Recruitment by mail survey had a 30% response rate, which is acceptable for mail surveys but low in terms of representation. Postal recruitment is most feasible across wider geographic areas in rural primary health, and avoids added burden on understaffed health facilities. (279, 280), While covering a wide area, the recruited population may not be representative. Exclusion of women who attended practices in surrounding smaller towns restricts awareness of intention and outcomes for these women. Due to literacy factors, confirmed by socio-economic evaluation of trial participants, it is unlikely that disadvantaged women are well represented in the randomised controlled trial data, although they are represented in the interview data to some degree.

Limitations have been noted with volunteer research. Those who volunteer for research “are more social” and tend to be more highly educated, in a higher social class, and have a higher need for approval. (366)(p 125) Continued breastfeeding is directly associated with socio-economic status (66) so the selected method is more likely to represent women with more social, educational and economic resources. Study participants were predominantly women with post-secondary vocational or academic qualifications (89%), a small subset of the diverse practice nurse clientele; they comprised about 10% of the mothers who gave birth during this period. However, the context is that very few Australians breastfeed exclusively for long, 18% reporting full breastfeeding for the recommended 6 months. (52)

Lower participation at the first data collection point, compared to the second may reduce the power of the study to detect small effects of the intervention at 4 months. Collection at multiple data points facilitated imputation for some women regarding breastfeeding since birth or any breastfeeding. (268)

Godwin et al. discuss the design of trials for primary care interventions, noting that potential confounders can be minimised with rigorous methodology but not all can be controlled. (397) Pragmatic research projects such as this have the benefit of improved external validity, according to Godwin et al. They measure the effectiveness, "the degree of beneficial effect in real clinical practice." Transition into routine practice is more straightforward than from trials conducted in a controlled environment. As described above in section 6.3.2, factors such as staff turnover and staff learning curve

243

may reduce intervention delivery to the intervention group. A related limitation is the absence of data about whether or not the intervention group actually received the intervention. Comprehensive monitoring and data collection about delivery of the intervention would have assisted the process evaluation.

In an attempt to avoid unnecessarily stimulating the interest of control practices in these topics, only brief details of the intervention were provided to practices prior to randomisation. The number of participants recruited was larger than the calculated required sample size, allowing adjustment for potential confounders in the regression analysis. In this study, randomisation was used for internal validity and clustering was used to reduce contamination. Outcome data collection was blinded for randomisation.

There was no log maintained by practice nurses against which contamination by attendance of control mothers or non-compliance by non-attendance of intervention mothers could be checked. This may have enabled adjustment for contamination and a clearer picture of intervention efficacy. (397)There was no control practices survey of nurse experience or training; they may have been providing skilled breastfeeding or motivational interviewing support, although this is unlikely. There was no assessment of general practitioner knowledge, skills, or attitudes regarding breastfeeding management, communication, or motivational interviewing. These could be confounding factors. Contamination at the provider level was not able to be controlled but was avoided among practice nurses during this trial; none of the trained practice nurses transferred to work at a control practice.

This study represents the outcome of a trial in general practices that were interested in improving their overall care of patients. These may not be representative of the larger general practice field. Use of practice nurses was unequally distributed across the two regions in this study, and population distribution was more centralised in the Riverina than in the Border region. Recruitment of mothers was initially greater in the Border region, however a greater proportion of these mothers were excluded due to attending a non-recruited general practice. These factors may also influence the degree to which these results may be representative. Practice nurse data was collected from intervention but not control general practices, and the assumption is untested that there was less support for ongoing breastfeeding in those practices, although this is likely.

244

Recommendation 25. Monitor compliance with a patient log, for practice nurses to record the attendance of eligible mothers and any intervention delivery, in future use of this intervention.

6.4.2 Community views This study makes minimal contribution to attitude change or structural supports outside the primary care health sector, factors that cause major hindrances to breastfeeding duration.

The depth of community antagonism to ongoing breastfeeding was unexpected in this study, although opposition has been reported. (105) Combined with low knowledge, this very apparent attitude may be a contributor to the initial reluctance of several practice nurses to promote ongoing breastfeeding. In a large Australian survey of community attitudes to breastfeeding (n=2502) 82.6% of the men and women telephoned agreed, "Bottle-feeding is more acceptable in public places". (95) Interviews with 29 Australian mothers in another study reported feedback they had received on continuing breastfeeding “Isn’t that baby old enough to, isn’t that toddler old enough to be off the breast now.” (Paula) This qualitative study found that women faced more opposition as their baby grew older. (105)

Conceptual structures have been developed to categorise the determinants of breastfeeding, areas of health promotion and points at which strategies address the needs of breastfeeding mothers. (411) The intervention in this study targeted individual level factors that influence breastfeeding, and health service orientation. (77) While the motivational interviewing approach may lead to empowerment of women in the face of system-level barriers to ongoing breastfeeding, such as unsupportive paid employment, workplace barriers need to be addressed directly for ongoing breastfeeding to become common. The Australian National Breastfeeding Strategy 2010-2015 has goals for breastfeeding mothers beyond eight weeks postnatal including "Increase the access to parental leave" and "Increase the number of model breastfeeding friendly workplaces, services and environments". (212)(p37) Similarly, expert opinion in the UK calls for extensive policy support and routine implementation of BFHI across hospital and community sectors. (412)

245

This study attempted to improve mothers' access to Australian Breastfeeding Association (ABA) groups and counsellors but no mother-to-mother links were made. Training processes increased practice nurse confidence in the support available to breastfeeding mothers through ABA, and the discourse data analysis in Chapter 5 indicates frequent encouragement for mothers to contact ABA. Involvement with community-based organisations for peer support is the final step of the BFHI, for both hospitals and the community, to support ongoing breastfeeding. Peer counselling significantly improves breastfeeding outcomes. (283)

Primary care services may obstruct ongoing breastfeeding through inadequate or erroneous welcome, information, advice, support, clinical care, or referral. The processes of this study addressed some of these barriers, through training of one group of staff members in general practices. Greater environment change is needed.

Recommendation 26. Work towards BFHI accreditation processes in Australia for primary care organisations, according to the Seven Point Plan.

6.4.3 The nature of general practice There are time and funding limitations to the introduction of interventions such as the Conversation Tool intervention. Overcoming such limitations in the longer term is best attempted with the involvement of the whole practice team, which was not a component of this study.

Public health is not well addressed in a fee for service model. Funding for practice nurse preventive health activities remains problematic for many health fields. A variety of organisational-level supports has been tested, including patient registries, reminders, and provider financial incentives. (413, 414) Some of these are currently used in Australian primary care such as for cervical cancer prevention and for infant immunisation.kk

It is difficult to involve practitioners in lengthy training programs, however longer training is more effective in creating attitudinal change and follow-up of trained personnel assists with attitude change. (Armstrong, cited p466 (415)) Government funding is needed to support training and to develop career pathways for trained staff,

kk http://www.medicareaustralia.gov.au/provider/incentives/pip/index.jsp

246

so that primary care research findings can be rolled out in practice. (414) Newly trained and knowledgeable practitioners need the capacity to introduce new services or change existing detrimental routines in their practices for them to have an impact on breastfeeding outcomes. (415)

Following recruitment, this study did not involve members of the practice team other than the practice nurses. At recruitment, management and medical personnel in general practice teams were involved, as the study needed consent regarding financial and organisational implications. Clustered trials require consent only from the team leader for participation, (285) however the researcher held individual on-site discussions with each practice nurse before the trial began. From that point, all interaction was solely with the practice nurses, and interaction with other team members regarding aspects of trial participation depended on the relationship between the practice nurse and the team.

Independent variables that can limit the success of practice nurse training are "the social system in the working environment including social support from superiors, and encouragement from colleagues in applying the newly acquired knowledge and skills in daily practice". (Francke, cited (386)(p130)) A positive work environment may increase nurses' use of new skills, enabling them better to assess and clarify patients' concerns. (Booth, cited (386)(p142)) However, nurses may be hindered in putting support skills learned during training into practice because of time pressure and negative attitudes on the part of colleagues. (Pool, cited (386)(p135)) Primary health care is a broad field involving teamwork, population health, health promotion and workforce, as well as important aspects of clinical care and equity. (416) There is a need to clarify the mechanisms of teamwork within each organisation and to build effective teamwork. (417)

Recruitment of general practices for research is known to be complex and time- consuming, (213) as illustrated in Figure 6.2. Being both a doctor and researcher was an advantage in gaining access to practices to seek research participation. However, numerous contacts were required to secure participation. This investment in time and contacts may have influenced the practices in their awareness of and allowances for the intervention, and may have become a contribution to the intervention that is part of the intervention efficacy. (396)

247

Practices were diverse with varying involvement of non-medical staff in decision- making such as non-clinical appointments for recruiting and in availability of the practice to participate in research. Practices varied in the degree of collaboration between doctors within the practice, and in the level of support from doctors for the interests of their practice nurses. Workforce pressures impacted differently, depending on changes in staffing and on town planning decisions such as location of new suburbs.

Figure 6.2 Notes on a practice recruitment attempt

 Planned to visit two practices during the lunch period, travelled 1.5 hours to town.  Overstayed at Practice A due to questions, telephoned Practice B about the delay, given reassurance by receptionist, then met with displeased first principal GP at Practice B.  Unexpectedly had to stay in town overnight and return to the practice the next morning to meet the second practice principal GP before the morning consulting started. Third principal GP was on annual leave.  Second principal GP agreed to ring the third and leave him a message to call me.  I had no calls and arranged finally to return and speak to third principal some weeks later.  Practice principals by now had discussed the project among themselves, and when I visited again I was told by third principal GP that the group decision was not to participate.  There were three reasons. The first was workload. Retirement of fourth principal GP was imminent. Replacement not yet found, recruitment efforts unsuccessful; remaining principals were concerned they would need to rely more on the (two part-time) practice nurses to manage the practice workload; did not want nurses to have less time available.  The second reason was financial commitments. With retirement of the fourth principal GP, administrative and nursing staff would be retained, increasing costs. In this setting, use of nursing time with an intervention that would not result in additional income was a deterrent.  The third reason was lack of remuneration. The practice was reluctant, in the setting of anticipated financial challenge, to release nurses for (paid) training time without income being generated by them in this (paid) time.

This intervention did not involve any training for general practitioner, management or reception staff, in contrast to the processes of the Baby Friendly Health Initiative (BFHI), which uses an environment approach to support ongoing breastfeeding. (343) Increased support for primary care Baby Friendly accreditation and monitoring of BFHI compliance are needed. (143) A broader environment approach is likely to also support practice nurses to continue using the Conversation Tool with mothers, as Reddin et al. found that graduate midwives entering even BFHI accredited hospitals needed more support within the workplace to continue supporting mothers with their breastfeeding. (49) Moreover, in one study of rural primary care practices that sustained health

248

promotions over decades, each practice featured a committed physician. The physician had the capacity to address barriers, "acting simultaneously as advocate and member of the entrenched power structure to bridge the gap between the program and opponents" (398)

6.4.4 Mothers' busy lives This study does not address any social needs of mothers that may limit their capacity to breastfeed. Mothers with large families were heard to say their lives were too busy for breastfeeding. This intervention provides support only to those mothers who are able to access the general practice and the practice nurse. (PN3)(Figure 5.2) Access to primary health care is not well surveyed, although there is the potential to monitor this through National Health Surveys. (418)

Some younger mothers and disadvantaged mothers, not trial participants, did make contact with intervention practice nurses although in the practice nurse discourse their heavy burden was noted.

Not so surprised with the mums who are having a hard time with it. Would you say low socio-economic? There's generally more kids, there's less income.

Support from trained practice nurses may be sufficient to help some disadvantaged women to continue breastfeeding, a project that needs to be tested. It may be however that these women need earlier and more extensive support. (246)

From the practice nurses' perspective, many women planned to use infant formula, or to wean earlier than recommended, for reasons other than ignorance. Their busy lives and their reliance on support from people without breastfeeding management knowledge were responsible.

At least while the intervention may not have been sufficient for some mothers, it most likely added no further burden. As discussed in Chapter 1 and confirmed by the practice nurses in Chapter 5 Part 1, mothers carry many responsibilities and the burden of their workload may restrict their ability to continue breastfeeding, either directly or by limiting access to support. For some of these women, discrediting the hazards of formula may be a way for them to down-regulate their emotional experience of lactation failure. This intervention took care not to ignore social determinants of health. Practice nurses acted to defuse conflict and to be responsive to mothers' needs

249

when supporting ongoing breastfeeding. "You're not coming home with me; you don't know what I'm going through", (419)(p8) as one mother responded to a breastfeeding advocate.

6.4.5 The mothers' perspective No satisfaction data was collected from mothers recruited for this trial, and mothers' responses to the intervention are recorded only from the perspective of the practice nurses. This thesis describes a trial of a system-level intervention, creating a change in usual practices within components of the health system. Mothers may be less aware of system-level supports or barriers compared to individual experiences, and may describe satisfaction with their care regardless of health outcomes. As described in Chapter 2, mothers in both the control and intervention groups of a large Danish study were equally satisfied with the care they received from health visitors, although subsequent breastfeeding duration differed. (252) Reliable and valuable measurement of patient satisfaction is complex, involving personal preference, expectations and experience of care, hence reflecting both the care and the individual. (420) Patient satisfaction "is always high" according to a review of communication training programs for nurses. (386)(p142) Lack of variability is problematic for analysis, although specific questions about communication such as information-giving and comprehension of messages in primary care have recorded levels of dissatisfaction. (420) It is anticipated that mothers who intend continuing would appreciate the intervention, as one mother commented in a focus group, (106)(p426)

It’s so wonderful for doctors to keep asking the questions even later on, even if there’s no problem. [They] only seem to ask the question when there is a problem.

Support must be demonstrable, not just from the provider's perspective, but also from the perspective of the receiver. (106)

Recommendations

27. Incorporate an assessment of the mothers' satisfaction with information content and presentation, correlated with current breastfeeding intentions, in future trials with the Conversation Tool intervention.

250

6.5 SUMMARY OF FINDINGS This thesis aimed to determine whether the introduction of motivational support from trained general practice nurses would lead to an increase in rates of exclusive breastfeeding and ongoing breastfeeding. The question was addressed using a systematic literature review, a randomised controlled trial and a process evaluation.

The systematic literature review found that breastfeeding education for health professionals delivering interventions was crucial. Successful interventions tended to use a counselling-based theoretical framework and effective communication.

Compared to the primary objectives (section 3.1.1) the randomised controlled trial was successful in demonstrating increased rates of full breastfeeding and exclusive breastfeeding at 4 months, among mothers who specified their employment plans for the first year after birth. The trial was not able to demonstrate change in "any breastfeeding", a measure of breastfeeding duration, at 4 months. There was no change found in exclusivity or duration at 6 months.

Compared to the secondary objectives (section 3.1.1) the randomised controlled trial revealed new evidence about the crucial influence of her first year's employment plans on a new mother's breastfeeding outcomes.

The process evaluation addressed other secondary objectives, exploring the experience of practice nurses. Their experience of education, mentoring, raising an emotive subject, listening to mothers, and offering support provides material for future research and indicates the feasibility of replicating this intervention more broadly.

Strengths and limitations of this research project provide guidance for the planning of future clinical trials in primary care, particularly considering the success of the Conversation Tool flowchart and the barriers of professional and community workload and time. A list of recommendations is provided.

Implications for practice are that breastfeeding exclusivity may be improved by periodic contact and intentional conversation with practice nurses. Community views must be considered. Practice nurses may provide effective support given eight hours of training, mentoring to develop comfort with new routines, and workplace support for them to undertake this role. Use of a flowchart assists in new workplace practices.

251

6.6 SUMMARY OF RECOMMENDATIONS 1. Disseminate trial findings to local and regional communities and to national bodies. 2. Introduce additional breastfeeding support for all mothers in order to increase exclusive and full breastfeeding at 5 months, in accordance with new indicators in the Australian National Breastfeeding Strategy 2010-2015. (212) 3. Explore reasons for the low rates of exclusive breastfeeding in developed countries at 6 months. 4. Increase community awareness of breastfeeding protections and strategies for mothers planning paid employment outside the home. 5. Provide additional support to pregnant and breastfeeding women who are planning to return to paid employment/study after the birth, in order to increase exclusive and full breastfeeding at 5 months. 6. Trial an adapted Conversation Tool intervention, with processes to improve and monitor receipt of the intervention, in a similar population, in order to increase duration of breastfeeding at 6 and 12 months (current practice). 7. Trial and evaluate the Conversation Tool intervention with relatively disadvantaged women. 8. Promote in Australia moves towards comprehensively supportive primary care health services, and evaluate primary care services in accordance with the Baby Friendly Seven Point Plan. 9. Explore novel approaches to increase the access of all breastfeeding women to established and skilled peer support. 10. Improve Australian family and community knowledge of the importance and management of breastfeeding beyond 6 or 12 months. 11. Address comprehensive health promotion campaigns towards community attitudes on breastfeeding in public. 12. Improve measures to protect the breastfeeding relationship for women in out-of- home employment. 13. Provide motivational interviewing training and a period of mentoring to all primary care health professionals who have regular contact with breastfeeding mothers. 14. Explore the use of an extension of the Conversation Tool intervention to pregnant women as a means of influencing breastfeeding intention and potentially duration.

252

15. Fund and encourage uptake of communication skills training for all primary health care professionals who have regular contact with breastfeeding mothers, consistent with the component of the BFHI 20-hour course used in this study. 16. Create a central body to provide cross-disciplinary breastfeeding research leadership and to give oversight to the National Breastfeeding Strategy. 17. Explore in depth the factors relating to reliance on formula by breastfeeding mothers in relatively privileged groups. 18. Educate health professionals about the risks of reliance on formula over breastfeeding. 19. Legislate to restrict approaches to the marketing of formula that minimise the health risks of weaning earlier than recommended. 20. Fund and encourage uptake of breastfeeding management training for all primary health care professionals who have regular contact with breastfeeding mothers, consistent with the BFHI 8-hour training package. 21. Build reflective practice activities, with opportunity for supportive feedback, into professional training programs to support ongoing breastfeeding. 22. Recognise the key role of practice nurses within the team, in promoting positive patient outcomes. 23. Fund and encourage more involvement of primary care professionals in health research through the provision of well-supported research capacity building programs. 24. Disseminate the Conversation Tool among other trainers so that they may consider adaptation or use of this flowchart by professionals they are training. 25. Monitor compliance with a patient log, for practice nurses to record the attendance of eligible mothers and any intervention delivery, in future use of this intervention. 26. Work towards BFHI accreditation processes in Australia for primary care organisations, according to the Seven Point Plan. 27. Incorporate an assessment of the mothers' satisfaction with information content and presentation, correlated with current breastfeeding intentions, in future trials with the Conversation Tool intervention.

253

REFERENCES

(1) Beasley A. Breastfeeding studies: culture, biomedicine and methodology. J Hum Lact. 1991;7(1):7-14. (2) Ip S, Chung M, Raman G, Chew P, Magula N, DeVine D, et al. Breastfeeding and maternal and infant health outcomes in developed countries. Evidence Report/Technology Assessment. Rockville, MD: U.S. Department of Health and Human Services; 2007 April. Report No.: 153. (3) Buttrose I, Adams P. Motherguilt. Camberwell, VIC: Penguin Books 2005. (4) World Health Organisation, UNICEF. Global strategy for infant and young child feeding; 2003. (5) Diaz S. The human reproductive pattern and the changes in womens roles. Conference on breastfeeding as a womens issue - A dialogue on health, family planning, work, and feminism; 1993 Sep 13; Washington, DC: Elsevier; 1993. p. S3-S9. (6) Brodribb W. Breastfeeding management 3rd ed. Melbourne: Australian Breastfeeding Association 2004. (7) 2 Gorgeous Girls. General > Australia's Flood Disaster > Next stupid Facebook / email status. 2011 12/01/2011, 03:19 PM [cited 2012 24 February]; Available from: http://www.essentialbaby.com.au/forums/lofiversion/index.php/t861244.html (8) Gijsbers B, Mesters I, Andre Knottnerus J, Legtenberg AHG, van Schayck CP. Factors influencing breastfeeding practices and postponement of solid food to prevent allergic disease in high-risk children: results from an explorative study. Patient Educ Couns. 2005 Apr;57(1):15-21. (9) The Parliament of the Commonwealth of Australia. The Best Start: Report on the inquiry into the health benefits of breastfeeding. Canberra: House of Representatives; 2007 August. (10) Gruich M. Life-changing experiences of a private practicing pediatrician: perspectives from a private pediatric practice. Pediatrics. 2006 May 2006;117(Supplement 4):S359-S64. (11) Hausman B. Women's liberation and the rhetoric of "choice" in infant feeding debates. Int Breastfeed J. 2008;3(1):10. (12) Van Esterik P. Breastfeeding and feminism. Int J Gynaecol Obstet. 1994 Dec;47 Suppl:S41-50; discussion S50-4. (13) Fallon A, Crepinsek M, Hegney D, O'Brien M. The Baby-Friendly Hospital Initiative and breastfeeding duration: Relating the evidence to the Australian context. Birth Issues. 2005;14(3):90-95. (14) Graham KI, Scott JA, Binns CW, Oddy WH. National targets for breastfeeding at hospital discharge have been achieved in Perth. Acta Paediatr. 2005;94:352-56. (15) Smith J, Dunstone M, Elliott-Rudder M. Health professional knowledge of breastfeeding: Are the health risks of infant formula feeding accurately conveyed by the titles and abstracts of journal articles? J Hum Lact. 2009 August;25(3):350-58. (16) Branca F, Ferrari M. Impact of micronutrient deficiencies on growth: the stunting syndrome. Ann Nutr Metab. 2002;46 Suppl 1:8-17. (17) Kramer MS, Chalmers B, Hodnett ED, Sevkovskaya Z, Dzikovich I, Shapiro S, et al. Promotion of Breastfeeding Intervention Trial (PROBIT): A randomized trial in the Republic of Belarus. JAMA. 2001 January 24, 2001;285(4):413-20. (18) Kramer MS, Aboud F, Mironova E, Vanilovich I, Platt RW, Matush L, et al. Breastfeeding and child cognitive development: new evidence from a large randomized trial. Archives of General Psychiatry. 2008 May;65(5):578-84. (19) Heinig MJ. Host defense benefits of breastfeeding for the infant. Effect of breastfeeding duration and exclusivity. Pediatr Clin North Am. 2001 Feb;48(1):105-23. (20) Nelson AM. A metasynthesis of qualitative breastfeeding studies. J Midwifery Womens Health. 2006;51(2):e13-e20.

254

(21) Smith J, Ellwood M. Feeding patterns and emotional care in breastfed infants. Social Indicators Research. 2011;101(2):227-31. (22) Chen A, Rogan WJ. Breastfeeding and the risk of postneonatal death in the United States. Pediatrics. 2004 May 1, 2004;113(5):e435-e39. (23) Morales AE. Maternal perceptions of closeness and attachment by infant feeding style: An examination of breastfeeding and bottle-feeding [M.A.]. United States -- New Jersey: Kean University; 2010. (24) Gross BA, Burger H, WHO Task Force on methods for the natural regulation of fertility. Breastfeeding patterns and return to fertility in Australian women. Aust N Z J Obstet Gynaecol. 2002 May;42(2):148-54. (25) Valdes V, Perez A, Labbok M, Pugin E, Zambrano I, Catalan S. The impact of a hospital and clinic-based breastfeeding promotion programme in a middle class urban environment. Journal of Tropical Pediatrics. 1993 Jun;39(3):142-51. (26) Short RV. Lactation - the central control of reproduction. Symposium on breast feeding and the mother. Ciba Foundation Symposium 45. London: Elsevier Scientific Publishing Company 1976:73-86. (27) Gunderson EP, Jacobs DR, Chiang V, Lewis CE, Feng J, Quesenberry CP, et al. Duration of lactation and incidence of the metabolic syndrome in women of reproductive age according to gestational diabetes mellitus status: A 20-year prospective study in CARDIA - The Coronary Artery Risk Development in Young Adults study. Diabetes. 2009. (28) McNiel M, Labbok M, Abrahams S. What are the risks associated with formula feeding? A re-analysis and review. Birth. 2010 Mar;37(1):50-58. (29) Labiner-Wolfe J, Fein SB, Shealy KR. Infant formula - handling education and safety. Pediatrics. 2008;122(Supplement_2):S85-90. (30) National Health and Medical Research Council. Dietary guidelines for children and adolescents in Australia incorporating the infant feeding guidelines for health workers. In: Commonwealth of Australia Department of Health and Ageing, ed.: Canberra 2003. (31) Minchin MK. Breastfeeding matters. 1st ed. North Sydney, Australia: Alma Publications and George Allen & Unwin 1985. (32) Klopffer W, Rippen G. Life Cycle Analysis and ecological balance: Methodical approaches to assessment of environmental aspects of products. Environment International. 1992;18(1):55-61. (33) Ball TM, Wright AL. Health care costs of formula-feeding in the first year of life. Pediatrics. 1999 April 1, 1999;103(4):870-76. (34) Smith JP, Thompson JF, Ellwood DA. Hospital system costs of artificial infant feeding: estimates for the Australian Capital Territory. Aust N Z J Public Health. 2002 Dec;26(6):543-51. (35) Smith JP. The contribution of infant food marketing to the obesogenic environment in Australia. Breastfeed Rev. 2007 Mar;15(1):23-35. (36) Shaw R. Lactating, feminists and breastfeeding advocacy: some complexities. New Zealand College of Midwives Journal. 2004;30:16-20. (37) Smith J, Ingham L. Mothers' milk and measures of economic output. Feminist Economics. 2005;11(1):41-62. (38) Bartlett A. Breastfeeding and time: In search of a language for pleasure and agency. In: Shaw R, Bartlett A, eds. Giving breastmilk: Body ethics and contemporary breastfeeding practice. Toronto, Canada: Demeter 2010:222-35. (39) Manderson L. Conflicting interests: Class, ethnicity, and infant feeding practice in Australian society. Healthright. 1982;2(1):37-42. (40) Bartlett A. Breastwork: rethinking breastfeeding. Sydney: University of New South Wales Press 2005. (41) Hausman B. On the topic of breastfeeding: A review of three books. Journal of Medical Humanities. 2009;30(1):77-81.

255

(42) Webb K, Marks GC, Lund-Adams M, Rutishauser IHE, Abraham B. Towards a national system for monitoring breastfeeding in Australia: recommendations for population indicators, definitions and next steps. Canberra: Australian Food and Nutrition Monitoring Unit; 2001. (43) Cattaneo A, Davanzo R, Ronfani L. Are data on the prevalence and duration of breastfeeding reliable? The case of Italy. Acta Paediatr. 2000;89(1):88-93. (44) Niegel S, Ystrom E, Hagtvet KA, Vollrath ME. Difficult temperament, breastfeeding, and their mutual prospective effects: The Norwegian Mother and Child Cohort Study. J Dev Behav Pediatr. 2008;29(6):458-62 10.1097/DBP.0b013e3181877a88. (45) Britton C, McCormick F, Renfrew M, Wade A, King S. Support for breastfeeding mothers. Cochrane Database Syst Rev. 2007(1):CD001141. (46) Li R, Fein SB, Grummer-Strawn LM. Association of breastfeeding intensity and bottle-emptying behaviors at early infancy with infants' risk for excess weight at late infancy. Pediatrics. 2008;122(Supplement_2):S77-84. (47) Bonuck KA, Trombley M, Freeman K, McKee D. Randomized, controlled trial of a prenatal and postnatal lactation consultant intervention on duration and intensity of breastfeeding up to 12 months. Pediatrics. 2005 Dec;116(6):1413-26. (48) Labbok M, Krasovec K. Toward consistency in breastfeeding definitions. Stud Fam Plann. 1990;21(4):226-30. (49) Reddin E, Pincombe J, Darbyshire P. Passive resistance: early experiences of midwifery students/graduates and the Baby Friendly Health Initiative 10 steps to successful breastfeeding. Women & Birth: Journal of the Australian College of Midwives. 2007 Jun;20(2):71-6. (50) Aarts C, Kylberg E, Hornell A, Hofvander Y, Gebre-Medhin M, Greiner T. How exclusive is exclusive breastfeeding? A comparison of data since birth with current status data. Int J Epidemiol. 2000 December 1, 2000;29(6):1041-46. (51) Binns CW, Fraser ML, Lee AH, Scott J. Defining exclusive breastfeeding in Australia. J Paediatr Child Health. 2009;45(4):174-80. (52) Donath SM, Amir LH. Breastfeeding and the introduction of solids in Australian infants: data from the 2001 National Health Survey. Aust N Z J Public Health. 2005 Apr;29(2):171-5. (53) Berry NJ, Gribble KD. Breast is no longer best: the World Health Organization, the Multi-centre Growth Reference Study and normal infant feeding. Aust N Z J Public Health. 2006 Aug 2006;30(4):387-89. (54) World Health Organization. Indicators for assessing infant and young child feeding practices. Washington DC USA; 2008 6-8 November 2007. (55) Bodnarchuk JL, Eaton WO, Martens PJ. Transitions in breastfeeding: Daily parent diaries provide evidence of behavior over time. J Hum Lact. 2006 May 1, 2006;22(2):166-74. (56) Hector DJ. Complexities and subtleties in the measurement and reporting of breastfeeding practices. Int Breastfeed J. 2011 17 May 2011;6(5). (57) Contento IR, Randell JS, Basch CE. Review and analysis of evaluation measures used in nutrition education intervention research. J Nutr Educ Behav. 2002 Jan-Feb;34(1):2-25. (58) Australian Institute of Health and Welfare. National breastfeeding indicators: workshop report. Canberra: AIHW; 2011 8 December. (59) Hector D, Webb K, Lymer S. Describing breastfeeding practices in New South Wales using data from the NSW Child Health Survey, 2001. NSW Public Health Bull. 2005;16(4):47-51. (60) Owen CG, Martin RM, Whincup PH, Smith GD, Cook DG. Effect of infant feeding on the risk of obesity across the life course: A quantitative review of published evidence. Pediatrics. 2005 May 1, 2005;115(5):1367-77.

256

(61) Australasian Society of Clinical Immunology and Allergy. ASCIA Infant feeding advice position statement. [Position Paper] 2010 [cited 2011 17 January]; Available from: http://www.allergy.org.au/content/view/350/287/ (62) Prescott SL, Smith P, Tang M, Palmer DJ, Sinn J, Huntley SJ, et al. The importance of early complementary feeding in the development of oral tolerance: Concerns and controversies. Pediatric Allergy and Immunology. 2008 Aug;19(5):375- 80. (63) Thorley V. Printed advice on initiating and maintaining breastfeeding in mid- 20th-century Queensland. J Hum Lact. 2003;19(1):77-89. (64) Wen LM, Baur L, Rissel C, Alperstein G, Simpson J. Intention to breastfeed and awareness of health recommendations: findings from first-time mothers in southwest Sydney, Australia. Int Breastfeed J. 2009;4(1):9. (65) Baxter J, Cooklin AR, Smith JP. Which mothers wean their babies prematurely from full breastfeeding? An Australian cohort study. Acta Paediatr. 2009 Aug;98(8):1274-7. (66) Amir LH, Donath SM. Socioeconomic status and rates of breastfeeding in Australia: evidence from three recent national health surveys. MJA. 2008 Sep 1;189(5):254-6. (67) Häggkvist A-P, Brantsæter AL, Grjibovski AM, Helsing E, Meltzer HM, Haugen M. Prevalence of breast-feeding in the Norwegian Mother and Child Cohort Study and health service-related correlates of cessation of full breast-feeding. . Public Health Nutr. 2010;13:2076-86. (68) Lande B, Andersen LF, Baerug A, Trygg KU, Lund-Larsen K, Veierod MB, et al. Infant feeding practices and associated factors in the first six months of life: The Norwegian Infant Feeding Survey. Acta Paediatr. 2003;92:152-61. (69) AIHW. Australian national infant feeding survey: indicator results. 2011 [cited 2012 January 25]; [Cat. no. PHE 156]. Available from: http://www.aihw.gov.au/publication-detail/?id=10737420927 (70) Nutbeam D, Wise M, Bauman A, Harris E, Leeder S. Goals and targets for Australia's health in the year 2000 and beyond. Canberra: AGPS 1993. (71) Donath S, Amir LH. Rates of breastfeeding in Australia by State and socio- economic status: Evidence from the 1995 National Health Survey. J Paediatr Child Health. 2000;36(2):164-68. (72) Australian Bureau of Statistics. Breastfeeding in Australia electronic delivery. [In HTML format.] 2003 [cited; Available from: http://www.abs.gov.au/ausstats/[email protected]/cat/4810.0.55.001 (73) Binns C. Encourage and support breastfeeding. Journal of the Home Economics Institute of Australia. 2004;11(1):28-38. (74) Hansard. Inquiry into the health benefits of breastfeeding. Standing Committee on Health and Ageing. Public Hearing June 4 ed. Sydney: Parliament of Australia 2007. (75) Commonwealth of Australia. Health Benefits of Breastfeeding. Proof Committee Hansard 2007 26 March 2007 [cited; 41]. Available from: http://www.aph.gov.au/hansard/reps/commttee/R10129.pdf (76) Hannan A, Li R, Benton-Davis S, Grummer-Strawn L. Regional variation in public opinion about breastfeeding in the United States. J Hum Lact. 2005 August 1, 2005;21(3):284-88. (77) Webb K, Stickney B, Heywood P. Guest editorial: Breastfeeding and the public's health. NSW Public Health Bull. 2005 Mar-Apr;16(3-4):37-40. (78) Palmer G. The politics of breastfeeding. 2nd ed. London: Pandora Press 1993. (79) Thorley V. Commercial interests and advice on infant feeding: Marketing to mothers in postwar Queensland. Health & History. 2003;5(1):65–89. (80) Thorley V. Middle-class mothers as activists for change: The Australian Breastfeeding Association. In: Andrea O'Reilly, ed. The 21st Century Motherhood Movement. Bradford, ON, Canada: Demeter Press 2011:219-32.

257

(81) Berry NJ, Jones SC, Iverson D. Circumventing the WHO Code? An observational study. Archives of Disease in Childhood 2011 [cited 16 January 2012]; Available from: http://adc.bmj.com/content/early/2011/06/29/adc.2010.202051.abstract (82) Berry NJ, Jones SC, Iverson D. Relax, you're soaking in it: Sources of information about infant formula. Breastfeed Rev. 2011 Mar;19(1):9-18. (83) Hambidge KM. The role of zinc and other trace metals in pediatric nutrition and health. Pediatr Clin North Am. 1977 February;24(1):95-106. (84) Bryant CA. The impact of kin, friend and neighbor networks on infant feeding practices. Cuban, Puerto Rican and Anglo families in Florida. Social Science and Medicine. 1982;16(20):1757-65. (85) Auerbach KG. Beyond the Issue of Accuracy: Evaluating Patient Education Materials for Breastfeeding Mothers. J Hum Lact. 1988;4(3):108. (86) Hausman BL. Mother's milk: Breastfeeding controversies in American culture. New York: Routledge 2003. (87) Cricco-Lizza RR. Thirst for the milk of human kindness: An ethnography of infant feeding decisions among Black women enrolled in WIC [Ph.D.]. New York, USA: New York University; 2002. (88) Abdulwadud O, Snow M. Interventions in the workplace to support breastfeeding for women in employment Cochrane Database Syst Rev. 2007 2007(3). (89) Hawkins SS, Griffiths LJ, Dezateux C, Law C, Millennium Cohort Study Child Health G. The impact of maternal employment on breast-feeding duration in the UK Millennium Cohort Study. Public Health Nutr. 2007 Sep;10(9):891-6. (90) Galtry J. The impact on breastfeeding of labour market policy and practice in Ireland, Sweden, and the USA. Soc Sci Med. 2003;57(1):167-77. (91) Fein SB, Mandal B, Roe BE. Success of strategies for combining employment and breastfeeding. Pediatrics. 2008;122(Supplement 2):S56-S62. (92) Baker M, Milligan K. Maternal employment, breastfeeding, and health: Evidence from maternity leave mandates. Journal of Health Economics. 2008;27(4):871-87. (93) Zinn B. Supporting the employed breastfeeding mother. J Midwifery Womens Health. 2000 May-Jun;45(3):216-26. (94) McLean S. A feminist critique of the role of medical discourse in the introduction of the breastfeeding amendment in the Anti-Discrimination Act 1977 NSW. Public Space: The Journal of Law and Social Justice. 2008;2(Art 5):1-29 (95) McIntyre E, Hiller JE, Turnbull D. Community attitudes to infant feeding. Breastfeed Rev. 2001 Nov;9(3):27-33. (96) Dykes F, Griffiths H. Societal influences upon initiation and continuation of breastfeeding. British Journal of Midwifery. 1998;6(2):76-80. (97) Dykes F, Moran V, Burt S, Edwards J. Adolescent mothers and breastfeeding: experiences and support needs - an exploratory study. J Hum Lact. 2003;19(4):391- 401. (98) Allen J, Heywood P. Infant feeding practices in Sydney 1976-77. Aust Paediatr J. 1979;15:113-17. (99) Van Esterik P. Medicalization and the infant formula controversy. In: Van Esterik P, ed. Beyond the Breast-Bottle Controversy: Rutgers University Press 1989:268. (100) Giashuddin MS, Kabir M. Duration of breast-feeding in Bangladesh. Indian J Med Res. 2004 June;119:267-72. (101) Haider R, Kabir I, Hamadani JD, Habte D. Reasons for failure of breast-feeding counselling: mothers' perspectives in Bangladesh. Bull World Health Organ. 1997;75(3):191-6. (102) Giles F. Fresh Milk. Crows Nest, Australia: Allen & Unwin 2003. (103) Dettwyler K. Beauty and the breast: the cultural context of breast-feeding in the United States. In: Stuart-Macadam P, Dettwyler K, eds. Breastfeeding: biocultural perspectives. New York: de Gruyter 1995:167-215.

258

(104) Rempel LA. Factors influencing the breastfeeding decisions of long-term breastfeeders. 11th Annual National Breastfeeding Seminar; 2001 Jun 07-08; Toronto, CANADA: Sage Publications Inc; 2001. p. 306-18. (105) Hauck YL, Irurita VF. Incompatible expectations: the dilemma of breastfeeding mothers. Health Care Women Int. 2003;24(1):62-78. (106) Dillaway HE, Douma ME. Are pediatric offices "supportive" of breastfeeding? Discrepancies between mothers' and healthcare professionals' reports. Clin Pediatr. 2004 June 1, 2004;43(5):417-30. (107) Bumgarner NJ. Mothering your nursing toddler. 3rd ed. Schaumburg, IL, USA: La Leche League International 2000. (108) Gribble KD. Long-term breastfeeding; changing attitudes and overcoming challenges. Breastfeed Rev. 2008 Mar;16(1):5-15. (109) Gribble KD. 'As good as chocolate' and 'better than ice cream': How toddler, and older, breastfeeders experience breastfeeding. Early Child Development and Care. 2007 23 November;1-16. (110) Mannien J, van den Brandhof WE, McIntyre E, Hiller JE. Breastfeeding articles in the Australian press: 1996-1999. Breastfeed Rev. 2002 Mar;10(1):5-10. (111) Henderson L, Kitzinger J, Green J. Representing infant feeding: content analysis of British media portrayals of bottle feeding and breast feeding. Br Med J. 2000;321(7270):1196. (112) Ward L, Merriwether A, Caruthers A. Breasts are for men: Media, masculinity ideologies, and men’s beliefs about women’s bodies. Sex Roles. 2006;55(9):703-14. (113) Brookes B. 'King, Frederic Truby - Biography' from the Dictionary of New Zealand Biography. 2010 1-Sep-10 [cited 2011 February 15]; Available from: http://www.TeAra.govt.nz/en/biographies/2k8/1 (114) Reiger KM. Our bodies our babies: The forgotten women's movement. Carlton South, VIC Australia: Melbourne University Press 2001. (115) Renfrew MJ, Lang S, Martin L, Woolridge MW. Feeding schedules in hospitals for newborn infants. Cochrane Database Syst Rev. 2000(2):CD000090. (116) Nutricia Australia Pty Ltd. Nutricia Infant Formulas 0-6 months. 2007 [cited 4 May 2007]; Available from: http://www.nutricia.com.au/infant00.asp (117) Cook PS. Childrearing, culture and mental health: exploring an ethnological- evolutionary perspective in child psychiatry and preventive mental health with particular reference to two contrasting approaches to early childrearing. Medical Journal of Australia Special Supplement. 1978 August 12;2:3-14. (118) Ames L. Especially for grandparents. Revised 2010, 3rd ed. East Malvern, VIC: Australian Breastfeeding Association 2006. (119) Lewis M, Ramsay D. Infant emotional and cortisol responses to goal blockage. Child Development. 2005;76(2):518-30. (120) Edwards A, Elwyn G, Covey J, Matthews E, Pill R. Presenting risk information - a review of the effects of "framing" and other manipulations on patient outcomes. Journal of Health Communication. 2001 Jan-Mar;6(1):61-82. (121) Brown LP, Bair AH, Meier PP. Does federal funding for breastfeeding research target our national health objectives? Pediatrics. 2003;111(4):e360-64. (122) Cantrill RM, Creedy DM, Cooke M. How midwives learn about breastfeeding. Australian Midwifery Journal. 2003 September 2003:25-30. (123) Freed GL, Clark SJ, Sorenson J, Lohr JA, Cefalo R, Curtis P, et al. National assessment of physicians' breast-feeding knowledge, attitudes, training, and experience. Jama. 1995 Feb 8;273(6):472-6. (124) Guise J-M, Freed G. Resident physicians' knowledge of breastfeeding and infant growth. Birth. 2000;27(1):49-53. (125) Hellings P, Howe C. Breastfeeding knowledge and practice of pediatric nurse practitioners. J Pediatr Health Care. 2004 Jan-Feb;18(1):8-14.

259

(126) Hillenbrand KM, Larsen PG. Effect of an educational intervention about breastfeeding on the knowledge, confidence, and behaviors of pediatric resident physicians. Pediatrics. 2002 November 1, 2002;110(5):e59-. (127) Register N, Eren M, Lowdermilk D, Hammond R, Tully MR. Knowledge and attitudes of pediatric office nursing staff about breastfeeding. J Hum Lact. 2000 August 1, 2000;16(3):210-15. (128) Boulton TN, Malacrida C. Women and Cosmetic Breast Surgery. Qual Health Res. 2011 April 1;22(4):511-23. (129) Baker JL, Michaelsen KF, Rasmussen KM, Sorensen TIA. Maternal prepregnant body mass index, duration of breastfeeding, and timing of complementary food introduction are associated with infant weight gain. Am J Clin Nutr. 2004;80(6):1579-88. (130) Harrod-Wild K. Infant nutrition - from professional to personal. Journal of Family Health Care. 2006;16(2):49-51. (131) Brodribb W, Jackson C, Fallon AB, Hegney D. Breastfeeding and the responsibilities of GPs: a qualitative study of general practice registrars. Aust Fam Physician. 2007 Apr;36(4):283-5. (132) Brodribb W, Fallon A, Jackson C, Hegney D. The relationship between personal breastfeeding experience and the breastfeeding attitudes, knowledge, confidence and effectiveness of Australian GP registrars. Matern Child Nutr. 2008 October 1, 2008;4(4):264-74. (133) Hussainy S, Dermele N. Knowledge, attitudes and practices of health professionals and women towards medication use in breastfeeding: A review. Int Breastfeed J. 2011;6(1):11. (134) Brodribb W, Fallon A, Jackson C, Hegney D. Breastfeeding and Australian GP registrars: Their knowledge and attitudes. J Hum Lact. 2008;24(4):422-30. (135) Abbott S, Renfrew MJ, McFadden A. 'Informal' learning to support breastfeeding: local problems and opportunities. Matern Child Nutr. 2006;2(4):232-38. (136) Hauck YL, Fenwick J, Dhaliwal SS, Butt J, Schmied V. The association between women's perceptions of professional support and problems experienced on breastfeeding cessation: A Western Australian study. J Hum Lact. 2011 February;27(1):49-57. (137) Hauck Y, Langton D, Coyle K. The path of determination: exploring the lived experience of breastfeeding difficulties. Breastfeed Rev. 2002 2002;10(2):5-12. (138) Anderson PO, Pochop SL, Manoguerra AS. Adverse drug reactions in breastfed infants: Less than imagined. Clin Pediatr. 2003 May 1, 2003;42(4):325-40. (139) Graffy J, Taylor J. What information, advice, and support do women want with breastfeeding? Birth. 2005;32(3):179-86. (140) Pelto GH, Santos I, Goncalves H, Victora C, Martines J, Habicht J-P. Nutrition counseling training changes physician behavior and improves caregiver knowledge acquisition. J Nutr. 2004 February 1, 2004;134(2):357-62. (141) Ekstrom A, Nissen E. A Mother's Feelings for Her Infant Are Strengthened by Excellent Breastfeeding Counseling and Continuity of Care. Pediatrics. 2006 August 1, 2006;118(2):e309-14. (142) Labbok M. Transdisciplinary breastfeeding support: Creating program and policy synergy across the reproductive continuum. Int Breastfeed J. 2008;3(1):16. (143) Merten S, Dratva J, Ackermann-Liebrich U. Do Baby-Friendly hospitals influence breastfeeding duration on a national level? Pediatrics. 2005 November 1, 2005;116(5):e702-08. (144) DiGirolamo AM, Grummer-Strawn LM, Fein SB. Effect of maternity-care practices on breastfeeding. Pediatrics. 2008;122(Supplement 2):S43-49. (145) O'Brien M, Buikstra E, Hegney D. The influence of psychological factors on breastfeeding duration. Journal of Advanced Nursing. 2008;63(4):397-408. (146) Kronborg H, Vaeth M. The influence of psychosocial factors on the duration of breastfeeding. Scandinavian Journal of Public Health. 2004 Jun;32(3):210-16.

260

(147) Chambers JA, McInnes RJ, Hoddinott P, Alder EM. A systematic review of measures assessing mothers' knowledge, attitudes, confidence and satisfaction towards breastfeeding. Breastfeed Rev. 2007 Nov;15(3):17-25. (148) Chin NP, Solomonik A. Inadequate: A metaphor for the lives of low-income women? Breastfeed Med. 2009 October 14;4(S1):S41-43. (149) Redman S, Watkins J, Evans L, Lloyd D. Evaluation of an Australian intervention to encourage breast-feeding in primiparous women. Health Promotion International. 1995;10(2):101-13. (150) Gatti L. Maternal perceptions of insufficient milk supply in breastfeeding. J Nurs Scholarsh. 2008;40(4):355-63. (151) Li R, Fein SB, Chen J, Grummer-Strawn LM. Why mothers stop breastfeeding: mothers' self-reported reasons for stopping during the first year. Pediatrics. 2008;122(Supplement 2):S69-76. (152) Hoddinott P, Pill R. Nobody actually tells you: a study of infant feeding. British Journal of Midwifery. 1999;7(9):558-65. (153) Murray L, Woolgar M, Murray J, Cooper P. Self-exclusion from health care in women at high risk for postpartum depression. Journal of Public Health Medicine. 2003 June 1, 2003;25(2):131-37. (154) Forster DA, McLachlan HL, Lumley J. Factors associated with breastfeeding at six months postpartum in a group of Australian women. Int Breastfeed J. 2006;1:18. (155) Taveras EM, Capra AM, Braveman PA, Jensvold NG, Escobar GJ, Lieu TA. Clinician support and psychosocial risk factors associated with breastfeeding discontinuation. Pediatrics. 2003 July 2003;112(1):108-15. (156) Donath SM, Amir LH. The relationship between prenatal infant feeding intention and initiation and duration of breastfeeding: a cohort study. Acta Paediatr. 2003;92:352 - 56. (157) Clifford J, McIntyre E. Who supports breastfeeding? . Breastfeed Rev. 2008;16(2):9-19. (158) Dennis C-L, Hodnett E, Gallop R, Chalmers B. The effect of peer support on breast-feeding duration among primiparous women: a randomized controlled trial. Canadian Medical Association Journal. 2002 Jan 8;166(1):21-8. (159) Graffy J, Taylor J, Williams A, Eldridge S. Randomised controlled trial of support from volunteer counsellors for mothers considering breast feeding. Br Med J. 2004 Jan 3;328(7430):26. (160) Stamp GE, Casanova HT. A breastfeeding study in a rural population in South Australia. Rural Remote Health. 2006 Apr-Jun 6(2):37KB. (161) Arora S, McJunkin C, Wehrer J, Kuhn P. Major factors influencing breastfeeding rates: Mother's perception of father's attitude and milk supply. Pediatrics. 2000 November 1, 2000;106(5):e67-. (162) Ganther JM, Wiederholt JB, Kreling DH. Measuring patients' medical care preferences: Care seeking versus self-treating. Medical Decision Making. 2001;21(2):133-40. (163) Hoddinott P, Pill R. A qualitative study of women’s views about how health professionals communicate about infant feeding. Health Expectations. 2000;3(4):224- 33. (164) McCarter-Spaulding D. Is breastfeeding fair? Tensions in feminist perspectives on breastfeeding and the family. J Hum Lact. 2008;24(2):206-15. (165) Harvey D. Understanding Australian rural women's ways of achieving health and wellbeing - a metasynthesis of the literature. Rural and Remote Health 2007. (166) Rich A. Of woman born: Motherhood as experience and institution. London: Virago 1977. (167) Kendall-Tackett KA. Violence against women and the perinatal period:The impact of lifetime violence and abuse on pregnancy, postpartum and breastfeeding. Trauma, violence & abuse. 2007 July;8(3):344-53.

261

(168) Schmied V, Barclay L. Connection and pleasure, disruption and distress: women's experience of breastfeeding. J Hum Lact. 1999 Dec;15(4):325-34. (169) Goodman PH. That's what they're for: Social support in the adaptation to breastfeeding. A health promotion model [Ed.D.]. Toronto, Canada: University of Toronto 1999. (170) Ingram J, Johnson D. A feasibility study of an intervention to enhance family support for breast feeding in a deprived area in Bristol, UK. Midwifery. 2004 2004;20(4):367-79. (171) Scott JA, Aitkin I, Binns CW, Aroni RA. Factors associated with the duration of breastfeeding amongst women in Perth, Australia. Acta Paediatr. 1999 Apr;88(4):416- 21. (172) Rempel LA, Rempel JK. Partner influence on health behavior decision-making: increasing breastfeeding duration. Journal of Social and Personal Relationships. 2004;21(1):92-111. (173) Baranowski T, Bee DE, Rassin DK, Richardson CJ, Brown JP, Guenther N, et al. Social support, social influence, ethnicity and the breastfeeding decision. Social Science and Medicine. 1983;17(21):1599-611. (174) Lu MC, Lange L, Slusser W, Hamilton J, Halfon N. Provider encouragement of breast-feeding: evidence from a national survey. Obstet Gynecol. 2001 Feb;97(2):290- 5. (175) Fletcher R. The Dad Factor: How father-baby bonding helps a child for life. Warriewood NSW, Australia: Finch Publishing 2011. (176) Tohotoa J, Maycock B, Hauck Y, Howat P, Burns S, Binns C. Dads make a difference: an exploratory study of paternal support for breastfeeding in Perth, Western Australia. Int Breastfeed J. 2009;4(1):15. (177) Tarkka M-T. Predictors of maternal competence by first-time mothers when the child is 8 months old. Journal of Advanced Nursing. 2003;41(3):233-40. (178) Sheehan A, Schmied V, Cooke M. Australian women's stories of their baby- feeding decisions in pregnancy. Midwifery. 2003 Dec;19(4):259-66. (179) Schmied V, Beake S, Sheehan A, McCourt C, Dykes F. Women’s Perceptions and Experiences of Breastfeeding Support: A Metasynthesis. Birth. 2011;38(1):49-60. (180) Kramer MS. "Breast is best" :The evidence. Early Human Development. 2010;86(11):729-32. (181) Wright AL. The rise of breastfeeding in the United States. Pediatr Clin North Am. 2001 Feb;48(1):1-12. (182) Binns C, Lee M. Hippocrates lives on: Nutrition and public health education in the 21st century. Asia Pac J Public Health. 2006 September 1, 2006;18(3):1-2. (183) McBride-Henry K, Shaw R. Giving breastmilk as being-with. In: Shaw R, Bartlett A, eds. Giving Breastmilk: Body ethics and contemporary breastfeeding practice Bradford, ON, Canada Demeter Press 2010:191-204. (184) Prochaska JO, Di Clemente CC. Trans-theoretical therapy - toward a more integrative model of change. Psychotherapy: Theory, Research & Practice. 1982 Fall;19(3):276-88. (185) Pollard D. Impact of a feeding log on breastfeeding duration and exclusivity. Maternal and Child Health Journal. 2010;15(3):395-400. (186) Wilhelm SL, Stepans MBF, Hertzog M, Rodehorst TKC, Gardner P. Motivational interviewing to promote sustained breastfeeding. JOGNN. 2006;35:340-48. (187) Ramji S. Breastfeeding – A 3D experience. Indian J Med Res. 2011 August;134(2):141-42. (188) Wallace L, Kosmala-Anderson J. Training needs survey of midwives, health visitors and voluntary-sector breastfeeding support staff in England. Matern Child Nutr. 2007;3(1):25-39. (189) Wolf JH. The social and medical construction of lactation pathology. Women & Health. 2000;30(3):93-110.

262

(190) Howard C, Howard F, Lawrence R, Andresen E, DeBlieck E, Weitzman M. Office prenatal formula advertising and its effect on breast-feeding patterns. Obstet Gynecol. 2000 Feb;95(2):296-303. (191) Wright CM, Waterston AJR. Relationships between paediatricians and infant formula milk companies. Arch Dis Child. 2006;91(5):383-85. (192) Li L, Zhang M, Scott JA, Binns CW. Factors associated with the initiation and duration of breastfeeding by Chinese mothers in Perth, Western Australia. J Hum Lact. 2004 May 1, 2004;20(2):188-95. (193) DiGirolamo AM, Grummer-Strawn LM, Fein SB. Do perceived attitudes of physicians and hospital staff affect breastfeeding decisions? Birth. 2003;30(2):94-100. (194) Dennis CL. Breastfeeding initiation and duration: a 1990-2000 literature review. J Obstet Gynecol Neonatal Nurs. 2002 Jan-Feb;31(1):12-32. (195) Thijs GA. GP's consult and health behaviour change project. Developing a programme to train GPs in communication skills to achieve lifestyle improvements. Patient Educ Couns. 2007 Aug;67(3):267-71. (196) Birden H. The researcher development program: how to extend the involvement of Australian general practitioners in research? Rural and Remote Health. 2007 14 August 2007;7(776). (197) Hailes JF, Wellard SJ. Support for breastfeeding in the first postpartum month: perceptions of breastfeeding women. Breastfeed Rev. 2000 Nov 8(3):5-9. (198) Jones A, Burgess TA, Farmer EA, Fuller J, Stocks NP, Taylor JE, et al. Building research capacity: an exploratory model of GPs' training needs and barriers to research involvement. Aust Fam Physician. 2003 2003 Nov;32(11):957 -60. (199) Houle C, Harwood E, Watkins A, Baum KD. What women want from their physicians: A qualitative analysis. Journal of Women's Health. 2007;16(4):543. (200) Hoddinott P, Pill R. A qualitative study of women's views about how health professionals communicate about infant feeding. Health Expectations. 2000;3(4):224- 33. (201) Chung M, Raman G, Trikalinos T, Lau J, Ip S. Interventions in primary care to promote breastfeeding: an evidence review for the U.S. Preventive Services Task Force. Annals of Internal Medicine. 2008 Oct 21;149(8):565-82. (202) de Oliveira MIC, Camacho LA, Tedstone AE. Extending breastfeeding duration through primary care: a systematic review of prenatal and postnatal interventions. J Hum Lact. 2001;17(4):326 - 43. (203) Spiby H, McCormick F, Wallace L, Renfrew MJ, D'Souza L, Dyson L. A systematic review of education and evidence-based practice interventions with health professionals and breast feeding counsellors on duration of breast feeding. Midwifery. 2009;25(1):50-61. (204) Forster D, McLachlan H, Lumley J, Beanland C, Waldenstrom U, Amir L. Two mid-pregnancy interventions to increase the initiation and duration of breastfeeding: a randomized controlled trial. Birth. 2004 Sep;31(3):176-82. (205) Guise JM, Palda V, Westhoff C, Chan BK, Helfand M, Lieu TA. The effectiveness of primary care-based interventions to promote breastfeeding: systematic evidence review and meta-analysis for the US Preventive Services Task Force. Ann Fam Med. 2003 Jul-Aug;1(2):70-8. (206) Bernaix LW. Nurses' attitudes, subjective norms, and behavioral intentions toward support of breastfeeding mothers. J Hum Lact. 2000 August 1, 2000;16(3):201- 09. (207) Weinstein P, Harrison R, Benton T. Motivating parents to prevent caries in their young children: One-year findings. The Journal of the American Dental Association. 2004;135(6):731-38. (208) Taveras EM, Li R, Grummer-Strawn L, Richardson M, et al. Mothers' and clinicians' perspectives on breastfeeding counseling during routine preventive visits. Pediatrics. 2004;113(5):1384.

263

(209) Stockdale J, Sinclair M, Kernohan WG, Keller JM, Dunwoody L, Cunningham JB, et al. Feasibility study to test DESIGNER BREASTFEEDING: a randomised controlled trial. Evidence Based Midwifery. 2008;6(3):76-82. (210) Golin C, Patel S, Tiller K, Quinlivan E, Grodensky C, Boland M. Start talking about risks: development of a motivational interviewing-based safer sex program for people living with HIV. AIDS and Behavior. 2007;11(0):72-83. (211) General Practice NSW. A Guide to understanding and working with General Practice in NSW. General Practice NSW, 2011:1-59. (212) Australian Health Ministers’ Conference 2009. The Australian National Breastfeeding Strategy 2010-2015. Canberra: Australian Government Department of Health and Ageing; 2009. (213) Goodyear-Smith F, York D, Petousis-Harris H, Turner N, Copp J, Kerse N, et al. Recruitment of practices in primary care research: the long and the short of it. Fam Pr. 2009 February 27, 2009:1-9. (214) Bonuck KA, Lischewski J, Brittner M. Clinical translational research hits the road: RCT of breastfeeding promotion interventions in routine prenatal care. Contemporary Clinical Trials. 2009;30(5):419-26. (215) Campbell MJ. Cluster randomized trials in general (family) practice research. Statistical Methods in Medical Research. 2000;9(2):81-94. (216) Eldridge SM, Ashby D, Feder GS, Rudnicka AR, Ukoumunne OC. Lessons for cluster randomized trials in the twenty-first century: a systematic review of trials in primary care. Clinical Trials. 2004;1(1):80-90. (217) Hordacre A, Howard S, Moretti C, Kalucy E. Moving ahead: Report of the 2006- 2007 Annual Survey of Divisions of General Practice. Adelaide: Department of General Practice, Flinders University, and Australian Government Department of Health and Ageing; 2008. (218) Jolly R. Practice nursing in Australia. Canberra: Parliament of Australia; 2007 17 September. (219) Willis E, Condon J, Litt J. Working relationships between practice nurses and general practitioners in Australia: a critical analysis. Nursing Inquiry. 2000;7(4):239-47. (220) Overland J, Hayes L, Yue DK. Social disadvantage: Its impact on the use of Medicare services related to diabetes in NSW. Aust N Z J Public Health. 2002;26(3):262-65. (221) Pearce C, Phillips C, Hall S, Sibbald B, Porritt J, Yates R, et al. Contributions from the lifeworld: quality, caring and the general practice nurse. Quality in Primary Care. 2009;17:5-13. (222) Bolton SC. Who cares? Offering emotion work as a 'gift' in the nursing labour process. Journal of Advanced Nursing. 2000;32(3):580-86. (223) Watts I, Foley E, Hutchinson R, Pascoe T, Whitecross L, Snowdon T. General practice nursing in Australia. Canberra: Royal Australian College of General Practitioners and Royal College of Nursing Australia; 2004 May. (224) Condon J, Willis E, Litt J. The role of the practice nurse. An exploratory study. Aust Fam Physician. 2000 Mar;29(3):272-7. (225) Keleher H, Joyce CM, Parker R, Piterman L. Practice nurses in Australia: current issues and future directions. MJA. 2007;187(2):108–10. (226) Hegney D, Price K, Patterson E, Martin-McDonald K, Rees S. Australian consumers' expectations for expanded nursing roles in general practice: choice not gatekeeping. Aust Fam Physician. 2004 Oct;33(10):845-9. (227) Halcomb EJ, Davidson PM, Patterson E. Promoting leadership and management in Australian general practice nursing: what will it take? Journal of Nursing Management. 2008 October;16(7):846-52. (228) Duckett S. The Australian health workforce: facts and futures. Aust Health Rev. 2000;23(4):60-77. (229) Rice VH, Stead LF. Nursing interventions for smoking cessation. Cochrane Database Syst Rev. 2008.

264

(230) Zwar N, Hermiz O, Hasan I, Comino E, Middleton S, Vagholkar S, et al. A cluster randomised controlled trial of nurse and GP partnership for care of Chronic Obstructive Pulmonary Disease. BMC Pulmonary Medicine 2008;8(1):8. (231) Halcomb EJ, Davidson PM, Salamonson Y, Ollerton R, Griffiths R. Nurses in Australian general practice: implications for chronic disease management. Journal of Nursing and Healthcare of Chronic Illness in association with Journal of Clinical Nursing. 2008 2 June;17(5a):6-15. (232) Brodribb W. Improving the breastfeeding knowledge and skills of GP registrars University of Queensland; 2009. (233) World Health Organisation, UNICEF. BFHI Section 3: Breastfeeding Promotion and Support in a Baby-friendly Hospital, a 20-hour course for maternity staff. 2006 [cited 2008 March 3]; Available from: http://www.who.int/nutrition/publications/infantfeeding/9789241595018_s3/en/print.html (234) Saadeh R, Akre J. Ten steps to successful breastfeeding: a summary of the rationale and scientific evidence. Birth. 1996 Sep;23(3):154-60. (235) Philipp BL, Merewood A. The Baby-Friendly way: the best breastfeeding start. Pediatr Clin North Am. 2004 2004 Jun;51(3):761-83. (236) Rea MF, Venancio SI, Martines JC, Savage F. Counselling on breastfeeding: assessing knowledge and skills. Bull World Health Organ. 1999;77(6):492-8. (237) King L, Hector D, Webb K. Promoting and supporting breastfeeding in NSW: case studies. Sydney: NSW Centre for Public Health Nutrition; 2005 3 August 2005. (238) Dykes F. Infant feeding initiative: A report evaluating the breastfeeding practice projects 1999-2002. London: Department of Health; 2003. (239) Vitolo MR, Bortolini GA, Dal Bo Campagnolo P, Feldens CA. Effectiveness of a nutrition program in reducing symptoms of respiratory morbidity in children: A randomized field trial. Prev Med. 2008;47(4):384-88. (240) Frank DA, Wirtz SJ, Sorenson JR, Heeren T. Commercial discharge packs and breast-feeding counseling: effects on infant-feeding practices in a randomized trial. Pediatrics. 1987 December 1, 1987;80(6):845-54. (241) Labarere J, Gelbert-Baudino N, Ayral AS, Duc C, Berchotteau M, Bouchon N, et al. Efficacy of breastfeeding support provided by trained clinicians during an early, routine, preventive visit: a prospective, randomized, open trial of 226 mother-infant pairs. Pediatrics. 2005 Feb;115(2):e139-46. (242) Kupratakul J, Taneepanichskul S, Voramongkol N, Phupong V. A randomized controlled trial of knowledge sharing practice with empowerment strategies in pregnant women to improve exclusive breastfeeding during the first six months Postpartum. Journal of the Medical Association of Thailand. 2010 September;93(9):1009 -18. (243) Albernaz E, Victora CG, Haisma H, Wright A, Coward WA. Lactation counseling increases breast-feeding duration but not breast milk intake as measured by isotopic methods. J Nutr. 2003 Jan;133(1):205-10. (244) Wambach KA, Aaronson L, Breedlove G, Domian EW, Rojjanasrirat W, Yeh H- W. A randomized controlled trial of breastfeeding support and education for adolescent mothers. Western Journal of Nursing Research. 2011 June 1;33(4):486-505. (245) Kemp L, Harris E, McMahon C, Matthey S, Vimpani G, Anderson T, et al. Child and family outcomes of a long-term nurse home visitation programme: a randomised controlled trial. Arch Dis Child. 2011;96(6):533-40. (246) Wen LM, Baur LA, Simpson JM, Rissel C, Flood VM. Effectiveness of an early intervention on infant feeding practices and "Tummy Time": A randomized controlled trial. Arch Pediatr Adol Med. 2011;165(8):701-07. (247) Chan-Yeung M, Manfreda J, Dimich-Ward H, Ferguson A, Watson W, Becker A. A randomized controlled study on the effectiveness of a multifaceted intervention program in the primary prevention of asthma in high-risk infants. Arch Pediatr Adolesc Med. 2000 July 1, 2000;154(7):657-63.

265

(248) Feldens CA, Vitolo MR, Drachler MdL. A randomized trial of the effectiveness of home visits in preventing early childhood caries. Community Dentistry & Oral Epidemiology. 2007 Jun;35(3):215-23. (249) Gathwala G, Singh B, Singh J. Effect of kangaroo mother care on physical growth, breastfeeding and its acceptability. Tropical Doctor. 2010 October;40(4):199- 202. (250) Neri I, Allais G, Vaccaro V, Minniti S, Airola G, Schiapparelli P, et al. Acupuncture treatment as breastfeeding support: preliminary data. The Journal of Alternative and Complementary Medicine. 2011 Feb 8;17(2):133-7. (251) Gijsbers B, Mesters I, Knottnerus JA, van Schayck CP. Factors associated with the duration of exclusive breast-feeding in asthmatic families. Health Educ Res. 2008 Feb;23(1):158-69. (252) Kronborg H, Vaeth M, Olsen J, Iversen L, Harder I. Effect of early postnatal breastfeeding support: a cluster-randomized community based trial. Acta Paediatr. 2007 Jul;96(7):1064-70. (253) Gunn J, Lumley J, Chondros P, Young D. Does an early postnatal check-up improve maternal health: results from a randomised trial in Australian general practice. British Journal of Obstetrics & Gynaecology. 1998;105(9):991-97. (254) Bartu A, Sharp J, Ludlow J, Doherty DA. Postnatal home visiting for illicit drug- using mothers and their infants: a randomised controlled trial. Aust N Z J Obstet Gynaecol. 2006 Oct;46(5):419-26. (255) McDonald SJ, Henderson JJ, Faulkner S, Evans SF, Hagan R. Effect of an extended midwifery postnatal support programme on the duration of breast feeding: A randomised controlled trial. Midwifery. 2008;26(1):88-100. (256) Pinelli J, Atkinson SA, Saigal S. Randomized trial of breastfeeding support in very low-birth-weight infants. Arch Pediatr Adolesc Med. 2001 May 1, 2001;155(5):548- 53. (257) McQueen KA, Dennis C-L, Stremler R, Norman CD. A pilot randomized controlled trial of a breastfeeding self-efficacy intervention with primiparous mothers. J Obstet Gynecol Neonatal Nurs. 2011 Jan-Feb;40(1):35-46. (258) Bolam A, Manandhar DS, Shrestha P, Ellis M, Costello AM. The effects of postnatal health education for mothers on infant care and family planning practices in Nepal: A randomised controlled trial. Br Med J. 1998 14 Mar 316(7134):805-11. (259) Gunn TR, Thompson JMD, Jackson H, McKnight S, Buckthought G, Gunn AJ. Does early hospital discharge with home support of families with preterm infants affect breastfeeding success? A randomized trial. Acta Pædiatrica. 2000;89(11):1358-63. (260) McLeod D, Pullon S, Benn C, Cookson T, Dowell A, Viccars A, et al. Can support and education for smoking cessation and reduction be provided effectively by midwives within primary maternity care? Midwifery. 2004 Mar;20(1):37-50. (261) Kools EJ, Thijs C, Kester ADM, van den Brandt PA, de Vries H. A breastfeeding promotion and support program a randomized trial in the Netherlands. Prev Med. 2005;40(1):60-70. (262) Pugh LC, Milligan RA. Nursing intervention to increase the duration of breastfeeding. Appl Nurs Res. 1998 Nov;11(4):190-4. (263) Minkovitz C, Strobino D, Hughart N, Scharfstein D, Guyer B, and the Healthy Steps Evaluation T. Early effects of the Healthy Steps for Young Children program. Arch Pediatr Adolesc Med. 2001 April 1, 2001;155(4):470-79. (264) Pugh LC, Milligan RA, Frick KD, Spatz D, Bronner Y. Breastfeeding duration, costs, and benefits of a support program for low-income breastfeeding women. Birth. 2002 Jun;29(2):95-100. (265) Petrova A, Ayers C, Stechna S, Gerling JA, Mehta R. Effectiveness of exclusive breastfeeding promotion in low-income mothers: A randomized controlled study. Breastfeed Med. 2009 2011/09/18;4(2):63-69.

266

(266) Bunik M, Shobe P, O'Connor ME, Beaty B, Langendoerfer S, Crane L, et al. Are 2 weeks of daily breastfeeding support insufficient to overcome the influences of formula? Academic Pediatrics. 2010 February;10(1):21-28. (267) Thurman SE, Allen PJ. Integrating lactation consultants into primary health care services: are lactation consultants affecting breastfeeding success? Pediatr Nurs. 2008;34(5):419-25. (268) Kemp L, Harris E, McMahon C, Matthey S, Vimpani G, Anderson T, et al. Miller Early Childhood Sustained Home-visiting (MECSH) trial: design, method and sample description. BMC Public Health. 2008;8(1):424. (269) Oddy W, Holt P, Sly P, Read A, Landau L, Stanley F, et al. Association between breast feeding and asthma in 6 year old children: findings of a prospective birth cohort study. BMJ. 1999 25 September;319(7213):815-9. (270) Gijsbers B. Promotion of exclusive breastfeeding for six months in asthmatic families [PhD Thesis]. Maastricht, The Netherlands: Maastricht University; 2007. (271) Bolam A, Manandhar DS, Shrestha P, Ellis M, Costello AM. The effects of postnatal health education for mothers on infant care and family planning practices in Nepal: a randomised controlled trial. BMJ. 1998 Mar 14;316(7134):805-11. (272) World Health Organization. Infant and young child feeding data by country The WHO Global Data Bank on Infant and Young Child Feeding 2012 [cited 2012 March 7]; Available from: http://www.who.int/nutrition/databases/infantfeeding/countries/en/index.html (273) Becker A, Watson W, Ferguson A, Dimich-Ward H, Chan-Yeung M. The Canadian asthma primary prevention study: outcomes at 2 years of age. Journal of Allergy & Clinical Immunology. 2004 Apr;113(4):650-6. (274) Taveras EM, Rifas-Shiman SL, Scanlon KS, Grummer-Strawn LM, Sherry B, Gillman MW. To what extent is the protective effect of breastfeeding on future overweight explained by decreased maternal feeding restriction? Pediatrics. 2006 Dec;118(6):2341-8. (275) ACOG Committee Opinion No. 423. Motivational Interviewing a tool for behaviour change. Obstetrics & Gynaecology. 2009 January;113(1):243-46. (276) Lawlor-Smith C, McIntyre E, Bruce J. Effective breastfeeding support in a general practice. Aust Fam Physician. 1997 1997 May;26(5):573-5. (277) Halcomb EJ, Davidson P, Daly J, Griffiths R, Yallop J, Tofler G. Nursing in Australian general practice: directions and perspectives. Aust Health Rev. 2005 May 2005;29(2):156-66. (278) Campbell MK, Elbourne DR, Altman DG. CONSORT statement: extension to cluster randomised trials. BMJ. 2004;328(7441):702-08. (279) Dillman DA. Mail and telephone surveys. New York: Wiley-Interscience 1978. (280) Dillman DA. The design and administration of mail surveys. Annu Rev Sociol. 1991;17:225-49. (281) Oakley A, Rajan L, Robertson P. A comparison of different sources of information about pregnancy and childbirth. J Biosoc Sci. 1990 Oct;22(4):477-87. (282) Rudman A, Waldenstrom U. Critical views on postpartum care expressed by new mothers. BMC Health Serv Res. 2007;7(178). (283) Perez-Escamilla R. Evidence based breast-feeding promotion: The Baby- Friendly Hospital Initiative. J Nutr. 2007 February 1, 2007;137(2):484-87. (284) UNICEF/WHO. Baby Friendly Hospital Initiative, revised, updated and expanded for integrated care. Section 1, Background and Implementation, Preliminary Version 2006. (285) Murphy AW, Esterman A, Pilotto LS. Cluster randomized controlled trials in primary care: an introduction. Eur J Gen Pract. 2006;12(2):70-3. (286) Australian Institute of Health and Welfare. Rural, regional and remote health: A guide to remoteness classifications AIHW cat. no. PHE 53. Canberra: AIHW; 2004 March.

267

(287) Greenhalgh T. Role of routines in collaborative work in healthcare organisations. BMJ. 2008 November 29;337:a2448. (288) Jakobsen M, Sodemann M, Molbak K, Alvarenga I, Aaby P. Promoting breastfeeding through health education at the time of immunizations: a randomized trial from Guinea Bissau. Acta Paediatr. 1999 Jul;88(7):741-47. (289) Benger JR, Pearce V. Simple intervention to improve detection of child abuse in emergency departments. BMJ. 2002 Mar 30;324(7340):780. (290) Miller W. Motivational interviewing with problem drinkers. Behavioural Psychotherapy. 1983;11:147-72. (291) Rogers CR. The necessary and sufficient conditions of therapeutic personality change. J Consult Clin Psychol. 1992 December;60(6):827-32. (292) Miller WR, Rollnick S. Motivational interviewing: preparing people for change (2nd ed ). New York: The Guilford Press 2002. (293) Suarez M, Mullins S. Motivational interviewing and pediatric health behavior interventions. J Dev Behav Pediatr. 2008 Oct;29(5):417-28. (294) Resnicow K, Davis R, Rollnick S. Motivational interviewing for pediatric obesity: Conceptual issues and evidence review. J Am Diet Assoc. 2006;106(12):2024-33. (295) Hecht J, Borrelli B, Breger RKR, DeFrancesco C, Ernst D, Resnicow K. Motivational interviewing in community-based research: experiences from the field. Ann Behav Med. 2005;29(2, suppl):29-34. (296) Miller ST, Beech BM. Rural healthcare providers question the practicality of motivational interviewing and report varied physical activity counseling experience. Patient Educ Couns. 2009;76(2):279-82. (297) Rubak S, Sandboek A, Lauritzen T, Christensen B. Motivational interviewing: a systematic review and meta-analysis. Br J Gen Pract. 2005 April 2005;55:305-12. (298) Moyers TB, Miller WR, Hendrickson SML. How does motivational interviewing work? Therapist interpersonal skill predicts client involvement within motivational interviewing sessions. J Consult Clin Psychol. 2005 Aug;73(4):590-8. (299) Gross SM, Caulfield LE, Bentley ME, Bronner Y, Kessler L, Jensen J, et al. Counseling and motivational videotapes increase duration of breast-feeding in African- American WIC participants who initiate breast-feeding. J Am Diet Assoc. 1998 Feb;98(2):143-48. (300) Freudenthal JJ, Bowen DM. Motivational interviewing to decrease parental risk- related behaviors for early childhood caries. J Dent Hyg. 2010;84(1):29-34. (301) Weinstein P, Harrison R, Benton T. Motivating mothers to prevent caries: confirming the beneficial effect of counseling. J Am Dent Assoc. 2006 Jun;137(6):789- 93. (302) Emmons KM, Hammond SK, Fava JL, Velicer WF, Evans JL, Monroe AD, et al. A randomized trial to reduce passive smoke exposure in low-income households with young children. Pediatrics. 2001 Jul;108(1):18-24. (303) Schwartz RP, Hamre R, Dietz WH, Wasserman RC, Slora EJ, Myers EF, et al. Office-based motivational interviewing to prevent childhood obesity: a feasibility study. Arch Pediatr Adolesc Med. 2007 May;161(5):495-501. (304) Nock MK, Kazdin AE. Randomized controlled trial of a brief intervention for increasing participation in parent management training. J Consult Clin Psychol. 2005 Oct;73(5):872-9. (305) Shaw DS, Dishion TJ, Supplee L, Gardner F, Arnds K. Randomized trial of a family-centered approach to the prevention of early conduct problems: 2-year effects of the family check-up in early childhood. J Consult Clin Psychol. 2006 Feb;74(1):1-9. (306) O'Connor AM, Legare F, Stacey D. Risk communication in practice: the contribution of decision aids. BMJ. 2003 Sep 27;327(7417):736-40. (307) Stacey D, Bennett CL, Barry MJ, Col Nananda F, Eden KB, Holmes-Rovner M, et al. Decision aids for people facing health treatment or screening decisions. Cochrane Database Syst Rev. 2011(10).

268

(308) Dennington V, Humeniuk R, Newcombe D, Ali R, Vial R. WHO ASSIST Phase III: Results from the Australian arm of an international RCT of a brief intervention for illicit drug use linked to the scores on the Alcohol, Smoking and Substance Involvement Screening Test (ASSIST). Parkside SA: Government of South Australia, South Adelaide Health Services; 2007. (309) New South Wales Health. SmokeCheck Project. Sydney: NSW Health; 2006. (310) Fiore M, Jaén C, Baker T, et al. Chapter 3 Clinical interventions for tobacco use and dependence. In: Public Health Service, ed. Clinical practice guideline treating tobacco use and dependence: 2008 update. Rockville, MD: U.S. Department of Health and Human Services 2009. (311) Cox S. The scientification of breastfeeding: help or hindrance to mothers who breastfeed their babies? Breastfeeding - Ancient Art - Modern Miracle ABA International Breastfeeding Conference; 2001 13-15 Sep; Brisbane: Australian Breastfeeding Association; 2001. p. 20-24. (312) Coreil J, Bryant CA, Westover BJ, Bailey D. Health professionals and breastfeeding counseling: client and provider views. J Hum Lact. 1995 December 1, 1995;11(4):265-71. (313) Valdes V, Pugin E, Schooley J, Catalan S, Aravena R. Clinical support can make the difference in exclusive breastfeeding success among working women. Journal of Tropical Pediatrics. 2000 Jun;46(3):149-54. (314) Elliott-Rudder ME. Support for ongoing breastfeeding: a randomised clinical trial endorsed by the APNA Policy & Research Committee. Primary Times. 2008 September;1:16-17. (315) Casey D. Using action research to change health-promoting practice. Nursing and health sciences. 2007 13 February;9(1):5-13. (316) Lewin S, Skea Z, Entwistle V, Zwarenstein M, Dick J. Interventions for providers to promote a patient-centred approach in clinical consultations. Art. No.: CD003267. Cochrane Database of Systematic Reviews. 2001(4). (317) Australian Breastfeeding Association. Breastfeeding Confidence. 2nd ed. Glen Iris, Victoria: Australian Breastfeeding Association 2004. (318) Hector D, King L, Webb K. Overview of recent reviews of interventions to promote and support breastfeeding: NSW Centre for Public Health Nutrition for NSW Health; 2004 September 2004. (319) McIntyre E, Pisaniello D, Gun R, Sanders C, Frith D. Balancing breastfeeding and paid employment: a project targeting employers, women and workplaces. Health Promotion International. 2002 Sep;17(3):215-22. (320) Clifford A, Shakeshaft A, Deans C. How and when health-care practitioners in Aboriginal Community Controlled Health Services deliver alcohol screening and brief intervention, and why they don't: A qualitative study. Drug and Alcohol Review. 2011:e1-7. (321) Noel-Weiss J, Rupp A, Cragg B, Bassett V, Woodend AK. Randomized controlled trial to determine effects of prenatal breastfeeding workshop on maternal breastfeeding self-efficacy and breastfeeding duration. J Obstet Gynecol Neonatal Nurs. 2006 Sep-Oct;35(5):616-24. (322) Byrne J, McIntyre E. A lactation course that is available for study by correspondence. Journal - Australian College of Midwives. 1997 Mar;10(1):3. (323) Allen J, Hector DJ. Benefits of Breastfeeding. NSW Public Health Bull 2005:42- 46. (324) Baxter J, Smith JP. Breastfeeding and infants’ time use. Melbourne: Australian Institute of Family Studies; 2009. (325) Scanlan JM, Care WD, Udod S. Unravelling the unknowns of reflection in classroom teaching. Journal of Advanced Nursing. 2002;38(2):136-43. (326) Elliott-Rudder ME. Enhancing women's motivation to breastfeed: our choice of words, her choice of action, is change possible? MumsRight Podcast 2007.

269

(327) Miller WR. Motivational factors in addictive behaviours. In: William R Miller KMC, ed. Rethinking substance abuse. 1 ed. New York: The Guilford Press 2006:134- 52. (328) Varga E. Predictors of breastfeeding duration in primiparous women receiving motivational interviewing [M.S. thesis]. United States - Wyoming: University of Wyoming; 2005. (329) Salzer MS, Rappaport J, Segre L. Mental health professionals' support of self- help groups. Journal of Community & Applied Social Psychology. 2001;11(1):1-10. (330) Hardiman E. Referral to consumer-run programs by mental health providers: A national survey. Community Mental Health Journal. 2007;43(3):197-210-10. (331) Gibbs G. Learning by Doing: A guide to teaching and learning methods. Oxford: Further Education Unit, Oxford Polytechnic 1988. (332) Reid B. `But we're doing it already!' Exploring a response to the concept of reflective practice in order to improve its facilitation. Nurse Education Today. 1993;13(4):305-09. (333) Armstrong HC. Breastfeeding promotion: training of mid-level and outreach health workers. Int J Gynaecol Obstet. 1990;31 Suppl 1:91-103; discussion 11-3. (334) Reason P. Three approaches to participative inquiry. In: Denzin NK, Lincoln YS, eds. Handbook of qualitative research. Thousand Oaks, California: Sage Publications 1994:324-39. (335) Smith LM. Biographical method. In: Denzin NK, Lincoln YS, eds. Handbook of qualitative research. Thousand Oaks, California: Sage Publications 1994:286-305. (336) Ruhe MB, Gotler RS, Goodwin MA, Stange KC. Physician and staff turnover in community primary care practice. Journal of Ambulatory Care Management Society and Ambulatory Care. 2004 July/August/September;27(3):242-48. (337) The Royal Australian College of General Practitioners. Breastfeeding. CHECK 1994. (338) The Royal Australian College of General Practitioners. Breastfeeding. CHECK 1999. (339) The Royal Australian College of General Practitioners. Breastfeeding. CHECK 2007. (340) Arbon S, Byrne J. The reliability of a breastfeeding questionnaire. Breastfeed Rev. 2001;9(2):23-32. (341) Cooklin AR, Donath SM, Amir LH. Maternal employment and breastfeeding: results from the longitudinal study of Australian children. Acta Paediatr. 2008 May;97(5):620-3. (342) Collins K, O'Cathain A. The continuum of patient satisfaction - from satisfied to very satisfied. Soc Sci Med. 2003;57(12):2465-70. (343) Martens PJ. Does breastfeeding education affect nursing staff beliefs, exclusive breastfeeding rates, and Baby-Friendly Hospital Initiative compliance? The experience of a small, rural Canadian hospital. J Hum Lact. 2000 Nov;16(4):309-18. (344) Coleman T, Wilson A. Anti-smoking advice in general practice consultations: general practitioners' attitudes, reported practice and perceived problems. Br J Gen Pract. 1996;46(403):87-91. (345) Greenhalgh T, Taylor R. Papers that go beyond numbers (qualitative research). Br Med J. 1997 Sep 20;315(7110):740-3. (346) Clark J. How to peer review a qualitative manuscript. In: Godlee F, Jefferson T, eds. Peer Review in Health Sciences. 2nd ed. London: BMJ Books 2003:219-35. (347) Kitto SC, Chesters J, Grbich C. Quality in qualitative research: Criteria for authors and assessors in the submission and assessment of qualitative research articles for the Medical Journal of Australia. MJA. 2008;188( 4):243-46. (348) Bazeley P. Mixed methods in management research. In: Thorpe R, Holt R, eds. The SAGE dictionary of qualitative management research. London: Sage 2008:133-36. (349) Silverman D. Doing qualitative research: a practical handbook. 3rd ed. London: Sage Publications 2010.

270

(350) Creswell JW, Shope R, Plano Clark VL, Green DO. How interpretive qualitative research extends mixed methods research. Research in the Schools. 2006 Spring;13(1):1-11. (351) Fernald DH, Duclos CW. Enhance your team-based qualitative research. Annals of Family Medicine. 2005;3(4):360-64. (352) Elliott-Rudder M. Researcher networking drives change: an autoenthnographic narrative analysis from medical graduate to primary health researcher. Aust J Prim Health. 2010 17 March;16(1):108-15. (353) Agboado G, Michel E, Jackson E, Verma A. Factors associated with breastfeeding cessation in nursing mothers in a peer support programme in Eastern Lancashire. BMC Pediatrics 2010:3. (354) Scott JA, Binns CW. Factors associated with the initiation and duration of breastfeeding: a review of the literature. Australian Journal of Nutrition and Dietetics. 1998 2 Jun 1998;55(2):51-61. (355) Bai Y. Factors influencing continuation of exclusive breastfeeding for six months: An application of the theory of planned behavior [PhD thesis]. United States -- Indiana: Indiana University; 2007. (356) Hawkins SS, Griffiths LJ, Dezateux C, Law C, Millennium Cohort Study Child Health G. Maternal employment and breast-feeding initiation: findings from the Millennium Cohort Study. Paediatric and Perinatal Epidemiology. 2007 May;21(3):242- 7. (357) Torvaldsen S, Roberts C, Simpson J, Thompson J, Ellwood D. Intrapartum epidural analgesia and breastfeeding: a prospective cohort study. Int Breastfeed J. 2006;1(1):24. (358) Hill PD, Ledbetter RJ, Kavanaugh KL. Breastfeeding patterns of low-birth- weight infants after hospital discharge. JOGNN. 1997;26(2):189-97. (359) Donath SM, Amir LH. Effect of gestation on initiation and duration of breastfeeding. Archives of Disease in Childhood Fetal & Neonatal Edition. 2008 Nov;93(6):F448-50. (360) Hall S, Vogt F, Marteau TM. A short report: survey of practice nurses' attitudes towards giving smoking cessation advice. Fam Pr. 2005 December 1;22(6):614-16. (361) Barclay L, Everitt L, Rogan F, Schmied V, Wyllie A. Becoming a mother - an analysis of women's experience of early motherhood. Journal of Advanced Nursing. 1997 Apr;25(4):719-28. (362) Achhra A. Health promotion in Australian general practice - A gap in GP training. Aust Fam Physician. 2009 4 August;38(8):605-08. (363) Rollnick S, Butler CC, Stott N. Helping smokers make decisions: the enhancement of brief intervention for general medical practice. Patient Educ Couns. 1997;31(3):191-203. (364) Rollnick S, Mason P, Butler C. Health behavior change: a guide for practitioners. London: Harcourt Brace and Co Ltd 1999. (365) Shah PS, Aliwalas LI, Shah V. Breastfeeding or breast milk for procedural pain in neonates. Cochrane Database Syst Rev. 2006;3:CD004950. (366) Bordens KS, Abbott BB. Research design and methods: a process approach. 2nd ed. Mountain View, California: Mayfield Publishing Company 1991. (367) Hoddinott P, Britten J, Pill R. Why do interventions work in some places and not others: a breastfeeding support group trial. Soc Sci Med. 2010 Mar;70(5):769-78. (368) Pallant J. SPSS survival manual: a step by step guide to data analysis using SPSS. 3rd ed. Sydney: Allen & Unwin 2007. (369) Hoddinott P, Pill R, Chalmers M. Health professionals, implementation and outcomes: reflections on a complex intervention to improve breastfeeding rates in primary care. Fam Pr. 2007 Feb;24(1):84-91. (370) Moran VH, Bramwell R, Dykes F, Dinwoodie K. An evaluation of skills acquisition on the WHO/UNICEF Breastfeeding Management Course using the pre-

271

validated Breastfeeding Support Skills Tool (BeSST). Midwifery. 2000 Sep;16(3):197- 203. (371) Larsen JS, Hall EOC, Aagaard H. Shattered expectations: when mothers' confidence in breastfeeding is undermined; a metasynthesis. Scand J Caring Sci. 2008;22(4):653-61. (372) Dozier AM. Community engagement and dissemination of effective breastfeeding programs. Breastfeed Med. 2010 Oct;5(4):215-16. (373) Amir LH, Donath S. A systematic review of maternal obesity and breastfeeding intention, initiation and duration. BMC Pregnancy & Childbirth. 2007;7:9. (374) Chezem J, Friesen C. Attendance at Breast-Feeding Support Meetings: Relationship to Demographic Characteristics and Duration of Lactation in Women Planning Postpartum Employment. J Am Diet Assoc. 1999;99(1):83-85. (375) Fallon AB, Hegney D, O'Brien M, Brodribb W, Crepinsek M, Doolan J. An evaluation of a telephone-based postnatal support intervention for infant feeding in a regional Australian city. Birth. 2005 Dec;32(4):291-8. (376) Donath SM, Amir LH. The introduction of breast milk substitutes and solid foods: evidence from the 1995 National Health Survey. Aust N Z J Public Health. 2002 Oct;26(5):481-84. (377) Wijndaele K, Lakshman R, Landsbaugh JR, Ong KK, Ogilvie D. Determinants of Early Weaning and Use of Unmodified Cow's Milk in Infants: A Systematic Review. J Am Diet Assoc. 2009;109(12):2017-28. (378) Baird J, Cooper C, Margetts BM, Barker M, Inskip HM. Changing health behaviour of young women from disadvantaged backgrounds: evidence from systematic reviews. Proc Nutr Soc. 2006;Forthcoming(-1):1-10. (379) Scott JA, Landers MC, Hughes RM, Binns CW. Factors associated with breastfeeding at discharge and duration of breastfeeding. J Paediatr Child Health. 2001 Jun;37(3):254-61. (380) Click ER. Developing a worksite lactation program. MCN, American Journal of Maternal Child Nursing. 2006 Sep-Oct;31(5):313-7. (381) McIntyre E, Turnbull D, Hiller JE. Breastfeeding in public places. J Hum Lact. 1999 Jun;15(2):131-5. (382) McIntyre E, Hiller JE, Turnbull D. Attitudes towards infant feeding among adults in a low socioeconomic community: what social support is there for breastfeeding? Breastfeed Rev. 2001 Mar;9(1):13-24. (383) Racine EF, Frick KD, Strobino D, Carpenter LM, Milligan R, Pugh LC. How motivation influences breastfeeding duration among low-income women. J Hum Lact. 2009;25(2):173-81. (384) O'Brien ML, Buikstra E, Fallon T, Hegney D. Strategies for success: a toolbox of coping strategies used by breastfeeding women. J Clin Nurs. 2009;18:1574-82. (385) Moodie R. The art and science of health promotion. Australian Journal of Primary Health Interchange. 2000 2000;6(3-4):61-8. (386) Kruijver IPM, Kerkstra A, Francke AL, Bensing JM, van de Wiel HBM. Evaluation of communication training programs in nursing care: a review of the literature. Patient Educ Couns. 2000;39(1):129-45. (387) AIHW. National breastfeeding indicators: workshop report. 2011 [cited 2012 January 21]; [Cat. no. PHE 146]. Available from: http://www.aihw.gov.au/publication- detail/?id=10737419577 (388) Wolf JB. Is breast really best? Risk and total motherhood in the National Breastfeeding Awareness Campaign. Journal of Health Politics, Policy & Law. 2007;32(4):595-636. (389) Heinig MJ. The burden of proof: A commentary on "Is breast really best: Risk and total motherhood in the National Breastfeeding Awareness Campaign". J Hum Lact. 2007 November 1, 2007;23(4):374-76.

272

(390) Hopkinson JM. Response to "Is breast really best? Risk and total motherhood in the National Breast-Feeding Awareness Campaign". Journal of Health Politics Policy and Law 2007 August 1, 2007;32(4):637-48. (391) Walker M. International breastfeeding initiatives and their relevance to the current state of breastfeeding in the United States. J Midwifery Womens Health. 2007 Nov-Dec;52(6):549-55. (392) World Health Organisation. International Code of Marketing of Breast-Milk Substitutes. 1981 [cited 4 May 2007]; Available from: http://www.who.int/nutrition/publications/code_english.pdf (393) Marley SC, Carbonneau K, Lockner D, Kibbe D, Trowbridge F. Motivational interviewing skills are positively associated with nutritionist self-efficacy. J Nutr Educ Behav. 2011 Jan-Feb;43(1):28-34. (394) Parker R, Keleher H, Forrest L. The work, education and career pathways of nurses in Australian general practice. Aust J Prim Health. 2011;17(3):227-32. (395) Creswell JW. Research design: Qualitative and quantitative approaches. Thousand Oaks CA: Sage Publications 1994. (396) Cohen DJ, Crabtree BF, Etz RS, Balasubramanian BA, Donahue KE, Leviton LC, et al. Fidelity versus flexibility: Translating evidence-based research into practice. Am J Prev Med. 2008;35(5, Supplement):S381-S89. (397) Godwin M, Ruhland L, Casson I, MacDonald S, Delva D, Birtwhistle R, et al. Pragmatic controlled clinical trials in primary care: the struggle between external and internal validity. BMC Med Res Methodol. 2003;3(1):28. (398) Wright DB. Care in the country: A historical case study of long-term sustainability in 4 rural health centers. Am J Public Health. 2009;99(9):1612-18. (399) Woolf SH. The meaning of translational research and why it matters. Jama. 2008 Jan 9;299(2):211-3. (400) Pigeot I, De Henauw S, Foraita R, Jahn I, Ahrens W. Primary prevention from the epidemiology perspective: Three examples from the practice. BMC Med Res Methodol. 2010;10(1):10. (401) Taylor LK, Lim K, Neville SE. Newborn feeding practices at the time of discharge from hospital in NSW in 2007: a descriptive study. NSW Public Health Bull. 2010;20(12):177-81. (402) Raupach JC, Pilotto LS. Randomised trials within the general practice evaluation program: why so few? Aust Fam Physician. 2001 2001 May;30(5):504 -7. (403) Wilkinson D, Beilby J, Pilotto LS. Are evaluation and research mutually exclusive for division of general practice? Aust Fam Physician. 2001;30(1):10. (404) Zwar NA, Weller DP, McCloughan L, Traynor VJ. Supporting research in primary care: are practice-based research networks the missing link? MJA. 2006;185(2):110-13. (405) Farmer E, Pilotto L. Capacity building in general practice research. Not a quick fix. Aust Fam Physician. 2001;40(40):309. (406) Macfarlane F, Shaw S, Greenhalgh T, Carter YH. General practices as emergent research organizations: a qualitative study into organizational development. Fam Pr. 2005;22(3):298-304. (407) Barnett L, Holden L, Donoghue D, Passey M, Birden H. What's needed to increase research capacity in rural primary health care? Aust J Prim Health. 2005;11(3):45-53. (408) Dickey LL, Gemson DH, Carney P. Office system interventions supporting primary care-based health behavior change counseling. Am J Prev Med. 1999 Nov;17(4):299-308. (409) Tobacco and Health Branch. "Let's take a moment" Quit smoking brief intervention - a guide for all health professionals. In: Health NDo, ed. Sydney: Better Health Centre 2005. (410) Eaton CB, McBride PE, Gans KA, Underbakke GL. Teaching nutrition skills to primary care practitioners. J Nutr. 2003;133(2):563S-66.

273

(411) Hector D, King L, Webb K, Heywood P. Factors affecting breastfeeding practices: applying a conceptual framework. NSW Public Health Bull 2005:52-55. (412) Dyson L, Renfrew MJ, McFadden A, McCormick F, Herbert G, Thomas J. Policy and public health recommendations to promote the initiation and duration of breastfeeding in developed country settings. Public Health Nutr. 2010 January;13(1):137-44. (413) Roski J, Jeddeloh R, An L, Lando H, Hannan P, Hall C, et al. The impact of financial incentives and a patient registry on preventive care quality: increasing provider adherence to evidence-based smoking cessation practice guidelines. Prev Med. 2003;36(3):291-99. (414) Douglas KA, Rayner FK, Yen LE, Wells RW, Glasgow NJ, Humphreys JS. Australia’s primary health care workforce — research informing policy. MJA. 2009;191(2):81-84. (415) Westphal M, Taddei J, Venancio S, Bogus C. Breast-feeding training for health professionals and resultant institutional changes. . Bull WHO. 1995;73:461-8. (416) Gardner K, Douglas KA, Gear A, Yen L, Sansoni E, McRae I, et al. Primary Health Care Position Statement: A scoping of the evidence An update for 2009. Canberra: Australian General Practice Network; 2009 5 Mar. (417) Delva D, Jamieson M, Lemieus M. Team effectiveness in academic primary health care teams. Journal of Interprofessional Care. 2008 December;22(6):598-611. (418) Comino EJ, Hermiz O, Flack J, Harris E, Powell Davies G, Harris MF. Using population health surveys to provide information on access to and use of quality primary health care. Aust Health Rev. 2006;30(4):485-95. (419) Heinig MJ. Are there risks to using risk-based messages to promote breastfeeding? J Hum Lact. 2009;25(1):7-8. (420) Sitzia J, Wood N. Patient satisfaction: A review of issues and concepts. Soc Sci Med. 1997;45(12):1829-43.

274

LIST OF APPENDICES

List of Appendices

Appendix 1 Recruitment of general practices...... 276

Appendix 2 Recruitment letter for general practices...... 278

Appendix 3 Mothers' information and consent ...... 279

Appendix 4 Mothers' reminder postcard...... 281

Appendix 5 Mothers' baseline questionnaire in pregnancy ...... 282

Appendix 6 Sources used for mothers' baseline questionnaire...... 287

Appendix 7 Birth data collection fields...... 289

Appendix 8 CATI outcomes survey for mothers at 4 and 6 months ...... 290

Appendix 9 Knowledge and attitudes practice nurse survey...... 295

Appendix 10 Sources for questions to practice nurses ...... 297

Appendix 11 Practice nurse semi-structured interview questions ...... 299

Appendix 12 Practice nurse telephone follow-up process...... 299

Appendix 13 Contents of resource folder for practice nurses...... 300

Appendix 14 Breastfeeding management course manual introduction ...... 302

Appendix 15 Delivery of breastfeeding management course...... 306

Appendix 16 The Conversation Tool Version 3...... 307

Appendix 17 PowerPoint slides used in training workshops ...... 308

275

[this page is blank]

276

Appendix 1 Recruitment of general practices

Recruitment of general practice clusters took place across two rural Divisions of General Practice, which comprised a total of 63 general practices..

Border Division of General Practice  9 practices were eligible for recruitment and 7 of these were recruited.  22 practices were ineligible, including 11 practice located in smaller rural towns. Riverina Division of General Practice and Primary Health  9 practices were eligible for recruitment and 8 of these were recruited.  23 practices ineligible, including 21 practice located in smaller rural towns.

Support for Ongoing Breastfeeding RCT participation Eligible Ineligible practices Total not Total recruited practices recruited 2-3 1 No PNs 0 Out of Other PN PN immunise PN town infants Approached Border 5 4 Riverina 6 3 Not recruited Border (2) (6) (3) (11) (2)* (24) Riverina (1) (1) (21) (1)^ (24) Recruited Border 3 4 7 Riverina 5 3 8 Total 8 7 PN practice nurse;*One Aboriginal Medical Service, one practice closing down; ^Aboriginal Medical Service

277

Appendix 2 Recruitment letter for general practices

6 May 2008

Dr Megan Elliott-Rudder PhD Candidate Telephone: Mobile: [email protected] RURAL CLINICAL SCHOOL

Faculty of Medicine

Dear Doctors

“Support for ongoing breastfeeding” is a randomised controlled trial based in General Practices in Wagga Wagga and Albury Wodonga. The trial aims to increase breastfeeding rates and mothers’ satisfaction by providing support for mothers from trained practice nurses.

We know that  exclusive breastfeeding to six months and ongoing breastfeeding beyond twelve months have long-term and short-term health benefits for mothers and infants  many mothers wean earlier, unless they have support  practice nurses see mothers at immunisation visits and often discuss infant feeding with mothers  nurses and doctors promote and support breastfeeding more if they have training or personal experience

Your practice is invited to join this trial which offers 8 hours of training to practice nurses, covering the basics of breastfeeding and of motivational interviewing. A resources package will be provided.

Practices will be randomised to either intervention or control, with training of the intervention group in June/July 2008 and the control group approximately January 2010. Intervention practice nurses will use several structured questions and responses to support breastfeeding mothers who attend for immunisations of two, four and six month old babies.

This project is being conducted through the Wagga Wagga and Albury Wodonga Campuses of the UNSW Rural Clinical School and has ethics approval from UNSW and the support of the Riverina Division of General Practice and Primary Health.

Please do not hesitate to contact me at any time regarding this project.

Yours sincerely, THE UNIVERSITYOF NEW SOUTH WALES RURAL CLINICAL SCHOOL PO BOX 5695, WAGGA WAGGA NSW 2650 AUSTRALIA

Telephone: +61 (2) 6933 5111

Facsimile: +61 (2) 6933 5100

Dr Megan Elliott-Rudder ABN 56 195 873 179 MBBS DRANZCOG GradDipRuralGP FRACGP

278

Appendix 3 Mothers' information and consent

Ethics reference number HREC 07246 Support for mothers’ choices about infant feeding You are invited to be part of a study, to find out how mothers feed their babies in their first year. We want to learn about mothers’ choices, and what sort of support is helpful for mothers. The aim is to lead to more support for mothers and their babies in the future. You are invited to be part of the study because you will be having a baby soon.

Description of study If you decide to be part of the study we need to have your consent (form enclosed). We would like you to fill out a questionnaire and to return it with the consent form (reply paid envelope enclosed). If you prefer to email, see below.

A researcher will telephone or email some mothers to fill in another questionnaire when their baby is 4, 6 and 9 months old. It will take about 20 minutes to answer the questions. Some mothers will also be interviewed after their baby is 6 months old. The interview will take about 45 minutes.

With your consent, a researcher will check hospital birth records to find out when and how your baby is born, your baby’s birth weight and sex, and how you choose to feed your baby at first. With your consent, we may also check with your General Practice about your feeding choices.

Your privacy is important to us. Whenever we use your information we will use a code instead of your name. This study has been approved by the Joint Hospitals’ Ethics Committee. A Wodonga Hospital staff member has mailed this letter to you.

Confidentiality and disclosure of information Any information that is collected for this study and that can identify you will remain strictly confidential. We plan to publish the results in a PhD thesis and in scientific journals. We will give a summary of the results to you, to local General Practices and to the Australian Breastfeeding Association, of which one of the research team is a member. In any publication, information will be provided in such a way that you cannot be personally identified.

Complaints If you have any concerns or complaints about this study, please contact the Ethics Secretariat, The University of New South Wales, Sydney 2052 Australia (phone 9385 4234, fax 9385 6648, email [email protected]). Any complaint you make will be investigated promptly and you will be informed of the outcome.

Your consent Your decision whether or not to participate will not prejudice, change or affect in any way your future relations with Wodonga Regional Health Service, your hospital, your doctors or nurses. If you decide to participate, you are free to change your mind and withdraw your consent at any time (form enclosed).

If you would prefer an email questionnaire or if you have any questions, please contact me: Megan Elliott-Rudder Phone 02 6933 5205, email [email protected]

(Signed) (Signed)

Megan Elliott-Rudder Professor Louis Pilotto Researcher Head, Rural Clinical School, UNSW

279

THE UNIVERSITY OF NEW SOUTH WALES/WODONGA REGIONAL HEALTH SERVICE HREC Approval No 07246

CONSENT FORM

Support for Mothers’ Choices about Infant Feeding

Sign here if you are willing to be part of this study. You are making a decision whether or not to participate. This page will be stored separately to your answers to other questions.

………………………………………………………………….……………… ……………………………………………………………..….……. Your signature as research participant Signature of witness

………………………………………………………………………………… …………………………………………………………………………. Please PRINT your full name Please PRINT name of witness

………………………………………………………………………………… …………………………………………………………………………. Your date of birth Nature of witness eg friend, relative

………………………………………………………………………………… Today’s date

I would like a summary when the study is completed Yes No

I am willing to be asked about an interview after Yes my baby is 6 months old No

My mailing address is: My telephone contacts are:

( ) …………………………………………………………………………….………………..… ………………………………………………………………………………… Street address or PO Box Home phone

……………………………………………………………………………….………………… ……………………………………………………………….……………… Suburb/Town Postcode Mobile phone

My email address is: ………………………………………………………………………………………………………………………….

If my phone number changes, the research team can ask for my new number:

…………………………………………………………………………..……… Name of person to ask

………………………………………………………………………..………… Please send this page, with the Their phone number questionnaire, in the enclosed envelope (No stamp needed)

280

Appendix 4 Mothers' reminder postcard

281

Appendix 5 Mothers' baseline questionnaire in pregnancy

We are interested in your views on feeding your baby

1. When is your baby due? (Please give your best guess if not sure) ___ / ___ / ______

2. How do you plan to feed your baby after the birth? (Please tick all that apply) 1 Bottle feed (formula) 2 Breastfeed for a few days 3 Breastfeed less than 1 month 4 Breastfeed for 1 - 3 months 5 Breastfeed for 4 - 6 months 6 Breastfeed for 7 - 9 months 7 Breastfeed for 10 - 12 months 8 Breastfeed for 12 months or longer 9 Breastfeed, but no plans for how long 10 Not sure about this

3. When do you think you will introduce solids? 1 Before 4 months 2 From 4 – 6 months 3 Around 6 months 4 After 6 months 5 Not sure about this

4. How many other children have you had? 1 This is my first baby (go to question 7) 2 One child 3 Two children 4 Three children 5 Four children 6 More than four children

5. How did you feed your other children? (Please tick all that apply) 1 I have bottle fed (formula) in the past 2 I have breastfed in the past

6. If you have breastfed, for how long? (Please tick once for each child) 1 a few days 2 less than 1 month 3 1 - 3 months 4 4 - 6 months 5 7 - 9 months 6 10 - 12 months 7 13 – 17 months 8 18 - 23 months 9 2 years or more Please turn over the page

282

7. Whether you think you will bottle feed or breastfeed your baby, to prepare for feeding have you… (Please tick all that apply) 1 looked for information about feeding 2 attended a hospital antenatal class, or enrolled to attend 3 attended another class, or enrolled to attend (Please describe) ______4 paid for or bought something (Please describe) ______5 borrowed something (Please describe) ______6 taken other action (Please describe) ______7 not taken any action to prepare for feeding

8. What sort of information have you received about bottle feeding (formula)? (Please tick all that apply) 1 I have spoken with friends/relatives 2 I have spoken with a GP 3 I have spoken with a nurse at my GP’s surgery 4 I have spoken with other health professionals such as a pharmacist or dietician 5 I have read at least one pamphlet or book on bottle feeding (Please describe) ______6 I have seen someone bottle feeding (formula) 7 I have seen at least one film, video or TV program on bottle feeding 8 I have received information from formula organisations 9 Other type of information (Please describe) ______10 I have not received any information

9. What sort of information have you received about breastfeeding? (Please tick all that apply) 1 I have spoken with friends/relatives 2 I have spoken with a GP 3 I have spoken with a nurse at my GP’s surgery 4 I have spoken with other health professionals such as a pharmacist or dietician 5 I have read at least one pamphlet or book on breastfeeding (Please describe) ______6 I have seen someone breastfeeding 7 I have seen at least one film, video or TV program on breastfeeding 8 I have received information from breastfeeding organisations such as Australian Breastfeeding Association 9 Other type of information (Please describe) ______10 I have not received any information

283

10 Were you breastfed as a baby? 1 Yes (If you know for how long, please describe) ______2 No 3 Don’t know

11. If you have a partner, was your partner breastfed as a baby? 1 Yes (If you know for how long, please describe)______2 No 3 Don’t know

12. If you have a partner, which of these statements do you feel most applies to your partner’s feelings about how you feed your baby 1 My partner would prefer me to breastfeed 2 My partner doesn’t mind how I feed the baby 3 My partner is supportive of my choice of feeding, no matter which I do 4 My partner would prefer me to bottle feed 5 I’m not sure

13. Which of these statements do you feel most applies to your family’s feelings about how you feed your baby? 1 My family thinks I should breastfeed 2 My family thinks I should bottle feed 3 My family are supportive of my choice of feeding, no matter which I do 4 I’m not sure what my family thinks

14. Which of these statements do you feel most applies to your friends’ feelings about how you feed your baby? 1 My friends think I should breastfeed 2 My friends think I should bottle feed 3 My friends are supportive of my choice of feeding, no matter which I do 4 I’m not sure what my friends think

The following questions are about your personal background. All your answers are confidential.

15. Are you … 1 Married 2 Living with your boyfriend or partner 3 Not living with your boyfriend or partner 4 Separated or divorced 5 Widowed 6 Single

Please turn over the page

284

16. Who do you live with? (Please tick all that apply) 1 Husband or partner 2 Other child or children 3 Mother 4 Relative (please describe) ______5 Someone else (please describe) ______6 Live alone

17. When did you leave school? 1 Completed secondary school to year 12 2 Did not complete secondary school 3 Attended primary school only 4 Did not attend primary school

18. Have you completed any further study since leaving school? 1 Completed a degree 2 Currently completing a degree 3 Completed a diploma 4 Currently completing a diploma 5 Completed an apprenticeship or traineeship 6 Currently completing an apprenticeship or traineeship 7 None of these 8 Other (please describe) ______

19. Is a pension or benefit the main income for your family? 1 Yes 2 No

20. What was the total before tax income of your household (all family members living at home) last year? This question is optional. 1 Less than $32,000 ($614 per week) 2 $32,000 - $101,000 ($615 - $1,938 per week) 3 More than $101,000 ($1,939 per week or more)

21. In which country were you born?

______

22. In which country was your mother born?

______

23. In which country was your father born?

______

285

24. If you were not born in Australia, in which year did you first settle in Australia?______

25. Is English your first language? 1 Yes 2 No

26. Are you … 1 Aboriginal 2 Torres Strait Islander 3 Both Aboriginal and Torres Strait Islander 4 None of the above

27. Do you have any plans for paid work or study after the birth, and if so, when? 1 No 2 Yes, to return part-time in the first 6 months 3 Yes, to return full-time in the first 6 months 4 Yes, to return part-time from 7 - 12 months 5 Yes, to return full-time from 7 - 12 months 6 Yes, but not for at least 12 months 7 Not sure about this

28. Where will you usually see a doctor (GP) after your baby is born? 1 Albury Wodonga Family Medical Centre, Elgin Boulevard 2 Central Medical Clinic, Beechworth Road 3 David Street Medical Clinic 4 Elmwood Medical Centre, Beechworth Road 5 Lavington Medical Centre, Griffith Road 6 Mate Street Medical Centre 7 Wodonga West Medical Clinic, Melbourne Road 8 Other General practice in Albury or Wodonga 9 General Practice in other town (Please write the name of the town) ______10 Not sure about this

29. Please write the name of the doctor, if you know his or her name.

______

Thank you for taking the time to answer these questions. We are grateful for your help with this study.

A researcher may phone you when your baby is about 4 months old.

286

Appendix 6 Sources used for mothers' baseline questionnaire

Question Source Comment 1 When is your baby due? Pilot Used in initial data management; analysis is based on actual birth data from hospital data 2 How do you plan to feed your baby Extended from ABFAB question with extended range ... ABFAB (15) of possible answers. Highest duration window in ABFAB answers is “6 months or longer” 3 When do you think you will Novel Theoretical framework relates planning introduce solids? to motivation. 4 How many other children have you Pilot Standard topic, parity had? 5 How did you feed your other Pilot Standard topic, past breastfeeding children? duration 6 If you have breastfed, for how Pilot Standard topic; mothers with a past long? history of difficulty breastfeeding tend to breastfeed for a shorter duration 7 ... to prepare for feeding have you Novel Theoretical framework suggests that ... the extent of motivation may be associated with level of planning 8 What sort of information have you Extended from ABFAB question with alternative but received about bottle feeding ABFAB related topic (formula)? 9 What sort of information have you ABFAB received about breastfeeding? 10 Were you breastfed as a baby? ABFAB 11 ... was your partner breastfed as ABFAB a baby? 12 ... your partner’s feelings about ABFAB how you feed your baby? 13 ... your family’s feelings about ABFAB how you feed your baby? 14 ... your friends’ feelings about Extended from ABFAB question with alternative but how you feed your baby? ABFAB related topic 15 [marital status] ABFAB (15) 16 Who do you live with? ABFAB 17When did you leave school? ABFAB

287

18 ... any further study ... ABFAB 19 Is a pension or benefit the main ABFAB income for your family? 20 ... total before tax income of your ABFAB household 21 In which country were you born? ABFAB 22 In which country was your mother ABFAB born? 23 In which country was your father ABFAB born? 24 ... in which year did you first settle ABFAB in Australia? 25 Is English your first language? ABFAB 26 [Aboriginality] Standard 27 ... plans for paid work or study ... Novel 28 Where will you usually see a Novel Data management: Assessed against doctor (GP) after your baby is born? exclusion criteria; randomisation 29 ... name of the doctor Novel Data management: Check on exclusion and randomisation; compared to Division of GP lists of doctors working at each medical practice.

288

Appendix 7 Birth data collection fields

Due Date

Mother surname

Mother first name/s

Mother DOB

Date of Delivery

Type of Delivery Vaginal, Vaginal, Caesarean, Caesarean, unassisted instrumental booked emergency

Gestation at Delivery (weeks)

Birth weight of baby (grams)

Systemic analgesia in labour, Yes No intramuscular, intravenous or subcutaneous Regional analgesia in labour, Yes No epidural or spinal Any breastfeeding Yes No

Any formula given Yes No

289

Appendix 8 CATI outcomes survey for mothers at 4 and 6 months

!DO ID ------QID,s,0,0 ------!DO INTRO1 ------QINTRO1,s,0,0 ------!DO INTRO2 ------QINTRO2,s,0,0 ------!clear !@ 10,0 say "DETERMINE THE RESPONDENT'S LANGUAGE(S) AND/OR COUNTRY OF ORIGIN" !@ 12,0 SAY "LANGUAGE(S)" GET LANG !@ 14,0 SAY "COUNTRY" GET COO !@ 18,0 say "Thank you for your time. " !READ ------langcob,x,0,0 ------Good morn/aftern/even my name is ..... calling on behalf of the University of New South Wales regarding the Support for Mothers & Infant Feeding survey.

May I speak to "+rtrim(b->mname)+ " please?

[IF RESPONDENT IS UNAVAILABLE OR BUSY REQUEST A SUITABLE TIME TO CALL BACK] # 1. YES, CONTINUE 2. REFUSAL BY RESPONDENT 15. NOT KNOWN AT NUMBER 24. NO ADULT AVAILABLE - SPOKE WITH CHILD 4. Not Now SPOKE TO RESPONDENT - CALLBACK ARRANGED (WITHIN SURVEY PERIOD) 44. CALLBACK - DID NOT SPEAK TO RESPONDENT 6. PERSON UNSUITABLE - EXPLAIN IN COMMENTS 66. PERSON AWAY FOR DURATION OF SURVEY 22. HOUSEHOLD REFUSAL - EXPLAIN IN COMMENTS 14. LANGUAGE PROBLEM 8. NOT IN AREA/QUOTA DONE 21. SOFT REFUSAL !if callnum>5

290

68. FIFTH CALL AFTER INTIAL CONTACT - INSTEAD OF 24 OR 44 ENTER 68 AND SAY [We won't trouble you again. If she would still like to participate in the [Support for Mothers & Infant Feeding survey please ring freecall 1800 355534 !endif ------qcontac,n,2,0,1,8,14,15,44,66,22,21,24,68 !if qcontac=3.or.qcontac=5.or.qcontac=7 ! skipto="qcontac" !else ! if qcontac>1 ! replace rrate with qcontac ! skipto="QENDBIT" ! endif !endif ------[WHEN REQUIRED PERSON IS ON PHONE REPEAT INTRODUCTION IF NECESSARY] Good morn/aftern/even my name is ..... calling on behalf of the University of New South Wales regarding the Support for Mothers & Infant Feeding survey.

You may remember completing a survey before your baby was born. [PAUSE] This is a follow-up call. I have just a few questions which will take about 5 minutes. Is it alright to talk to you now?

# 0. OR 1. CONTINUE 2. REFUSAL BY RESPONDENT 4. Not Now SPOKE TO RESPONDENT - CALLBACK ARRANGED (WITHIN SURVEY PERIOD) 6. PERSON UNSUITABLE - EXPLAIN IN COMMENTS 66. RESPONDENT UNAVAILABLE FOR SURVEY PERIOD (COMMENTS) 14. LANGUAGE PROBLEM 8. NOT IN AREA !if callnum>5 67. FIFTH CALL AFTER INTIAL CONTACT - INSTEAD OF 4 ENTER 67 AND SAY[We won't trouble you again. If you would still like to participate in the [Support for Mothers & Infant Feeding survey please ring freecall 1800 355534

This research is to find out what foods are given to babies at different ages. We know that families make choices for many different reasons, and we are not making any judgments.

291

Your participation is voluntary and you can ask me to stop at any time. All information you provide is confidential and you will not be identified in the results.

Can I confirm that you are "+rtrim(b->mname)+" and that your baby's date of birth is "+rtrim(b->babybd)+".

# 1. Yes 2. No [8. DON'T KNOW 9. REFUSED] ------qnbdb,n,1,0,1,2,8,9 if qnbdb<>1 skipto qendbit ------Q1. Could you tell me your baby's name? # ------q1,c,20,0,0,0 ------Q2. Since this time yesterday, has "+rtrim(q1)+" been breastfed or been fed breastmilk? # 1. Yes 2. No [8. DON'T KNOW 9. REFUSED] ------q2,n,1,0,1,2,8,9 if q2<>1 skipto q14 ------Q3. I am going to read to you a list of various drinks and foods which are sometimes given to children, can you tell me, Since this time yesterday, did "+rtrim(q1)+" receive any of the following ...... 1. Yes 2. No [8. DON'T KNOW 9. REFUSED] # Plain water? # Sweetened or flavoured water? # Fruit juice? # Tea? # Infant formula? (if yes, skip 11 and 13) # Tinned, powdered or fresh milk? # Solid or semi-solid food? (if yes, skip 12, 13, 14) # Anything else? ------q3,n,1,0,1,2,8,9 q4,n,1,0,1,2,8,9 q5,n,1,0,1,2,8,9

292

q6,n,1,0,1,2,8,9 q7,n,1,0,1,2,8,9 q8,n,1,0,1,2,8,9 q9,n,1,0,1,2,8,9 q10,n,1,0,1,2,8,9 ------IF q10=1 What else did "+rtrim(q1)+" receive? # ------q10c,c,180,0,0,0 ------IF Q7<>1 Q11. I am going to ask you to think back. In the last week, did "+rtrim(q1)+" receive any infant formula? # 1. Yes 2. No [8. DON'T KNOW 9. REFUSED] ------q11,n,1,0,1,2,8,9 ------IF q9<>1 Q12. In the last week, did "+rtrim(q1)+" receive any solid or semi-solid food? # 1. Yes 2. No [8. DON'T KNOW 9. REFUSED] ------q12,n,1,0,1,2,8,9 if q12=1 skipto q15 ------IF q9<>1 .and. q12<>1 Q13. Now I am going to ask you to think back further. Since birth, has "+rtrim(q1)+" ever been given any solids or liquids other than breastmilk? # 1. Yes 2. No [8. DON'T KNOW 9. REFUSED] ------q13,n,1,0,1,2,8,9 ------IF Q2<>1 Q14. At what age was "+rtrim(q1)+" last fed breastmilk? # Weeks(0 = Never Breastfed, Maximum=26) [88. DON'T KNOW 99. REFUSED] ------q14,n,2,0,0,26,88,99

293

------Q15. Have you had any contact with the Australian Breastfeeding Association? It is a voluntary organization that some mothers make contact with. # 1. Yes 2. No [8. DON'T KNOW 9. REFUSED] ------q15,n,1,0,1,2,8,9 ------Thank you very much. We hope that this research will lead to more support for mothers. We may call you again in about 2 months, thank you again for your time.

294

Appendix 9 Knowledge and attitudes practice nurse survey

Knowledge questions 1) Babies who are formula fed are not as healthy as babies who are breastfed.

2) If a baby has diarrhoea or is vomiting, breastfeeding should continue.

3) Growth patterns of breastfed babies differ from those of formula fed babies.

4) Women who breastfeed have a lower risk of breast and ovarian caner.

5) If a mother has mastitis, she should keep breastfeeding.

6) Nearly all medications appear in significant amounts in breastmilk.

7) In the first few weeks a breastfed baby will usually feed 8 - 12 times in 24 hours.

8) A baby will commonly need breastfeeding during the night up to 12 months of age or older.

9) A mother needs to drink lots of fluids to be able to produce enough breastmilk.

10) A mother can be reassured she is producing enough breastmilk if her baby has pale odourless urine

11) The most common cause of insufficient milk supply is not enough effective breastfeeding.

12) A woman with nipple and breast pain during and following a breastfeed may have a thrush infection of the nipple.

13) A baby is breastfeeding well if swallowing can be observed and the mother is in no pains.

14) The let-down reflex is only working effectively if the mother can feel tingling in her breasts.

15) The best visual way to check that a baby is correctly attached is to ensure the baby's whole body is facing the mother's body, baby's mouth is wide open and baby's lips are turned out.

16) An exclusively breastfed baby will need water during hot weather.

17) A mother should be advised to stop breastfeeding if her baby is 12 months or older.

18) A mother should be advised to stop breastfeeding if she b3ecomes pregnant.

295

19) Expressed breastmilk can be stored safely in a refrigerator up to 5 days and in a freezer up to 12 months.

20) A father has a lot of influence on how long the mother will breastfeed.

21) It's good for a baby to be exclusively breastfed if the mother smokes.

22) It's good for a baby to be exclusively breastfed if the mother eats a lot of junk food.

Attitudes questions

23) Breastfeeding helps a mother and her baby develop close feelings.

24) Breastfeeding makes a mother feel good about herself.

25) Breastfeeding allows a mother to go places and do things outside the home easily.

26) Formula feeding is the better choice if a mother plans to go out to work.

27) Formula feeding helps a mother and her baby develop close feelings.

28) Breastfeeding the baby is a good thing for the mother's partner.

29) I would encourage my own friends to breastfeeding their babies.

30) I would be comfortable (not embarrassed) if I saw a woman breastfeeding her baby in a public place.

31) Women who are uncertain should be encouraged by our staff to breastfeed their babies.

32) Women who initially choose to formula feed should be informed about the importance of breastfeeding.

33) I would encourage breastfeeding mothers to contact a support group.

34) I prefer not to discuss continuing to breastfeed with mothers unless they raise the subject.

35) My advice can be effective in persuading some mothers to continue breastfeeding.

36) Discussing continuing to breastfeed with mothers can be rewarding.

296

Appendix 10 Sources for questions to practice nurses

Knowledge questions Question Topic covered Source for Status question 1 Importance - Baby PM Adapted 2 Health – Mx – Baby JB Original 3 Health – growth JI/WB Original 4 Health – mother WB Adapted 5 Health – Mx – Mo WB Adapted 6 Health – Mx – Mo JB/related in WB Original 7 Supply – Mx – Frequency WB Original 8 Supply – Mx – Older/Sleep MER New 9 Supply – Mx – Mo WB Adapted 10 Supply – Mx – Baby JB Original 11 Supply – Frequency/attachment WB Adapted 12 Pain – thrush WB/related in JI Original 13 Attachment – pain/check JB Original 14 Attachment – let-down JB Adapted 15 Attachment – check JB Original 16 Exclusivity – water WB Adapted 17 Duration - expected MER New 18 Challenges – pregnancy WB Adapted 19 Challenges – storage EBM MER New 20 Challenges – Father JB Original 21 Risk – smoking PM Adapted 22 Risk – Mother's diet PM Adapted

PM: Patricia Martens (1); JB: Jennifer Byrne (2); JI: Jenny Ingram (3); WB: Wendy Brodribb, (4); MER novel question. Mx management; Mo mother; EBM expressed breastmilk

297

Attitude questions Question Topic covered Source for Status question 23 Attitude – bonding/BF PM/related Original question also in WB 24 Attitude – Mo self-esteem PM Original 25 Attitude – convenience PM/related in WB Original 26 Attitude – employment/formula PM/related in ET Original 27 Attitude – bonding/formula PM Original 28 Attitude – father’s role PM/related in WB Original 29 Attitude PM Original 30 Attitude – in public PM Original 31 Attitude/Role – promotion/unsure PM/related in ET Original Mo 32 Attitude/Role – promotion/resistant PM Original Mo 33 Attitude/Role – support group EO Adapted referral 34 Attitude/Role – promotion/initiation TC/related in ET Adapted 35 Attitude/Role – promotion/efficacy TC Adapted 36 Attitude/Role – TC Adapted promotion/satisfaction

PM: Patricia Martens (1); WB: Wendy Brodribb (4); ET Elsie Taveras (5); EO Eme Owoaje (6); TC Tim Coleman. (7) BF breastfeeding; Mo mother;

298

Appendix 11 Practice nurse semi-structured interview questions

Have you seen any mums for baby vaccinations? How are you going with Conversation Tool, using it with mums? How do you go about asking about feeding? What about the recommendations? How are you finding their responses? How are you handling that? Can you figure out which colour box they're at? How do you find it with giving them the response for that? What about offering the resources? Have you run out of Breastfeeding Confidence?

Interviews also aimed to: Elicit reflection on the training process and on their experience delivering the intervention Facilitate practice nurse problem-solving

Appendix 12 Practice nurse telephone follow-up process

Contacting the practice nurse Nurses were interviewed in the context of their workplace, when the nurse had time between patients, during dedicated telephone time for patient to call for test results, during a lunch break, or at the close of business.

A telephone call to the practice number would usually be answered by a medical receptionist, and occasionally by the practice nurse or practice manager. The researcher introduced herself as a medical doctor, calling regarding "the breastfeeding research project that your practice is involved in". The receptionist would determine whether the nurse was available, and take a message if not. Repeat calls were made when phone calls were not returned. Practice nurses or receptionists were asked to nominate suitable times to call or to return the call.

299

Appendix 13 Contents of resource folder for practice nurses

All materials used with permission

Intervention Tool  The Conversation Tool (version 2) see Appendix 16

Resources for mothers and practice nurses  Booklet: Breastfeeding Confidence  Leaflets: Starting family foods; Balancing breastfeeding and work  Directory of local health contacts: services provided by Child Health or Early Childhood Nurses including clinic locations and day-stay parent and child units  Directory of local Australian Breastfeeding Association contacts: services including mothers group bulletins; local breastfeeding counsellors; local breastfeeding education classes; Breastfeeding Welcome Here list of businesses and rest rooms; telephone helpline; website for email counselling, information and forum.  Posters for display in the practice as desired: "Sole Food", A4 laminated, black and white close-up photograph, baby's face while breastfeeding, promoting exclusive breastfeeding for six months with ABA helpline and website details; "Whatever you do, keep a good thing going", A3 colour, photo-boards of women in variety of workplaces, regarding support available through ABA.

Training materials for practice nurses  Program: Training session topics and timetable  Notes: training session PowerPoint presentation slides in handout format  Breastfeeding Management participant material: for the Breastfeeding Management in a Baby Friendly Health Service (BMBFHS) 8 hour structured course for health professionals working with mothers and babies in maternity settings, 5th ed., Australasian Lactation Services, 2008(2).  quick reference sheets “LACTFACTS Information for health professionals only” produced by the Australian Breastfeeding Association(8);  Supplement to BMBFHS participant material: BMBFHS presenter’s material on specific topics: breast and nipple thrush, engorgement, sore and cracked nipples  Supplement to BMBFHS participant material: the paper "Benefits of Breastfeeding" by Jane Allen and Debra Hector (9) from the NSW Public Health Bulletin  Article for further reading: "Australian women’s stories of their baby-feeding decisions in pregnancy" (10), about mothers’ motivation to breastfeed, by Athena

300

Sheehan, Virginia Schmied and Margaret Cooke published in Midwifery Best Practice Volume 4 2006 ed Sara Wickham  Instruction page: Reflective practice  Information leaflet: Is your baby sleeping safely? Included to enable the Practice nurse to respond consistently and reassuringly to frequently asked questions. We aimed to normalise biological baby sleeping patterns and to address concerns about SIDS (11, 12). “Mothers are given a wide range of seemingly different advice” as the Australian parliamentary inquiry observed in 2007, whereas “promotion of confidence in the ability to breastfeed... is likely to increase duration of breastfeeding among women”.(13)

Process materials for the conduct of the trial  Researcher contact details  Introductory letter: Research plan for Practice nurses

301

Appendix 14 Breastfeeding management course manual introduction

Introduction

This course is designed to be an integral part of your hospital achieving and maintaining the goal of the Baby Friendly Health Initiative (BFHI). BFHI provides accreditation for hospitals that implement best practice in the care of breastfeeding mothers and babies.

Course Aim and Objectives

This course aims to assist hospitals in equipping their staff with the knowledge and management skills necessary to help mothers breastfeed successfully. The objectives of this course are to provide a common foundation of basic breastfeeding knowledge and management among doctors, midwives, nurses and other maternity staff which will lay the basis for the implementation of the Ten Steps to Successful Breastfeeding. This course will build on the current breastfeeding knowledge of the staff using an adult learning approach.

Length of course

The length of the course is 8 hours. This may be presented as either a full day session or as two 4 hour sessions. It is intended that every hospital staff member who has direct patient care responsibility for mothers and their babies will attend the course .

Course materials

Printed material

The following documents have been included in the package and should be made available to both the facilitator and participants of the course:  Evidence for the ten steps to successful breastfeeding, WHO/CHD/98 .9, 1998  "The key to successful breastfeeding" A4 poster and brochures (plus order form), R Glover, 2005 Infant feeding guidelines for health workers, NHMRC - this can be downloaded from www.nhmrc.gov.au/publications/synopses/dietsyn.htnt

302

Video

The video "Follow Me Mum: The Key to Successful Breastfeeding" by Rebecca Glover, a Lactation Consultant in WA is a valuable visual tool to use for both staff and parents.

Participant material (Appendix)

Master copies of the following have been included for photocopying and distribution to course participants

 assessment tool  hospital self assessment  breastmilk component table to be completed  case studies with discussion questions  copies of pertinent overhead transparencies  Lactfacts, ABA (note these are only for use by health professionals)  useful websites  evaluation form

Overhead transparencies (At end of each section. Master copies of B&W transparencies in Appendix)

A complete set of overhead transparencies sheets are provided for use in the workshops. These highlight the main points of each section plus include coloured transparencies depicting key breastfeeding issues. Overhead transparencies are designated throughout these notes using underscored numbers corresponding to the sections they represent eg 1 .1 OT Mothers breastfeeding is the first transparency to be used in section 1 . Transparencies for each section are included at the end of that section. Master copies of all B&W transparencies are included in the appendix.

Assessment tool (section 10)

This tool consists of a series of true/false questions with answers. It can be used as either a means of determining what participants know about breastfeeding and lactation, how well participants have understood the course material or determine whether participants have gained knowledge from attending the workshop. The assessment should ideally be taken at the beginning and reviewed by participants again at the conclusion of the workshop. This is then followed by a discussion of the results. Answers with rationale are included.

303

Certificate of Completion (Appendix)

A Certificate of Completion is included that can be reproduced and awarded to participants who have correctly answered 80% of the true / false questions at the end of the workshop. It is the responsibility of the facilitator to check the assessments and to issue the Certificates. If health professionals have not obtained an 80% pass rate, they may wish to repeat the assessment following discussion of incorrect answers with facilitator and/or further self study of the course material.

Self Study module (Section II)

While ideally, health professionals are encouraged to attend this workshop, this is not always possible. Hence the content of the course has been structured so that it can be used as a self study module for those who are unable to attend the workshop. The study guide questions are the basis of this module. They encourage participants to work through a series of exercises that will ensure they have covered the same content as that covered in the workshop. In conclusion, participants can assess their level of understanding by undertaking the assessment as described above and will be awarded a Certificate of Completion if they obtain an 80% pass for this assessment.

Preparing for the workshop

 To help make the workshop as effective as possible, the following should be included in the preparation and planning of the workshop.  Presenter to read facilitators notes and review overhead transparencies before commencing workshop  Use a room that is large enough to seat all participants and that there is sufficient room for chairs to be rearranged for small group work  Make sure chairs are comfortable and room is well ventilated  Arrange chairs in a semi-circle so everyone can see each other  Ensure all participants can see the overhead projector and video and that the room can be darkened if necessary  Ensure participants have something to write on eg desk, table  Ensure sufficient copies of the handout material are available  Ensure all resource material is available throughout the workshop  Provide name tags for all participants  Apply for professional credit points if/where necessary (see below)

304

Presenting the workshop

 Welcome all participants  Explain how workshop will be conducted - briefly describe the program  Encourage participants to actively participate through asking relevant questions, and participating in tasks set by facilitator  Use the transparencies as the guide to the presentation . It is not possible to cover all the material in the background notes.  Ensure program timelines are maintained  Provide breaks at suitable times throughout  Ensure sufficient time is allowed for the assessment and evaluation to be completed

CERPs, CPDs

During the planning stages of the workshop, the facilitator may wish to contact the following for accreditation of education points. The addresses below refer to Australian organisations. There may be similar ones in your area.

For Lactation Consultants: CERPS - Continuing Education Recognition Points 1BLCE PO Box 1533 Oxenford QLD 4210 www.iblce.edu.au

For General Practitioners: CPD - Continuing Professional Development points RACGP Quality Assurance and Continuing Education Program www.racgp.org.au

For Midwives: CPD - Continuing Professional Development points (MidPLUS) Australian College of Midwives Po Box 666 Canberra ACT 2601 www.midwives .org.au

305

Workshop outline (with approximate times for each session) This workshop has been designed to be conducted as either a full day session with a one hour lunch break or as 2 four hour sessions.

First 4 hour session (including 15 min break)

 Pre-workshop assessment (20 min)  Introduction - BFHI and implementing the Ten Steps to Successful  Breastfeeding and the 7 Point Plan for Community Health Centres (75 min)  Benefits of breastfeeding/ hazards of formula feeding (30 min)  Anatomy and physiology of breastfeeding (20 min)  Breastmilk composition (20 min)  Management of breastfeeding (60 min) - includes 20 min video

Second 4 hour session (including 15 min break)

 Management of common problems (75 min)  Issues impacting on breastfeeding (45 min)  Case studies to consolidate information (60 min)  Community resources (in each community) (15 min)  Post-workshop assessment and evaluation (30 min)

Appendix 15 Delivery of breastfeeding management course

The ALC course includes a range of resources including:  DVD “Follow me mum” produced by Rebecca Glover.  self-completion exercise on breastmilk components.  group exercises on management of common breastfeeding challenges.  case studies for more complex problem-solving exercises . To these for this practice nurse training program were added.  photographs of mothers breastfeeding older babies, from the Australian Breastfeeding Association photolibrary, used with permission.  Close-up photographs of the latch process, from BFHI online teaching materials, used with permission. (14)

306

Appendix 16 The Conversation Tool Version 3

307 Page 308

Appendix 17 PowerPoint slides used in training workshops

Page 309 Page 310

Page 311 Page 312

Page 313 Page 314

Page 315 Page 316

Page 317 Page 318

Page 319 Page 320

Page 321 Page 322

Page 323

REFERENCES FOR THE APPENDICES

(1) Martens PJ. Does breastfeeding education affect nursing staff beliefs, exclusive breastfeeding rates, and Baby-Friendly Hospital Initiative compliance? The experience of a small, rural Canadian hospital. J Hum Lact. 2000 Nov;16(4):309-18. (2) Byrne J, McIntyre E. A lactation course that is available for study by correspondence. Journal - Australian College of Midwives. 1997 Mar;10(1):3. (3) Ingram J, Johnson D. A feasibility study of an intervention to enhance family support for breast feeding in a deprived area in Bristol, UK. Midwifery. 2004 2004;20(4):367-79. (4) Brodribb W. Improving the breastfeeding knowledge and skills of GP registrars University of Queensland; 2009. (5) Taveras EM, Li R, Grummer-Strawn LM, Richardson M, Marshall R, Rego V, et al. Opinions and practices of clinicians associated with continuation of exclusive breastfeeding. Pediatrics. 2004;113(4):e283-e90. (6) Owoaje ET, Oyemade A, Kolude OO. Previous BFHI training and nurses' knowledge, attitudes and practices regarding exclusive breastfeeding. African Journal of Medicine & Medical Sciences. 2002 Jun;31(2):137-40. (7) Coleman T, Wilson A. Anti-smoking advice in general practice consultations: general practitioners' attitudes, reported practice and perceived problems. Br J Gen Pract. 1996;46(403):87-91. (8) Lactation Resource Centre ABA. Lactfacts - Information for health professionals. 2nd ed: Australian Breastfeeding Association 2007:1-13. (9) Allen J, Hector DJ. Benefits of Breastfeeding. NSW Public Health Bull 2005:42- 46. (10) Sheehan A, Schmied V, Cooke M. Australian women's stories of their baby- feeding decisions in pregnancy. Midwifery. 2003 Dec;19(4):259-66. (11) Ball H. Breastfeeding, bed-sharing, and infant sleep. Birth. 2003 September 2003;30(3). (12) Ball HL, Ward-Platt MP, Heslop E, Leech SJ, Brown KA. Randomised trial of infant sleep location on the postnatal ward. Arch Dis Child. 2006 Dec;91(12):1005-10. (13) The Parliament of the Commonwealth of Australia. The Best Start: Report on the inquiry into the health benefits of breastfeeding. Canberra: House of Representatives; 2007 August. (14) UNICEF. Section 2: Strengthening and sustaining the Baby-friendly Hospital Initiative BFHI training materials Updated 2009 [cited 2011 February 2]; Available from: http://www.unicef.org/nutrition/files/BFHI.section2.2009_slides.pdf (15) Forster D, McLachlan H, Lumley J, Beanland C, Waldenström U, Harris H, et al. ABFAB. Attachment to the breast and family attitudes to breastfeeding. The effect of breastfeeding education in the middle of pregnancy on the initiation and duration of breastfeeding: a randomised controlled trial [ISRCTN21556494]. BMC Pregnancy Childbirth 2003.

324