Journal of the New Zealand Medical Association Vol 131 | No 1480 | 17 August 2018 First, do no harmony: an examination of attitudes to music played in operating theatres

Antipsychotic prescribing in New Zealand between 2008 and 2015 Ethnic disparities in community antibacterial dispensing in New Zealand, 2015

A genome project for Māori and Pasi ka: charting Is there a place for day Provincial or Tertiary Surgical a path to equity in genomic hospitals in older persons Treatment for Rectal Cancer in medicine for Aotearoa mental health care? New Zealand? Publication Information published by the New Zealand Medical Association

NZMJ Editor NZMA Chair Professor Frank Frizelle Dr Kate Baddock

NZMJ Production Editor NZMA Communications Manager Rory Stewart Diana Wolken

Other enquiries to: NZMA To contribute to the NZMJ, fi rst read: PO Box 156 www.nzma.org.nz/journal/contribute The Terrace Wellington 6140 © NZMA 2018 Phone: (04) 472 4741

To subscribe to the NZMJ, email [email protected]

Subscription to the New Zealand Medical Journal is free and automatic to NZMA members. Private subscription is available to institutions, to people who are not medical practitioners, and to medical practitioners who live outside New Zealand. Subscription rates are below. All access to the NZMJ is by login and password, but IP access is available to some subscribers. Read our Conditions of access for subscribers for further information www.nzma.org.nz/journal/subscribe/conditions-of-access If you are a member or a subscriber and have not yet received your login and password, or wish to receive email alerts, please email: [email protected] The NZMA also publishes the NZMJ Digest. This online magazine is sent out to members and subscribers 10 times a year and contains selected material from the NZMJ, along with all obituaries, summaries of all articles, and other NZMA and health sector news and information.

Subscription rates for 2018 New Zealand subscription rates Overseas subscription rates Individuals* $306 Individual $426 Institutions $530 Institutions $571 Individual article $25 Individual article $25

*NZ individual subscribers must not be doctors (access is via NZMA Membership) New Zealand rates include GST. No GST is included in international rates. Note, subscription for part of a year is available at pro rata rates. Please email [email protected] for more information. Individual articles are available for purchase by emailing [email protected]

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 2 www.nzma.org.nz/journal CONTENTS

EDITORIALS 32 8 What keeps Northland general A genome project for Māori and practitioners working in Pasika: c harting a path to equity Northland? in genomic medicine for Aotearoa Tanya Quin, Kyle Eggleton Martin A Kennedy 38 11 A media content analysis of Is there a place for day hospitals in New Zealand’s district health older persons mental health care? board Population-Based Funding Matthew Croucher Formula Aaron N Chester, Erin C Penno, Robin DC 14 Gauld Provincial or Tertiary Surgical Treatment for Rectal Cancer in 50 New Zealand? Ethnic disparities in community Ian Bissett antibacterial dispensing in New Zealand, 2015 ARTICLES Naomi Whyler, Andrew Tomlin, Murray 17 Tilyard, Mark Thomas Psychogeriatric day hospital 61 reduces depression and anxiety Antipsychotic prescribing in New symptoms and improves quality of Zealand between 2008 and 2015 life Sam Wilkinson, Roger T Mulder Petra Ann Hoggarth 68 23 First, do no harmony: an A multicentre, benchmarking examination of attitudes to music study of rectal cancer played in operating theatres management in provincial New Anantha Narayanan, Andrew R Gray Zealand hospitals Michael J O’Grady, Josephine O’Grady, 75 Rebecca Shine, Gerard Bonnet, An audit of regular medication Tim Eglinton compliance prior to presentation for elective surgery Daniel Wood, Nicholas Lightfoot

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 3 www.nzma.org.nz/journal CONTENTS

VIEWPOINT LETTER 81 94 Genomic medicine must reduce, A new report on the effects of a not compound, health inequities: policy used to restrict access to the case for hauora-enhancing elective inguinal hernia surgery genomic resources for New Zealand Philip Bagshaw Stephen P Robertson, Jennie Harré Hindmarsh, Sarah Berry, Vicky A METHUSELAH Cameron, Murray P Cox, Ofa Dewes, 96 Robert N Doughty, George Gray, Jessie Five-year risk of stroke a er TIA or C Jacobsen, Albert Laurence, Elizabeth Matisoo-Smith, Susan Morton, Andrew minor ischemic stroke N Shelling, Dianne Sika-Paotonu, Anna 100 YEARS AGO Rolleston, Jonathan R Skinner, Russell G Snell, Andrew Sporle, Cristin Print, Tony R 97 Merriman, Maui Hudson, Philip Wilcox A Case of Adherent Meckel’s Diverticulum Causing Fatal CLINICAL CORRESPONDENCE Strangulation 90 Redundant laparoscopic adjustable gastric band tubing causing internal hernia and small bowel obstruction Erika Fernandes, James Tan, Glenn Farrant, Karl Kodeda

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 4 www.nzma.org.nz/journal SUMMARIES

Psychogeriatric day hospital reduces depression and anxiety symptoms and improves quality of life Petra Ann Hoggarth Day hospitals are outpatient clinics where groups of patients attend for several hours and receive comprehensive treatment from an interdisciplinary team of health professionals. Only one psychogeriatric day hospital exists in New Zealand, at the Canterbury District Health Board. Self-report measures of depression, anxiety, and life satisfaction and quality were completed at intake and discharge in a sample of 185 older adults attending he day hospital. Statistically and clinically signifi cant improvements were found on all measures. This is the fi rst step in demonstrating that New Zealand psychogeriatric day hospitals may be an effective way of treating older adults. This is an important consideration given the increased need for effective and effi cient mental health services for the burgeoning older adult population. A multicentre, benchmarking study of rectal cancer management in provincial New Zealand hospitals Michael J O’Grady, Josephine O’Grady, Rebecca Shine, Gerard Bonnet, Tim Eglinton This study compared patients who received care for their rectal cancer in provincial verses main centre hospitals. It shows that patients who receive treatment for rectal cancer in provincial hospitals achieve similar results to those in the main centres. This is important because some international studies have shown inferior outcomes in smaller hospitals. What keeps Northland general practitioners working in Northland? Tanya Quin, Kyle Eggleton In response to understanding Northland’s GP workforce issues, this paper used face-to-face interviews with Northland GPs to explore why Northland GPs work in Northland. To answer this question the interview information was grouped together to fi nd common reasons. Devel- oping a personal connection to the region explained why Northland GPs work in Northland. This paper explores that connection. A media content analysis of New Zealand’s district health board Population-Based Funding Formula Aaron N Chester, Erin C Penno, Robin DC Gauld The fi ndings suggest general discontent with the PBFF model across the DHB sector and a sense that the PBFF has failed to address various challenges facing DHBs. The geographic imbalance in reporting volume suggests that frustration with the PBFF is particularly keenly felt in the South Island. Although the PBFF is a lightning rod for frustrations over limited health funding, the fi ndings point to the need to improve transparency and dialogue around the formula and to monitor of the impact of PBFF allocations throughout the country.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 5 www.nzma.org.nz/journal SUMMARIES

Ethnic disparities in community antibacterial dispensing in New Zealand, 2015 Naomi Whyler, Andrew Tomlin, Murray Tilyard, Mark Thomas Compared with many other countries, the level of antibiotic consumption by people in the community in New Zealand is very high, encouraging the spread of antibiotic resistant bacteria in New Zealand, and threatening the effectiveness of many antibiotic medicines in the coming years. One of the goals of the Antimicrobial Resistance Action Plan, developed by the NZ Ministry of Health, is to optimise the use of antimicrobial medicines in human health by encouraging appropriate antbiotic prescribing and discouraging inappropriate prescribing. However, there are large differences between the ethnic groups in New Zealand with regard to the rates of various infectious diseases, including rheumatic fever, pneu- monia, and skin infections, and interventions to improve antibiotic prescribing need to take account of these ethnic differences in the rates of infectious diseases. This study has shown that Pacifi c and Maori people, who have very much higher rates of many infectious diseases, have only moderately higher rates of dispensing of antibiotics, when compared with people of European, Middle Eastern, Latin American or African, or Asian ethnicity. General practi- tioners and other health care workers caring for patients in the community need to reduce antibiotic prescribing for all population groups, but particularly for those groups with lower rates of serious infectious diseases. Antipsychotic prescribing in New Zealand between 2008 and 2015 Sam Wilkinson, Roger T Mulder We looked at antipsychotic prescribing for the whole population of New Zealand using Ministry of Health data. Antipsychotic prescribing rose by 49% over the years 2008 to 2015. Most of this change was in prescribing atypical antipsychotics. Young Maori males and older European females were most likely to receive antipsychotics. First, do no harmony: an examination of attitudes to music played in operating theatres Anantha Narayanan, Andrew R Gray Music is commonly played in operating theatres worldwide. This paper examines the use of and perceptions towards music being played during operations, by operating theatre staff. 72% of respondents liked it or did not mind music being played. We found that easy listening was the most frequently played, and most frequently preferred. Music was preferable for familiar, non-urgent procedures, at a low to medium volume from a CD or MP3. Surgeons were the group of staff most empowered to choose the music. While most respondents felt music improved calmness, mood, team and surgical performance, most felt it worsened communication and is distracting during crises.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 6 www.nzma.org.nz/journal SUMMARIES

An audit of regular medication compliance prior to presentation for elective surgery Daniel Wood, Nicholas Lightfoot An audit was conducted looking at whether people coming in for planned surgery had not taken some of their usual medications and why. Some medications should be stopped but it is better if certain others like heart medication are taken to help prevent problems arising at the time of surgery. We found many people missed some or all of their medications and this was mostly due to them not understanding the instructions they were given before surgery. There are methods of improving the communication (verbal and written) of instructions given to patients to help people take the medications they should be taking. Improving the way we do this should help prevent some avoidable complications during and after surgery. Genomic medicine must reduce, not compound, health inequities: the case for hauora-enhancing genomic resources for New Zealand Stephen P Robertson, Jennie Harré Hindmarsh, Sarah Berry, Vicky A Cameron, Murray P Cox, Ofa Dewes, Robert N Doughty, George Gray, Jessie C Jacobsen, Albert Laurence, Elizabeth Matisoo-Smith, Susan Morton, Andrew N Shelling, Dianne Sika-Paotonu, Anna Rolleston, Jonathan R Skinner, Russell G Snell, Andrew Sporle, Cristin Print, Tony R Merriman, Maui Hudson, Philip Wilcox The genomes of people living in Aotearoa are derived from populations that have disparate and distinct ancestries. As genomic medicine increases in its reach and signifi cance, the accurate and equitable interpretation of genetic variation in the healthcare context is critical. Misinterpretation or inaccurate analysis of genetic data by using datasets mismatched to the populations being studied or served has contributed to disparities in health outcomes else- where, such in the US and Canada. We suggest that this will become a signifi cant problem in Aotearoa unless an appropriate and relevant understanding of our genomic diversity is estab- lished. Aotearoa New Zealand needs to assemble a genomic resource for use in healthcare settings to aid in the accurate interpretation of genomic data for all New Zealanders, most especially for Māori and Pacifi c peoples.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 7 www.nzma.org.nz/journal EDITORIAL

A genome project for Māori and Pasi ka: charting a path to equity in genomic medicine for Aotearoa Martin A Kennedy

ver since completion of the human ge- genetic variation that can be discerned in nome project nearly two decades ago,1 every individual are shaped by the environ- Ethe application of genomic medicine ments in which our ancestors lived, and the has promised a great deal. Over the past de- migratory journeys they undertook. Our cade, this promise has begun to be realised. genomes therefore not only embody and The ability to affordably sequence all 20,000 refl ect our ancestries and ethnic origins, genes in the human genome, a process but they also infl uence our health, and to a called whole exome sequencing (WES), and greater or lesser degree chart the course of more recently to sequence entire human ge- our lives. How then to ensure that genomic nomes, is changing the way genetic diseases medicine has equal value for all New are diagnosed or understood.2 Zealanders, regardless of their ancestral Genome wide association studies (GWAS), histories? using methods that provide genotypes for In order to answer this question, it needs hundreds of thousands of single nucle- to be appreciated that genomic techniques otide polymorphisms (SNPs) throughout all depend for their success on the mass the genome, have identifi ed thousands of sharing of genomic data derived from large genetic variants that contribute to many cohorts, in public or controlled-access data- complex traits (from Alzheimer’s disease bases, that provide detailed information to zinc levels in serum).3 The genes and about the identity and frequency of genetic regulatory circuits identifi ed by GWAS are variants (mainly SNPs).4 The majority of radically improving our fundamental under- these data are, however, derived from standing of human development, physiology European, Asian and African populations.5 and disorders, and opening doors to the And therein lies a potential problem for repurposing of drugs or development of people whose ancestors voyaged the oceans novel treatments.3,4 There now seems little and founded Aotearoa, and for many more doubt that our expanding knowledge of recent migrants to this country. Although human genomics will steadily change the many fi ndings in genomic medicine will be way we understand, diagnosis, manage, and equally applicable to all people, it is also perhaps ultimately prevent, the common clear that there will be some signifi cant illnesses that have the greatest impact on genetic differences of medical relevance that human suffering and healthcare costs. may be of greater importance to those of However, a fundamental concern that Māori or Polynesian descent. Two important may limit the widespread, equitable appli- examples are SNPs that affect the function 6 cation of genomic methods in New Zealand of the pharmacogene CYP2C19 and the 7 is the relative dearth of detailed knowledge CREBRF gene, which infl uences body mass about the nature and frequency of specifi c index. While the existing genome data- genetic variants that may infl uence the bases may be suffi cient to enable genomic health and well-being of Māori and Poly- medicine applications based on genetic nesian people. While we all share essentially variants shared by all New Zealanders, the the same genome, the precise patterns of relative paucity of baseline genome data

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 8 www.nzma.org.nz/journal EDITORIAL

for Māori and Polynesian people will limit Many opportunities and challenges lie development of applications tailored to ahead in order to realise this promise of a people of these ancestral origins. “genome project” for Māori and Pasifi ka A viewpoint paper in this issue provides people. The opportunities include the 9 a robust blueprint to address this chal- Genomics Aotearoa platform which lenge, by clearly laying down the rationale provides an infrastructure and government and ground-rules for an ambitious plan to funding for the proposed project; the work develop a repository of genomic data drawn carried out to establish culturally appro- from Māori and Polynesian volunteers.8 The priate guidelines for genomic research and 10 key elements of this proposed framework biobanking in Māori ; and the database are that governance and management will itself, which it is hoped, will drive discovery be led by Māori and Pacifi c representa- research that ultimately leads to improved tives, the work will be informed by cultural health outcomes for its primary stake- concepts and values, it will be acknowledged holders. In addition, the scale of this that DNA storage, utilisation and interpre- project will provide opportunities for tation is a culturally signifi cant activity, and workforce development around the scien- that the primary use of the data generated tifi c and medical aspects of genomic data. should be to drive improved understanding Māori and Pasifi ka community leaders of the genetic contributors to health and education groups have done well to outcomes of Māori and Pasifi ka. support training and development of many lawyers, business people, and doctors, but There is some very important context we need more scientists drawn from these for the rationale suggested by Robertson et communities who are versed in genetics, al.8 While to some extent the approaches bioinformatics, and big data analyses. proposed in this paper may seem contrary to the principles of open science and On the other hand, challenges include extensive sharing of data (with partic- garnering wide support for a national ipant consent) that have underpinned resource from the many Māori and Pasifi ka this fi rst decade of medical genomics, it communities yet to be engaged, generating is important to remember that human the genomic data on-shore and ensuring its genetics has a chequered history including storage in a secure and culturally appro- major missteps that have engendered priate manner, and establishing procedures great mistrust amongst indigenous people that maximise the health gains from these around the globe.8 Add to this the centrality data while limiting the perceived risks of whakapapa (genealogy) to the Māori posed. The last of these would seem to world view, the signifi cance of DNA and present the greatest challenge, but once genomic data as a taonga (treasure) linked overcome, will be the greatest source of to whakapapa, and the explicit obligations opportunity. The framework proposed by 8 in the Treaty of Waitangi that the rights, Robertson et al, provides a clearly stated interests and taonga of Māori be protected, and culturally appropriate path to a future and the foundations of this proposal become where genomic medicine will contribute to abundantly clear. reductions rather than increases in health disparities, in Aotearoa/New Zealand.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 9 www.nzma.org.nz/journal EDITORIAL

Competing interests: Nil. Author information: Martin A Kennedy, PhD, Research Professor and Head of Department, Pathology and Bio- medical Science, University of Otago, Christchurch. Corresponding author: Martin A Kennedy, Research Professor and Head of Department, Pathology and Biomedical Science, University of Otago, Christchurch, P.O. Box 4345, Christchurch. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2018/vol-131-no-1480- 17-august-2018/7657

REFERENCES: 1. Kennedy MA. What 5. Lek M, Karczewski KJ, New Zealand. N Z Med J. does the human genome Minikel EV, Samocha KE, 2018;131(1480):82-90. project mean for Banks E, Fennell T, et al. 9. Dodson M, Williamson medicine? N Z Med J. Analysis of protein-cod- R. Indigenous peoples 2001;114(1130):190-2. ing genetic variation in and the morality of the 2. Biesecker LG, Green 60,706 humans. Nature. Human Genome Diversity RC. Diagnostic clinical 2016;536(7616):285-91. Project. J Med Ethics. genome and exome 6. Gladding P, White H, 1999;25(2):204-8. sequencing. N Engl J Webster M. Prasugrel, 10. Unknown. Genomics Med. 2014;371(12):1170. Maori, and personalised Aotearoa 2018. Available 3. Visscher PM, Wray NR, medicine in New from: https://www.otago. Zhang Q, Sklar P, McCarthy Zealand. N Z Med J. ac.nz/biochemistry/ MI, Brown MA, et al. 10 2010;123(1310):86-90. research/themes/ Years of GWAS Discovery: 7. Minster RL, Hawley NL, otago673820.html Biology, Function, and Su CT, Sun G, Kershaw EE, 11. Hudson M, Russell K, Translation. Am J Hum Cheng H, et al. A thrifty Uerata L, Milne M, Wilcox Genet. 2017;101(1):5-22. variant in CREBRF strongly P, Port RV, et al. Te Mata 4. Manolio TA, Fowler DM, infl uences body mass Ira—Faces of the Gene: Starita LM, Haendel MA, index in Samoans. Nat Developing a cultural foun- MacArthur DG, Biesecker Genet. 2016;48(9):1049-54. dation for biobanking and LG, et al. Bedside Back to 8. Robertson Stephen P et al. genomic research involving Bench: Building Bridges Genomic medicine must Māori. AlterNative: An between Basic and Clinical reduce not compound, International Journal Genomic Research. Cell. health inequities:the case of Indigenous Peoples. 2017;169(1):6-12. for hauora-enhancing 2016;12(4):341-55. genomic resources for

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 10 www.nzma.org.nz/journal EDITORIAL

Is there a place for day hospitals in older persons mental health care? Matthew Croucher

n this issue of the Journal, senior clini- in respect of the funding and planning of cal psychologist Dr Hoggarth presents services such as this day hospital to quan- Iencouraging clinical outcomes data for titative data since the kinds of studies that an older persons mental health day hospital would need to be mounted to compare that has been running in the Canterbury interventions are seldom feasible in the District Health Board since 1987.1 Her paper New Zealand funding environment. Indeed, replicates in more detail previous fi ndings some of these questions suit themselves using a simpler measurement instrument better to qualitative data and some are best that demonstrated a robust benefi cial effect answered by the application of common for depression and a moderate effect size for sense and experience to complex systems: anxiety and other relevant psychiatric symp- that somewhat maligned property of key toms, excluding cognitive impairment and service leaders we call ‘expertise’. 2 substance misuse. This day hospital ‘works’ The CDHB operated a geriatric day (disclaimer: I have been working in this day hospital from 1984 until a phased closure hospital since January 2018). from 2008 -2012, since when it shifted to The day hospital routinely collects providing all community and, crucially, all outcomes data in respect of some key post-discharge multi-disciplinary care by clinical domains via the Ministry of teams that could visit people in their own Health-mandated “Health of the Nation homes. Stroke care was the fi rst to move Outcomes Scale – 65 plus”3 as well as with followed by the remainder of older persons some more fi nely grained measures that health care not served by the medical outpa- had previously been tested and found tient clinic system. The initial formation of to be suitable in this service setting. It this day hospital was a giant step forward hardly needs to be argued that outcomes from the previous model of care not only measurement should be routine for all because it was a brand new service addition clinical services in order to establish the but because the interdisciplinary healthcare effectiveness of the service for a specifi c it facilitated was more effective and because population, to monitor potential harms, it shortened hospital admissions and some- to identify service gaps, to help assess a times avoided them entirely, with better service’s value for money, and hopefully to rehab outcomes. help identify which patients will be most The Canterbury health system evolved or least benefi tted so that referrals can be beyond this geriatrics day hospital model better triaged to optimise outcomes. after twenty fi ve years because a strong However, formal metrics are not the only argument was made for the enhanced source of information on which to base setting and achievement of rehabilitation decisions about the ‘worthwhile-ness’ of goals in a person’s own home with their own services. How do health interventions stack care partners, as well as for removing the up against other means of achieving the attendance barriers posed by a centralised same goals in terms of costs, accessibility, day hospital. This movement was supported and acceptability? How well do services by good data (it also ‘works’) but the decision fi t into the overall healthcare web for the was also informed by expert opinion, a very specifi c population concerned? It is review of model services elsewhere, and not appropriate to limit considerations listening to the voices of the patients and

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 11 www.nzma.org.nz/journal EDITORIAL

their care partners. An important key word are key elements of the treatment. Reha- was “empowering”. Nevertheless, the costs bilitation does meaningfully take place in from this change of strategy include the this artifi cial environment, in no small part expense of transport for staff doing home because it is not ‘real life’. The day hospital visits, the loss of fi t-for-purpose facilities, patients and staff do indeed become a and the opportunity costs from staff time “tangata whaiora” – a community of people spent in transit rather than in clinical work, moving towards health. the loss of in-the-moment access to the range Psychotherapists recognise the importance of interdisciplinary team members, loss of of the relationship and context of therapy to the benefi ts of being part of a temporary its success and this is no less true of a group community of care with other like people, process. In addition, the complexity of the and the loss of respite from the realities of inter-disciplinary assessment and treatment the home environment for both the patient process is usually high for the psyche and their live-in care partners. whereas it is much less for some discrete What then for psychogeriatric day problems in the soma, adding weight to hospitals in the early twenty-fi rst century? a day hospital’s one-stop shop approach. My own appraisal of this service mode has Although the service emphatically does not changed signifi cantly after listening to and provide ‘day care’, I have also observed understanding more about the population it that the respite function looms large as a serves. My previous more ambivalent view treatment in itself for both the day hospital’s of day hospitals per se was, regrettably, not patients and their care partners, because the supported by data, quantitative outcomes or core problems people are facing affect not otherwise – as Dr Hoggarth has pointed out, only what they do but very much who they there hasn’t been anything much until now.1 are. It was more guided by a bias that given the The questions I am now left with are resource intensiveness of a day hospital how to best ensure that the older people service and the general swing toward for whom this level of intervention is community-based health services (“better, essential can make their way to it (and sooner, more convenient”, anyone?), surely less well matched people can be referred the opportunity costs must be too high. on elsewhere), and how to best mitigate My impression has changed in the face the potential negatives and undeniable of the facts. I have seen for myself that opportunity costs of this centralised, with older persons anxiety and depressive resource-intensive service mode? disorders in particular, but also with In addition to everything I knew about cognitive impairment and psychosis at the collecting and analysing outcomes metrics milder end of the spectrum, being part of a with rigorous quantitative methods, joining community of care is not just a nice-to-have this service has opened my eyes to the but is critical for some folk to gain the importance of taking a thoughtful look trust required to form a meaningful thera- at what staff, patients and care partners peutic alliance. It provides a ready-made say; and of contextualising this within safe place in which to practise new ways an overview of services as a whole when of thinking and behaving, and the mutual coming to a view about the merit of a encouragement and more realistic goal- service. setting that are facilitated by a group context

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 12 www.nzma.org.nz/journal EDITORIAL

Competing interests: Matthew Croucher is a part time consultant psychiatrist of old age and team member at the Burwood Day Clinic, the day hospital of the CDHB's Older Persons Mental Health service. Author information: Matthew Croucher, Psychiatrist of Old Age, Canterbury District Health Board, Christchurch, New Zealand. Corresponding author: Matthew Croucher, Older Persons Mental Health, Burwood Hospital, Private Bag 4708, Christchurch 8140, New Zealand. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2018/vol-131-no-1480- 17-august-2018/7658

REFERENCES: 1. Hoggarth, P.A. Psychogeri- 2. Gee, S.B., Croucher, M.J., 3. Burns, A., Beever, A., atric day hospital reduces & Beveridge, J. (2010) Lelliott, P., Wing, J., Blakey, depression and anxiety Measuring outcomes in A., et al. (1999a) Health of symptoms and improves mental health services for the Nation Outcome Scales quality of life. N Z Med J. older people: an evaluation for elderly people (HoNOS 2018;131(1480):18-23. of the HoNOS65+. Interna- 65+). British Journal of tional Journal of Disability, Psychiatry, 174, 424-427. Development and Educa- tion, 57(2), 155-174.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 13 www.nzma.org.nz/journal EDITORIAL

Provincial or Tertiary Surgical Treatment for Rectal Cancer in New Zealand? Ian Bissett

ne of the key elements of the size. There were some differences in the New Zealand Health Strategy re- populations with more patients of a higher leased by the Ministry of Health is ASA in the tertiary centres and somewhat O 1 ‘Closer to home’ or ‘Ka aro mai ki te kāinga’. greater numbers in the lower third of the Providing surgery for patients with cancer rectum in the provincial hospitals. at the District Health Board of domicile There were no differences in the hard meets this aspiration. Internationally, how- endpoints of 30-day mortality (0.8% versus ever, there has been much debate around 1.4%) and length of hospital stay (12 versus the importance of centralising diffi cult oper- 12.5 day). Overall complications and anas- ations to hospitals that carry out a high-vol- tomotic leaks also occurred at similar 2-5 ume of these procedures. Much of the rates. The laparoscopic approach was international data comes from North Amer- more likely to be used in the provincial ica where surgical outcomes appear to vary hospital perhaps refl ecting the fact that greatly across different settings. It is uncer- the both tertiary centres were involved in tain whether the results obtained from these the laparoscopic versus open randomised countries, with their own particular health controlled trial of rectal resection at this systems, are applicable in smaller countries time. There were differences identifi ed in with different health systems and approach- some of the other quality indicators that 4 es to surgical training. The authors of the represented the multidisciplinary nature of study ‘A multicentre, benchmarking study rectal cancer management. These included of rectal cancer management in provincial fewer completed colonoscopies (91% New Zealand hospitals’ have attempted versus 99%), fewer pathological specimens to take one relatively common and com- with 12 or more lymph nodes identifi ed plex condition, rectal cancer, and identify (41% versus 76%) and fewer reports of the whether clinical outcomes are comparable quality of mesorectal dissection (62% versus between smaller hospitals and large tertiary 77%) in the provincial centres. Perhaps 6 institutions in the New Zealand context. the most important measure of multidis- In this issue of the Journal O’Grady et al ciplinary decision making and surgical have presented the results of a retrospective quality, the incidence of positive margins study of those who underwent elective rectal on the resected specimen, was not different cancer resection in six provisional New between the hospital sites (2% versus 4%). Zealand hospital in 2012 and 2013. These A further study by Keane et al has recently results have been compared with data from been published reporting quality indicators two tertiary hospitals over the same time of treatment for rectal cancer across the period. There were 124 patients with rectal whole New Zealand population for the year cancer from provincial hospitals and 145 of 2015.7 The methodology of this study from tertiary hospitals. Although there were differed in that it obtained the data from a many more procedures performed in the full year’s rectal cancer patients from the tertiary hospitals overall, the actual number National Cancer Registry. Each pathology of resections undertaken by individual report for a resected rectal cancer specimen surgeons annually did not differ by hospital was viewed and data relating to surgeon,

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 14 www.nzma.org.nz/journal EDITORIAL

operating hospital, laboratory, type of report Overall it is heartening to note that the (structured or not), mesorectal quality, management of rectal cancer nationally lymph node number, and circumferential is improving and in terms of the quality margin (CRM) status were recorded. This indicators measured compares well with is the most recent complete national data international studies. Mortality and compli- available and includes both publicly and cation rates are comparable between privately performed procedures. This tertiary and provincial centres and improve- dataset has identifi ed higher national rates ments in the other markers (lymph node of the reporting of mesorectal quality (82%) yield and mesorectal quality) should be the and CRM status (87%) than those reported focus of service improvement initiatives by O’Grady et al6; however, the rate of CRM nationally. This could be achieved if all involvement by cancer was similar at 6%. laboratories reported every rectal cancer Although the national fi gures for these resection specimen in a structured format. quality indicators are encouraging and have At least in the case of rectal cancer shown improvement since the PIPER study, management, the studies by O’Grady et al6 which included patients from 2007 and and Keane et al7 have shown that patients 2008, there is still a wide range of outcomes who receive their care closer to home are affecting some individual hospitals and mostly having comparable outcomes to 7,8 laboratories. The reporting of lymph node those treated in tertiary centres.6 Relatively numbers is not just of theoretical interest simple initiatives could also decrease the as it has been demonstrated that patients few areas were differences have been iden- with node negative stage and fewer than tifi ed. A concerted effort to close this gap 12 lymph nodes identifi ed have signifi - should be our priority. This should be moni- cantly worse outcomes than those with 12 tored by further national audits completed or more examined. Complete mesorectal at regular intervals to ensure that those who dissection is associated with lower rates of are managed in provincial centres get care local recurrence and incomplete recording not only closer to home but also as good as of this measure may mask potentially poorer in larger centres. quality treatment.

Competing interests: Nil. Author information: Ian Bissett, Department of Surgery, University of Auckland, Auckland. Corresponding author: Ian Bissett, Department of Surgery, University of Auckland, Private Bag 92019, Auckland 1142 [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2018/vol-131-no-1480- 17-august-2018/7659

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 15 www.nzma.org.nz/journal EDITORIAL

REFERENCES: 1. New Zealand Health cancer surgery. Surgery. Eglington T. A multicentre, Strategy 2016 [12 August 2016;159(3):736-48. benchmarking study of 2018]. Available from: 4. Leonard D, Penninckx F, rectal cancer management https://www.health.govt.nz/ Kartheuser A, Laenen A, in provincial New Zeland about-ministry/what-we-do/ Van Eycken E, Procare. hospitals. N Z Med J. new-zealand-health-strat- Effect of hospital volume 2018; 131(1480):24-32. egy-update. on quality of care and 7. Keane C, Lin AY, Kramer 2. Archampong D, Borowski outcome after rectal N, Bissett I. Can patho- D, Wille-Jorgensen P, cancer surgery. Br J Surg. logical reports of rectal Iversen LH. Workload 2014;101(11):1475-82. cancer provide national and surgeon’s special- 5. Xu Z, Becerra AZ, Justiniano quality indicators? ty for outcome after CF, Boodry CI, Aquina CT, ANZ J Surg. 2018. colorectal cancer surgery. Swanger AA, et al. Is the 8. Firth MJ, Sharples KJ, Cochrane Database Syst Distance Worth It? Patients Hinder VA, Macapagal Rev. 2012(3):CD005391. With Rectal Cancer Travel- J, Sarfati D, Derrett 3. Aquina CT, Probst CP, ing to High-Volume Centers SL, et al. Methods of Becerra AZ, Iannuzzi JC, Experience Improved a national colorectal Kelly KN, Hensley BJ, et Outcomes. Dis Colon cancer cohort study: the al. High volume improves Rectum. 2017;60(12):1250-9. PIPER Project. N Z Med outcomes: The argument 6. O’Grady M, O’Grady J. 2016;129(1440):25-36. for centralization of rectal J, Shine R, Bonnet G,

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 16 www.nzma.org.nz/journal ARTICLE

Psychogeriatric day hospital reduces depression and anxiety symptoms and improves quality of life Petra Ann Hoggarth

ABSTRACT AIMS: To measure changes in depression and anxiety symptoms, as well as quality and satisfaction with life in older adults attending a psychogeriatric day hospital. METHODS: One hundred and eighty-five consecutive patients (24% male) provided self-report data at intake and discharge. RESULTS: Patients showed significant reductions in anxiety and depressive symptoms and significantly improved quality of life and satisfaction with life. All changes were of moderate e ect size. CONCLUSION: Attending a psychogeriatric day hospital that provides time-limited personalised care is associated with statistically significant and clinically meaningful reductions in anxious and depressive symptoms for patients, as well as increasing subjective satisfaction and quality of life. Given the projected rise in the number of older adults in coming decades, the establishment of more psychogeriatric day hospitals should be considered as an alternative to expensive and stretched inpatient services.

he population of older adults in devel- for people with dementia.3,4 This is quite a oped countries is increasing markedly, different model from day hospital treatment Twith estimates that at least 25% of primarily for psychiatric disorder in a people in these countries will be aged 60 and majority cognitively intact cohort. 1 over by 2050. Social and medical services Only one study could be found that will be stretched and providing effi cient ser- measured treatment outcomes for a psychi- vices will be essential. Psychogeriatric care atric population. Bramesfeld et al5 ran an provided in a day hospital may be an effec- observational study that assessed 44 older tive option for service delivery for patients adults (mean age 68.9) in a German day with more needs than can be met by outpa- clinic. They found a reduction in depressive tient visits alone, but who do not require an symptoms after 11 weeks of treatment. inpatient admission. The needs of older adults are different Research into the effectiveness of psychi- from the needs of younger people. atric day hospitals is scarce, and research Therefore, planners need evidence as to that focuses on older adults is rarer still. The whether a psychogeriatric day hospital evidence for the effectiveness of psychogeri- can provide mental health benefi ts that atric day hospitals has been judged as “very could make it a viable treatment option weak” due to the low number of studies for the burgeoning older adult population. and lack of controlled and randomised The current study investigates whether 2 designs. The dearth of research is likely in treatment at a day hospital was associated part related to day hospitals having different with decrease in depressive and anxious functions and goals. Commentators note that symptoms, and increases in satisfaction with many psychogeriatric day hospitals in the and quality of life. UK focus on providing day or respite care

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 17 www.nzma.org.nz/journal ARTICLE

time of the study was always a consultant Methods psychiatrist (mostly psychiatrists of old In New Zealand there is only one psycho- age), an experienced senior medical offi cer geriatric day hospital, run by the Canterbury or a supervised registrar doing advanced District Health Board (one of 20 district training in psychiatry of old age. Patients health boards across the country). The also have access to additional health day hospital was fi rst introduced as a pilot workers as required, including an inconti- in 1987 at Sunnyside Hospital in Christ- nence nurse and pharmacist. church (now Hillmorton Hospital). In 1997, Patients referred to the Burwood Day the day hospital service was moved to Clinic come from a larger pool of older The Princess Margaret Hospital. In 2016, adults referred to the Older Persons Mental it was moved to purpose-built facilities at Health service. All come following an Burwood Hospital and was renamed the assessment from some part of this wider Burwood Day Clinic. The day hospital is service, including a doctor, nurse or allied part of a larger older adult mental health health assessment following discussion in service, providing for all adults 65 and over the community team IDT, the psychiatric in the catchment area (with a few excep- inpatient ward, the memory assessment tions). The older adult mental health service clinic or from a medical ward through the includes the day hospital, a community consult liaison team. Patients are referred team, a consult liaison team, two inpatient to the Burwood Day Clinic when their wards and a memory assessment clinic.. acuity is believed to be higher than can be Current staffi ng of the day clinic consists managed well with outpatient appointments of one full-time equivalent (FTE) clinical through the community team, but lower manager, three FTE nurses, one FTE doctor, than required for an inpatient admission. one FTE clinical psychologist (shared with Often the purpose of the admission is to two inpatient wards), one FTE social worker, clarify diagnosis, including that of cognitive one FTE occupational therapist, 0.75 FTE impairment, providing treatment in an occupational therapist assistant, one FTE IDT environment and for determining the physiotherapist, one FTE physiotherapist level of assistance required for successful assistant (each shared with two inpa- living in the community, or for determining tient wards) and one FTE administrator. whether the person wants/needs to move to The medical cover for the unit during the aged residential care.

Figure 1: The lounge, dining and kitchen areas at the Burwood Day Clinic.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 18 www.nzma.org.nz/journal ARTICLE

Figure 2: Large room for family meetings and group treatment.

The day hospital operates fi ve days a week The Clinically Useful Depression Outcome from 10am to 3pm. Patients are provided Scale (CUDOS)6 is a self-report measure that with transport, lunch, and morning and rates the severity of all DSM-IV symptom afternoon teas. Most patients attend one criteria for major depressive and dysthymic day per week, with fl exibility to increase disorders. It also includes an item rating the the number of days in response to symptom degree of disruption caused by symptoms exacerbations or increased risk level. of depression, and an item rating quality Patients take part in a range of planned of life. The Geriatric Anxiety Inventory group activities and meet individually with (GAI)7 was developed specifi cally for older members of the interdisciplinary team, adults using plain language, a simple yes/ depending on their individualised treatment no response format, and it omits somatic plan. Facilities include a combined lounge, symptoms that are common in all older dining and kitchen area; an outside adults. The Satisfaction with Life Scale courtyard and seating area (Figure 1); (SWLS)8 assesses a subjective judgement interview rooms; a treatment room; a large of satisfaction with life. This scale consists meeting room for family meetings and of fi ve questions measured on a 7-point group treatments (Figure 2); and a gym for scale. Two research reviews by the authors physical therapy. identify several studies that have used the 8,9 Beginning in late 2014, three self-report SWLS with diverse groups of older adults. measures began to be routinely used for Day hospital patients were administered every patient at intake and discharge as part the questionnaires within their fi rst few of an ongoing auditing process of the effec- visits following admission. When nurse tiveness of day clinic intervention, as well case managers believed that the patient as to provide useful information for patients was competent to complete the scale and their case-managers. The measures independently they provided the scales to were selected to assess symptoms of the complete and return. Very unwell, cogni- most frequent diagnoses seen at the clinic tively impaired or literacy challenged (depression and anxiety disorders), as well patients completed the questionnaires as measures of quality and satisfaction with with the nurse case manager who read life as a proxy for functionally signifi cant the questions and response options for the symptom change. patient to respond to verbally. The scales

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 19 www.nzma.org.nz/journal ARTICLE

Table 1: Comparison of median depression, anxiety, and satisfaction with life scores from intake to discharge.

Self-report scale n* Intake Discharge Z P-value E ect size (r) Clinically Useful Depression 176 21 8 -9.17 <.001 0.49 Outcome Scale (CUDOS) total score

CUDOS q.17** 183 2 1 -8.73 <.001 0.46

CUDOS q.18*** 181 2 1 -7.39 <.001 0.39

Geriatric Anxiety Inventory 183 13 5 -8.69 <.001 0.45

Satisfaction with Life Scale 180 18 23 -6.27 <.001 0.33

*Number of participants with both intake and discharge data for comparison for this variable. **Degree of disruption caused by depression symptoms. ***Quality of life. The question is reverse scored meaning that a lower score equates to a higher quality of life. were completed again at discharge in the As shown in Table 1, scores on all same manner. Statistical analyses consisted measures signifi cantly improved from of descriptive statistics and a comparison intake to discharge. All improvements were of intake and discharge scores using the of moderate effect size (r >.30). Wilcoxon Signed-ranks test. This is similar to a paired-samples t-test, used when data Discussion is non-parametric (in this case ordinal). This is the largest published study to show Patients missing intake or discharge data an improvement in psychiatric symptoms were excluded from analysis. A small and functioning for older adults during number of individual missed items on attendance at a day hospital. Statistically measures led to case-wise exclusion of that and clinically signifi cant improvement was patient (deletion just from that particular seen on all measures, with a reduction in comparison). depressive and anxious symptom severity, This study has been approved by the reduced distress caused by symptoms and University of Otago Human Research Ethics improved quality and satisfaction with Committee as an audit study (reference life. This study supports the claim that a HD16/037). day hospital providing care intermediate to outpatient and inpatient treatment may Results produce clinical benefi ts for older adults From November 2014 to December 2017 with mental health diagnoses. there were 308 patients who completed at There are a number of limitations to least one of the assessments. Many at the this study. Most notably is that the study is beginning of the time period completed observational and did not have a control discharge assessments only since the study group, therefore causal statements cannot had not begun at the time of their intake. be made, as an unknown amount of Many other patients at the end of this time variance in symptom reduction will be due period had intake data only as they were to factors that were not measured, such as still currently attending the day clinic. One natural recovery over time. Since there was hundred and eighty-fi ve patients (24% no comparison to another active treatment male) had both intake and discharge scores we also cannot compare how treatment available and were included in the analysis. at the day hospital compared to standard The mean length of admission was 22 weeks outpatient or inpatient care. Observational (min = 2, max = 70). At admission patients and pilot studies play an important role ranged in age from 57 to 93 years (mean = in the beginning of data collection into 76.6). Age distribution was 0.5% below age emerging areas of care and treatment. This 65, 14.1% in the 65–69 group, 25.4% aged is only the second study published on this 70–74, 24.9% aged 70–74, 21.1% aged 80–84, topic, and comes 17 years after the original 12.4% aged 85–89 and 1.6% aged 90+. study by Bramesfeld et al,5 with a sample

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 20 www.nzma.org.nz/journal ARTICLE

size four times larger. Studies such as this will be recruited over time, which will allow provide evidence that the effort involved in for fi ner-grained investigation of who is recruiting a control group may be a justi- benefi tting from the day hospital model, fi able next step to determining the success eg, is there a difference in outcome by age of day hospitals. group or symptom severity? Data collection Another limitation is that the data will also be revised to include both ethnicity collected does not record specifi c diagnoses data and whether patients were living in the for each patient. Given that most patients community or in residential care. who attend the clinic are given a diagnosis This paper has not investigated the of major depressive episode and/or an economic viability of establishing or anxiety disorder, this may not be too much running a day hospital on a per patient of an issue as we are directly measuring basis, but can provide some information as these symptoms. However, it would be to costs of the day clinic. The clinic has bene- possible when sample sizes are larger to use fi ted from a purpose-built facility in which cut-off scores to delineate diagnostic groups staff were able to have signifi cant input and compare how they fared at the end of into the design. However, a purpose-built treatment. Data on physical and cognitive facility is not a prerequisite and the clinic health conditions are also not systematically was run from 1987 to 2016 without such collected. It is not uncommon for patients an advantage. The largest cost is staffi ng. to have major or minor surgical procedures Other signifi cant costs include transport while attending the day hospital. Strokes for patients to and from the clinic (this is and heart attacks and new diagnoses of currently performed with a mix of hospital serious physical illness occasionally occur. A and taxi transport), meals (provided by the minority of patients are diagnosed with mild hospital) and food supplies, staff training cognitive impairment or dementia during and education, admin and stationary costs, their time at the clinic. Because data about and activity and craft supplies. This is the these events was not collected for this study, fi rst study since Bramesfeld et al5 to provide their impact on the psychiatric symptoms any data about psychiatric treatment of and quality of life cannot be explored in older adults in a day clinic. This study post-hoc analyses. provides additional evidence that day As data collection is now part of standard hospital treatment for older adults may clinical practice, an increasing sample size indeed be an effective method for treating psychiatric disorder.

Competing interests: Dr Hoggarth works at the Burwood Day Clinic as a clinical psychologist. Author information: Petra Ann Hoggarth, Older Persons Mental Health, Burwood Hospital, Burwood; Canterbury District Health Board, Christchurch. Corresponding author: Dr Petra Hoggarth, Burwood Day Clinic, Older Persons Mental Health, Burwood Hospital, 300 Burwood Road, Burwood, Christchurch 8083. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2018/vol-131-no-1480- 17-august-2018/7660

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 21 www.nzma.org.nz/journal ARTICLE

REFERENCES: 1. United Nations, Depart- 3. Fasey C. The day hospital 7. Pachana NA, Byrne GJ, ment of Economic and in old age psychiatry: The Siddle H, Koloski N, Harley Social Affairs, Population case against. Int J Geriatr E, Arnold E. Development Division. World Popula- Psychiatry. 1994; 9:519–23. and validation of the tion Prospects: The 2015 4. Rosenvinge HP. The role Geriatric Anxiety Inven- Revision, Key Findings of the psychogeriatric tory. Int Psychogeriatr. and Advance Tables. day hospital. Psychiatr 2007; 19:103–14. Working paper number: Bull. 1994; 18:733–36. 8. Pavot W, Diener E. ESA/P/WP.241, 2015. 5. Bramesfeld A, Adler G, Review of the satisfaction 2. Draper B, Low L. What is Brassen S, Schenitzler M. with life scale. Psychol the effectiveness of old-age Day-clinic treatment of late- Assess. 1993; 5:164–72. mental health services? life depression. Int J Geriatr 9. Diener E. Review of the Copenhagen, WHO Region- Psychiatry. 2001; 16:82–7. satisfaction with life scale. al Offi ce for Europe [Cited 6. Zimmerman M, Chelminski Assessing Well-Being: 29 April 2017.] Available I, McGlinchey JB, Posternak The Collected Works from URL: http://www. MA. A clinically useful of Ed Diener. 2009. euro.who.int/__data/assets/ depression outcome Springer Netherlands. pdf_fi le/0008/74690/E83685. scale. Compr Psychiatry. pdf 2008; 49:131–40.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 22 www.nzma.org.nz/journal ARTICLE

A multicentre, benchmarking study of rectal cancer management in provincial New Zealand hospitals Michael J O’Grady, Josephine O’Grady, Rebecca Shine, Gerard Bonnet, Tim Eglinton

ABSTRACT AIM: The quality of rectal cancer management within New Zealand provincial hospitals is largely unknown. This study was conducted to appraise and benchmark the management of rectal cancer in provincial New Zealand centres as compared to specialist tertiary centres. METHOD: Retrospective data was collected for all patients who underwent elective rectal cancer resection in six provincial New Zealand hospitals from January 2012 to December 2013. This was then compared with data from two tertiary hospitals over the same time period. The complete management pathway was evaluated. RESULTS: A total of 124 provincial and 145 tertiary rectal cancer resections were analysed. Completeness of preoperative staging was comparable between provincial and tertiary centres, as was type of surgical procedure performed and rates of clear surgical margins. A statistically significant di erence was observed in mean number of lymph nodes analysed (10.3 v 17.2), reporting of mesorectal grade (61% v 77%), and completion colonoscopy rates (91% v 99%), all of which were lower in provincial hospitals. Multidisciplinary team discussion, rates of neoadjuvant therapy and post-operative parameters such as 30-day mortality (0.8% v 1.4%), length of stay (11.9 v 12.4 days), anastomotic leak (7% v 5%) and return to theatre (8% v 8%), were similar. CONCLUSION: Management of rectal cancer in provincial hospitals is comparable to specialist centres, however lymph node harvest, reporting of mesorectal grade and complete colonoscopy were factors identified which were lower in the provincial group. Provincial rectal cancer management remains an important resource for patients living outside major centres.

ew Zealand has one of the highest MRI to assess tumour and lymph node stage incidences of colorectal cancer in as well as imaging of the chest and abdomen the world, with approximately 30% to detect metastases. Information is then N 1,2 of these cancers arising in the rectum. collated and discussed at a multidisciplinary Surgical resection remains the primary meeting (MDM) where treatment pathway curative treatment, however optimum ther- is determined. Typically, locally advanced apy requires a complex, multidisciplinary tumours will receive neoadjuvant therapy approach and specialised care. followed by surgery whereas early cancers The standard of management in New can proceed straight to surgery. Zealand begins with accurate preoperative Total mesorectal excision is the accepted diagnosis and workup. Colonoscopy is the technique for resection of the rectum. predominant means of diagnosis.2 Staging Complete excision with clear margins is is determined through utilisation of pelvic a signifi cant prognostic factor for local

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 23 www.nzma.org.nz/journal ARTICLE

recurrence and metastatic disease, and 1 January 2012 and 31 December 2013 were is an important predictor of long-term identifi ed by authors MO, JO and RS through survival.4,5 The number of lymph nodes hospital coding (ICD-10 code C-20), and harvested and subsequently analysed through prospective hospital audit data- from an excision specimen should meet an bases where available. Clinical records were acceptable threshold, with 12 or more lymph reviewed retrospectively to obtain clinical nodes commonly accepted as an adequate outcomes. Data collectors were not able to analysis.6–8 The ratio of anterior resection to be blinded to the surgeons involved. abdominoperineal resection has been used MRI reports were studied to confi rm as a discriminator of surgical technique distance from anorectal junction, classifi ed with some studies reporting higher rates as lower (≤7cm), mid (8–12cm) or upper of sphincter preservation in high-volume rectum (12–15cm). Where MRI reports did 11,25 hospitals. Post-operative complications not provide distance from anorectal junction are also surrogate markers of surgical measurements were taken from, in order technique, the most important being 30-day of preference, rigid sigmoidoscopy, oper- mortality, reoperation rate and anastomotic ation notes and MDM records. Demographic leak rate. data was recorded as well as selected key There is ongoing debate about whether performance indicators derived from patho- effective management of rectal cancer can logical data, preoperative work-up including be provided outside of specialist colorectal details of neoadjuvant therapy, operative centres. International literature has shown parameters and postoperative outcomes. an association between surgeon and hospital These variables were identifi ed from the volume, and patient outcomes with higher New Zealand Guidelines Group document: rates of positive circumferential margins, Management of Early Colorectal Cancer3 and non-restorative procedures and periop- other international guidelines.6,20–22 erative mortality demonstrated in lower The data analysis for this paper was 11–15 volume hospitals. generated using SAS software© (SAS In New Zealand, geography and popu- Institute Inc., Cary, NC, USA). Values lation distribution dictate the provision were expressed as Mean ± SD or relative of rectal cancer services through a combi- frequency. Age, hospital post-operative nation of specialist and provincial centres, length of stay and number of nodes and while it is of benefi t for patients to harvested were analysed by Mann-Whitney receive care in their local area, avoiding U test as data did not meet the normality costs of travelling and dislocation from assumption (Shapiro-Wilks test for support services, this care needs to be of continuous variables). Chi-square or Fisher’s comparable standard to that provided in exact test were employed to compare tertiary centres. The aim of this research, frequencies distributions between groups therefore, is to examine in detail the rectal for the remainder variables. Signifi cance set cancer management pathway, directly at p<0.05. comparing provincial to tertiary hospitals. Review was sought from The Health and Disability Ethics Committee. Formal review Method was not required as this was a straight- Six provincial hospitals and two tertiary forward audit which did not seek further hospitals were included for analysis in this information from patients. Māori consul- study. Hospitals included were those at tation was undertaken and locality approval which the authors were employed. There was sought from participating hospitals. were no exclusion criteria. Individual hospital data including total number of Results rectal cancer resections performed during Between January 2012 and December the observed time period and number of 2013, a total of 124 patients in provincial surgeons undertaking these procedures was centres and 145 patients in tertiary centres recorded. underwent elective surgery for rectal All patients undergoing elective resection cancer. These procedures were performed for adenocarcinoma of the rectum between by nine surgeons over six provincial

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 24 www.nzma.org.nz/journal ARTICLE

Table 1: Individual hospital data.

Number of resections Number of surgeons Mean number of resections per surgeon per year Provincial hospitals 1 19 1 9.5

2 11 2 5.5

3 27 2 6.8

4 44 2 11

5+6 23 2 5.8

Tertiary hospitals 1 72 6 6

2 73 4 9.1 hospitals and 10 surgeons over two tertiary higher proportion of patients from tertiary hospitals (Table 1). The average district centres categorised ASA 3 compared to the health board (DHB) population base served provincial group (30% vs 18%, p 0.0312). In by a provincial hospital included in this addition, tumour was recorded as being in study was 106,613. This is compared to the lower third in a higher proportion of an average of 113,837 for all New Zealand provincial patients, 63% compared with 54% DHBs not served by a tertiary hospital.10 for tertiary patients (p 0.0252). Demographic parameters and individual Preoperative data patient characteristics are described in Signifi cantly less colonoscopies were Table 2. Patients from both centres were completed preoperatively, or within one found to be comparable in mean age and year of surgery for patients treated at male:female ratio, however there was a provincial centres as compared to tertiary signifi cant difference in ASA score with a centres (91% vs 99%, p 0.0036) (Table 3).

Table 2: Demographic distribution of patients from provincial versus tertiary centres.

Provincial (N=124) Tertiary (N=145) p value Age Mean (SD) 68.9 (9.8) 66.9 (12.2) 0.3596

Gender Female 40 (33%) 44 (30%) 0.6686 Male 82 (67%) 101 (69%)

ASA score 1 32 (30%) 26 (18%) 0.0312 2 54 (51%) 74 (51%)

3 19 (18%) 44 (30%)

Rectal level Lower third 76 (63%) 77 (54%) 0.0252 Middle third 33 (28%) 60 (42%)

Upper third 11 (9%) 6 (4%)

Tumour stage T1 8 (7%) 13 (9%) 0.5479 T2 27 (22%) 40 (28%)

T3 77 (64%) 79 (55%)

T4 9 (7%) 11 (8%)

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 25 www.nzma.org.nz/journal ARTICLE

Table 3: Analysis of preoperative workup and therapy for provincial versus tertiary centres.

Provincial (N=124) Tertiary (N=145) p value Colonoscopy 107 (91%) 143 (99%) 0.0036

CT 119 (99%) 145 (100%) 0.4528

MRI 113 (97%) 143 (99%) 0.412

Discussed at MDM 116 (97%) 143 (99%) 0.4148

Neoadjuvant therapy None 27 (22%) 49 (34%) 0.0991 Short course 14 (11%) 13 (9%)

Long course 81 (66%) 82 (56%)

No further signifi cant differences were tertiary patients undergoing harvest of 12 observed in preoperative work-up or or more lymph nodes (110 vs 47, p<0.0001). therapy with CT staging, MRI staging, multi- Reporting of mesorectal grading was lower disciplinary meeting (MDM) discussion in provincial hospitals (62% vs 77%, p rates and courses of neoadjuvant therapy 0.0158). Positive circumferential and distal otherwise comparable between groups. margins did not differ signifi cantly between Operative data groups. Surgical access was gained via an open Post-operative data technique in signifi cantly more tertiary No post-operative outcomes were found patients (70%) than provincial patients to reach statistical signifi cance in their (54%) (Table 4). There was no signifi cant variability between groups (Table 5). difference in type of procedure performed. Mean length of stay, 30-day mortality rate When pathology reports were studied, a and the incidence of surgical and medical lower mean number of lymph nodes were complications were all demonstrated to be recorded for provincial patients (10.3 vs comparable between provincial and tertiary 17.2, p<0.0001) with substantially more patients.

Table 4: Analysis of operative data for provincial versus tertiary centres.

Provincial (N=124) Tertiary (N=145) p value Surgical entry Laparoscopic 47 (39%) 18 (13%) <0.0001 Laparoscopic assisted 0 (0%) 23 (16%)

Convert 9 (7%) 3 (2%)

Open 65 (54%) 100 (70%)

Procedure type Low anterior resection 75 (62%) 96 (66%) 0.0502 Abdominoperineal resection 33 (27%) 42 (29%)

Hartmann’s procedure 13 (11%) 4 (3%)

Panproctocolectomy 1 (1%) 3 (2%)

Mean number of lymph nodes harvested (SD) 10.3 (7) 17.2 (9) <0.0001

12 or more lymph nodes harvested 47 (41%) 110 (76%) <0.0001

Positive circumferential margins (≤1mm) 2 (2%) 6 (4%) 0.2047

Positive distal margins (≤1cm) 1 (1%) 2 (1%) >0.9999

Mesorectal grading reported 55 (62%) 111 (77%) 0.0158

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 26 www.nzma.org.nz/journal ARTICLE

Table 5: Analysis of post-operative outcomes for provincial versus tertiary centres.

Provincial (N=124) Tertiary (N=145) p value Mean length of stay (days, SD) 11.94 (8.5) 12.43 (9.8) 0.748

30 day mortality rate 0.8% 1.4% >0.9999

Surgical complications Anastomotic leak 8 (7%) 7 (5%) 0.5409

Returned to theatre 10 (8%) 12 (8%) 0.9813

Wound infection 6 (5%) 11 (8%) 0.3737

Prolonged ileus 16 (13%) 21 (15%) 0.7472

Abdominal pelvic collection 8 (7%) 5 (4%) 0.2396

Superficial wound dehiscence 2 (2%) 1 (1%) 0.4633

Deep wound dehiscence 1 (1%) 3 (2%) 0.4025

Small bowel obstruction 3 (3%) 3 (2%) 0.8304

Venous thromboembolism 1 (1%) 2 (1%) >0.9999

Haemorrhage 2 (2%) 3 (2%) >0.9999

Total 49 (40%) 62 (43%) 0.6683

Medical complications Cardiac 3 (3%) 6 (4%) 0.5153

Chest infection 7 (6%) 8 (6%) 0.9379

Urinary tract infection 5 (4%) 1 (1%) 0.0962

Total 16 (13%) 27 (19%) 0.2223

demonstrated to apply to both surgeons and Discussion pathology laboratories in New Zealand by We have conducted a retrospective Keane et al. This study examined pathology study comparing preoperative, oper- reports for all rectal cancer resections ative and post-operative indicators for in 2015 and demonstrated high-volume patients undergoing rectal cancer surgery laboratories reported an adequate lymph at selected provincial and tertiary New node count in 74% of specimens compared Zealand hospitals. The majority of key with 54.6% for low volume, and respec- performance indicators were comparable, tively for high-volume surgeons (69.1%) however several important differences were to low-volume surgeons (59.8%).26 Other identifi ed. factors known to affect lymph node count A signifi cant fi nding was the lower such as patient age, tumour stage and number of lymph nodes analysed in rates of preoperative radiotherapy can be provincial patients compared to tertiary discounted as confounders as there were patients (10 versus 17). Not only was there a no observable differences in these baseline lower number in the provincial group, this characteristics. Laparoscopic surgery, value fell below the commonly accepted which was more common in the provincial threshold for accurate staging. Patient, group, is associated with lower counts and disease, surgical and pathology staff factors may have contributed to this difference. have been shown to affect lymph node Given the likely multifactorial nature of this count.16,17 Effect of volume was recently difference it is an important area for further examination and improvement.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 27 www.nzma.org.nz/journal ARTICLE

Our study found the rates of preoperative MDM discussion and indeed for adminis- and completion colonoscopy were signifi - tration of adjuvant therapy in provincial cantly lower in provincial hospitals. It is settings. In this subgroup of provincial interesting to note that the PIPER study hospitals three had resident oncolo- did not fi nd colonoscopy rates were signifi - gists, while the remaining had a visiting cantly infl uenced by travel distance and oncology service and in all cases neoad- deprivation, however other factors such as juvant therapy was provided by the nearest provincial centres’ resource constraints on tertiary hospital. This has been shown to be providing timely endoscopy or follow-up a problem in the past with lower rates of practices, such as dedicated colorectal MDM discussion in rural patients 2 and was nurse specialist follow-up, require further identifi ed as a focus area in the New Zealand investigation.2 National Rectal Cancer Summit, convened 19 Mesorectal grading was reported in a in August 2013. It is therefore encouraging lower percentage of provincial patients (62% that there were no observed differences. v 77%). The fi gures are higher than those Moreover, the rates of MDM discussion were reported in the PIPER study (45%), however much higher than those seen in the PIPER still lower than required, given its prog- study. The PIPER study includes data from nostic signifi cance. Routine reporting, as 2007 and 2008 and hopefully this increase part of synoptic reporting, has been a target is refl ective of the entire country, signifying area for improvement.19 The Royal College routine discussion of all patients in this of Pathologists of Australasia has developed forum. a standardised reporting template for In this study there were no signifi cant colorectal cancer, which includes mesorectal differences in key surgical performance grading. Its use is not currently mandated, indicators such as clear resection margins, however Keane et al has shown a high rate anastomotic leak, unplanned return to of adoption with 89.5% of all New Zealand theatre and 30-day mortality. Sphincter rectal cancers reported synoptically in 2015. preservation has been shown in several Further analysis in this paper demonstrated publications to be affected by hospital a difference in use of synoptic reporting, volume,11,14,15 however no signifi cant and reporting of mesorectal grading, in high- difference was seen in this study even versus low-volume laboratories (97% versus though provincial patients had a higher 81.1% and 86.5% versus 76.5% respectively) proportion of lower third tumours. It should suggesting room for further improvement be noted that classifi cation of tumour level particularly in low volume/provincial is notoriously diffi cult to standardise and laboratories.26 therefore this result may, or is likely to, There was a statistically signifi cant refl ect a difference in interpretation of difference in surgical entry between defi nition. groups, however, this data should be inter- This study was limited by its retrospective preted with caution. Both tertiary centres nature, the non-randomised selection of had surgeons who were participating hospitals and the relatively small sample in the ALaCaRT trial,18 where patients size with the potential for type two error. were randomised to open or laparoscopic Data on referral of complex patients was dissection of the mesorectum. Patients not captured in this study and may be randomised to open surgery could either a potential confounder. Examination of have the entire operation completed open baseline characteristics shows there was or undergo hybrid techniques of laparo- a higher proportion of patients in the scopic mobilisation of the splenic fl exure tertiary group classifi ed ASA 3, however and lymphovascular ligation with open there was no signifi cant difference in T4 mesorectal dissection—recorded as laparo- tumours (a proxy for patients requiring scopic assisted in this study. Interpretation more complex surgery). While it is unlikely of this difference is therefore not possible as higher ASA classifi cation would be indi- a result. cation for referral to a tertiary centre, it Rates of discussion at an MDM and could infl uence decision making in what subsequent neoadjuvant therapy were not is a multifactorial consideration. A further signifi cantly different. Barriers do exist for limitation of this study is the absence

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 28 www.nzma.org.nz/journal ARTICLE

of long-term data, specifi cally rates of to 150 cases per annum, and similarly local recurrence and survival. These are high volume surgeons ranging from fewer important oncological outcomes to which than 10 through to annual caseloads in further research could be addressed. Future excess of 40. A Cochrane review concluded replications of this study may also benefi t higher hospital, and particularly surgeon from the addition of ethnicity data. volume, was associated with improved There is limited direct analysis of the outcomes after rectal surgery but was quality of rectal cancer management unable to recommend a defi nition of high 11 within New Zealand provincial hospitals, versus low volume. A recent review of however two relevant case series were 71 Dutch hospitals found no difference in published in 2007. The fi rst, an analysis outcomes following rectal cancer surgery of colorectal surgery performed in Nelson between low (<20 per annum), medium 27 Hospital, found anastomotic leak following (20–40) or high volume hospitals (>40). anterior resection, 30-day reoperation and According to the Dutch classifi cation fi ve mortality rates, and anterior to abdomino- of six provincial hospitals in this study perineal resection ratio were comparable are low volume, the sixth being medium to published literature at the time.23 The volume as are both tertiary hospitals. While second series from Invercargill Hospital the surgeon numbers in our study may found anastomotic leak and local recur- be underestimates as cases performed in rence following anterior resection was also private hospitals were not captured, all but comparable to published data.24 In addition, one of the hospitals would have surgeon the recently published PIPER study has volumes classifi ed as low volume (if the contributed signifi cantly to the body of lowest defi nition for high volume surgeons knowledge around colorectal cancer demo- of 10 is used). New Zealand has a small graphics and management in New Zealand. medical community with regionalised In this report comparisons were made training of general surgeons meaning ready between rural, independent urban and dissemination of skills and knowledge. urban patients, based on their residential Post fellowship subspecialty training is address, although it cannot be ascertained if also common place. Like the Dutch system, these patients received care in provincial or our medical model utilises medium sized urban hospitals. No difference was observed hospitals for the majority of rectal cancer in stage at diagnosis, or in complete staging work, with referral to tertiary centres (colonoscopy/CT/MRI) rates by rurality. Rural reserved for complex cases. These factors patients were more likely to receive neoad- perhaps explain the fi ndings of this and juvant chemotherapy and radiotherapy, but other New Zealand studies mentioned, that were noted to be of younger age and lower in general, rectal cancer management is of comorbidity. Regarding operative indicators a high standard throughout New Zealands synoptic reporting was more common in provincial and tertiary hospitals. urban versus rural patients (54% v 50%), and the proportion of patients having 12 or Conclusion more lymph nodes examined was 48% for This study shows that the management of rural areas, compared with 38% for inde- patients undergoing rectal cancer surgery in pendent urban areas, and 53% for urban provincial hospitals is generally comparable areas (p=0.004). There was no difference in to tertiary hospitals, demonstrating patients complete excision, return to theatre, anasto- are receiving high-quality care in their local motic leak rate or 30- and 90-day mortality. hospital. Lymph node analysis, colonoscopy Comparison with international data is rates and mesorectal grade reporting were diffi cult due to signifi cant heterogeneity found to differ signifi cantly in provincial in studies and particularly in defi nition of hospitals, highlighting important areas for volume, with high hospital volume defi - resource allocation and further research. nition ranging from fewer than 20 through

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 29 www.nzma.org.nz/journal ARTICLE

Competing interests: Nil. Author information: Michael J O’Grady, Registrar, General Surgery, Canterbury District Health Board, Christchurch; Josephine O’Grady, Registrar, Orthopaedic Surgery, Hawkes Bay District Health Board, Hastings; Rebecca Shine, Registrar, General Surgery, Waitemata District Health Board, Auckland; Gerard Bonnet, Consultant Surgeon, General Surgery, Whanganui District Health Board, Whanganui; Tim Eglinton, Associate Professor, General Surgery, Christchurch District Health Board, Christchurch. Corresponding author: Michael O'Grady, Department of Surgery, Christchurch Hospital, 2 Riccarton Ave, Christchurch, New Zealand. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2018/vol-131-no-1480- 17-august-2018/7661

REFERENCES: 1. International Agency for 6. National Comprehensive new-zealand-health-system/ Research on Cancer. GLOB- Cancer Network. NCCN my-dhb OCAN 2012: Estimated Clinical Practice Guide- 11. Archampong D, Borowski Cancer Incidence, Mortality lines inOncology (NCCN D, Wille-Jørgensen P, and Prevalence Worldwide Guidelines™). Colon/ Rectal Iversen LH. Workload in 2012. [Online] November Cancer. Available from: and surgeon’s special- 2013. [Cited: 23 November http://www.nccn.org/ ty for outcome after 2017.] http://globocan.iarc. professionals/physician_ colorectal cancer surgery. fr/Pages/updates.aspx gls/f_guidelines.asp#site Cochrane Database Syst 2. Firth MJ, Sharples KJ, 7. Washington MK, Berlin J, Rev. 2012 Mar 14; (3). Hinder VA, et al. Methods of Branton P, et al. Protocol 12. Gietelink L, Henneman a national colorectal cancer for the examination of D, van Leersum NJ, et al; cohort study: the PIPER specimens from patients Dutch Surgical Colorectal Project. N Z Med J. 2016 with primary carcinoma Cancer Audit Group. The Aug 19; 129(1440):25–36. of the colon and rectum. Infl uence of Hospital 3. New Zealand Guide- Arch Pathol Lab Med. 2009 Volume on Circumfer- lines Group. Clinical Oct; 133(10):1539–51. ential Resection Margin practice guidelines for 8. Chen HH, Chakravarty KD, Involvement: Results of the the management of Wang JY, et al. Pathological Dutch Surgical Colorectal early colorectal cancer. examination of 12 regional Audit. Ann Surg. 2016 Wellington: New Zealand lymph nodes and long- Apr; 263(4):745–50. Guidelines Group; 2011. term survival in stages 13. Gietelink L, Henneman 4. Adam IJ, Mohamdee MO, I-III colon cancer patients: D, van Leersum NJ, et al; Martin IG, et al. Role of an analysis of 2,056 Dutch Surgical Colorectal circumferential margin consecutive patients in two Cancer Audit Group. The involvement in the local branches of same institu- Infl uence of Hospital recurrence of rectal cancer. tion. Int J Colorectal Dis. Volume on Circumfer- Lancet. 1994; 344:707–711. 2010 Nov; 25(11):1333–41. ential Resection Margin 5. Nagtegaal ID, Marijnen 9. Chen SL, Bilchik AJ. More Involvement: Results of the CA, Kranenbarg EK, et al. extensive nodal dissection Dutch Surgical Colorectal Circumferential margin improves survival for Audit. Ann Surg. 2016 involvement is still an stages I to III of colon Apr; 263(4):745–50. important predictor of cancer: a population-based 14. Aquina CT, Probst CP, local recurrence in rectal study. Ann Surg. 2006 Becerra AZ, et al. High carcinoma: not one milli- Oct; 244(4):602–10. volume improves meter but two millimeters 10. Ministry of Health Manatu outcomes: The argument is the limit. Am J Surg Hauora, My DHB 2016 for centralization of rectal Pathol. 2002; 26:350–357. Available from: https:// cancer surgery. Surgery. www.health.govt.nz/ 2016 Mar; 159(3):736–48.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 30 www.nzma.org.nz/journal ARTICLE

15. Leonard D, Penninckx F, cancer: future direc- management in the provin- Kartheuser A, Laenen A, tions and priorities for cial New Zealand setting Van Eycken E; PROCARE. treatment, research and of Nelson. ANZ J Surg. Effect of hospital volume policy in New Zealand. 2007 Nov; 77(11):1004–8. on quality of care and N Z Med J. 2014 Jun 6; 24. Samson PB, Ngaei outcome after rectal cancer 127(1395):63–72. G. Colorectal resection in surgery. Br J Surg. 2014 20. Australian Cancer Network peripheral New Zealand: Oct; 101(11):1475–82 Colorectal Cancer Guide- workload, outcomes and 16. Li Destri G, Di Carlo I, lines Revision Committee. its future. ANZ J Surg. 2007 Scilletta R, et al. Colorectal Guidelines for the preven- Nov; 77(11):999–1003. cancer and lymph nodes: tion, early detection and 25. Baek JH, Alrubaie A, the obsession with the management of colorectal Guzman EA, et al. The asso- number 12. World J cancer. Sydney, Austra- ciation of hospital volume Gastroenterol. 2014 Feb lia: The Cancer Council with rectal cancer surgery 28; 20(8):1951–60. Australia and Australian outcomes. Int J Colorectal 17. McColl RJ, McGahan CE, Cancer Network; 2005. Dis. 2013 Feb; 28(2):191–6. Cai E, et al. Impact of 21. Association of Coloproc- 26. Keane C, Lin AY, Kramer hospital volume on quality tology of Great Britain N, Bissett I. Can patho- indicators for rectal cancer and Ireland. Guidelines logical reports of rectal surgery in British Colum- for the management of cancer provide national bia, Canada. Am J Surg. colorectal cancer. London, quality indicators? ANZ 2017 Feb; 213(2):388–394. England: Association of J Surg. 2018 Mar 23. doi: 18. Stevenson AR, Solomon Coloproctology of Great 10.1111/ans.14440. [Epub MJ, Lumley JW, et al. Britain and Ireland; 2007. ahead of print] PubMed Effect of Laparoscopic-As- 22. Glimelius B, Oliveira J; PMID: 29569820. sisted Resection vs Open ESMO Guidelines Working 27. Jonker FH, Hagemans JA, Resection on Pathological Group. Rectal cancer: ESMO Burger JW, et al. The infl u- Outcomes in Rectal Cancer: clinical recommendations ence of hospital volume The ALaCaRT Randomized for diagnosis, treatment on long-term oncological Clinical Trial. JAMA. 2015 and follow-up. Ann Oncol. outcome after rectal cancer Oct 6; 314(13):1356–63. 2009 May; 20 Suppl 4:54–6. surgery. Int. J. Colorectal 19. Jackson C, Ehrenberg N, 23. O’Grady G, Secker Dis. 2017; 32:1741–7. Frizelle F, et al. Rectal A. Colorectal cancer

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 31 www.nzma.org.nz/journal ARTICLE

What keeps Northland general practitioners working in Northland? Tanya Quin, Kyle Eggleton

ABSTRACT AIMS: The aim of this study was to understand why general practitioners continue living and practicing in Northland. METHODS: The study was qualitative in nature. Ten Northland GPs were interviewed. Interviews were analysed using an evolved grounded theory approach based on the sociological perspective of symbolic interactionism. RESULTS: Results from the study indicated that doctors whose values, and those of their developing families, are compatible with the accessible and a ordable experiences that the physical environment of Northland provides, will stay in Northland. This connection intertwines with the value of family commitments and opportunities in the region. Both will not be shaken by professional or practice factors that, if positive, will only reinforce the depth of commitment to the region. CONCLUSIONS: When selecting students or GP registrars to provincial areas or recruiting new GPs to those areas, the values, family connections and lifestyle desires of the candidate and their family should match the environment o ered by the region in order to improve retention in that region.

ultiple factors infl uence the choice satisfaction while practice factors decrease of general practice as a career. satisfaction and there is little impact of MCommon external factors include personal factors. In contrast a number positive exposure of medical students to of Australian studies on rural GPs have general practice1,2 and the prestige (or lack recognised the importance of personal of) of general practice.3 Personal factors, in- factors in job satisfaction and have recom- fl uencing a career choice of general practice, mended focusing on integrating rural GPs include better work-life balance, personal and their families into rural communities.8,9 ambition (as compared to career ambition) Kamien8 noted the importance of personal and patient orientation.4 Having a rural factors for doctors when deciding to leave background is also likely to infl uence indi- their rural practices. Bogue9 noted the value viduals choosing to work and live in rural of quality personal time as a statistically areas as general practitioners (GPs).5 important satisfaction factor. Choosing There are a number of competing factors general practice as a career and remaining that may impact on retention as a general working in a general practice is therefore a practitioner. Sibbald, Bojke and Gravelle6 complex process. This complexity between suggested that higher job satisfaction is job satisfaction factors, education, life expe- associated with less likelihood of exiting rience, background and individual traits general practice. Sibbald et al6 also noted may make workforce planning diffi cult. that doctors’ personal and practice charac- The Royal New Zealand College of General teristics did not explain their job satisfaction Practitioners (RNZCGP) has highlighted factors, rather it was their working lives a number of workforce concerns.10 One that contributed to retention. Similarly, of the major areas affected by workforce Van Ham, Verhoeven and Groenier et al7 pressure is Northland.11 Northland’s popu- suggest that professional factors increase lation is signifi cantly older than the national

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 32 www.nzma.org.nz/journal ARTICLE

average, has one of the most deprived popu- lations and is the least urbanised region in Methods New Zealand. These demographic challenges This study used evolved grounded theory are exacerbated by potential GP shortages. (GT), a methodological approach proposed Northland GP shortages were fi rst identifi ed by Strauss and Corbin18 and adapted by more than decade ago, in the 2001 review of Charmaz.19 This version of GT has its founda- Medical and Health Workforce Recruitment tions in the sociological theory of symbolic and Retention in Northland by the Council interactionism (SI), a theory that recognises of Medical Colleges in New Zealand.12 This not only interactions but also meaning.20 review made a number of recommenda- The principles and procedures based on the tions for workforce planning, including theory of SI generate concepts to provide supporting the now defunct Northern a consistent explanation of the social Regional General Practice Consortium to phenomenon studied.18 This theoretical increase locum supply in the region. Other perspective presupposes that behaviour recommendations included expansion of (observable external action and internal medical training for local graduates and experience) is guided by an individual’s post-graduate year one placements in defi nition of reality. These defi nitions in general practice in Northland. The review turn are derived from social interactions also identifi ed a number of job satisfaction in which active individuals exert mutual factors for Northland GPs and the impor- infl uence and meaning is found. tance of workload balance. Despite this The interviewer (TQ) was known by acknowledgement of Northland general most of the participants as a Northland GP practice workforce factors, 16 years on there colleague and committee member of the is ongoing concern about workforce issues Northland Faculty RNZCGP. The beliefs of and the ageing Northland GP population. the interviewer and the existing relation- A number of workforce surveys have high- ships that TQ had with her colleagues in lighted the ageing population of Northland interviewing them are not easily cleaved GPs with the 2015 RNZCGP workforce from the interview process. This propo- survey reporting that 44% of Northland sition thus fi ts more comfortably with the GPs were aged over 5510 and the 2014 SI perspective of adapted GT methods as survey suggesting that 38% of Northland described by Charmaz19 that “we are part of GPs intended on retiring in 10 years.13 High the world we study and the data we collect”. retirement rates are also reported in other Charmaz19 proposed that researchers provincial areas. Workforce surveys by construct theory through their past and the RNZCGP show New Zealand GP 10-year present involvement and interactions with retirement numbers of 36.4% 2014 to people, perspectives and research prac- 44% 2016,13,14 with Northland as one of the tices. In accordance with the concept, of regions most affected by rural vacancies a continually constituted refl exive self,21 and hotspots.15 Wong,16 in reviewing the TQ documented her emotions and actions RNZCGP 2014 workforce survey, suggests a within memos of interviews, personal “looming workforce crisis in rural general refl ection and discussions with KE. This practice”. 2003 NZMA subcommittee enabled TQ to challenge her interpretation analysis of the New Zealand GP workforce of the data, to seek further data and improve expressed concern about a crisis regarding theoretical sensitivity by recognising 8% of GPs intending to retire over four years nuances, extracting relevant elements and and falling new GP numbers.17 By 2016 the reconstructing meaning from the data.21 intentions of GPs to retire over 1–5 years Convenience sampling was used to recruit was 23%. participants. Participants were recruited The complexity of the numerous factors through email via the Northland Faculty impacting on retention of GPs as well as RNZCGP and also via announcement at the the chronic workforce crisis in Northland annual conference of the Northland Faculty. and other rural areas is the focus of this Inclusion criteria were GPs who had been study. In this paper we present a qualitative living and working in Northland for at least exploration of what keeps GPs working in fi ve years. Exclusion criteria were GPs who Northland.16 lived more than a one-hour drive away from TQ.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 33 www.nzma.org.nz/journal ARTICLE

The data were obtained through indi- Values vidual, open, semi-structured face-to-face The theme of values describes how GPs interview. Each person attended one make connections to Northland based on interview of 30–60 minutes duration. Partic- congruence between their values (and their ipants were interviewed in either their family’s values) and the perceived values offi ce or their homes. Each interview was of Northland. Often this matching of values recorded and transcribed. Demographic and was something that they did not have in professional data were not obtained. previous locations that they lived. The interviews were based on the “I didn’t have that connection like I have following guiding questions: “Tell me your had up here.” (Bob) story…”, “Why do you continue to practice GPs frequently discussed how their values in Northland?” and “How did you get related to the environment of Northland here…?”. New questions, based on evolving and what this environment represented. theoretical concepts, were added to clarify For some GPs the environment had a social responses. A list of known general prac- aspect in which good friendships were titioner satisfaction factors, based on the created. These social activities embedded GPs literature, was available during interviews within the communities that they lived in. to explore the research question in more depth as decided by the interviewer. Data “We made some very good friends in collection was maintained until all partici- Northland; and have enjoyed it ever since and pants who volunteered were interviewed. never been tempted to go and live anywhere Data saturation was achieved by the last two else really.” (Gareth) interviews, and interviews continued both “Plenty of opportunities for involvement out of courtesy for those who wanted to in community and social outlets, music, the participate and also to reassure the inter- sailboat. Encouragement for exercise.” (Doug) viewer that indeed saturation had been For many GPs the spatial environment achieved. matched their values. For example, The collection and analysis of data were Northland was perceived as safe and undertaken simultaneously as recom- natural. mended by GT. The data were analysed “we wanted to take somewhere to grow up through open, axial and selective coding in a good environment.” (Wayne) in accordance with evolved GT and the Other spatial characteristics included the constant comparative method.18 The initial rurality of Northland. transcripts were independently coded by TQ and KE; the codes were compared between “was the right size town for us and it just both authors and refi ned. TQ then coded the felt right.” (Bob) remainder of the interviews. The evolving “So, it was kind of natural that I would theoretical schema was reviewed by both want to—I like living where there’s country.” authors to verify internal consistency for (Erica) gaps in logic. The environment was not only important Ethics approval was granted by the in matching the values of GPs but also in the University of Auckland Human Partici- lifestyle afforded by the environment. This pants Ethics Committee—reference number led to the second theme of ‘Lifestyles’. 018783. Lifestyles Affordable and accessible interactions Results with the physical environment of Northland In total, 10 general practitioners partici- characterised the theme of ‘Lifestyles’. pated in this study. Four broad themes were The physical environment of Northland’s derived from the data and are illustrated beaches, ocean activities, natural beauty by anonymised examples below. The most and the lifestyle options were highly important themes were those of values, desired by GPs. lifestyle and family. These three themes “Good friends, got beautiful beaches, you’ve underpin the phenomena of connections to got a nice climate generally, you’ve got beau- Northland. Less important was the theme of tiful scenery and lovely open spaces.” (Doug) practice relationships and workplace respect.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 34 www.nzma.org.nz/journal ARTICLE

“I mean that’s paradise. Why would you factors like family and all the rest of it that live anywhere else and still be able to feel keeps you here.” (Gareth) like you’re contributing and improving This quote by Gareth refl ects a common or working to achieving wellness in the pattern—that practice and professional community that you reside in.” (Sarah) factors enhanced satisfaction for GPs but did An accessible lifestyle involved not just not alter their intention to continue to live in physical access but social access to the activ- Northland. The main aspects of practice and ities as well. profession that GPs commonly mentioned “A huge asset to living in Northland is were professional relationships and work- what’s at your doorstep; the coastal juxta- place respect. posed with rural.” (Jo) Professional relationships and Affordability was also important, not just workplace respect in terms of cost but also the quality of life- This theme encompassed practice and style afforded. professional aspects of a balanced workload “For us it is lifestyle. So certainly, I would with positive respected relationships with never be able to afford this size property in colleagues, patients and the district health Auckland at all.” (Erica) board. This lifestyle was appreciated by the “Because as an employed physician you’re partners of GPs as well. gonna be willing to do more sacrifi ce if you’ve “Because my husband likes to fi sh, this is got a lot of respect and some infl uence.” (Jo) why we came here.” (Trudy) A couple of participants had undertaken While the importance of families their postgraduate training in Northland connecting to the environment and enjoying prior to entering general practice. The rela- the lifestyle offered by Northland infl u- tionships that they had with specialists and enced GPs’ experience of Northland, there GP colleagues had infl uenced their decision was another aspect of families that was to live in Northland. commonly discussed. “That was a big thing too, is being inspired by those people and deciding well that’s Family where I wanted to go.” (Ed) The theme of ‘Family’ describes how GPs created family connections or reconnected Financial concerns and after-hours with family in Northland. Many GPs had commitment did not feature as reasons why family ties to Northland, either directly or GPs lived in Northland. through their partner. “I’ve got family here so that connects me.” Discussion (Brett) When undertaking the analysis it was Having family connections created a clear from the content of the conversation deeper sense of ‘belonging’ to Northland. and the developing themes that lifestyle and family were important and mean- “I guess I feel inherently like I belong in ingful. When the researchers reviewed Northland.” (Sarah) the interview content, to understand what Social and economic opportunities for was symbolic about family and lifestyle, it partners were important and assisted became apparent that values refl ected how families in feeling contented with where the physical and social environment has they lived. These deeper connections with meaning to those who enjoy it compared Northland and the happiness of GPs’ families to those who don’t. The participants had to outweighed any negative factors that related value this above other interactions or (as to the general practice workplace. one participant said) ‘material things’. “I wouldn’t go anywhere else in New This study proposing that personal factors, Zealand now, to me this is home, and it rather than professional and practice factors, always will be for our family; but I don’t think determine retention of GPs is in contrast to a that’s a decision based on anything to do with number of other studies.6,7 One explanation work or anything to do with general practice for this mismatch is suggested by Hays.22 necessarily; it’s more probably those external Hays points out that most research studies

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 35 www.nzma.org.nz/journal ARTICLE

why doctors leave; consequently there is retention in rural communities related to little known about why doctors stay in posi- the age and stage of their families with older tions. Hays’22 work showed that the personal families being more likely to move away factor of being linked to a community was from rural areas. Noonan et al,24 however, an important factor in staying in a rural also acknowledged the importance of the area. His recommendation for recruitment physical connection to a region, in common targeting was to focus on integrating GPs into with this study, that infl uenced GP retention. 8 the region of work. Kamien noted the social This study has a number of weaknesses. enjoyment and esteem of being involved One weakness is that this study does not in a rural community was a more common make an association with how satisfi ed the value in the group of GPs that stayed in rural GP is, just why they stay. Another weakness practice over 10 years compared to those is that there may be a selection bias with who left. Both of these studies highlight the participants volunteering if they had theme of connecting to a community that positive intentions in staying in Northland. this study shares. Based on this study our recommenda- The link to the community, with many of tions are to adjust the selection criteria the participants, was embedded socially and for students, GP registrars and GPs physically and not out of a sense of loyalty to entering rural locations. An undergraduate a community for the services provided. This programme that selected medical students study would agree with other studies on the based partially on connections and rural importance of integrating into a community origin is the Pūkawakakawa programme, 22 when addressing retention. However, we run in conjunction with Northland DHB also suggest that integration alone might and the University of Auckland.1 The not be enough and that matching values Pūkawakawa programme demonstrated during the recruitment process may be more that 62% of participants continued to work important in order to ensure a ‘link’. What in rural and regional centres after gradu- this study does is to go further in under- ation. Likewise, we would recommend that standing what that ‘link’ is and what it means one of the criteria for selecting registrars for to GPs who stay living and practicing in regional programmes is choosing trainees Northland, New Zealand. The nature of that whose values, family connections and life- ‘link’ has not been explored in other studies. styles match the environment offered by Previous studies mostly confi ne the the region. Rather than simply attempting personal factors to marital relations and to integrate new GPs or amending commu- to children.9,23 For example in a study by nities, organisations and practices should Noonan, Arroll, Thomas and Elley24 an actively recruit GPs who have concordance important observation was made that GP with the factors suggested in this study.

Competing interests: Dr Quin reports grants from Auckland Faculty Charitable Trust Board, from null, during the conduct of the study. Dr Eggleton reports grants from Auckland Faculty Charitable Research Trust RNZCGP during the conduct of the study. Author information: Tanya Quin, General Practitioner, Auckland University, Okaihau; Kyle Eggleton, Senior Lecturer, Department of General Practice and Primary Health Care, University of Auckland, Auckland. Corresponding author: Dr Tanya Quin, General Practitioner, 127 Forest Rd, Okaihau 0475. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2018/vol-131-no-1480- 17-august-2018/7662

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 36 www.nzma.org.nz/journal ARTICLE

REFERENCES: 1. Matthews C, Bagg W, Yield- fi es pressure points and 17. New Zealand Medical er J, Mogol V, Poole P. Does reveals life practices of Association Workforce Pukawakawa (the region- highly satisfi ed doctors. Subcommittee. An Anal- al-rural programme at the The Physicians Executive. ysis of the New Zealand University of Auckland) 2006; Nov/Dec:30–39. General Practitioner infl uence workforce choice. 10. Royal New Zealand College Workforce. A Report from N Z Med J. 2015; 128:35–43. of General Practitioners. the New Zealand Medical 2. Block SD, Clark-Chiarelli 2015 Workforce Survey. Association. 2004; May. N, Peters AS, Singer JD. Wellington: The Royal 18. Strauss A, Corbin J. Basics Academia’s chilly climate New Zealand College of of qualitative research: for primary care. JAMA. General Practitioners; 2015 Grounded theory proce- 1996; 276:677–682. 11. Health Workforce New dures and techniques. 3. Olid A, Zurro AM, Villa JJ Zealand. Health of the Thousand Oaks: Sage; 1998. et al. Medical students’ health workforce 2015. 19. Charmaz K. Construct- perceptions and attitudes Wellington: Ministry ing grounded theory about family practice: of Health; 2015. - a practical guide a qualitative research 12. Council of Medical Colleges through qualitative synthesis. BMC Med in New Zealand. Review analysis. London: Sage Educ. 2012; 12:1–16. of medical and health Publications; 2006. 4. Kiolbassa K, Miksch A, workforce recruitment 20. Benzies KM, Allen MN. Hermann K, et al. Becom- and tretention in North- Symbolic interactionism ing a general practitioner land. Wellington: Council as a theoretical perspec- - Which factors have most of Medical Colleges in tive for multiple method impact on career choice New Zealand; 2001. research. J Adv Nurs. of medical students. BMC 13. Royal New Zealand College 2001; 33:541–547. Fam Pract. 2011; 12:25. of General Practitioners. 21. Chamberlain-Salaun J, 5. Laven G, Wilkinson D. 2014 Workforce Survey. Mills J, Usher K. Linking Rural doctors and rural Wellington: Royal New symbolic interactionism backgrounds: how strong Zealand College of General and grounded theory meth- is the evidence? A system- Practitioners; 2014. ods in a research design: atic review. Aust J Rural 14. Royal New Zealand College From Corbin and Strauss’ Health. 2003; 11:277–284. of General Practitioners. assumptions to action. 6. Sibbald B, Bojke C, Gravelle 2016 Workforce Survey. SAGE Open. 2013; 3:1–10. H. National survey of job Wellington: Royal New 22. Hays R, Wynd S, Veitch satisfaction and retirement Zealand College of General C, Crossland L. Getting intentions among general Practitioners; 2016. the balance right? GPs practitioners in England. 15. Adair A, Coster H, Adair who chose to stay in BMJ. 2003; 326:22. V. Review of international rural practice. Australian 7. Van Ham I, Verhoeven AA, and New Zealand liter- Journal of Rural Health. Groenier KH, Groothoff ature relating to rural 2003; 11:193–198. JW, De Haan J. Job satis- models of care, workforce 23. Alexander C. Why doctors faction among general requirements and oppor- would stay in rural practice practitioners: A systematic tunities for the use of new in the New England health literature review. Eur J Gen technologies. Cambridge: area of New South Wales. Pract. 2006; 12:174–180. The New Zealand Institute Australian Journal of Rural 8. Kamien M. Staying in or of Rural Health; 2012. Health. 1998; 6:136–139. leaving rural practice: 1996 16. Wong D, Nixon G. The 24. Noonan T, Arroll B, Thomas outcomes of rural doctors’ rural medical generalist D, Elley R. When should I 1986 intentions. Med J workforce: The Royal do rural general practice? Aust. 1988; 169:318–321. New Zealand College of A qualitative study of job/ 9. Bogue RJ, Guarneri G, General Practitioners’ life satisfaction of male Reed M, Bradley K, Hughes 2014 workforce survey rural GPs of differing J. Secrets of physician results. J Prim Health ages in New Zealand. satisfaction study identi- Care. 2016; 8:196–203. NZMJ. 2008; 121:59–67.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 37 www.nzma.org.nz/journal ARTICLE

A media content analysis of New Zealand’s district health board Population-Based Funding Formula Aaron N Chester, Erin C Penno, Robin DC Gauld

ABSTRACT AIM: The Population-Based Funding Formula (PBFF) has a significant impact on health funding distribution between New Zealand’s 20 district health boards (DHBs) yet is subject to little independent oversight or public scrutiny. There has been widespread dissatisfaction among DHBs with the allocation process; however, there are limited formal avenues available for DHBs and the public to discuss the PBFF. As such, the news media has become a key platform for voicing concerns. This study aims to gain a better understanding of how the PBFF is portrayed in the news media and of perceptions of funding allocations across the country. METHOD: We conducted thematic analyses of 487 newspaper articles about the PBFF, published over 13 years from 2003–2016. We then identified trends in the data. RESULTS: Typically presented in a negative light, the PBFF was commonly framed against a background of financial struggle and resultant impacts on health services and sta . The e ect of factors driving DHB allocations and the PBFF process itself were also key themes. There were significant regional and temporal variations in reporting volume, with most articles focusing on South Island DHBs and occurring during the introduction of the PBFF and at the time of the most recent review. CONCLUSIONS: The findings suggest general discontent with the PBFF model across the DHB sector and a sense that the PBFF has failed to address various challenges facing DHBs. The geographic imbalance in reporting volume suggests that frustration with the PBFF is particularly keenly felt in the South Island. Although the PBFF is a lightning rod for frustrations over limited health funding, the findings point to the need to improve transparency and dialogue around the formula and to monitor of the impact of PBFF allocations throughout the country.

ew Zealand’s Population-Based account for diseconomies of scale related to Funding Formula (PBFF) was imple- rurality, overseas visitors and unmet need. Nmented by the Fifth Labour Gov- The formula controls almost three quarters ernment on 1 July 2003, as part of major of Vote Health, the main source of funding health system reforms at the turn of the for New Zealand’s health system, making millennium.1 The Ministry of Health (MoH) it one of the single largest determinants of uses the PBFF to distribute funding among government expenditure ($11.72 billion the nation’s 20 district health boards (DHB), in 2015/16).1,2 The PBFF is reviewed by the which in turn provide or fund health ser- MoH every fi ve years, with the most recent vices within their districts. The allocation review taking place in 2015 in concert with that each DHB receives is determined by the belated 2013 Census.1,2 Despite periodic the number of people in their catchment internal review, there have been calls for areas, their ethnicity, sex, age and relative a comprehensive and independent review deprivation. Adjustors are also included to of the PBFF amid concern over a lack of

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 38 www.nzma.org.nz/journal ARTICLE

transparency surrounding the PBFF.1,3 It has spondingly, its considerable political and been argued that such a review would allow social relevance, there is a need for a greater greater public discourse around this issue of understanding of patterns and content of national interest, as occurs in other nations the media discourse surrounding the PBFF. which have greater transparency around This study aims to fi ll this gap by exploring their health funding formulae.4 how the PBFF has been represented in the Since the formula’s advent there has media throughout New Zealand for the 13 been widespread dissatisfaction among years from its introduction in 2003 through DHBs with the allocation process.5 This is to 2016. Specifi cally, this study aims to fuelled in part by considerable variations explore variations in regional coverage in fi nancial performance across the DHB of the PBFF over time and to identify key sector, coupled with growing disparities commonalities and differences in the issues in access to health services.6 These issues reported across the DHB sector. In doing are compounded by a lack of transparency so, this study contributes to a growing surrounding the methodology under- body of research surrounding popula- pinning the formula, along with limited tion-based health funding models within 1,4,13–16 formal avenues available for DHBs to New Zealand, offering insights into the express their concerns and a corresponding context, perceptions and impact of the PBFF limit on public engagement.1 Thus, media throughout the country. reporting on the PBFF provides DHBs with an important social platform to voice their Materials and concerns about how the formula is oper- ating.7–9 It is also the key avenue for the methods public to gain knowledge about a salient New Zealand print media ownership is issue which has considerable impact on the dominated by two public companies, NZME level of healthcare services they receive.9 and Fairfax, which hold a duopoly on the Since health is a key political concern market at 89.3% of circulation. Allied Press, among the New Zealand public, the PBFF a privately-owned company with a focus on has received widespread coverage in the the Southern regions, constitutes the bulk of media. However, New Zealand’s print media the remaining market share at 8.4% of circu- is dominated by regional news outlets. lation. Collectively, Fairfax, NZME and Allied These print sources tend to offer regionally Press own all of New Zealand’s major news- 17 focused views refl ecting how favourable the papers. We searched for stories related to PBFF has been to their respective DHB.6,10–12 the PBFF in the Newztext and Factiva online newspaper databases, which collectively Fragmented media coverage presents cover these major print sources. The fi rst signifi cant challenges to understanding both author led data collection and analyses with the perceived and real impacts of the PBFF oversight from the second and third authors. model across the DHB sector. The focus on regional issues means it is diffi cult to gain The search parameters included a prag- a complete perspective of the key issues matic timeframe from 1 January 2003 to surrounding PBFF and the common threads 1 October 2016. This allowed us to focus emerging with respect to health funding on the established formula rather than its throughout the country. At the same time, development or previously used formulae. media coverage has been shown to play a Our search terms were: ‘population-based key role in steering public conversations and funding formula’ OR ‘DHB AND funding perceptions of health funding decisions.6,7,9 AND formula’ OR ‘population based funding’ This in turn drives political discourse and OR ‘needs based funding’ OR ‘capitation’ OR ultimately infl uences the policy decisions ‘rural adjuster’. Our criterion for analysis surrounding resource allocation.7,9 However, was broad and included any story related to news media tends to sensationalise health the PBFF. Articles which did not specifi cally funding issues and does not necessarily mention New Zealand’s DHB PBFF were drive health policy in the direction of excluded from the study. effi cacy and pragmatism.9 We used a qualitative descriptive Given the signifi cant pecuniary impact of approach to our analysis, which focused the PBFF on the health sector and, corre- on the content of the data to identify key

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 39 www.nzma.org.nz/journal ARTICLE

themes within the articles.18,19 Coding was 36.3% of articles. Of 487 articles, 438 were led by the fi rst author. Coding was data news articles and 49 were opinion pieces. driven with codes for each idea developed Key themes with oversight from the second and third Over the 13-year period, media coverage authors. We then reviewed these codes largely centred around four major themes: to identify overarching themes and went DHB fi nancial positions, the impact of on to discuss each of them in turn.18,20 We the PBFF on DHBs, factors affecting PBFF also recorded the DHB at the focus of each allocations and the PBFF policy process. article and the status of that DHB’s share Table 1 shows these themes and the major of Vote Health, be it increased, decreased sub-themes identifi ed in our analysis. or unstated. We recorded any attitudes expressed towards the formula, which were Financial position of DHBs gauged as positive, negative or neutral. The PBFF was typically mentioned in the The attitude assigned to each article was context of fi nancial constraints or diffi culties based on comments of key interest groups faced by a given DHB and the resultant therein, with DHBs’ comments taking impact on availability of health services and precedence over other groups (ie, if a DHB staffi ng. DHB defi cits were the most signif- had misgivings with the PBFF, yet the MoH icant factor driving media coverage on the gave a glowing appraisal, the attitude was PBFF and were mentioned in over a quarter recorded as negative). We noted which of articles. Defi cits were a common theme groups or individuals were commenting throughout the country. However, they on the formula and recorded any relevant received particular attention in the South comments therein. Island where 88.5% of articles mentioned defi cits. Discussion of defi cits was frequently Results accompanied by information regarding the share of funding or the size of budget Search results increase a DHB received, often couched The Newztext and Factiva databases held in comparison to other DHBs across the 487 articles relevant to the PBFF, published country. For example, early coverage (2004) between 1 January 2003 and 1 October 2016. of the PBFF model in The Christchurch Press The Newztext database search returned 435 noted: articles; 12 were duplicates and 113 were “Reduced funding from July will severely irrelevant, giving 310 relevant articles in squeeze an already cash-strapped Canterbury total. The Factiva database search returned District Health Board (CDHB), health bosses 197 articles; 20 were irrelevant, leaving 177 warn…despite a projected $5.2 million budget relevant articles. blowout at Christchurch Hospital—it faced Regional and temporal trends an even bigger challenge from July, when its The number of articles published about annual funding increases start to shrink. The the PBFF reduced over the 13 years, from Ministry of Health estimates that the DHB a maximum of 93 in 2003 to a low of 19 in is $42m over-funded under the new popula- 2013. In 2015, there was a sharp spike of 47 tion-based funding model. That means it will articles published. Figure 1 shows there was get a smaller share of funding over the next a clear regional bias in reporting, with the six years as other DHBs are paid more to South Island’s fi ve DHBs the focus of 78.6% catch up to Canterbury.”21 of all articles; whereas the North Island’s Likewise, more recent coverage (2015) 15 DHBs were the focus of only 21.4% of in the Taranaki Daily News prefaced a articles. The Southern DHB and its prede- comment around Taranaki DHB’s $2.32 cessors (Otago and Southland) were the million defi cit by noting that: focus of almost half (47.4%) of all articles. “Taranaki’s share of the nation’s health Correspondingly, most articles were printed budget is likely to be reduced after changes by South Island-based newspapers (81.7%). were made to the formula used to allocate the Fairfax-owned newspapers printed 56.7% money. Health funding is distributed among of articles, compared with NZME at 7.0%. the 20 district health boards (DHBs), but after Allied Press-owned newspaper, the Otago changes to the population-based formula, Daily Times (ODT), accounted for remaining four DHBs—Taranaki, Nelson Marlborough,

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 40 www.nzma.org.nz/journal ARTICLE

Figure 1: The number of newspaper articles published about New Zealand’s Population-Based Funding Formula per district health board between 1 January 2003 and 1 October 2016.

*Otago and Southland DHBs were amalgamated to form Southern DHB in 2010.

Auckland and Wairarapa—are likely to DHBs (48.1% of articles assigned this receive a smaller share from July 1. Seven theme). The rising cost of staff wages were DHBs are likely to get an increase.”12 frequently tied to fi nancial crises faced by Impact of PBFF allocations on DHBs, which in turn were linked to budget allocations. In order to balance the books, DHBs DHBs were reported to be grappling with The impact or relationship between job cuts, under-staffi ng or cutting back health funding allocations and DHB perfor- health services. The pressure placed on mance was also a persistent theme. Health doctors due to insuffi cient staffi ng was also funding crises linked to funding alloca- cited, especially in rural areas such as South tions were frequently cited as the cause Canterbury or Ashburton. of health services being cut or stretched. The abolition or reduced funding of aged Factors a ecting PBFF allocations care services was a prominent example of Technical descriptions of the PBFF in this, with headlines such as “Knife Taken media articles were limited and, where To Elderly Services” and “Elderly Health present, tended to be presented in rela- Care Fears.”22,23 The impact the formula had tively simple terms. However, discussion on DHB staffi ng was also a major theme, surrounding the factors underpinning in particular for the Canterbury (24.7% of PBFF allocations, such as population size articles assigned this theme) and Southern and composition and their relationship

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 41 www.nzma.org.nz/journal ARTICLE

with funding allocations were identifi ed lysed considerable discussion in the South, as issues in a number of DHBs. Population with the MoH reviews of the PBFF featuring age was cited as an issue related to demand prominently in the ODT and the Southland for health services for six DHBs, including Times. The bulk of these articles (34.1%) Canterbury, MidCentral, Southern, South were printed in 2015 and coincided with Canterbury, Nelson Marlborough and Wait- the release of the Government’s fi ve-yearly emata. With the exception of Waitemata, review of the formula. these DHBs had older populations than the Attitudes and Vote Health national average.24 Although also included Most attitudes expressed towards the as factors within the PBFF, concerns formula were negative (n=84, 96.6%). Many surrounding deprivation and ethnicity of the more antipathetic comments came appeared less frequently. Similarly, although from politicians. For example, one politician also adjusted for within the formula, unmet argued that “the government should reject need and the effect of overseas visitors this population-based funding formula that received less attention. clearly isn’t working for our communities In contrast, rurality was a commonly and fund services to the current need.”12 raised issue, particularly among South At the same time, many DHBs displayed Island DHBs. As with DHBs, political parties resigned acceptance of the PBFF’s realities, often cited rurality as an issue facing the “Board chairman Syd Bradley insisted the PBFF, with the relationship between rurality CDHB [Canterbury DHB] would learn to live and diffi culty recruiting medical staff within its new budget…”26 frequently linked to funding. An enduring The three positive attitudes all coincided opinion was that the PBFF model did not fi t with an increase in the share of Vote Health the unique needs of DHBs with large rural for the DHB concerned. In most articles areas such as the West Coast or the Southern (59.1%), the DHB’s share of Vote Health District Health Board. For example, the decreased, or there were issues with under- Southland Times wrote: funding. Most (55.2%) of these articles “King [New Zealand Labour Party Health were published in the two years after the Spokeswomen] said the current popula- formula’s introduction. In 2003, 12 (13.8%) tion-based funding formula was not providing negative attitudes were expressed towards enough funding for big rural areas such as the formula. This number decreased and the Southern District Health Board, which plateaued over time, excepting a peak (n=19, faced massive defi cits and was replaced with 21.8%) in 2015, which coincided with the a Commissioner this winter.”25 MoH review of the PBFF. South Island DHBs Population change was also identifi ed as a were the focus of most (83.3%) articles problem. DHBs with small, relatively static expressing negative attitudes towards the populations were portrayed as struggling formula. Opinion pieces contained most due to the comparatively small increases negative attitudes (55.3%), compared with in funding under the PBFF model, which news articles (13.2%). was seen as a problem confronting South Interest groups Island DHBs in particular. Conversely, rapid We identifi ed 11 key interest groups in population growth in the Auckland region the media coverage, most prominently DHBs was seen as creating a “gap” between DHBs, the MoH and politicians—both in increased demand for health services and government and in opposition (Table 2). funding allocations. In terms of volume, the vast majority of PBFF policy comments on the formula were offered A perceived lack of transparency by individuals associated with DHBs, such surrounding the PBFF model and allocation as DHB board members, Chief Executive process was discussed in 26 articles. Twen- Offi cers or Planning and Funding staff. ty-two (84.6%) of these articles mentioned Many of these comments implicated the the Southern DHB or its predecessors, formula in fi nancial woes. For example, one suggesting transparency was a particular DHB board member blamed “The board’s concern among DHBs in the South Island. funding shortfall [on] general under- Potential changes to the PBFF model and, funding and the ‘adjusters’ used in the correspondingly, DHB allocations, also cata- Health Ministry’s population-based funding

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 42 www.nzma.org.nz/journal ARTICLE

Table 1: Salient themes in newspaper articles about the Population-Based Funding Formula.

Theme Meaning n=487 % Financial position of DHBs Vote Health share A change in a DHB’s share of Vote Health was mentioned. 153 31.4

DHB deficit DHB deficits were mentioned. 131 26.8

Impact of PBFF factors on DHBs Services stretched DHB health services were stretched or compromised in relation to funding. 104 21.3

Services cut DHB health services were abolished in relation to funding. 94 19.3

Impact on sta DHB sta were impacted (eg, job loss) by changes to funding or underfunding. 81 16.6

Doctor recruitment The recruitment of doctors was impacted by the formula. 18 3.7

Factors a ecting PBFF allocations Age PBFF and health costs were a ected by the age of a population. 71 14.5

Rurality PBFF and health costs were a ected by rural populations. 54 11.1

Population Population size and changes to that size a ected PBFF. 51 10.5

Unmet need There was a level of unmet need in healthcare stemming from PBFF allocations. 22 4.5

Ethnicity Ethnicity was an issue in relation to PBFF. 17 3.5

Deprivation Deprivation was an issue in relation to PBFF. 13 2.7

Overseas visitors Overseas visitors were a ecting PBFF. 6 1.2

PBFF policy Review Review of the PBFF was mentioned or advocated for. 40 8.2

Transparency There was concern over a limited transparency around the PBFF and its workings. 26 5.3

Abbreviations: PBFF, population-based funding formula; DHB, district health board.

Table 2: Groups commenting on the Population-Based Funding Formula.

Interested party Number of articles group Percentage of articles group commented on (n=442) commented on (n=487) * District health boards 211 43.3

Ministry of Health 77 15.8

Politicians 50 10.3

Contracted providers 23 5.2

Professional groups 18 3.6

Medical professionals 14 2.8

Patient groups 14 2.8

Other 14 2.8

Academics 12 2.5

Government (other than MoH) 6 1.2

Local body politicians 2 0.4

*Percentages do not add to 100%. Abbreviations: MoH, Ministry of Health.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 43 www.nzma.org.nz/journal ARTICLE

method, particularly the rural adjuster, and the impact and success of health which included patient transport. They had policies, including healthcare funding.32,33,35 not kept up with infl ation and the adjuster The distribution of healthcare resources did not cover costs.”27 However, there was is of signifi cant public interest in New a dearth of explanations as to how the Zealand and is often the subject of intense formula could be improved. Phrasing also public debate.6,36 News media coverage appeared to erroneously imply that certain has previously been used to explore the DHBs’ funding had been reduced in absolute effects of variations in health policy deci- terms. On the other hand, DHB representa- sions across the DHB sector,6 lending tives also often emphasised efforts to live insights into the factors and circumstances within budgetary restrictions or refused to affecting nationally consistent provision of comment. Comment from the MoH tended health services. Viewing the PBFF through to emphasise that funding allocations were a news media lens, we sought to explore never cut and were increased annually. the contextual issues and perceptions Political aspects surrounding the DHB funding allocations, developing a national picture of the impacts Following its introduction in 2003, of the PBFF based on regional perspectives. the PBFF has remained in place largely unchanged during the period of analysis, The patterns and content of media bridging both Labour- and National-led reporting on the PBFF have several implica- governments. Correspondingly, consec- tions. First, they suggest general discontent utive Labour and National-led governments with the PBFF model across the DHB sector. have defended the formula, though both Typically presented in a negative light, the parties have also criticised it during their PBFF was most commonly framed against time in opposition. While in opposition, a background of fi nancial struggle and National MPs highlighted the strain popula- resultant impacts on health services and tion-based funding placed on DHBs. Many staff. The underlying reasons centred on of National’s comments focused on the the failure of the health funding model to formula’s effect on health services, where adequately account for the pressures placed these services were generally being cut on DHBs as a result of various geographic or compromised.28 After being elected in and demographic characteristics. These 2008, National appeared to change their issues tended to be characterised as idio- rhetoric in defence of the formula. When syncratic challenges facing a given DHB; facing criticism over the formula, National however, aging populations and diffi culties often rebutted that a Labour government associated with either rapid or static popu- introduced it.29 The Labour-led Government lation growth and rurality were common (1999–2008) advocated the formula as a concerns. Although much was made of the method of fairly distributing funding across balance sheets of DHBs, contrasting funding the country. While in opposition, Labour allocations across the DHB sector created has called for comprehensive review of the a tacit atmosphere of ‘winners’ and ‘losers’ formula and has criticised the funding levels under the PBFF within the media coverage. received by certain DHBs and rural areas. Second, the geographic imbalance in In recent years, both National and Labour reporting volume suggests that frustration have acknowledged a lack of transparency with the PBFF is particularly keenly felt surrounding the PBFF and health funding in the South Island. One explanation for allocations and supported the need for this regional bias may lie in the legacy of greater clarity and public discussion around historic funding arrangements. The PBFF funding models.30,31 was introduced in an attempt to address historic imbalances in health funding Discussion between regions and to push newly estab- lished DHBs towards a position of funding The media has a well-established role equity. Funding allocations were imposed in framing the debate on health policy, gradually with the proviso that no DHB acting as a vehicle for both information and would ever receive an absolute reduction opinion on key policy issues.32–34 Interna- in funding. However, those DHBs deemed tionally, analysis of news media coverage to be over-funded under the new formula has been used to evaluate attitudes towards

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 44 www.nzma.org.nz/journal ARTICLE

received progressively smaller shares of reporting on the formula. Unmet need funding relative to other DHBs. For some features prominently outside of the context regions, such as Canterbury and Otago, this of the PBFF, though it likewise received generated considerable concern around the relatively little attention in the context of the impacts of a sinking fi scal lid in an already funding formula.50 37,38 strained fi nancial environment, as Third, the weight given to opinions of key refl ected in the media coverage. However, interest groups in the media points to the although contributing to the comparatively relative roles of different interest groups high volume of reporting in the South, framing media coverage as well as implying fi nancial pressures do not fully explain partiality in media coverage. The domi- variations in media coverage across the nance of opinions from those within the country. The focus on Vote Health share DHB sector suggests that DHBs themselves reductions was concentrated early on in the play an important role in infl uencing the analysis period and decreased over time as media’s narrative.32,51 Although often critical 39 funding allocations reached target levels, of the PBFF, many DHB comments were suggesting DHBs adapted to the new funding also comparatively conservative, possibly arrangements. Furthermore, while many pointing to acceptance of the PBFF within articles covering the PBFF mentioned DHB the sector or alternatively a symptom of defi cits in the South Island, they appeared political restraint. In comparison, the more far less frequently in articles about North incendiary comments offered by politicians Island DHBs, despite these DHBs posting illustrate the protean nature of politics, 40 numerous defi cits over the past 13 years. with political rhetoric linked to power. The regional variation in media coverage News media outlets may themselves have was likely also infl uenced by the rurality of an interest in infl uencing health policy. certain DHBs. Rurality is known to affect One indication of this is the dominance of the accessibility of health services in New coverage by the Otago Daily Times (ODT) Zealand,41,42 and creates a diseconomy of over the analysis period in combination scale for smaller DHBs (on a per capita with a particular focus on key themes basis).1,43 Some costs associated with rurality including rurality, transparency and the are adjusted for within the PBFF,2 though PBFF review. The ODT is New Zealand’s the numerous articles and comments iden- only nationally-owned newspaper and has tifying rurality as an issue suggests there a strong focus on policy issues affecting the may have been dissatisfaction with the southern regions. While the large volume of level or manner of adjustment for these articles in the ODT further highlights a sense South Island DHBs. There also appeared to of misgiving surrounding the PBFF model be a mismatch between those DHBs iden- in the southern regions, it also supports tifying rurality as an issue and those DHBs the notion of mass media acting as policy receiving the highest shares of the rurality contributors in the PBFF debate, rather than adjuster. For example, Northland and merely a conduit for the healthcare visions Waikato DHBs received among the highest of other key interest groups.36,52 Notably, shares,2 yet neither identifi ed issues with there was limited comment from inde- rurality in media coverage. These DHBs pendent commentators in the media, which have enjoyed relatively strong fi nancial may create barriers to impartial debate.32 health, which could have mitigated funding Furthermore, in combination with a focus on pressures associated with rurality.44,45 In fi nancial and health services diffi culties, the contrast, South Canterbury and Taranaki attention given by the media to voices with a DHBs receive relatively small shares of vested interest indicates a tendency towards the rurality adjustor, yet both identifi ed sensationalism in reporting on the PBFF.51 rurality as an issue. The prominence of the Fourth, temporal patterns in media rurality theme is perhaps also related to reporting suggest DHBs and politicians and the dominance of Southern reporting. As the media may be using pivotal points in the with rurality, ethnicity and deprivation are policy process to advocate for change to the factors which may affect the access and use PBFF model. The fi rst peak in media attention 46–49 of health services in New Zealand, yet coincided with the introduction of the PBFF. neither have featured prominently in media Nationally, this period also corresponded

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 45 www.nzma.org.nz/journal ARTICLE

with the greatest emphasis on Vote Health share reductions in the media, drawing Conclusion public attention to the potentially detrimental At the nexus of the themes emerging from effects of the new policy on DHBs. media coverage of the PBFF lie perceptions of fairness of the model. Although the role The second spike in the number of articles of the media in steering policymaking is and negative attitudes recorded coincided controversial it plays a crucial role in infl u- with the release of the Government’s most encing perceptions of the public.33,51,54 The recent review of the PBFF; over a third of PBFF may be intended as an impartial mech- articles mentioning a review were published anism for the distribution of health funding in that year alone. This period represented throughout the country,55 but public accep- a key window in which change to the PBFF tance of resource allocation decisions hinge was under consideration. The previous on value judgements and a sense that they review, conducted in 2007, failed to elicit result from a fair process.56,57 Our fi ndings a comparable spike in media attention, show that media coverage consistently links perhaps indicating rising interest and levels the PBFF to the cause and effects of fi nancial of unease with the current healthcare pressures experienced by DHBs throughout funding paradigm. the country, creating a sense that funding allocations are inequitable and that DHBs Strengths and are not entirely comfortable with the PBFF limitations model and its impacts on DHB income. The Our fi ndings contribute to a growing persistently high volume of reporting in the body of research on funding allocations in southern regions implies that the formula New Zealand,1,4,13–16 capturing a compre- has been particularly poorly perceived in hensive picture of reporting on the PBFF the South Island. At a national level, the over the majority of its lifetime and across pattern of media reporting and the themes the country. The combination of qualitative identifi ed suggest the PBFF has failed to and quantitative data offers insights into keep pace with the challenges facing DHBs the key patterns and themes surrounding such as aging populations, the diseconomies the PBFF as it is reported in the media, of scale related to rurality and health work- enhancing our understanding of the state of force recruitment and retention. health funding in New Zealand and how the While the PBFF may act as somewhat of a PBFF has been presented to the public. Our lightning rod for frustrations over fi nancial methodology is comparable to other media strain in the DHB sector, the media narrative content analyses.6,33 While some authors points to a number of lessons for policy- have argued that codes should only be makers. First, it highlights the need for created a priori to avoid the introduction of transparency around the PBFF model and observer bias,53 our justifi cation for a data- the process and principles underpinning driven method of creating codes is the need PBFF allocations. Second, the dearth of for fl exibility in a poorly researched area independent comment within the media is where existing literature could not guide an argument for policymakers to support us on the themes and issues we were likely the development of independent expertise to encounter. Although our study included on the PBFF model in order to provide all available news media stories over the credible and balanced viewpoints on the 2003–2016 period, scarce media coverage of distribution of health funding.33 Lastly, in some North Island DHBs makes it diffi cult to light of ongoing fi nancial disparities across identify issues and trends in those regions. the DHB sector, it reinforces the impor- Furthermore, the relatively small amount tance of monitoring of the impact of PBFF of research on the PBFF creates diffi culties allocations throughout the country. With when comparing our data and conclu- increasing healthcare costs combined with sions with other studies. Nevertheless, our an ageing population, a growing prevalence fi ndings are consistent with international of chronic illness and persistent disparities literature, demonstrating the importance of in healthcare access and outcomes, it will the news media in framing health funding become increasingly important that Vote policy debates.32,51,54 Health is distributed as fairly and as effec-

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 46 www.nzma.org.nz/journal ARTICLE

tively as possible. A greater understanding fi nancial strife would be useful in furthering of the interplay between the PBFF and this goal and in optimising New Zealand’s other factors which may contribute to DHB healthcare system.

Competing interests: Nil. Acknowledgements: This study was funded by the University of Otago Dunedin School of Medicine. Author information: Aaron N Chester, Wellington School of Medicine, University of Otago, Dunedin; Erin C Penno, Otago Business School, University of Otago, Dunedin; Robin DC Gauld, Otago Business School, University of Otago, Dunedin. Corresponding author: Aaron N Chester, 88 Melrose Road, Island Bay, Wellington. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2018/vol-131-no-1480- 17-august-2018/7663

REFERENCES: 1. Penno E, Audas R, Gauld 6. Derrett S, Cousins K, Gauld ‘rough-cut’. Otago Daily R. The state of the art?: R. A messy reality: An Times. 2012 June 19. An analysis of New analysis of New Zealand’s 12. Harvey H. Taranaki is Zealand’s Population-Based elective surgery scoring likely to get less money Funding Formula for system via media sources, for health. Taranaki Daily health services: Centre 2000–2006. Int J Health News. 2016 February 11. for Health Systems, Plan M. 2013; 28(1):48–62. 13. Penno E, Gauld R. How Department of Preventive 7. Benelli E. The role of the are New Zealand’s district and Social Medicine, media in steering public health boards funded University of Otago; 2012. opinion on healthcare and does it matter if we 2. Ministry of Health. issues. Health Policy. can’t tell? N Z Med J. Population-Based Fund- 2003; 63(2):179–86. 2013;126(1376):71-84. ing Formula review 8. McLean E. So much to 14. Mellsop G, McClintock – 2015 Technical report. review, so little to count K, Dutu G. (2007). Wellington; 2016. on. Otago Daily Times. May Population-based fund- 3. Population-Based Funding 25, 2015 Dunedin: Otago ing formulae cultural Formula. New Zealand Daily Times; [updated May coeffi cients: seeking a Medical Association 25, 2015. Available from: more equitable distri- Website. [Available from: http://www.odt.co.nz/ bution of health dollars http://www.nzma.org. news/dunedin/so-much- in New Zealand. New nz/advocacy/?a=39429 review-so-little-count Zealand Medical Journal, 4. Penno E, Gauld R, Audas R. 9. Rachul C, Caulfi eld T. 120(1251):U2465. How are population-based The media and access 15. Langton J, Crampton P. funding formulae for issues: Content analysis (2008). Capitation funding healthcare composed? A of Canadian newspaper of primary health organisa- comparative analysis of coverage of health policy tions in New Zealand: are seven models. BMC Health decisions. Orphanet J enrolled populations being Serv Res. 2013; 13(1):1–13. Rare Dis. 2015; 10(1):1. funded according to need? 5. Ashton T. Recent develop- 10. Milne A. Ministry mum New Zealand Medical ments in the funding and on health funding formu- Journal, 121(1272):47–58. organisation of the New la. Southland Times. 16. Cumming J, Mays N. (2011). Zealand health system. 2009 September 28. New Zealand’s Primary Aust New Zealand Health 11. Goodwin E. Criteria Health Care Strategy: early Policy. 2005; 2(1):1–8. for funding hospitals effects of the new fi nancing

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 47 www.nzma.org.nz/journal ARTICLE

and payment system for Phones Clyde man about of Otago District Health general practice and future no-confi dence vote. Otago Board. New Zealand challenges. Health Econom- Daily Times. 2015 June 5. Parliament; 2007. ics, Policy and Law, 6:1–21. 30. Gibb J. Health board 40. Summary fi nancial 17. Myllylahti M. New problems raised at reports. Ministry of Zealand media owner- meeting. Otago Daily Health Web site [updated ship 2016. Auckland: Times. 2014 March 1. September 2, 2011; cited Centre for Journalism, 31. Kirk S. Minister says 2017 6 February]. Avail- Media and Democracy, health board changes able from: http://www. Auckland University unlikely. Taranaki Daily health.govt.nz/new-zea- of Technology; 2016. News. 2015 July 29. land-health-system/ key-health-sector-or- 18. Braun V, Clarke V. Using 32. Daw JR, Morgan SG, Collins ganisations-and-people/ thematic analysis in PA, Abelson J. Framing district-health-boards/ psychology. Qual Res incremental expansions accountability-and-funding/ Psychol. 2006; 3(2):77–101. to public health insurance summary-fi nancial-reports 19. Neergaard MA, Olesen F, systems: the case of Andersen RS, Sondergaard Canadian pharmacare. J 41. Brabyn L, Skelly C. J. Qualitative description– Health Polit Policy Law. Modeling population the poor cousin of health 2014; 39(2):295–330. access to New Zealand public hospitals. Int J research? BMC Med Res 33. Haas M, Chapman S, Viney Health Geogr. 2002; 1(1):3. Methodol. 2009; 9(1):52. R, Hall J, Ferguson A. The 20. Clarke V, Braun V. Thematic news on health care fund- 42. Brabyn L, Barnett R. analysis. Encyclopedia ing: a study of reporting Population need and of critical psychology: in the Australian print geographical access to Springer; 2014:1947–52. media for 1996. J Health general practitioners in rural New Zealand. N Z 21. Ross T. Health’s funding Serv Res Policy. 2001; 6(2). Med J. 2004; 117(1199). woes set to worsen. The 34. Shanahan EA, McBeth MK, Press. 2004 April 3. Hathaway PL, Arnell RJ. 43. Moore D, Blick G, Whelen C. Review of the rural 22. Hayman K. Knife taken Conduit or contributor? and tertiary adjusters. to elderly services. The The role of media in policy Ministry of Health; 2015. Press. 2004 November 6. change theory. Policy Sciences. 2008; 41(2):115. 44. Ministry of Health. DHB 23. Bailey E. Elderly health Net Results for 2015/2016 care fears. The Timaru 35. Fredriksson M, Tiainen for the period ended Herald. 2007 October 1. A, Hanning M. Regional media coverage infl uences 30/04/2016. Wellington: 24. My DHB. Ministry of Health the public’s negative Ministry of Health; 2016. Web site [updated 29 attitudes to policy imple- 45. The Treasury. Effi ciency September 2016. Avail- mentation success in Analysis of District able from: http://www. Sweden. Health Expect. Health Boards - Response health.govt.nz/new-zea- 2015; 18(6):2731–41. to Offi cial Information land-health-system/ Request (OIA 20170064) my-dhb 36. Gabe J, Chamberlain K, Norris P, Dew K, Madden 2017 [Available from: 25. Harding E. King casts doubt H, Hodgetts D. The http://treasury.govt.nz/ on funding. The Southland debate about the funding publications/oia-response/ Times. 2015 Aug 24. of Herceptin: A case effi ciency-analysis-dis- 26. Hayman K. Patients lose study of ‘countervailing trict-health-boards-re- out under funding plan. powers’. Soc Sci Med. sponse-offi cial-informa- The Press. 2004 July 1. 2012; 75(12):2353–61. tion-request-oia-20170064 27. Sharpe M. A tale of two 37. Health Committee. 2005/06 46. Cumming J, Stillman S, stories. The Dominion Financial review of Otago Liang Y, Poland M, Hannis Post. 2008 February 22. District Health Board. 2007. G. The determinants of GP visits in New Zealand. 28. Collins S. After 18 years 38. Committee H. 2001/02 Aust N Z J Public Health. of development, it’s Financial review of 2010; 34(5):451–7. back to old Maori way. Canterbury District New Zealand Herald. Health Board. 2003. 47. Barnett R, Malcolm March 17, 2006. L. Practice and ethnic 39. Health Select Committee. variations in avoidable 29. Goodwin E. Instant 2005/06 Financial review response from Minister; hospital admission rates

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 48 www.nzma.org.nz/journal ARTICLE

in Christchurch, New stakeholders’ perspec- review. Implementation Zealand. Health Place. tives in the newsprint Science. 2017; 12(1):52. 2010; 16(2):199–208. media. Int J Health Plan 55. Ministry of Health. 2004. 48. Ministry of Health. New M. 2014; 29(1):70–89. Population-based Funding Zealand health survey: 52. Light DW. Countervail- Formula 2003. Wellington: Annual update of key fi nd- ing powers. The Wiley Ministry of Health. ings 2012/13. Wellington: Blackwell Encyclopedia of 56. Rogge J, Kittel B. Who shall Ministry of Health; 2013. Health, Illness, Behavior, not be treated: Public atti- 49. Minko N, Rains M, Bridg- and Society: Wiley-Black- tudes on setting health care ford P. Inpatient discharges well; 2014:325–32. priorities by person-based demographic profi les. 53. Macnamara JR. Media criteria in 28 nations. PLoS What do they tell us? content analysis: Its uses, One. 2016; 11(6):e0157018. Health Care Inform Rev benefi ts and best practice 57. Robinson S, Williams I, Online. 2009; 13(2):2–8. methodology. Asia-Pa- Dickinson H, Freeman T, 50. Gauld R, Raymont A, cifi c Public Relations Rumbold B. Priority-setting Bagshaw PF, Nicholls MG, Journal. 2005; 6(1):1. and rationing in health- Frampton CM. The impor- 54. Bou-Karroum L, El-Jardali care: evidence from the tance of measuring unmet F, Hemadi N, Faraj Y, Ojha English experience. Soc Sci healthcare needs. N Z Med U, Shahrour M, et al. Med. 2012; 75(12):2386–93. J. 2014; 127(1404):63–7. Using media to impact 51. Popa AE. Hospital decen- health policy-making: an tralisation in Romania: integrative systematic

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 49 www.nzma.org.nz/journal ARTICLE

Ethnic disparities in community antibacterial dispensing in New Zealand, 2015 Naomi Whyler, Andrew Tomlin, Murray Tilyard, Mark Thomas

ABSTRACT AIMS: There are significant ethnic disparities in the incidence of various infectious diseases in New Zealand. Antimicrobial stewardship interventions which ignore these disparities may have negative e ects on the health of some ethnic groups. We aimed to determine the relationship between ethnicity and community antimicrobial dispensing in New Zealand, to inform the development of antimicrobial stewardship interventions in New Zealand. METHODS: Demographic data on all patients registered with a general practice in New Zealand and on all community pharmacy antibacterial dispensings during 2015 were obtained from national healthcare databases. The rates of dispensing were measured as the number of dispensings per 1,000 population per day and as defined daily doses per 1,000 population per day. RESULTS: The rate of community antibacterial dispensing for the total population surveyed was 3.01 dispensings per 1,000 population per day, and was 3.49 for Pacific, 3.23 for Māori, 3.02 for European, 2.70 for Middle Eastern, Latin American and African, and 2.35 for Asian people. In all ethnic groups the rate of community antibacterial dispensing increased with increasing socioeconomic deprivation. Seasonal variation in antibacterial dispensing ranged between 34% in Asian people and 24% in European people. CONCLUSIONS: The ethnic disparities in the rates of antibacterial dispensing in New Zealand are consistent with, but less marked than, the ethnic disparities in the incidence of infectious diseases in New Zealand. Improved community-wide understanding of both the benefits and the harms of antibacterial medicines is necessary to support improved antibacterial use in New Zealand in the future.

he relentless spread of antibacterial Antibacterials can, however, provide resistance poses major threats for hu- major benefi ts in the treatment of many Tman health.1,2 The main infl uence on infections. Strategies to reduce unnecessary the speed with which antibacterial resistant antibacterial prescribing therefore need bacteria spread within a nation is the level to be suffi ciently nuanced so that antibac- of antibacterial consumption within that na- terials continue to be prescribed for those tion.3 The total level of community antibac- infections in which they provide signif- terial consumption is high in New Zealand icant benefi ts. In New Zealand, many of when compared with many other nations.4,5 the infections for which antibacterials are Community dispensing, mostly for prescrip- benefi cial are more common in Māori and tions written by family doctors, comprises Pacifi c people than people of other ethnic- approximately 95% of antibacterial dispens- ities. For example, over the fi ve-year period ing in New Zealand, a higher proportion than 2004–2008, the age-standardised rate of in most other nations.6 Many studies in other admissions to hospital for an infectious nations have suggested that approximately disease was 2.35 times higher in Pacifi c half of all antibiotic prescriptions dispensed people and 2.15 times higher in Māori than in the community are not indicated, and pro- in all other ethnic groups.9 From 2012 to vide nil or minimal benefi t to the patient.7,8 2014 the rate of admission to hospital for a

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 50 www.nzma.org.nz/journal ARTICLE

fi rst episode of acute rheumatic fever was 63 each patient.13 Each patient was allocated times higher in Pacifi c people and 27 times to one of fi ve deprivation quintiles (quintile higher in Māori than in all other ethnic 1 = the least deprived, quintile 5 = the most groups.10 One might therefore conclude that deprived). antibacterials should be dispensed more Data on all community antibacterial often to Māori and Pacifi c people than to dispensings from 1 January 2015 to 31 people from other ethnic groups. However, December 2015 were obtained from some recent studies have suggested that the National Pharmaceutical Collection. antibacterials are less frequently dispensed Pharmaceuticals in this dataset are clas- to Māori when compared with most other sifi ed under the Anatomical Therapeutic ethnic groups. For example, Norris et al Chemical (ATC) system.14 Data for each measured antibacterial dispensing in the dispensing included the antibacterial agent, Tairawhiti region of New Zealand during its formulation, daily dose and the quantity 2011 and found that Māori, particularly dispensed. Records from the two national those living rurally, were less likely to be datasets were linked for each patient using 11 dispensed an antibacterial than non-Māori. an encrypted form of their National Health More recently, Williamson et al measured Index (NHI) code. antibacterial dispensing throughout New Antibacterial dispensing was measured Zealand during 2014 and found that the as the number of prescription items per highest rate of dispensing was to Pacifi c 1,000 population per day, and as the number people with the lowest rates to Māori and of defi ned daily doses (DDDs) per 1,000 Asian people.5 population per day.14 For prescriptions The aim of our present study was to that resulted in repeat dispensings, meas- measure the rates of community anti- urements of prescription items per 1,000 bacterial dispensing for the major ethnic population per day included only the initial groups in New Zealand during 2015. We dispensing, but measurements of DDDs anticipated that this information would help per 1,000 population per day included the to inform antibiotic stewardship programs total amount dispensed inclusive of repeat in New Zealand. dispensings. We compared the total level of community Methods antibacterial dispensing in New Zealand Data for the study were obtained from during 2015 with similar recent data from two national healthcare databases managed other developed nations. by the New Zealand Ministry of Health. We determined the effect of age and Information on all patients registered with ethnicity on the rate of dispensing for each a New Zealand general practice during the of the major antibacterial classes, and the fi rst quarter of 2015 was obtained from the effect of socio-economic deprivation on Primary Health Organisation Enrolment the rate of total antibacterial dispensing Collection. This database records each for each ethnic group. We also measured patient’s date of birth, gender, ethnicity, esti- the seasonal variation in total antibacterial mated level of socioeconomic deprivation dispensing by calculating the percentage and the general practice with which they increase in antibacterial dispensing during are registered. Based on their self-reported the winter months (April–September) 12 ethnicity, patients were assigned to one of compared with antibacterial dispensing six ethnic groups: Asian, European, Māori, during the summer months (January– MELAA (Middle Eastern, Latin American March plus October–December). (Seasonal and African), Pacifi c and Other. The New variation = [dispensings (winter months)/ Zealand Deprivation Index, which relates a dispensings (summer months)-1]X100.) person’s place of residence to data contained We used Poisson regression to test for a in the New Zealand 2013 Census of Popu- linear trend in antibacterial dispensing rates lation and Dwellings, was used to determine with increasing deprivation after adjusting the level of socio-economic deprivation for for patient ethnicity.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 51 www.nzma.org.nz/journal ARTICLE

people had the highest rate of dispensing Results (3.49), followed by Māori (3.23), European A total of 4,315,871 people were registered (3.02), MELAA (2.70), Other (2.63) and with New Zealand general practitioners Asian people (2.35) (Figure 1A) (Table 1). In during 2015. These patients comprised contrast, when measured as DDDs per 1,000 93.9% of the total estimated New Zealand population per day, the rate of dispensing of population of 4,596,700 at 30 June 2015.15 penicillins (15.99) and tetracyclines (13.17) Patient ethnicity was recorded as: Asian were similar, comprising 46% and 38% (10.1%), European (65.2%), Māori (14.7%), respectively of total antibacterial dispensing Pacifi c (7.3%), MELAA (1.2%) and Other or (34.97). Europeans had the highest rate unspecifi ed (1.5%). Māori or Pacifi c ethnicity of dispensing (37.41), followed by Other was strongly associated with high levels of (35.96), MELAA (33.76), Pacifi c (33.01), Māori socioeconomic deprivation. The proportion (30.69) and Asian people (26.99) (Figure 1B). of patients within the most deprived quintile Measurements of antibacterial dispensing of the New Zealand population was 53% in DDDs per 1,000 population per day are for Pacifi c people, 40% for Māori, 22% for misleading when applied to children and MELAA, 15% for Asian, 14% for Other and signifi cantly underestimate the number 10% for European. of children who have been dispensed an In total, 5,204,143 antibacterial courses antibacterial medicine. As dispensing for were dispensed by community phar- children comprises a large proportion of macies, with 4,741,239 (91.1%) of these total antibacterial dispensing, we used dispensed to 1,882,966 patients registered dispensings per 1,000 population per day for with a New Zealand general practice. The all further analyses. overall rate of community antibacterial dispensing for practice-registered patients Antibacterial dispensing by age was 4,741,239/4,315,871 (1.1 dispensings and ethnicity per person per year), with one or more Total community antibacterial dispensing antibacterial medicines dispensed to 43.6% followed a ‘u-shaped’ distribution, (1,882,966/4,315,871) of all patients. regardless of ethnicity, with the highest rates of dispensing in people <5 years and Antibacterial dispensing by >80 years of age (Figure 2A). The overall ethnicity rate of dispensing in New Zealand during The method used to measure antibac- 2015 was broadly similar to that in Australia terial dispensing had a signifi cant impact on during 2014 (personal communication K comparisons between ethnic groups. When Meleady, Australian Commission on Safety measured as the number of prescription and Quality in Healthcare),16 and the US items dispensed per 1,000 population per during 2011,17 but much greater than that day, penicillins (1.72) comprised 57% of in Sweden during 2014.18 The distributions total antibacterial dispensing (3.01). Pacifi c

Figure 1: Amount of antibacterial medicines dispensed in the community in New Zealand, in relation to ethnicity, during 2015; (A) dispensings per 1,000 population per day, (B) defi ned daily doses (DDDs) per 1,000 population per day.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 52 www.nzma.org.nz/journal ARTICLE

Table 1: Community antibacterial dispensing of medicines within each antibacterial class (dispensings per 1,000 population per day) for each ethnic group in New Zealand during 2015.

Pacific Māori European MELAA Asian Other Not specified Total Penicillins 2.39 2.05 1.62 1.65 1.44 1.51 1.51 1.72

Cephalosporins 0.12 0.12 0.13 0.10 0.10 0.09 0.10 0.12

Tetracyclines 0.16 0.23 0.40 0.28 0.21 0.34 0.29 0.34

Macrolidesa 0.26 0.27 0.30 0.23 0.23 0.25 0.24 0.28

Sulphonamidesb 0.19 0.20 0.21 0.14 0.11 0.15 0.17 0.20

Quinolones 0.03 0.04 0.08 0.04 0.03 0.05 0.05 0.06

Other agentsc 0.11 0.14 0.15 0.13 0.10 0.12 0.14 0.14

Topical agentsd 0.23 0.19 0.14 0.13 0.13 0.12 0.11 0.15

Total 3.49 3.23 3.02 2.70 2.35 2.63 2.63 3.01 (95% CI) (3.48–3.50) (3.23–3.24) (3.02–3.03) (2.68–2.72) (2.34–2.36) (2.61–2.66) (2.58–2.67) (3.01–3.01)

Abbreviations: MELAA, Middle Eastern, Latin American, African; CI, confidence intervals. a Includes lincosamides. b Includes trimethoprim (alone or in combination with sulphamethoxazole). c Includes nitroimidazoles, nitrofurantoin, vancomycin, fusidic acid and colistin. d Includes fusidic acid and mupirocin.

of total antibacterial dispensing by age for (Figure 3G) and topical antibacterials (Figure ethnic groups in New Zealand were approx- 3H). There were high rates of dispensing imately parallel (Figure 2B) with the rate of tetracyclines for European, MELAA and of dispensing generally highest for Pacifi c Asian people aged 15–19 years (Figure people, and lowest for Asian people. 3C) and of other antibacterials (predomi- Similar ‘u-shaped’ distributions, with nantly nitroimidazoles and nitrofurantoin) varying degrees of accentuation at the for Māori, Pacifi c and European people extremes of age, were seen for dispensing aged 20–29 years (Figure 3G). Quinolone of penicillins (Figure 3A), cephalosporins dispensing increased with age across all (Figure 3B), macrolides and lincosamides ethnic groups with the highest rates in (Figure 3D), and sulphonamides and European, Māori and MELAA people aged trimethoprim (Figure 3E). However, the >80 years (Figure 3F). Rates of dispensing patterns of dispensing in relation to age of topical antibacterials were highest for differed for tetracyclines (Figure 3C), Pacifi c and Māori children <5 years of age quinolones (Figure 3F), other antibacterials (Figure 3H).

Figure 2: Amount of antibacterial medicines dispensed in the community (dispensings per 1,000 popu- lation per day) in relation to age; (A) in New Zealand during 2015, Sweden during 2014,17 the US during 2011,16 and Australia during 2015 [personal communication K Meleady, Australian Commission on Safety and Quality in Healthcare], (B) by ethnicity, in New Zealand during 2015.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 53 www.nzma.org.nz/journal ARTICLE

Figure 3: Amount of antibacterial medicines dispensed in the community (dispensings per 1,000 population per day) in relation to age and ethnicity, in New Zealand during 2015; (A) penicillins, (B) cephalosporins, (C) tetracyclines, (D) macrolides and lincosamides, (E) sulphonamides and trimethoprim, (F) quinolones, (G) other antibacterials and (H) topical antibacterials.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 54 www.nzma.org.nz/journal ARTICLE

Figure 4: Amount of antibacterial medicines dispensed in the community (dispensings per 1,000 popula- tion per day) in relation to ethnicity and level of socioeconomic deprivation, in New Zealand during 2015.

In all age groups the rate of dispensing day). The overall rate of dispensing was 48% of penicillins was higher in Māori and in (95% CI, 48–49%) higher in Pacifi c people Pacifi c people than in other ethnic groups. (3.49 dispensings per 1,000 population per The rate of dispensing of penicillins to those day) than in Asian people (2.35 dispensings most at risk of rheumatic fever—Pacifi c and per 1,000 population per day). Māori people aged 5–20 years—was 45% Antibacterial dispensing by (95% CI, 44–46%) higher in Pacifi c people (2.22 dispensings per 1,000 population per ethnicity and season The seasonal variation in the rate of day), and 24% (95% CI, 23–25%) higher in antibacterial dispensing for the total New Māori (1.90 dispensings per 1,000 population Zealand population was 26% (95% CI, per day), than in people of the same age in 26–27%) (Figure 5). The seasonal variation other ethnic groups (1.53 dispensings per in the rate of dispensing was 29% (95% CI, 1,000 population per day). 29–30%) in Pacifi c people, 28% (95% CI, Antibacterial dispensing by 27–28%) in Māori, 25% (95% CI, 24–25%) in ethnicity and level of socio- European, 31% (95% CI, 28–33%) in MELAA economic deprivation and 35% (95% CI, 34–36%) in Asian people. The impact of socio-economic deprivation on rates of antibacterial dispensing was rela- Discussion tively consistent in all ethnic groups (Figure We found that the rate of total community 4). There was a 3.6% increase in dispensing antibacterial dispensing in New Zealand rate per unit increase in deprivation quintile during 2015 (3.01 dispensings per 1,000 across all patients after adjusting for ethnic population per day) was 3.3 times higher group (P<0.001). This linear trend with than in Sweden during 2014 (0.90),18 2.1 increasing deprivation was also found times higher than in Denmark during 2015 within each ethnic group (P<0.001). The (1.45),19 1.7 times higher than in Canada overall rate of dispensing was 18% (95% CI, during 2011 (1.79)20 and 1.3 times higher 18–19%) higher for all people in quintile 5 than in the US during 2010–2011 (2.31),17 (3.37 dispensings per 1,000 population per but 14% lower than in Australia during day) than it was for all people in quintile 1 2014 (3.50).16 These fi ndings emphasise the (2.85 dispensings per 1,000 population per pressing needs, to strengthen public health

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 55 www.nzma.org.nz/journal ARTICLE

Figure 5: Amount of antibacterial medicines dispensed in the community (dispensings per 1,000 population per day) in relation to ethnicity and season, in New Zealand during 2015.

programs, incorporating both pharmaceu- in more socio-economically deprived areas, tical and non-pharmaceutical approaches, supporting the fi ndings of two previous New that will reduce the incidence of infectious Zealand studies that also found higher rates diseases in New Zealand, and also to more of antibacterial dispensing in communities effectively deliver community antimicrobial with higher levels of socio-economic depri- stewardship in New Zealand. vation.22,23 A large US study also found that Overall rates of community antibacterial community antibacterial dispensing rates dispensing were higher in Pacifi c and Māori were highest in counties with the lowest per 17 people than in people of other ethnicities. A capita income. previous study found lower rates of antibac- Higher rates of antibacterial consumption terial dispensing in Māori, when compared by Pacifi c and Māori people may result in with non-Māori, in a geographically isolated higher prevalence of antibiotic resistance region of New Zealand during 2005–2006,11 in the bacteria that commonly cause infec- and a national study found that the rates of tious diseases in Pacifi c and Māori people. dispensing in Māori, during 2008 to 2014, A study of Staphylococcus aureus infections were less than in Pacifi c, European and in 16,249 patients admitted to Auckland City MELAA people, but more than in Asian Hospital between 2001 and 2011 found the people.5 Other recent studies found antibac- proportion of infections caused by non- terial dispensing rates in Māori and Pacifi c multidrug resistant methicillin resistant S. people in New Zealand to be greater than in aureus (nmMRSA) to be 2.41 times higher other ethnic groups.21,22 Ethnic disparities in in Pacifi c people and 1.48 times higher community antibacterial dispensing have in Māori people than in people of other also been demonstrated in the US where ethnicities.24 High rates of dispensing of community antibacterial dispensing rates topical fusidic acid for Pacifi c and Māori during 2011 were found to be higher in children under fi ve years of age (Figure 3H) counties with a greater proportion of black may have contributed to the high proportion residents.17 We also found antibacterial of S. aureus infections in Pacifi c and Māori dispensing rates to be higher in people living people that were caused by nmMRSA.25

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 56 www.nzma.org.nz/journal ARTICLE

Other somewhat unexpected effects of The management of patients with phar- antimicrobial consumption are widely yngitis due to Streptococcus pyogenes poses reported. Frequent antibiotic treatment in a particular problem with regard to anti- early childhood is strongly associated with microbial stewardship in New Zealand. childhood obesity,26 possibly as the conse- Between 2000 and 2014, 2,108 patients were quence of inhibiting the growth of intestinal admitted to hospital for fi rst episodes of bacteria that help to maintain the normal acute rheumatic fever in New Zealand: 1,068 thickness and function of the mucin on the (51%) were Māori, 821 (39%) were Pacifi c intestinal surface.27 The rate of antibac- people, and 219 (10%) were people of Other terial dispensing in children less than fi ve ethnicity, with the overwhelming majority years old was 52% (95% CI, 51–53%) higher of patients aged 5–20 years.10 The incidence in Pacifi c children (6.60 dispensings per of hospitalisation for a fi rst episode of acute 1,000 population per day), and 29% (95% rheumatic fever during 2012–2014 was 63 CI, 28–30%) higher in Māori children (5.60 times greater in Pacifi c people, and 27 times dispensings per 1,000 population per day), greater in Māori, than in people of other than in children of all other ethnicities (4.33 ethnicities.10 However, we found that the dispensings per 1,000 population per day). rate of dispensing of penicillins for patients The higher rates of antibacterial dispensing 5–20 years of age was only 45% higher in for Pacifi c and Māori children may have Pacifi c people, and 24% higher in Māori, contributed to obesity rates approximately than in people of all other ethnicities. It four times higher in Pacifi c children and seems likely that a large proportion of the 1.6 times higher in Māori children than penicillins dispensed to patients of other in children of all other ethnicities in New ethnicities aged 5–20 years was inappro- Zealand.28 priately prescribed for people either with Higher rates of antibacterial dispensing pharyngitis but minimal risk of rheumatic 31 for Pacifi c and Māori people are to be fever, or with self-limiting respiratory tract expected in the light of their higher inci- infections for which antibacterial therapy is 32 dence of infectious diseases. From 2004 almost never indicated. to 2008, rates of admission to hospital for In previous studies seasonal variation in infectious diseases were 2.35 times greater antibacterial dispensing has been used as in Pacifi c and 2.15 times greater in Māori a surrogate measure of prescribing quality. people than in people of other ethnicities.9 Recent studies have found the seasonal vari- Even greater ethnic disparities were found ation in community antibacterial dispensing in the rates of hospital admission for to be 12% in Denmark, 14% in the UK, 36% staphylococcal skin and soft tissue infec- in Germany and 39% in Italy during 2015,33 tions during 2000–2011, which were fi ve and 25% in the US between 2006 and 2010.34 times greater in Pacifi c and three times We found that the seasonal variation in greater in Māori people than in Europeans,29 antibacterial dispensing ranged from 25% and for rates of notifi cation of invasive in European people to 35% in Asian people, pneumococcal disease in 2015, which were suggesting that, based on this indicator, 4.3 times greater in Pacifi c and 3.8 times the quality of antibacterial prescribing is greater in Māori people than in people of relatively consistent across ethnic groups other ethnicities.30 However, the overall in New Zealand. However, the relatively rate of community antibacterial dispensing high levels of seasonal variation in antibac- in New Zealand during 2015 was only 19% terial dispensing in all ethnic groups in New higher in Pacifi c people, and 11% higher in Zealand, when compared with the overall Māori than in people of all other ethnicities. levels in Denmark and the UK, do suggest These fi ndings suggest that antibacterial that there is considerable scope to reduce prescribing may be appropriate more often prescribing in New Zealand for patients with for Māori and Pacifi c people than for people winter coughs and colds. from other ethnic groups, and there may be The strengths of this study include the more opportunities to reduce inappropriate inclusion of data for more than 90% of prescribing when treating people from other all community antibacterial dispensings, ethnic groups. the high rate of self-reported ethnicity in

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 57 www.nzma.org.nz/journal ARTICLE

the general practice patient records, and intended to reduce community antibac- the ability to include other demographic terial dispensing in New Zealand need factors including age and socio-economic to be suffi ciently nuanced so that they deprivation in our analyses. Limita- do not lead to reduced rates of antibac- tions of the study include the inability to terial treatment for infections that require include antibacterials dispensed directly treatment. Our study suggests that the risk by doctors from supplies provided to them of this occurring will be greatest in Pacifi c on practitioner supply orders, or to include and Māori people. Unfortunately, higher antibacterials dispensed by nurses within rates of antibacterial consumption by Pacifi c school-based public health programmes and Māori people will inevitably result in including the sore throat management correspondingly higher rates of adverse programme. Such dispensing is more likely consequences, including higher preva- to have occurred for socio-economically lence of antibacterial resistance and of the deprived patients, of whom a very large metabolic consequences that are caused by proportion would have been of Māori or alterations in the gut microbiome. Further Pacifi c ethnicity. research is required to better understand In conclusion, we found the overall rate the factors that lead to ethnic differences in of community antibacterial dispensing in the levels of antibacterial dispensing in New New Zealand to be high, with the highest Zealand. Reducing inappropriate antibac- rates in Pacifi c and Māori people. The terial consumption is at least as important higher rates of dispensing in Pacifi c and for Pacifi c and Māori people as it is for other Māori people refl ect their higher incidence population groups. of various infectious diseases. Programmes

Competing interests: Nil. Acknowledgements: Lisa Couldrey kindly assisted with design of the fi gures. Author information: Naomi Whyler, Adult Infectious Diseases Department, Auckland City Hospital, Auckland; Andrew Tomlin, Best Practice Advocacy Centre, Dunedin; Murray Tilyard, Best Practice Advocacy Centre, Dunedin; Mark Thomas, Adult Infectious Diseases Department, Auckland City Hospital, Auckland; Department of Molecular Medicine and Pathology, University of Auckland, Auckland. Corresponding author: Associate Professor Mark Thomas, Department of Molecular Medicine and Pathology, University of Auckland, Park Rd, Grafton, Private Bag 92019, Auckland 1142. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2018/vol-131-no-1480- 17-august-2018/7664

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 58 www.nzma.org.nz/journal ARTICLE

REFERENCES: 1. WHO. Worldwide country among US ambulatory June 2015. situation analysis: response care visits, 2010–2011. Available at: http://www. to antimicrobial resistance. JAMA 2016; 315:1864–73. stats.govt.nz/browse_ 2015. ISBN9789241564946. 9. Baker MG, Telfar Barnard for_stats/population/ Available at: http://www. L, Kvalsvig A, et al. estimates_and_projections/ who.int/antimicrobial-re- Increasing incidence of NationalPopulationEsti- sistance/publications/ serious infectious diseases mates_HOTPAt30Jun15. situationanalysis/en/ and inequalities in New aspx 2. The Review on Antimicrobi- Zealand: a national 16. Australian Commission al Resistance. Antimicrobial epidemiological study. on Safety and Quality in Resistance: Tackling a crisis Lancet 2012; 379:1112–9. Healthcare. AURA 2016 for the health and wealth 10. Jack S, Williamson D, - First Australian report of nations. 2014. Available Galloway Y, et al. Interim on antimicrobial use and at: https://amr-review.org/ evaluation of the sore resistance in human health. 3. Austin DJ, Anderson throat component of the Available at: https://www. RM. Studies of antibiotic rheumatic fever prevention safetyandquality.gov.au/ resistance within the programme – quantitative publications/aura-2016- patient, hospitals and the fi ndings. The Institute of fi rst-australian-report-on- community using simple Environmental Science and antimicrobial-use-and-re- mathematical models. Research Ltd. Porirua, New sistance-in-human-health/ Phil Trans R Soc Lond B. Zealand; 2015. Available at: 17. Hicks LA, Bartoces MG, 1999; 354(1384):721–38. http://www.health.govt.nz/ Roberts RM, et al. US outpa- 4. Thomas MG, Smith A, publication/interim-evalu- tient antibiotic prescribing Tilyard M. Rising antimi- ation-sore-throat-manage- variation according crobial resistance: a strong ment-component-new-zea- to geography, patient reason to reduce excessive land-rheumatic-fever-pre- population, and provider antimicrobial consumption vention-programme specialty in 2011. Clin in New Zealand. N Z Med 11. Norris P, Horsburgh S, Infect Dis 2015; 60:1308–16. J 2014; 127(1394):72–84. Keown S, et al. Too much 18. Swedres-Svarm 2014. 5. Williamson DA, Roos and too little? Prevalence Consumption of antibi- RF, Verrall A. Antibiotic and extent of antibiotic use otics and occurrence of consumption in New in a New Zealand region. antibiotic resistance in Zealand, 2006–2014. The J Antimicrob Chemother Sweden. Solna/Uppsala Institute of Environmental 2011; 66:1921–6. ISSN 1650-6332. Available Sciences and Research 12. Ministry of Health. at: http://www.sva.se/en/ Ltd. Porirua, New Ethnicity data protocols for antibiotics/svarm-reports Zealand. 2016. Available the health and disability 19. Danmap 2015 – Use of at: https://surv.esr.cri.nz/ sector. 2004. Available at: antimicrobial agents and PDF_surveillance/Antibi- http://www.health.govt.nz/ occurrence of antimicrobial oticConsumption/2014/ publication/ethnicity-da- resistance in bacteria from Antibiotic_Consump- ta-protocols-health-and-dis- food animals, food and tion_Report_Final.pdf ability-sector humans in Denmark. ISSN 6. Duffy E, Ritchie S, Metcalfe 13. Salmond C, Crampton 1600-2013. Available at: S, et al. Antibacterials P. Development of New http://www.danmap.org. dispensed in the commu- Zealand’s deprivation index 20. Canadian Integrated nity comprise 85–95% of (NZDep) and its uptake Program for Antimicrobial total human antibacterial as a national policy tool. Resistance Surveillance. consumption. J Clin Pharm Can J Pub Health 2012; Human antimicrobial use Ther 2018; 43:59–64. 103 (Suppl. 2):S7–S11. report, 2011. Available 7. Grijalva CG, Nuorti JP, 14. Collaborating Centre for at: http://publications. Griffi n MR. Antibiotic Drug Statistics Meth- gc.ca/collections/collec- prescription rates for acute odology. Guidelines for tion_2014/aspc-phac/ respiratory tract infections ATC classifi cation and HP40-109-2014-eng.pdf in US ambulatory settings. DDD assignment 2013. 21. Metcalfe S, Laking G, JAMA 2009; 302:758–66. Oslo, 2012. Available Arnold J. Variation in 8. Fleming-Dutra KE, Hersh at: http://www.whocc. the use of medicines by AL, Shapiro DJ, et al. no/atc_ddd_index/ ethnicity during 2006/07 Prevalence of inappropri- 15. Statistics NZ. National in New Zealand: a ate antibiotic prescriptions population estimates: at 30 preliminary analysis. N Z Med J 2013; 126:14–41.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 59 www.nzma.org.nz/journal ARTICLE

22. Hobbs MR, Grant CC, et al. Akkermansia Heart Foundation of New Ritchie SR, et al. Anti- muciniphila and improved Zealand. Available at: http:// biotic consumption by metabolic health during assets.heartfoundation.org. New Zealand children: a dietary intervention nz/shop/heart-healthcare/ exposure is near universal in obesity: relationship non-stock-resources/ by the age of 5 years. J with gut microbiome gas-sore-throat-rheumat- Antimicrob Chemother richness and ecology. ic-fever-guideline.pdf 2017;72:1832-40. Gut 2016; 65:426–36. 32. Best Practice Advocacy 23. Walls G, Vandal AC, du 28. Ministry of Health. Annual Centre. 2015. Respiratory Plessis T, et al. Socioeco- update of key results tract infections (self-limit- nomic factors correlating 2015/16: New Zealand ing) – reducing antibiotic with community antimicro- Health Survey. Welling- prescribing. BPAC. Avail- bial prescribing. N Z Med ton: Ministry of Health. able at: http://www.bpac. J 2015; 128(1417):16–23. Available at: http://www. org.nz/guidelines/1/docs/ 24. Williamson DA, Lim health.govt.nz/system/fi les/ Respiratory-tract-in- A, Thomas MG, et al. documents/publications/ fections-%28self-limit- Incidence, trends and annual-update-key-results- ing%29-reducing-anti- demographics of Staphy- 2015-16-nzhs-dec16-v2.pdf biotic-prescribing.pdf lococcus aureus infections 29. Williamson DA, Zhang 33. European Centre for in Auckland, New Zealand, J, Ritchie SR, et al. Disease Prevention and 2001–2011. BMC Infectious Staphylococcus aureus Control. Quality indicators Diseases 2013; 13:569. infections in New Zealand, for antibiotic consump- 25. Williamson DA, Monecke 2000–2011. Emerg Infect tion in the community. S, Heffernan H, et al. High Dis 2014; 20:1157–62. Available at: https://ecdc. usage of topical fusidic acid 30. Institute of Environmental europa.eu/en/antimicrobi- and rapid clonal expansion Science and Research Ltd. al-consumption/database/ of fusidic acid-resistant Invasive pneumococcal quality-indicators Staphylococcus aureus: a disease in New Zealand, 34. Suda KJ, Hicks LA, Roberts cautionary tale. Clin Infect 2015. Porirua: ESR; 2017. RM, et al. Trends and Dis 2014; 59:1451–4. Available at: https://surv. seasonal variation in 26. Saari A, Virta LJ, Sankil- esr.cri.nz/PDF_surveillance/ outpatient antibiotic ampi U, et al. Antibiotic IPD/2015/2015IPDAn- prescription rates in the exposure in infancy and nualReport.pdf United States, 2006 to 2010. risk of being overweight in 31. Heart Foundation of Antimicrob Ag Chemo- the fi rst 24 months of life. New Zealand. Group A ther 2014; 58:2763–6. Pediatrics 2015; 135:617–26. streptococcal sore throat 27. Dao MC, Everard A, management guideline. Aron-Wisnewsky J, 2014 update. Auckland,

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 60 www.nzma.org.nz/journal ARTICLE

Antipsychotic prescribing in New Zealand between 2008 and 2015 Sam Wilkinson, Roger T Mulder

ABSTRACT AIM: To examine antipsychotic prescribing trends in New Zealand adults from 2008–2015. Methods: Antipsychotic prescribing data was sourced via the Ministry of Health. Data were examined by year, type of drug, ethnicity, gender, age and location of district health board. RESULTS: All individuals dispensed an antipsychotic were included. Rates of antipsychotic prescribing rose from 1.88% to 2.81%, an increase of 49% over the seven years. Most of the increase was in atypical antipsychotics (particularly quetiapine and olanzapine), which accounted for 82% of total antipsychotics in 2015. Māori were prescribed more antipsychotics than non-Māori. Asian people had the lowest rate of prescribing (0.86%). The highest rate of antipsychotic use was in European females aged 65 plus. CONCLUSION: Rates of antipsychotic prescription are increasing. Most of this change is in prescribing atypical antipsychotics. Young Māori males and elderly European females are most likely to receive antipsychotics.

ntipsychotics are a heterogeneous population census data, including New class of medications used to treat psy- Zealand. However, their study was primarily Achotic symptoms—such as delusions, a between-country comparison. hallucinations or paranoia—principally in schizophrenia and bipolar disorder, but also Objectives in major depression and anxiety. Typical The objectives of this study were to antipsychotics are ‘fi rst-generation’ antipsy- examine antipsychotic prescribing trends chotic drugs—discovered in the 1950s—that in New Zealand adults from 2008 to 2015, act predominantly by reducing dopamine by antipsychotic type, age, ethnicity, levels in the brain. Atypical (or ‘second-gen- gender and district health board location. eration’) antipsychotics were developed Our hypothesis was that antipsychotic more recently; they act on a range of recep- prescribing rates in New Zealand are tors such as acetylcholine, noradrenaline increasing. and serotonin in addition to dopamine. The fi rst atypical antipsychotic, clozapine, was discovered in the 1960s. Method Atypical drugs have been gradually Antipsychotic prescribing data was replacing typical antipsychotics over time. sourced via the Ministry of Health. The Verdoux et al1 reviewed pharmaco-epide- dataset contains every person in New miological studies and reported a marked Zealand who had been prescribed antipsy- 3 increase in antipsychotic prescription in chotic medication. This data is collected by most countries since the introduction of The Pharmaceutical Management Agency of atypical antipsychotics. New Zealand (PHARMAC) via the National Health Index (NHI) number—a unique To our knowledge, there has been no identifi er assigned to every person who uses detailed analysis of antipsychotic use in health services in New Zealand. The number New Zealand. Hálfdánarson et al2 reported of prescriptions with an NHI number rose to general trends in antipsychotic use from 90% in 2008, and to 98% in 2015. 2005–2014 in 16 countries, 10 of which were

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 61 www.nzma.org.nz/journal ARTICLE

The dataset lists the number of patients Statistics collecting prescriptions for antipsychotic Microsoft Excel was used to analyse the medication within each district health data. Data are descriptive and presented as board (DHB), fi ve-year age bracket, gender population prevalence. and ethnicity (Māori, Pacifi c, Asian and ‘Other’). The ‘Other’ category includes the following ethnic groups: European, Middle Results Eastern, Latin American, African and ‘Other National rates Ethnicity’. For simplicity we named this In 2015, 2.81% of all New Zealanders category ‘European/Other’. Data for eight aged 15 and over were prescribed an anti- consecutive years—2008 to 2015—was psychotic (2.94% of females and 2.67% of examined by ethnicity. males), an increase of 49% from the 2008 rate of 1.88%. The prescription rates were broken down into each antipsychotic class, and then Ethnicity broken down into individual drugs within Figure 1 shows the distribution of anti- each class. The data is presented over the psychotic prescribing by ethnicity. 3.37% four-year period from 2012 to 2015 because of Māori were prescribed antipsychotics in this data is more complete. 2015, compared to 3.15% of the European/ For the national prescribing rates, four Other category, 1.87% of Pacifi c Islanders age brackets were used: 15–24, 25–44, 45–64 and 0.86% of Asians. and 65+. The exception was individual DHB Since 2008, adult antipsychotic prescribing comparisons, where all ages were included, has increased across all ethnicities. Māori as only the total DHB populations were rates increased 60%, Pacifi c rates by 50%, available. European/Other rates increased by 52% and Population data were obtained from the Asian rates by 53%. Ministry of Health to calculate dispensing Gender rates.4 As age group data was available for Figures 2 and 3 show the distribution each ethnicity, the population aged 15 and of antipsychotic prescribing by gender, over could be calculated. The ethnic groups ethnicity and age in 2015. Female and male used are based on the ethnic groups from antipsychotic use varied with ethnicity the PHARMAC database.3 and gender. The highest rate of antipsy- Bias chotic use in females was European/Other This is a census of all New Zealand (3.38%) followed by Māori (3.16%). This prescribing data so there is no sampling bias. trend was reversed for males, where Māori were dispensed at the highest rate (3.61%) followed by European/Other (2.92%).

Figure 1: Percentage of New Zealand adults dispensed antipsychotics over time, by ethnicity.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 62 www.nzma.org.nz/journal ARTICLE

Figure 2: Percentage of females dispensed antipsychotics in 2015, by age bracket and ethnicity.

Age and ethnicity Geographic distribution The highest user of antipsychotics in Figure 4 shows the distribution of anti- 2015 was European/Other females aged 65+ psychotic prescribing by DHB in 2015. (5.04%), followed by Māori males aged 25–44 Signifi cant geographic differences were (4.77%) and Māori females aged 65+ (4.09%). found. The highest rate was in the West For males, Māori dispensing rates were Coast: 2.1 times the lowest rate in Counties higher in each age bracket except 65+, Manukau. where European/other had the higher rate. Drug classes For females, rates were highest for Māori Figure 5 shows the total prescriptions by between ages 25 and 64, and highest for antipsychotic class in New Zealand over European/other in the 15–24 and 65+ ranges. the years 2012 to 2015. The total number In 2015, the greatest ethnic difference in of prescriptions for antipsychotics has dispensing rates was between Māori and increased from 439,000 in 2012 to 501,000 Asian males. Māori males received antipsy- in 2015, a 14% increase in three years. chotics at 7.2 and 5.0 times that of Asian There was a 15.2% increase in atypical anti- males in the 25–44 and 15–24 age brackets, psychotics and a 6.1% increase in typical respectively. For people aged 45+, females antipsychotics. received more antipsychotics than males Quetiapine and olanzapine are the across all ethnicities. two drugs that largely accounted for the

Figure 3: Percentage of males dispensed antipsychotics in 2015, by age bracket and ethnicity.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 63 www.nzma.org.nz/journal ARTICLE

Figure 4: Antipsychotic use by DHB in 2015.

Note the all-ages DHB populations were used instead of the 15+ population, so rates are lower. increasing numbers prescribed atypical anti- 13%, which was a slight drop from 2012. psychotics; they increased by 24% and 22%, Māori and Pacifi c males were dispropor- respectively, over the three-year period. tionately prescribed clozapine. Clozapine Dispensing of risperidone and clozapine was prescribed at 3.9 times for Māori decreased by 4.7% and 14% over this period. and 2.3 times for Pasifi ka compared to Two drugs accounted for 69% of the typical European/Other males in 2015. Māori and antipsychotics in 2015: haloperidol (49%) Pacifi c females were also more likely to and levopromazine (20%). be prescribed clozapine. Clozapine was Atypical antipsychotics accounted for 87% prescribed at 2.7 times for Māori and 2.0 of the total antipsychotics dispensed in 2015 times for Pasifi ka compared to European/ and typical antipsychotics accounted for Other females in 2015.

Figure 5: Total prescriptions by antipsychotic class in New Zealand 2012–2015.

Antipsychotic medication included ‘typical’ antipsychotics, ‘atypical’ antipsychotics and lithium carbonate. ‘Typical’ antipsychotics: butyropherones (haloperidol and haloperidol decanoate), phenothiazines (chlorpromazine hydrochloride, fluphenazine decanoate, levomepromazine maleate, pericyazine, pipothiazine palmitate and trifluoperazine hydrochloride) and thioxanthines (flupenthixol decanoate, zuclopenthixol decanoate and zuclopenthixol hydrochloride). ‘Atypical’ antipsychotics: clozapine, aripiprazole, olanzapine, paliperidone, quetiapine, risperidone, ziprasidone and amisulpride (a benzamide).

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 64 www.nzma.org.nz/journal ARTICLE

likely to be prescribed antipsychotics than Discussion European/Other people (1.87% vs 3.15%), National rates although Pacifi c rates are increasing at the The number of New Zealanders taking same rate as European/Other. Asian people antipsychotic medication has risen to 1 in were far less likely to be prescribed anti- 36 New Zealand adults (2.81%) in 2015, a psychotics than other ethnicities (0.86% in 49% increase since 2008. Overall antipsy- 2015). Māori and Pacifi c males and females chotic use varies greatly among countries. were prescribed clozapine at rates that were Hálfdánarson et al2 reported rates in 2014 disproportionately higher than the total were highest in Taiwan at 7.8% and lowest antipsychotic rates. in Lithuania (1.2%) and Colombia (0.3%). Gender, age and ethnicity In Australia rates were 1.7% in 2014, and in European/Other and Asian females the US publicly insured patients had a rate received antipsychotic medication at 1.16 of 4.0%. The New Zealand rate therefore and 1.25 times the male rate, respectively. appears to sit in the middle of countries However, the opposite was true for Māori where data is available. Our rate of increase and Pacifi c people, where males received appears quite high. The Australian increase antipsychotics at 1.14 and 1.13 times the was 36% between 2006 and 2014, for female rates, respectively. Overall, females example. were prescribed antipsychotics at 1.10 times Antipsychotic drug classes the male rate. Māori males aged 25–44 The absolute number of prescriptions received more antipsychotics than any increased by 13% in three years, but the other male subgroup. For Māori females, proportion of the total prescriptions made European/Other, Asians and Pacifi c people, up by each drug class is stable. Atypical those aged 65+ received antipsychotics at antipsychotics increased from 86.4% in 2012 the highest rate. It is likely that much of to 87.4% in 2015. Typical antipsychotics this prescribing would be off-label low-dose decreased from 13.6% in 2012 to 12.6% of medication for sleep. the total prescriptions in 2015. Gender and age differences in prescribing Quetiapine, olanzapine, clozapine and have been reported in other countries. risperidone made up 93% of all atypical Trifi ro et al,6 examining prescribing patterns antipsychotics in 2015. Quetiapine and olan- in Italian general practitioners, found that zapine largely account for the increasing women and patients aged over 65 were numbers of atypical antipsychotics, with more likely to receive an antipsychotic medi- 24% and 22% increases in dispensing cation. Marston et al7 reported that females, from 2012–2015, respectively. Risperidone older people and people from the most dispensing decreased by 4.7% over the three deprived areas in the UK were more likely to years. be prescribed an antipsychotic. A similar pattern was reported in Hálfdánarson et al2 reported that 48.6 Australia by Stephenson et al.5 They noted out of 1,000 New Zealanders (4.9%) aged a 218% increase in atypical antipsychotic 65+ (compared with 3.0% aged 20-64) were dispensing from 2000 to 2011. The drugs dispensed an antipsychotic in 2014. Taiwan showing the largest increases were queti- had the highest use with 146/1,000 (14.6%) of apine and risperidone, but olanzapine was those aged 65+ receiving an antipsychotic in the most prescribed atypical at 36%. In 2013. The number of Australians prescribed contrast, there was a 61.2% decrease in the an antipsychotic in 2014 rose from 1.8% for dispensing of typical antipsychotics. 20–64 year-olds to 3.3% for those aged 65+. Ethnicity Geographic distribution Signifi cant ethnic differences were Prescribing trends differed geographically. evident. Māori were more likely to be For example, West Coast DHB prescribed prescribed antipsychotics than European/ antipsychotics at more than double the rate Other people (3.37% vs 3.15% in 2015), of Counties Manukau DHB in 2015. The and the rate of prescribing for Māori is reasons for this variability in prescribing are increasing faster than other ethnic groups: unknown. It seems unlikely that numbers of 60% in seven years. Pacifi c people were less individuals suffering from psychoses vary

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 65 www.nzma.org.nz/journal ARTICLE

by up to two-fold across different DHBs. are often prescribed ‘off-label’ for sleep and Differences reported more likely refl ect local anxiety,9 which probably explains the high prescribing practices. rate in older women. Generalisability Our results suggest some obvious ques- The results of this study are generalisable tions. Why are rates of antipsychotic drug to New Zealand and most other developed prescribing rising steadily? There is no countries, especially those with a colonised evidence that rates of psychosis in New native population and other diverse ethnic Zealand are increasing so the rise appears populations. to be related to other indications—low doses for anxiety and sleep and high doses Limitations for behavioural control perhaps? Why do Although a patient is prescribed a medi- Māori and Pacifi c Island males aged 25–44 cation, they may not have taken it. It is likely receive more antipsychotics than other that many antipsychotics are prescribed males, and why are they much more likely off-label for sleep (particularly low dose to receive clozapine? Are they more likely to quetiapine). be psychotic or is their ethnicity part of the reason? Conclusion These questions need to be explored and The prevalence and distribution of anti- ongoing systematic monitoring of antipsy- psychotic prescribing appears somewhat chotic prescribing is of major public health arbitrary. Rates of prescribing vary dramat- importance. Our data suggest we need to ically across countries and generally appear more critically examine when and why we to be increasing. In New Zealand, rates vary prescribe antipsychotics. Given their toxicity by geographical location, ethnicity, gender and the lack of evidence around much of the and age. This arbitrariness is concerning. prescribing we may need to look at health Antipsychotics have signifi cant adverse policy changes to ensure more rational use effects, and long-term safety and/or effec- of these drugs. tiveness data are lacking.8 Antipsychotics

Competing interests: Nil. Acknowledgements: Sam Wilkinson wishes to thank the University of Otago, School of Medicine for the support to carry out this research as part of the Trainee Intern elective. Matthew Dwyer from the Ministry of Health provided invaluable assistance in obtaining the relevant data. Funding for some of the data sourcing was provided by the Department of Psychological Medicine, University of Otago, Christchurch. Author information: Sam Wilkinson, Canterbury District Health Board, Christchurch; Roger T Mulder, Department of Psychological Medicine, University of Otago, Christchurch. Corresponding author: Roger Mulder, Department of Psychological Medicine, University of Otago, PO Box 4345, Christchurch. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2018/vol-131-no-1480- 17-august-2018/7665

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 66 www.nzma.org.nz/journal ARTICLE

REFERENCES: 1. Verdoux H, Tournier M, Mental Health Population. in UK primary care: a Begaud B. Antipsychotic Updated in November cohort study. BMJ Open. prescribing trends: a 2016. Received 28/3/2017. 2014; 4:e006135. review of pharmaco-ep- 5. Stephenson CP, Karanges E, 8. Murray RM, Quattrone idemiological studies. McGregor IS. Trends in the D, Natesan S, et al. Acta Psychiatr Scand. utilisation of psychotropic Should psychiatrists be 2010; 121:4–10. medications in Australia more cautious about the 2. Hálfdánarson O, Zoega from 2000 to 2011. Aust NZ long-term prophylactic H, Aagaard L, et al. J Psychiatry. 2013; 47:74–87. use of antipsychotics? Br J International trends in 6. Trifi ro G, Spina E, Brignoli Psychiatry. 2016; 209:361–5. antipsychotic use: A study O, et al. Antipsychotic 9. Carton L, Cottencin O, in 16 countries, 2005–2014. prescribing pattern among Lapeyre-Mestre M, et al. Eur Neuropsychophar- Italian general practi- Off-Label Prescribing of macol. 2017; 27:1064–76. tioners: a population-based Antipsychotics in Adults, 3. Ministry of Health. study during the years Children and Elderly Pharmaceutical Collection. 1999–2002. Eur J Clin Individuals: A Systematic Job number 2015–1601. Pharmacol. 2005; 61:47–53. Review of Recent Prescrip- Extracted 22/4/2016. 7. Marston L, Nazareth I, tion Trends. Curr Pharm 4. Ministry of Health. Petersen I, et al. Prescrib- Des. 2015; 21:3280–97. Analytical Services ing of antipsychotics

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 67 www.nzma.org.nz/journal ARTICLE

First, do no harmony: an examination of attitudes to music played in operating theatres Anantha Narayanan, Andrew R Gray

ABSTRACT AIM: It is common for music to be played in operating theatres around the world. The benefits or harms of this music on practitioners and patients are not well known. The aim of this paper is to examine the attitudes and perceptions of theatre sta towards the presence of music during operations on the theatre environment. METHOD: This observational study was conducted in a single-centre at a tertiary teaching hospital in New Zealand. Over a two-week period in 2014, the entire theatre sta population were sent an online survey. We gathered data on demographics (eg, age, gender, theatre role), perceived frequency of music being played in operating rooms, types of music, disposition to music in di erent situations, the perception of an e ect on the theatre team, and on with whom the decision to have music played lay with. Appropriate statistical analyses were applied. RESULTS: A hundred and six responses were received (45% response rate). 98% of people said that music is played in their operating rooms, with 50% saying it was played more than 50% of the time. 60% liked having it, whereas 30% disliked it. Easy Listening was the genre most played closely followed by Pop and Classical. Easy Listening was the most preferred and classical music the second most. Music was preferred for longer, more familiar and non-urgent procedures, at a low to medium volume, preferably from a CD or MP3 player, though 84% felt it was a distraction in a crisis situation. Surgeons were the most empowered group, and anaesthetic technicians the least when it came to choosing music. Most respondents felt music improved calmness, overall mood, overall team performance and surgeon’s performance, though worsened communication. CONCLUSION: This study, while limited in size, demonstrates that theatre sta generally like music in the operating theatre, and most believe it has a positive impact on several aspects of the theatre environment, though a negative influence on communication. There have been no clinical impact studies on patient and this could be an avenue for future research.

As far as we are aware, objective data on n operating theatres (OTs) around the the effect of having music playing in the world, it is commonplace for music to be theatre has not been reported. The literature Iplayed on the sound system both before in this area currently comprises studies that and during the procedure. Studies conduct- have collected subjective survey answers ed in Israel and Scotland found music was and opinions from theatre staff. Ullman et regularly played in 63–72%1,2 of operating al, found that two-thirds of the doctors and theatres. However, there is disagreement in nurses they surveyed believed that music the literature about the benefi ts or harms has a favourable effect, and 77% reported it that music can have on the theatre environ- made them feel calmer and more effi cient. ment. Moreover, 63% felt that music improved communication between staff.1

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 68 www.nzma.org.nz/journal ARTICLE

Some studies show that there may be an Survey increase in a surgeon’s performance when music is played, particularly if the music is Demographics Demographic (age, sex, ethnicity) as well of a familiar or rhythmic nature.3,4 as their speciality and role in OT were Music can also impact on patients in collected. We asked about the number theatre. Surgery is a very stressful time for of hours spent in an OT per week as well patients, and patients frequently experience as the number of years they have spent anxiety before and after an operation. It is working in OTs. Finally we enquired about possible that music could be benefi cial to the the frequency with which the participant patient in these circumstances. A number listened to music in their free time. of studies suggest that peri-operative music may increase patients’ pain thresholds, Music – the survey decrease anxiety scores and even reduce We asked participants to estimate the dosage for induction.5,6 percentage of time they would work in an However, in a survey of anaesthetists in operating theatre and music is played. We Glasgow, one quarter of those surveyed felt inquired as to whether they liked this or that music reduced vigilance and impaired they thought it was appropriate. communication, and one in 10 felt that the Genres of music was asked in two ways: music could distract them from alarms. which is personally preferred and which Over half thought that music was distracting are played most frequently. We listed to during times of crisis.2 eight genres: Classical, Jazz, Rock, Hip-Hop/ In light of these confl icting accounts of Rap/R&B, Folk, Pop, Reggae and Easy music in the operating theatre, in this study Listening. There was an option to put No we aimed to explore the experiences of Preference, or to specify another genre. We and attitudes to music from a wide range inquired to preferred volume and format of operating theatre personnel in a New (eg, CDs vs radio). Zealand context. The next part of the survey related to the disposition to listen to music in different Methods situations (length of procedure, amount of experience with the procedure, emergency, Sampling time of day). This was a single-centre study, in a tertiary We asked about the participants’ percep- teaching hospital in New Zealand, servicing tions on the effect of music on seven an urban and rural population of around different aspects of the theatre environment 120,000. The population of interest was all using a fi ve-point scale; communication with staff who worked in any operating theatre in other staff, individual vigilance, calmness this centre during the two-week period that and focus, mood, the theatre team’s perfor- the survey was run (12 June – 26 June 2014). mance and the surgeon’s performance. We This included all anaesthetists (consul- also asked about views on the effect of music tants and registrars), surgeons and trainees on the patient. from all surgical specialty departments, theatre nurses and anaesthetic techni- Finally we asked who made the decisions cians. The email addresses of all the staff about music in theatre, and whether the were obtained from the respective author- status quo was acceptable. ities. The entire theatre staff population of Procedure 234 were sent an invitation to participate The entire current theatre staff population in the study. Information was provided were invited to participate in this study and consent was taken on the fi rst page of in July 2014. An initial email was sent to the online survey. This study was granted participants, which included information category A ethics by the University of Otago about the study and a link to the online Human Ethics Committee (F13/016), and survey. The survey was open two weeks consultation was undertaken with the Māori and a further two reminder emails were research committee. sent six to seven days apart. The survey

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 69 www.nzma.org.nz/journal ARTICLE

was designed and run on SurveyMonkey®. graphic, OT workload and music-related Information collected from the survey was variables. Analyses were performed using kept confi dential and participants were not Stata 13.1 and p<0.05 was considered statis- identifi able. No incentive was offered for tically signifi cant. participation. Data analysis Results Summary statistics were calculated for Demographic information all outcomes of interest. Tests of proportion A hundred and six theatre staff responded were used to compare the fraction feeling to the survey, a response rate of 45%. Five music improved some attribute versus those responders did not fi ll-in the survey to feeling it worsened that attribute. Logistic completion. regression was used to model associations Slightly more than half of respondents between demographic, OT workload, and were aged 45 and above (55%) and male music-related variables and being happy (57%). All four roles were represented with with the current situation regarding music at least 10 respondents but there were rela- in the OT. Kruskal-Wallis tests were used tively few anaesthetic technicians compared to compare liking music and fi nding music to the other roles. The interpolated median appropriate in the OT between demo- level of experience was 14.4 years and

Table 1: Demographic and behavioural data.

Variable Total 101

Sex Male 58 Female 43

Age <35 18 35–44 27

45–54 27

55+ 29

Theatre role Anaesthetist 29 Surgeon 37

Nurse 25

Anaesthetic technician 10

Experience in OT <6 14 (years) 6–15 39

>15 48

Time per week in 0–20 40 the OT (hours) 21–40 42

>40 19

Listen to music in All the time 9 freetime O en 48

Sometimes 21

Occasionally 22

Never 1

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 70 www.nzma.org.nz/journal ARTICLE

the interpolated median time per week in Preferences and experiences of OT was 26.3 hours. Over half of respon- music in OT dents listened to music often or all the time The most preferred forms of music were in their spare time (57%) with only one Easy Listening (54%), Classical (36%), respondent never listening to music (1%). Pop/1980s (28%), Rock (27%), and Jazz/Blues Respondent attitudes (20%). Least popular were Reggae (6%) and 98% of respondents indicated music is Hip-Hop/Rap/R&B/Dance/Electronica (9%). played in the operating theatres in which The most commonly heard forms of music they work, 49% reported that music was were Easy Listening (52%), Pop/1980s (48%), played over 50% of the time. Classical (45%), and Rock (30%). Despite While over 70% of respondents reported being preferred by one in fi ve, Jazz/Blues liking music or being neutral to it (72%), was one of the least commonly heard forms 28% did not. However, a smaller percentage (3%), along with Reggae (0%), Hip-Hop/ (6%) felt music was inappropriate in the Rap/R&B/Dance/Electronica (3%), and folk operating theatre. Roughly half of respon- (4%) (Figure 1). dents were happy with how music was There was a strong preference towards managed (52%), with roughly a quarter music being played at either a low (61%) each indicating they were unhappy (23%) or medium (39%) volume. Radio was not a or neutral on the issue (26%). Music was popular choice for the source of music (2%). more preferred for longer, more familiar, Respondents reported that surgeons non-urgent procedures than shorter, less were the most likely group to be involved familiar, urgent procedures and music was in choosing whether music was played felt to be a distraction in a crisis by 84% of (91%) as well as what music is played (96%), respondents. There did not appear to be any followed by anaesthetists then nurses, strong preference regarding time of day. technicians and patients (Figure 2). One in Music was seen by respondents as more fi ve respondents indicated that patients are favourable to patients (53% positive and 2% involved in the decision whether music is negative) than to the theatre staff (38% felt played (23%), what music is played (19%), they performed worse and 5% better). and 23% said they regularly asked patients about the type of music they would like.

Figure 1: Genres of music.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 71 www.nzma.org.nz/journal ARTICLE

Figure 2: Decisions about music in the OT (%).

Table 2: Specifi c effects of music (%). Effect on Signi�i- Slightly No Slightly Signi�i- cantly wors- effect im- cantly wors- ened proved im- ened proved Own communication 3 27 64 7 0

Own vigilance 1 10 79 11 0

Own calmness 1 8 43 46 3

Own focus 4 13 58 25 1

Mood in OT 0 3 12 64 22

Team’s performance overall 2 4 44 44 7

Surgeon’s performance 0 6 59 34 2

Specific e ects of music Fifty-three percent of respondents thought Ignoring those responding that music that music had a positive effect on patients had no effect, music was seen as improving peri-operatively. calmness (84%, more than 50% of the Associations by demographic non-neutral responses, test for proportion None of the demographic predictors were P<0.001), mood (97%, P<0.001), the team’s associated with liking music (all Krus- overall performance (89%, P<0.001) and kal-Wallis p≥0.066). There was only one the surgeon’s performance (86%, P<0.001). predictor for fi nding music appropriate in However, 81% thought that communi- the operating theatre with higher ratings of cation was worsened (P<0.001). There was appropriateness among those who listened no statistically signifi cant evidence of a to music more in their free time (Krus- direction of effect for vigilance or own focus. kal-Wallis p=0.006).

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 72 www.nzma.org.nz/journal ARTICLE

Some signifi cant associations were Our study shows similar preference to identifi ed for satisfaction with the current Dalton’s recommendations in the Journal of practice. Younger respondents were more Occupational Ergonomics for performance likely to be happy with current practice in such tasks as comprehension, attention (logistic regression p=0.034) as were those and vigilance. He suggests that a moderate who spend more time in the operating volume of music may facilitate perfor- theatre each week (logistic regression mance of tasks that require high levels of p=0.014). concentration and attention, however it can be distracting during comprehension Discussion tasks (such as communication).7 Arousal theory proposes that vigilant performance This study has collected self-reported is optimal when background music is data on the way in which music is used and discontinuous and contains elements of perceived in the operating theatre envi- uncertainty,8 corroborated in a study of ronment by staff. Music is played often in participants using the Da Vinci surgical operating theatres, with Easy Listening robot, where performance was improved being the most commonly played and while listening to either Hip-Hop or preferred genre. Music was preferred Jamaican music.4 This is a potential avenue for longer, more familiar and non-urgent for future interventional research. procedures, at a low to medium volume, preferably from a CD or MP3 player, though A variety of studies have demonstrated a majority felt it was a distraction in a that music can reduce patients’ peri-op- 5,6 crisis situation. Surgeons were the most erative fear and anxiety, and one paper empowered group, and anaesthetic tech- proposed a positive relationship between nicians the least when it came to choosing patients’ satisfaction with their surgical music. Respondents felt that music generally experience after listening to music peri-op- 5 improved calmness, overall mood, overall eratively. A small randomised control trial team performance, the surgeon’s perfor- in America and Lebanon suggests that even mance and the patient’s experience. Most when ambient OT noises are controlled for, respondents felt it worsened communi- having intra-operative music decreased 9 cation, which could be a potential safety a patient’s propofol requirements, while concern. another proposed that there was a reduction in post-operative intubation time.10 With regards to perceptions of music in the operating theatre, Ullman et al, found that two-thirds of the doctors and Limitations nurses they surveyed believed that music The data collected in this survey gives an has a favourable effect, and 77% reported insight into the views and perceptions of it made them feel calmer and more effi - theatre staff towards music. However, there cient. However contrary to our fi nding, in are no objective measurements of music Ullman’s study population, 63% felt that frequency, type, effect nor a quantifi cation music improved communication between of clinical endpoints. 1 staff. A survey of 200 anaesthetists in Our study is likely to be impacted by Glasgow found that 62.5% of anaesthetists responder bias, where those who have liked music in theatre. However, 26% of stronger opinions would be more likely to the sample felt that music reduced their take the time to fi ll in the questionnaire. vigilance and impaired communication, There was an under-representation of and 11.5% felt the music had the potential theatre nurses and theatre technicians— to distract them from the alarms. Fifty-one who are staff who would spend a greater percent felt the music was distracting when proportion of their clinical work time within 2 a problem was encountered during surgery, the operating theatres—and this may have compared to 83% in our study. caused some skew particularly in the ques- tions regarding the decisions around music.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 73 www.nzma.org.nz/journal ARTICLE

communication and is viewed negatively Conclusion in the context of critical theatre situations Music is a common and favourable part and urgent procedures. There have been no of the theatre environment; in this demo- impact studies of music on clinical outcomes graphic Easy Listening and Classical genres or perceived experience for patients under- were the most popular. Music is seen as going surgery, and this would be an area for improving calmness, mood, surgeons’ and future research. overall team performance, but can impair

Competing interests: Nil. Acknowledgements: Thanks to Dr Mathew Zacharias, Dunedin School of Medicine, University of Otago, Dunedin for his contribution to this research. Author information: Anantha Narayanan, Dunedin School of Medicine, University of Otago, Dunedin; Andrew R Gray, Dunedin School of Medicine, University of Otago, Dunedin. Corresponding author: Anantha Narayanan, Dunedin School of Medicine, University of Otago, Dunedin. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2018/vol-131-no-1480- 17-august-2018/7666

REFERENCES: 1. Ullmann Y, Fodor L, surgical performance. 8. Corhan CM, Gounard Schwarzberg I, Carmi N, Surgical innovation BR. Types Of Music, Ullmann A, Ramon Y. The 2010; 17(4):306–11. Schedules Of Background sounds of music in the 5. Johnson B, Raymond Stimulation, And Visual operating room. Injury S, Goss J. Perioperative Vigilance Performance. 2008; 39(5):592–97. Music or Headsets to Perceptual and motor 2. Hawksworth C, Asbury Decrease Anxiety. Journal skills 1976; 42(2):662–62. AJ, Millar K. Music in of PeriAnesthesia Nursing 9. Ayoub C, Rizk L, Yaacoub C, theatre: not so harmonious. 2012; 27(3):146–54. Gaal D, Kain Z. Music and A survey of attitudes to 6. Wang S-M, Kulkarni L, ambient operating room music played. Anaesthe- Dolev J, Kain ZN. Music noise in patients under- sia 1997; (52):77–83. and Preoperative Anxiety: going spinal anesthesia. 3. Allen K, Blascovich A Randomized, Controlled Anaesthesia and Analgesia J. Effects of Music on Study. Anesthesia & Anal- 2005; 5(100):1316–9. Cardiovascular Reac- gesia 2002; 94(6):1489–94. 10. Twiss E, Seaver J, McCaf- tivity Among Surgeons. 7. Dalton BH, Behm DG. frey R. The effect of music Jama 1994; 272(11). Effects of noise and listening on older adults 4. Siu KC, Suh IH, Mukherjee music on human and task undergoing cardiovascular M, Oleynikov D, Stergiou performance: A system- surgery. Nursing in critical N. The effect of music on atic review. Occupational care 2006; 11(5):224–31. robot-assisted laparoscopic Ergonomics 2007; 7:143–52.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 74 www.nzma.org.nz/journal ARTICLE

An audit of regular medication compliance prior to presentation for elective surgery Daniel Wood, Nicholas Lightfoot

ABSTRACT AIM: Patients presenting for elective surgery frequently omit one or more regular medications. There is evidence that withholding some medication classes may lead to harm in the peri-operative period. We aim to quantify medication non-compliance in an adult, elective surgical population and identify the reasons for this non-compliance. METHOD: Data was collected over a four-week period at an elective surgical facility on adult patients presenting for surgery. Surgery and anaesthesia type, the names of regular medications and reasons for omission were recorded. This information was matched to pharmacy data for six months prior to the audit. RESULTS: Six hundred and ninety-two patients underwent surgery during the audit period. Data for 631 patients (91%) was collected. Three hundred and sixty-six (58%) were prescribed regular medications. A er exclusion of diabetic medication, 109 patients (30%) had missed one or more regular medications. Thirty- four patients (31%) were non-compliant with medication due to fasting. Twenty-seven patients (25%) were confused. Ten percent and 21% of patients prescribed beta-blockers or calcium channel blockers respectively were non-compliant on the day of surgery. CONCLUSIONS: A significant number of patients were non-compliant with their regular medication on the day of surgery. This represents a preventable source of risk that could be addressed by improved patient education or improved organisational pathways.

pproximately half of all adult patients admission. The reason for this non-com- presenting for non-cardiac surgery pliance may be misunderstandings arising are prescribed regular medications from their fasting status, or from general A 1,2 for indications unrelated to their surgery. uncertainty over which medications should The indications are diverse, but between be taken. This is compounded by the lack of 40–50% of these may be cardiovascular standardised advice provided through the drugs, including anti-hypertensive, statins, anaesthesia clinic or the admitting surgical anti-anginal or anti-platelet agents.1,2 One services. study reported 49% of all medications were While some classes of medication should omitted on the day of surgery with the most be routinely withheld in the fasting patient, 1 common reason given as patient fasting. withholding others may be harmful.3,4 There is anecdotal evidence that patients Contraindications to administration may presenting for surgery from the community exist for oral diabetic medications in the to the Manukau Surgical Centre (MSC) in fasted patient, anticoagulants and anti- Auckland, New Zealand, frequently omit platelet medications. All regular medications their regular medications prior to their should be assessed on a patient by patient

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 75 www.nzma.org.nz/journal ARTICLE

basis prior to the day of surgery, reasons for attending clinic, a non-medical member any planned omissions should be docu- of the admitting service administrative mented and instructions clearly provided to staff contacted the patient by telephone the patient.5 and written information was mailed to the In line with previous guidance we patient. This advice was not standardised consider the standard of treatment is that all across services at the time of this review, patients receive their regular medication on however consultation was available with the the day of surgery unless a specifi c contra- anaesthesia service through the anaesthesia indication has been identifi ed.1 To assess clinic or the on-call Specialist Anaesthetist. compliance with this recommendation we Pharmacies within the Auckland area reviewed adult patients presenting to MSC provide medication dispensing information, on the day of surgery over a four-week which is accessible through a patient’s elec- period. We assessed what proportion had tronic record. The CMDHB Clinical Decision not taken their usual regular medication on Support Service electronically collected all the day of surgery and assessed the reasons community pharmacy dispensing records for non-compliance. for the six months prior to the conclusion of the audit for patients who were enrolled Method in the audit. This information was matched to collected data using a patients National The audit was registered with the Health Index (NHI) number. Counties Manukau District Health Board (CMDHB) Research Offi ce (Application 102) Descriptive statistics were used to describe and National Health and Disability Ethics the demographics of the audited population. Committee (HDEC) out of scope status was This included the number and percentage formally confi rmed (20 July 2016). of patients taking regular medications, those missing non-diabetic medications, the The audit was conducted over a four-week reasons given for omission and those who period at the MSC from 26 September 2016 subsequently received their medication to the 21 October 2016. This period was before transfer to the operating theatre. expected to be adequate to provide an inci- Pharmacy dispensing data was used to dence of patient non-compliance. A data determine the number of patients on each of collection form was inserted in the notes several medication classes and to calculate of every adult patient (18 years or older) the percentage who omitted one of these presenting for a procedure (excluding medications. endoscopy) on day of surgery. Pre-operative nurses fi lled in the form for each patient based on information in the notes and direct Results patient enquiry. During the audit period, 692 patients Samples of completed data forms and underwent elective surgery. Nurses a contact number for any problems were completed data forms for 91% (631 of 692) provided to the nurses in the main pre-oper- of patients. Demographic information is ative area. Data forms were collected when presented in Table 1. Information for the the patient proceeded from the pre-oper- remaining 61 patients is not included in the ative area to the operating theatre, then subsequent analysis. stored in a secure fi ling cabinet within the Of the patients for whom data was department of anaesthesia until data entry available, 58% (366 of 631) were prescribed and analysis took place. one or more regular medications. Of Medication advice was provided to those on regular medications, 30% (109 patients at the anaesthesia pre-admission of 366) missed one or more of these drugs clinic or via the admitting service clinic (excluding diabetic agents). Of the patients prior to admission. An electronically who missed one or more non-diabetic generated form was provided to each medications, 24% (26 of 109) had been patient, medication instructions were typed instructed not to take their medication in free text but were not standardised across (Figure 1). Thirty-one percent (34 of 109) services or between patients. For those not were non-compliant with medication due

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 76 www.nzma.org.nz/journal ARTICLE

Table 1: Patient demographics.

Patient demographic data Number (%)

Age 18–40 years 139 (22%)

41–60 years 192 (31%)

61–80 years 239 (38%)

81+ years 59 (9%)

Gender Female 391 (62%)

Male 240 (38%)

Anaesthesia General anaesthesia 400 (63%)

Monitored anaesthesia care 231 (37%)

Surgical sub-speciality General surgery 102 (16%)

Orthopaedic surgery 82 (13%)

Hands 65 (10%)

Plastics 90 (14%)

Gynaecology 98 (16%)

ORL 51 (8%)

Urology 25 (4%)

Renal 2 (<1%)

Ophthalmology 116 (18%)

RCRI surgical severity Low 455 (72%)

Intermediate/high 175 (28%)

(RCRI, revised cardiac risk index.)

Figure 1: Summary of medication compliance and causes.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 77 www.nzma.org.nz/journal ARTICLE

to fasting. A further 25% (27 of 109) were unsure of their instructions or forgot their Discussion medication. During our study period, 30% (109 of 366) of elective patients presenting on the day From pharmacy dispensing records of surgery were non-compliant with their the percentage of patients omitting each regularly prescribed non-diabetic medica- of several important medication classes tions. This included non-compliance rates of was calculated (Table 2). Ten percent (7 of between 10–23% for important cardiovas- 70) of patients prescribed beta-blockers cular medications. were non-compliant on the day of surgery whereas 20%, 21% and 23% were non-com- After the exclusion of diabetic medi- pliant with their angiotensin II receptor cations, 56% (61 of 109) of all regular blocker (ARB), angiotensin converting medications omitted were due to patient enzyme inhibitor (ACEi) and calcium fasting, confusion over instructions or channel blocker (CCB) respectively. forgetting to take medications. Only 24% (26 of 109) specifi cally reported having been Across these medication classes, the told to not take their medication. There was majority of omissions occurred due to substantial variability in practice regarding fasting status 36% (19 of 53), the patient whether patients who had missed a dose of being unsure of the instructions provided beta-blocker, CCB, ACEi or ARB subsequently or forgetting to take medication 36% (19 of received their medication prior to theatre. 53). While many clinicians may routinely instruct patients to omit their ACEi or Our rates of non-compliance are compa- ARB, we found only 10% (3 of 29) of the rable with previous studies (49%) where patients who were non-compliant with these pre-operative medication instructions reported being instructed to omit these were given in a non-standardised manner.1 medications. Where patient medication instructions were generated from electronic medical Of those who had omitted one of their records compliance rates of 84% have regular cardiovascular medications, 21% been reported, this increased to 90% when (11 of 53) received an oral dose prior to these instructions were standardised for all transfer to theatre based on the anaesthe- patients.6 tist’s instructions. This included 14% (1 of 7) of patients who had omitted a beta-blocker, Improvements in our current service 29% (2 of 7) omitting ARB, 14% (3 of 22) could include standardisation of instruc- omitting ACEi and 29% (5 of 17) omitting tions and guidelines available to clinicians CCB. and support staff providing pre-operative advice to patients. Specifi c guidelines

Table 2: Non-compliance for specifi c medication classes and causes of non-compliance.

Medication Non-com- Prescribed % Instruct- NBM Unsure/ Other class pliant Non-com- ed not to forgot pliant take β-Blockers 7 70 10 0 3 2 2

CCB 17 79 21 1 5 7 4

ACEi 22 96 23 3 9 7 3

ARB 7 35 20 0 2 3 2

(CCB, calcium channel blocker. ACEi, angiotensin converting enzyme inhibitor. ARB, angiotensin II receptor blocker.)

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 78 www.nzma.org.nz/journal ARTICLE

should be prepared and disseminated to was collected on more than 90% of patients the surgical subspecialty and anaesthesia undergoing surgery during the study period. clinic practitioners. This would reinforce the A possible bias arises from the reliance on importance of the continuation of medica- nursing staff to complete the audit forms tions when there are no contraindications and the additional work this entailed. This and highlight circumstances where discon- would most likely result in a lower number tinuation may be warranted. of omitted medications being recorded, Factors infl uencing patients’ retention hence the results may underestimate the of instructions can be catagorised as those number of patients who are non-compliant. relating to the clinician (use of jargon), Errors in the electronic assessment of those relating to the information (written dispensed medication may introduce bias. or verbal, use of pictures) and patient Patients may no longer have been taking factors (expectations and anxiety levels).7,8 a dispensed medication due to side effects Simple written instructions and pictographs or discontinuation by their supervising increase the rates of patient recall and clinician. Dispensed medication may not compliance.7–9 Utilising simple language, have been recorded if the prescription was repetition and explicit categorisation such fi lled outside of our hospital catchment area. as dividing medications into groups to be These scenarios could lead to inaccuracies in taken or omitted under clear headings has the rates of non-compliance. been shown to improve retention of infor- In conclusion, large numbers of patients 7,10 mation. These principles will be applied to omit medications on the day of elective improve the electronic patient advice form surgery, predominantly due to misunder- in order to enhance retention of instructions standing the pre-operative advice provided. and compliance. Consideration does need A number of these omissions may lead to to be given to variability in patient demo- adverse events. Through standardisation graphics and ethnic differences which exist of advice across surgical specialities and within the South Auckland population, as improving our verbal and written patient medication compliance rates and prefer- instructions we could improve medication ences for receiving medication information compliance and decrease interpatient 11,12 vary for Māori and Pacifi c populations. variability. This may lead to improve- The limitations of this study include its ments in patient-centric outcome measures relatively small sample size compared to the should complications of non-compliance total number of patients presenting each be avoided. Introduction of such measures year for surgery and that only one of the two would require consensus between anaes- elective surgical venues at our district health thesia and referring clinicians such that board was surveyed. Reassuringly, data discrepancies in advice are minimised.

Competing interests: Dr. Lightfoot reports personal fees from Merck - MSD, outside the submitted work. Author information: Daniel Wood, Consultant Anaesthetist, Department of Anaesthesia and Pain medicine, Middlemore Hospital, Auckland; Nicholas Lightfoot, Consultant Anaesthetist, Department of Anaesthesia and Pain Medicine, Middlemore Hospital, Auckland. Corresponding author: Dr Daniel Wood, Consultant Anaesthetist, Department of Anaesthesia and Pain medicine, Middlemore Hospital, 100 Hospital Road, Atahuhu, Auckland, 2025. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2018/vol-131-no-1480- 17-august-2018/7667

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 79 www.nzma.org.nz/journal ARTICLE

REFERENCES: 1. Kluger MT, Gale S, ESC/ESA Guidelines on information. J R Soc Plummer JL, Owen H. non-cardiac surgery: Med 2003; 96:219–22. Peri-operative drug cardiovascular assess- 9. Watson PW, McKinstry B. A prescribing pattern and ment and management. systematic review of inter- manufacturers’ guidelines. European Heart Journal. ventions to improve recall An audit. Anaesthesia. 2014; 35:2383–2431. of medical advice in health- 1991 Jun; 46(6):456–9. 5. ANZCA Guidelines on care consultations. J R Soc 2. Kennedy JM, van Rij AM, the Pre-Anaesthesia Med 2009; 102:235–43. Spears GF, Pettigrew RA, Consultation and Patient 10. Ley P. Memory for medical Tucker IG. Polypharmacy Preparation 2016. information. BrJ Soc Clin in a general surgical unit 6. Pfeifer K, Slawski B, Manley Psychol 1979; 18:245–55. and consequences of drug AM, et al. Improving 11. Treharne J, Richardson A, withdrawal. Br J Clin preoperative medication Neha T, et al. Education Pharmacol. 2000; 49(4):353. compliance with stan- preferences of people with 3. Fleisher LA, Fleischmann dardized instructions. gout: Exploring differences KE, Auerbach AD, et al. Minerva Anestesiol between indigenous and 2014 ACC/AHA guideline 2016 Jan; 82(1):44–9. nonindigenous peoples on perioperative cardio- 7. Vetter T, Downing M, from rural and urban vascular evaluation and Vanlandingham S, et locations. Arthr itis Care management of patients al. Predictors of patient Res 2017 May 2 [EPub undergoing noncardiac medication compliance ahead of print]. surgery: a report of the on the day of surgery and 12. Gu Y, Warren J, Kennelly American College of the effects of providing J, et al. Cardiovascular Cardiology/American Heart patients with standardized disease risk management Association Task Force on yet simplifi ed medication for Māori in New Zealand Practice Guidelines. Circu- instructions. Anesthesiolo- general practice. J lation. 2014; 130(24):e278. gy 2014 July ; 121(1):29–35. Prim Health Care 2014 4. Kristensen SD, Knuuti 8. Kessels RP. Patients’ Dec 1; 6(4):286–94. J, Saraste A, et al. 2014 memory for medical

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 80 www.nzma.org.nz/journal VIEWPOINT

Genomic medicine must reduce, not compound, health inequities: the case for hauora-enhancing genomic resources for New Zealand Stephen P Robertson, Jennie Harré Hindmarsh, Sarah Berry, Vicky A Cameron, Murray P Cox, Ofa Dewes, Robert N Doughty, George Gray, Jessie C Jacobsen, Albert Laurence, Elizabeth Matisoo-Smith, Susan Morton, Andrew N Shelling, Dianne Sika-Paotonu, Anna Rolleston, Jonathan R Skinner, Russell G Snell, Andrew Sporle, Cristin Print, Tony R Merriman, Maui Hudson, Philip Wilcox

ABSTRACT Precision medicine seeks to draw on data from both individuals and populations across disparate domains to influence and support diagnosis, management and prevention in healthcare at the level of the individual patient and their family/whānau. Central to this initiative is incorporating the e ects of the inherent variation that lies within genomes and can influence health outcomes. Identifying and interpreting such variation requires an accurate, valid and representative dataset to firstly define what variants are present and then assess the potential relevance for the health of a person, their family/whānau and the wider community to which they belong. Globally the variation embedded within genomes di ers enormously and has been shaped by the size, constitution, historical origins and evolutionary history of their source populations. Māori, and more broadly Pacific peoples, di er substantially in terms of genomic variation compared to the more closely studied European and Asian populations. In the absence of accurate genomic information from Māori and Pacific populations, the precise interpretation of genomic data and the success and benefits of genomic medicine will be disproportionately less for those Māori and Pacific peoples. In this viewpoint article we, as a group of healthcare professionals, researchers and scientists, present a case for assembling genomic resources that catalogue the characteristics of the genomes of New Zealanders, with an emphasis on peoples of Māori and Polynesian ancestry, as a healthcare imperative. In proposing the creation of these resources, we note that their governance and management must be led by iwi and Māori and Pacific representatives. Assembling a genomic resource must be informed by cultural concepts and values most especially understanding that, at a physical and spiritual level, whakapapa is embodied within the DNA of a person. Therefore DNA and genomic data that connects to whakapapa (genealogy) is considered a taonga (something precious and significant), and its storage, utilisation and interpretation is a culturally significant activity. Furthermore, such resources are not proposed to primarily enable comparisons between those with Māori and broader Pacific ancestries and other Aotearoa peoples but to place an understanding of the genetic contributors to their health outcomes in a valid context. Ongoing oversight and governance of such taonga by Māori and Pacific representatives will maximise hauora (health) while also minimising the risk of misuse of this information.

enomic medicine is poised to generate of variation across the human genome,2 cou- substantial health benefi ts for both pled with an expanding capability to defi ne the individual and at the population and measure the subset of genetic variants G 1 health level. The increasingly detailed that impact upon health traits, has driven understanding of the structure and breadth the development of multiple diagnostic tests

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 81 www.nzma.org.nz/journal VIEWPOINT

in recent years. Together with the develop- human populations dispersed around the ment of individualised digital health records world, their demographic histories, inter- and the implementation of linked ‘big data’ actions with their environment and their resources, personalised genomic informa- interactions with other peoples have shaped tion stands as a central pillar of what has their genomes in distinctive and diverse come to be known as ‘precision medicine’.3 fashions.9 Much of this genetic diversity This term, popularised by President Obama remains poorly mapped and understood in his State of the Union address in 2015, em- simply because the scientifi c hegemony bodies the assertion that the inherent vari- of the western world has placed excessive ability of many of the multiple contributing emphasis on the study of numerically and factors that impinge on health will be both economically dominant European and resolvable and measurable. Moreover, the Asian populations.10 As a deeper under- claim is that this approach will guide health- standing of the genetic underpinnings of so care at the level of the individual as well as many disease traits has been defi ned, it has populations in the foreseeable future.4,5 This become increasingly apparent that specifi c view of healthcare is starting to supplant the genetic variants can be considered of greater paradigm of choosing diagnostic, treatment, or lesser importance to health outcomes management and prevention options guided depending on the populations being studied by evidence obtained from the study of and the ancestral history of their genomes.11 populations, but agnostic to the inherent bi- Several notable instances exist where ological variability between individuals and misinterpretation of genetic data has led to subgroups within these populations. signifi cant and adverse health outcomes for 12–14 The accurate interpretation of genetic minority populations. data will be pivotal to the aspiration of The genomes of individuals living in precision medicine and will require precise Aotearoa refl ect the ancestral origins of their and affordable laboratory and analytical forebears. Our population structure and capability over a diverse set of countries pattern of variation continues to be moulded and health jurisdictions. Indeed interna- and infl uenced by the degree of admixture tional umbrella organisations such as the of individuals of different ancestry. An Global Alliance for Genomics and Health important context for Aotearoa New Zealand have begun to systemically address many is the migration of Pākeha following fi rst of these challenges to arrive at solutions European contact. Subsequent waves of that could be generally applicable. Some migration have introduced genomes from issues, however, will remain as distinctly across the globe, most notably from the near local concerns and require the development Pacifi c (Western and Eastern Polynesia) of tailored solutions to make genomic as well as South and South East Asia10,15 medicine accurate, benefi cial, acceptable, resulting in differing levels of admixture ethical and equitable for regional popu- even between groups and complex and lations.6 The promise of emerging health diverse patterns of ancestral self-identi- technologies can be inequitably realised, as fi cation that in some cases do not refl ect the applicability and access often focuses self-declared ethnicities. The net result is a on well-resourced nations and individuals.7 modern-day admixture that is a mosaic of The cost of this technology and the infor- DNA of diverse and multifarious origins. mation systems on which they are founded Depending on the historical origins and could therefore result in genomic medicine genetic adaptations that these genomes have becoming a contemporary manifestation of undergone prior to their arrival in Aotearoa, Hart’s inverse care law.8 they will have been shaped by their migra- A key aspect for the effective application tions across Europe, Asia, Island Southeast 10,16 of genomic medicine is the interpretation Asia and the Pacifi c. Sequence changes of genetic variation in the context of the arise due to many forces, including selective populations from which patients originate.9 pressures imposed by the environments Genomic epidemiology has mapped wide- with which these genomes have inter- spread and pervasive differences in the acted, which may in turn have subsequent nature, distribution and frequency of genetic medical consequences in modern Aotearoa 17 variants in genomes across the globe.2,10 As environments.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 82 www.nzma.org.nz/journal VIEWPOINT

These infl uences will have shaped the population groups are well-studied and genomes of Māori and Pacifi c peoples during understood from analysis of their source their ancestral migrations to Aotearoa—the populations.17 Consequently, there is a last temperate landmass to be inhabited by lesser imperative to study European and human beings. The genomes of Māori and Asian genomes in the local context simply Pacifi c peoples are likely to have acquired because this has already been comprehen- unique variants that have been positively sively achieved elsewhere. In sharp contrast, selected as their tīpuna (ancestors) traversed knowledge of genomic factors that impact diverse geographical regions.18 Additionally on the health of people with Māori and other genetic variants will have increased Pacifi c ancestry is acutely lacking. Despite or decreased in frequency through combi- this, the limited research undertaken to date nations of genetic drift, bottlenecking events indicates that genomes of individuals and and introgression from other populations. populations with Māori ancestry are signifi - Some of these functional genomic adapta- cantly different from western European tions are likely to have signifi cant relevance and Asian genomes.21–23 Moreover, there is for health. Thus it is imperative to under- evidence for differences between Eastern stand these genetic adaptations and how they (Cook Island Māori, Hawaiian, Tahitian and interact with the modern environment to Aotearoa New Zealand Māori) and Western impact on health in contemporary Aotearoa. (Samoa, Tonga, Niue, Tokelau, Tuvalu, The challenge before us is to generate Rotuma) Polynesian genomes and those of 24 capability and knowledge that will benefi t other Pacifi c populations. These differences those peoples in New Zealand who are most include variants at either lower or higher under-represented in the current interna- frequencies than observed elsewhere, tional genomic knowledge base. To avoid together with novel variations that are not perpetuating health inequity, this under-rep- present at all in large catalogues of genomic resentation needs to be addressed since variation derived from other populations 22,23 this knowledge gap will lead to disparities (TM, SR unpublished). Even more impor- in the application and clinical utility of tantly, these studies have shown that some genomic technologies9,11 and reduce the genetic differences may explain some of the capacity for research to generate future differential susceptibility to conditions such genomic medicine strategies relevant to as gout and diabetes that are more common these population groups.11 Further, this in Māori and Pacifi c relative to European 22,23 effort must be conducted locally, as there is populations. Similarly, there is apparent little incentive for globally dominant inter- potential to identify genetic resilience national genomics institutions to undertake factors (variants that confer a protective this work due to the distinctiveness of this effect) for conditions that are under-rep- country’s population genetics in addition to resented in Māori such as rheumatological the cultural concept of DNA as a precious disorders and some cancers. taonga. More importantly, it is an imper- Illustrating these observations, a prevalent ative derived from Article 2 of Te Tiriti and apparently population-specifi c, variant o Waitangi that the rights, interests and in the Samoan population—a so-called taonga of Māori be protected and that this favourable adiposity genetic variant in extends to ancestral lore and whakapapa,19 CREBRF—has recently been associated with hence encompassing DNA and its associated increased rates of obesity, while simul- information.20 Additionally, Article 3, by taneously conferring a protective effect extending to Māori the rights and privileges on the development of type 2 diabetes.22 of British subjects mandates an equal share These data suggest that some novel genetic in the technological dividend offered by factors localised to the South Pacifi c region genomic medicine. contribute distinct pathogenic mecha- New Zealanders with ancestral origins nisms for highly prevalent complex disease 23 located in Western Europe and South East traits. Such data may be relevant for Asia are currently poised to benefi t the both medical treatment and reduction of most from precision medicine incorporating risk through preventative strategies such 13 genomic information. The variation and as the prescription of appropriate drugs structure of the genomes of these major and/or other interventions such as ‘green’ prescriptions.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 83 www.nzma.org.nz/journal VIEWPOINT

Having uncharacterised genetic differ- and cardiac disease.12 Similar inequi- ences embedded throughout the genomes of table outcomes are distinct possibilities in New Zealanders has three principal reper- Aotearoa unless genomic medicine is estab- cussions for their healthcare. The fi rst and lished along with the appropriate resources most immediate concern is that the inter- that facilitate accurate and appropriate pretation of genetic test results could be analysis for all New Zealanders, particularly inaccurate because the reference genomic Māori and Pacifi c peoples. data being used as a comparator is derived A second ramifi cation of our lack of from individuals of entirely different knowledge of the Māori and Pacifi c compo- ancestries. A striking example has been nents of the Aotearoa genome relates to recently recognised during genetic testing understanding the causes and natural carried out under the auspices of The New history of disease states that confer signif- Zealand Cardiac Inherited Disease Registry, icant morbidity and mortality in Aotearoa. which aims to uncover the cause of sudden These studies will inevitably involve unexpected deaths in New Zealand. It is addressing the complex interplay between known that highly lethal, yet very treatable, genes and environment in determining familial conditions are common causes of outcomes. Knowing the differences in 25 sudden unexpected death in the young. A susceptibility present within populations variant in the cardiac sodium channel gene can inform interventions to minimise risk SCN5A known as R1193Q causes Brugada and target scarce health resources.28,29 A lack syndrome or long QT syndrome and is very of appreciation of the full gamut of genetic rare in Western European populations. variation embedded within New Zealanders’ Despite this link, and the observation that genomes will inhibit the development of this it confers distinct functional cellular abnor- understanding, limit its potential to improve 26 malities in vitro, it is present in 10% of the health30 and may perpetuate health ineq- Han Chinese. In Aotearoa, this variant is uities. Emerging fi ndings in this area point present among some Māori who have died to the potential of adopting this as a health from sudden unexplained death in early research priority.31,32 adulthood.25 Presently it is unclear if this A third repercussion of the lack of genetic variant can be considered causal knowledge of Māori and Pacifi c genomes is of these presentations and hence whether the impaired ability to undertake research other family members are at risk or not. studies with Māori and Pacifi c cohorts to Genomic data from Māori and Pacifi c popu- the same level of precision as is possible in lations are needed, with some urgency, to other major Aotearoa population groups. answer this question. The relative paucity of genetic studies in A parallel situation also exists in indig- minority populations is now recognised enous Americans: in a recent report internationally.33 Representation of Indig- reviewing pharmocogenetics of indigenous enous Americans, Australian Aborigines and North American populations, the authors Pacifi c peoples in catalogued genome-wide stated “Not only do Indigenous groups often association studies (GWAS) has fallen since have different allele frequencies compared 2009.33 In Aotearoa, Māori are yet to be to other global populations but marked represented in a single such study. The only differences in allele frequencies can also be GWAS conducted to date in a Polynesian found between subcultures within a given population (addressing obesity; in Samoans geographical region” in regard to genes living in Samoa and American Samoa) 27 with important clinical outcomes. Several had to suffi ce with a genotyping platform examples are now recorded internationally designed for use in European, African and where genetic tests have been misinter- East Asian populations. Consequently, it was preted and individuals misdiagnosed and not possible to comprehensively survey the mismanaged because their genomic ancestry genomic complexity of the study population was not factored into analysis of their owing to the unavailability of an appro- 9,12 genetic tests. Most notable are instances priate genomic dataset that refl ected the where individuals of African-American constitution of the local population.23 origin have had adverse health outcomes A strong imperative therefore exists relating to diagnosis of drug side effects to establish hauora-enhancing genomic

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 84 www.nzma.org.nz/journal VIEWPOINT

resources for Aotearoa that are focused on Therefore, appropriate safeguards must be Māori and Pacifi c health priorities. Such established and adhered to, to ensure that resources should survey and catalogue the the utilisation of this resource remains true distribution and extent of genomic vari- to its intent. Recently explicit tikanga-based ation present in a large cohort of modern guidelines (Te Mata Ira) for conducting day New Zealanders, oversampled for genomic research with Māori have been peoples of Māori and Pacifi c ancestry. developed with some iwi by a Māori-led The construction of such resources will research team and they should be widely be central to the equitable deployment of adopted as best practice in the context of genomic medicine for all New Zealanders. these research partnerships.20 The utility of these resources for Māori and An important issue relating to the Pacifi c peoples will be most immediately generation of genomic resources is that evident in the clinic where individuals surrounding sovereignty of the data and whānau who undergo genetic testing generated. The Te Mata Ira guidelines will have their results interpreted with explicitly affi rm the view of Māori that DNA 34 appropriate precision. It also will be samples and the data generated from them evident where whānau, hapū/iwi and their have intrinsic links with whakapapa.20 The healthcare providers use precise genetic contiguous relationship between these three information about causal mechanisms that domains exists because they are expressions contribute to highly prevalent diseases in of the same wairua—the spirit that enjoins their communities to mitigate or reduce the ancestral, physical and metaphysical their risks of developing these diseases, components of an individual’s being. Conse- in addition to informing their diagnosis quently, DNA samples and genomic data and treatment. Ongoing research on the are considered intertwined as a whole and genetic contributors to major and signif- to be a taonga and therefore both need to icant diseases such as cancer, cardiovascular be subject to appropriate kaitiaki (guard- disease and type 2 diabetes will be greatly ianship) arrangements.20 Moreover, a facilitated by a deeper and more accurate related set of considerations are currently understanding of the genomes of the source being undertaken by Te Mana Raraunga, 11,23 populations under study. This work will the Māori Data Sovereignty Network (www. be best enabled if engagement with Māori temanararaunga.maori.nz) in the context and Pacifi c communities occurs at all levels, of international and other national Indig- including addressing the acute capacity enous Data Sovereignty projects (eg, http:// issue in the Māori and Pacifi c healthcare fnigc.ca/). The Te Mana Raraunga work and research workforce. seeks to formulate and enunciate Māori and Having envisioned this future dividend, also iwi-specifi c perspectives on the gover- it is important to acknowledge that genetic nance and management of data resources, research in Aotearoa has a chequered including genomic datasets. The views of history, specifi cally concerning lack of different iwi and hapū, and even whānau respectful and positive relationships with within them, may vary in many of these Māori and their associated communities. respects, not only in terms of the tikanga Notable successes that have had signifi cant around these resources but also the regu- translational outcomes have occurred,35,36 lation of the collection, storage and use while less enlightened initiatives have of the data.20,34 While a unifi ed national misapplied genomics and simultaneously genomic resource may offer the most diag- failed to be based on genuine partnerships nostic and scientifi c utility to deliver health with iwi or Māori in general to improve benefi ts, this will require Māori commu- health outcomes as a primary objective.37,38 nities to agree to have genomic data derived Such studies had negative outcomes for from their people managed collectively Māori, mirroring similar experiences with under agreed protocols. These conversations other technologies.39 Any research of this have yet to occur and they will be essential nature carries with it the risk of multiple before plans to progress a model of a single, negative outcomes—not the least being as opposed to multiple, genomic resources the reinforcement of negative stereotypes, can be developed. inequities and related power relationships.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 85 www.nzma.org.nz/journal VIEWPOINT

We contend that the best way to achieve there is a need to increase responsiveness this is a programme of work that is of non-Māori researchers and clini- explicitly co-led, co-curated and governed cians, including a readiness for them to by iwi-mandated40 and Pacifi c represen- acknowledge, understand and enact the tatives. This will include involvement of tikanga that will govern this work, espe- these representatives in how genetic and cially initially when the number of Māori genomic resources are used in diagnostics and Pacifi c experts will be outnumbered and the science that it potentiates. This by tauiwi (non-Māori/non-Pacifi c) experts. oversight will extend to how the results of From the outset it must be clearly under- research are interpreted, disseminated and stood that facilitating research designed to explained to the general public and, above primarily make comparisons between popu- all and as a priority, with the communities lations, rather than understand the genetic with whom the research was conducted. In factors for ill-health within populations, is genuinely moving towards this goal, a more not the primary goal of the creation of these substantial and explicit effort must be made resources and the aspiration of precision to grow future leaders in genomics/genetics/ medicine that it supports. bioinformatics who are Māori and/or from We foresee that the generation of Pacifi c communities and can both lead this hauora-enhancing genomic resources research and its governance, and who also for Aotearoa New Zealand is essentially have the skills to form relationships with the formal creation and cataloguing of their communities and to share and commu- a unique national treasure—a taonga— nicate research fi ndings and knowledge that will be key to equitably delivering with them. We think that the development positive healthcare outcomes. As genomic of this proposed genomic resource needs medicine gathers pace, we have the oppor- to include upskilling Māori and Pasifi ka tunity, if not an imperative seated in Te to ensure that future needs and applica- Tiriti o Waitangi, to ensure that it does not tions can be addressed within a community contribute to more healthcare disparities on health provision environment. In addition, the basis of ancestry.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 86 www.nzma.org.nz/journal VIEWPOINT

Competing interests: Dr Doughty reports grants from NZ Heart Foundation, Health Research Council and Healthier Lives National Science Challenge during the conduct of the study. Dr Cameron reports grants from MBIE Healthier Lives National Science Challenge and Heart Foundation of New Zealand outside the submitted work. Author information: Stephen P Robertson, Professor, Department of Women’s and Children’s Health, Dunedin School of Medicine, University of Otago, Dunedin; Jennie Harré Hindmarsh, Research Coordinator, Ngāti Porou Hauora Charitable Trust, Te Puia Springs, Tairāwhiti, Gisborne; Sarah Berry, Senior Research Fellow, Centre for Longitudinal Research - He Ara ki Mua and Growing Up in New Zealand, The University of Auckland, Auckland; Vicky A Cameron, Professor, Christchurch Heart Institute, Department of Medicine, University of Otago, Christchurch; Murray P Cox, Professor, Statistics and Bioinformatics Group, Institute of Fundamental Sciences, Massey University, Palmerston North; Ofa Dewes, Research Fellow, Maurice Wilkens Centre for Molecular Biodiscovery, Department of Molecular Medicine and Pathology, Faculty of Medical and Health Sciences, The University of Auckland, Auckland; School of Nursing, Faculty of Medical and Health Sciences, The University of Auckland, Auckland; Robert N Doughty, Professor, Department of Medicine, University of Auckland, Auckland; George Gray, Public Health Physician, Planning and Funding, Bay of Plenty District Health Board, Tauranga; Jessie C Jacobsen, Research Fellow, Centre for Brain Research and School of Biological Sciences, The University of Auckland, Auckland; Albert Laurence, Te Kaika, Caversham, Dunedin; Elizabeth Matisoo-Smith, Professor, Department of Anatomy, University of Otago, Dunedin; Susan Morton, Senior Research Fellow, Centre for Longitudinal Research - He Ara ki Mua and Growing Up in New Zealand, The University of Auckland; Andrew N Shelling, Professor, Department of Obstetrics and Gynaecology, The University of Auckland, Auckland; Dianne Sika-Paotonu, Associate Dean (Pacifi c), Senior Lecturer Pathology & Molecular Medicine, Wellington School of Medicine and Health Sciences, University of Otago, Wellington; Honorary Research Associate, Victoria University of Wellington, Wellington; Honorary Research Associate Telethon Kids Institute, Perth, Western Australia; Affi liate Investigator, Maurice Wilkins Centre for Molecular Biodiscovery, University of Auckland, Auckland; Anna Rolleston, Clinical Director, The Centre for Health, Manawa Ora Centre, Tauranga; Jonathan R Skinner, Director, Cardiac Inherited Disease Group, Auckland City Hospital and Green Lane Paediatric and Congenital Cardiac Services, Starship Children’s Hospital, Auckland; Russell G Snell, Professor, Centre for Brain Research and School of Biological Sciences, The University of Auckland, Auckland; Andrew Sporle, Senior Research Fellow, Department of Statistics, The University of Auckland, Auckland; Cristin Print, Professor, Department of Medicine, University of Auckland, Auckland; Professor and Director, The Bioinformatics Institute, and the Genomics Into Medicine Programme University of Auckland, Auckland; Tony R Merriman, Professor, Department of Biochemistry, University of Otago, Dunedin and Principal Investigator, Maurice Wilkins Centre for Molecular Biodiscovery; Maui Hudson, Associate Professor, Māori and Indigenous Governance Centre, University of Waikato, Hamilton; Philip Wilcox, Senior Lecturer, Department of Mathematics and Statistics, University of Otago, Dunedin. Corresponding author: Professor Stephen Robertson, Department of Women’s and Children’s Health, Dunedin School of Medicine, University of Otago, PO Box 56, Dunedin 9054. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2018/vol-131-no-1480- 17-august-2018/7668

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 87 www.nzma.org.nz/journal VIEWPOINT

REFERENCES: 1. Green ED, Guyer MS. medicine to populations politics/treaty/read-the- Charting a course for with African ancestry. Nat treaty/english-text (2017). genomic medicine from Commun 2016; 7:12521. 20. Hudson M, Beaton A, Milne base pairs to bedside. 12. Manrai AK, Funke BH, M, Port W, Russell K, Smith Nature 2011; 470:204–13. Rehm HL, et al. Genetic B, Toki V, Uerata L, Wilcox 2. Lek M, Karczewski KJ, Mini- Misdiagnoses and the P, Te Mata Ira. Guidelines kel EV, et al. Analysis of Potential for Health for Genomic Research with protein-coding genetic vari- Disparities. N Engl J Maori, University of Waika- ation in 60,706 humans. Med 2016;375: 655–65. to Te Mata Hautū Taketake Nature 2016; 536:285–91. 13. Kaneko A, Lum JK, – Māori & Indigenous 3. Collins FS, Varmus H. A Yaviong L, et al. High Governance Centre, Hamil- new initiative on preci- and variable frequencies ton, New Zealand, 2017. sion medicine. N Engl J of CYP2C19 mutations: 21. Benton M, Macartney-Cox- Med 2015; 372:793–5. medical consequences son D, Eccles D, et al. 4. Ashley EA. Towards of poor drug metabolism Complete mitochondrial precision medicine. Nat in Vanuatu and other genome sequencing Rev Genet 2016; 17:507–22. Pacifi c islands. Pharmaco- reveals novel haplotypes genetics 1999; 9:581–90. in a Polynesian population. 5. Blair V, Kahokehr A, PLoS One 2012; 7:e35026. Sammour T. Cancer in 14. Masimirembwa C, Dandara Maori: lessons from C, Leutscher PD. Rolling 22. Tanner C, Boocock J, prostate, colorectal and out Efavirenz for HIV Stahl E, et al. Population gastric cancer and progress Precision Medicine in specifi c resequencing in hereditary stomach Africa: Are We Ready associates the ATP Binding cancer in New Zealand. for Pharmacovigilance Cassette Subfamily C ANZ J Surg 2013; 83:42–8. and Tackling Neuropsy- Member 4 (ABCC4) gene chiatric Adverse Effects? with gout in New Zealand 6. Rid A, Johansson MA, Omics 2016;20:575–80. Māori and Pacifi c men. Leung G, et al. Towards Arthritis and Rheuma- Equity in Health: Research- 15. Edinur HA, Dunn PP, tology 2017; 69:1461–9. ers Take Stock. PLoS Med Hammond L, et al. HLA 2016; 13:e1002186. and MICA polymorphism 23. Minster RL, Hawley NL, in Polynesians and New Su CT, et al. A thrifty 7. Tehranifar P, Neugut AI, Zealand Maori: implica- variant in CREBRF strongly Phelan JC, et al. Medical tions for ancestry and infl uences body mass advances and racial/ health. Hum Immunol index in Samoans. Nat ethnic disparities in 2013; 74: 1119–29. Genet 2016; 48:1049–54. cancer survival. Cancer Epidemiol Biomarkers 16. Vernot B, Tucci S, Kelso 24. Shipley GP, Taylor DA, Prev 2009; 18:2701–8. J, et al. Excavating Nean- Tyagi A, et al. Genetic dertal and Denisovan structure among Fijian 8. Hart JT. The inverse care DNA from the genomes of island populations. J Hum law. Lancet 1971; 1:405–12. Melanesian individuals. Genet 2015; 60:69–75. 9. Petrovski S, Goldstein DB. Science 2016; 352:235–9. 25. Bagnall RD, Weintraub RG, Unequal representation 17. Elhaik E, Tatarinova Ingles J, et al. A Prospective of genetic variation across T, Chebotarev D, et al. Study of Sudden Cardiac ancestry groups creates Geographic population Death among Children healthcare inequality in structure analysis of and Young Adults. N Engl the application of preci- worldwide human popu- J Med 2016; 374:2441–52. sion medicine. Genome lations infers their Biol 2016; 17:157. 26. Huang H, Zhao J, Barrane biogeographical origins. FZ, et al. Nav1.5/R1193Q 10. Skoglund P, Posth C, Nat Commun 2014;5:3513. polymorphism is associated Sirak K, et al. Genomic 18. Gosling AL, Matisoo-Smith with both long QT and insights into the peopling E, Merriman TR. Hyper- Brugada syndromes. Can of the Southwest Pacifi c. uricaemia in the Pacifi c: J Cardiol 2006; 22:309–13. Nature 2016; 538:510–3. why the elevated serum 27. Henderson LM, Claw 11. Kessler MD, Yerges-Arm- urate levels? Rheumatol KG, Woodahl EL, et al. strong L, Taub MA, et al. Int 2014; 34:743–57. P450 Pharmacogenetics Challenges and disparities 19. Kawharu, H, Te Tiriti o in Indigenous North in the application of Waitangi, Available at American Populations. personalized genomic https://nzhistory.govt.nz/ J Pers Med 2018; 8.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 88 www.nzma.org.nz/journal VIEWPOINT

28. Do R, Xie C, Zhang X, et of the lipoprotein recep- 37. Wensley D, King M. al. The effect of chromo- tor-related protein 2 gene Scientifi c responsibility some 9p21 variants on with gout and non-additive for the dissemination and cardiovascular disease interaction with alcohol interpretation of genetic may be modifi ed by consumption. Arthritis research: lessons from the dietary intake: evidence Res Ther 2013; 15:R177. “warrior gene” controversy. from a case/control and 33. Popejoy, AB and Fuller- J Med Ethics 2008; 34:507–9. a prospective study. PLoS ton, SM Genomics is 38. Lea R, Chambers G. Med 2011; 8:e1001106. failing on diversity. Monoamine oxidase, 29. Natarajan P, Young R, Stitz- Nature 2016; 538:161–4. addiction, and the iel NO, et al. Polygenic Risk 34. Port RV, Arnold J, Kerr D, “warrior” gene hypothesis. Score Identifi es Subgroup et al. Cultural enhance- N Z Med J 2007; 120:U2441. With Higher Burden ment of a clinical service 39. Mead A, Ratuva S. Pacifi c of Atherosclerosis and to meet the needs of Genes and Life Patents: Greater Relative Benefi t indigenous people; genetic Pacifi c experiences and From Statin Therapy in the service development in analysis of the commod- Primary Prevention Setting. response to issues for ifcation and ownership Circ Arrhythm Electro- New Zealand Maori. Clin of life. (Call of the Earth physiol 2017;135: 2091–101. Genet 2008; 73:132–8. Llamado de la Tierra and 30. Pearce N, Foliaki S, Sporle 35. Guilford P, Hopkins J, the United Nations Univer- A, et al. Genetics, race, Harraway J, et al. E-cad- sity Institute of Advanced ethnicity, and health. herin germline mutations Studies., Wellington, 2007). Bmj 2004; 328:1070–2. in familial gastric cancer. 40. Beaton A, Hudson M, Milne 31. Dalbeth N, House ME, Nature 1998; 392:402–5. M, et al. Engaging Maori in Gamble GD, et al. Popu- 36. Robertson SP, Twigg SR, biobanking and genomic lation-specifi c infl uence Sutherland-Smith AJ, et al. research: a model for of SLC2A9 genotype on Localized mutations in the biobanks to guide cultur- the acute hyperuricaemic gene encoding the cytoskel- ally informed governance, response to a fructose etal protein fi lamin A cause operational, and commu- load. Ann Rheum Dis diverse malformations in nity engagement activities. 2013; 72:1868–73. humans. Nat Genet 2003; Genet Med 2017; 19:345–51. 32. Rasheed H, Phipps-Green A, 33:487–91. Topless R, et al. Association

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 89 www.nzma.org.nz/journal CLINICAL CORRESPONDENCE

Redundant laparoscopic adjustable gastric band tubing causing internal hernia and small bowel obstruction Erika Fernandes, James Tan, Glenn Farrant, Karl Kodeda

42-year-old female farmer present- and mortality rates, is minimally invasive, ed to a provincial hospital in New almost completely reversible and preserves A Zealand with severe colicky central the anatomy of the gastrointestinal tract.1 abdominal pain. Her past medical history Despite its safety, it is not without risks. included insertion of a laparoscopic adjust- Early complications include gastric wall able gastric band (LAGB) seven years prior. perforation secondary to technical error. Associated symptoms concerning for small Late complications include; port infection, bowel obstruction (SBO) were present. Full tube disconnection, dislodgement of the blood count and biochemistry were unre- access port, leak of reservoir, band erosion markable, excepting a mild leukocystosis of or slippage and skin ulceration at port site. 12.1x109/L. Computer tomography (CT) scan These have been well described.2,3 Small revealed a SBO without a clear transition bowel obstruction caused by band tubing is point and raising the possibility of a closed uncommon. Review of the English language loop obstruction (Figure 1). CT demonstrated literature reveals eight cases of SBO caused satisfactory position of the gastric band. by an internal hernia secondary to band 4 The patient underwent an urgent lapa- tubing. Two cases of caecal volvulus caused 5,6 rotomy, which revealed a small bowel by tubing have also been published. All obstruction secondary to herniation at the cases describe redundancy in the tubing as root of the mesentry through a redundant being the cause of the obstruction. loop of gastric band tubing with congested Strobos and Hamed et al also postulate mesentery and small bowel (Figure 2). that a lack of omental covering was also a The tubing required disconnection from contributing factor. Hamed et al elegantly the subcutaneous access port in order to summarise these cases and include details relieve the obstruction (Figure 3). No bowel about the operative interventions used to resection was required. The band was prevent recurrence.4 Some groups sutured defl ated, tubing reconnected and the device the tubing to the anterior abdominal wall; left in situ. The redundant tubing was simply two removed the band entirely, two simply placed back on top of the bowel. The post-op- rerouted the tubing while others, including erative recovery was uneventful, the patient ourselves, took no additional measures discharged on post-operative day two and after freeing the obstruction. We are the the limited follow-up to date of two months only group to describe disconnecting the without complications or recurrence. tubing from the access port in order to Although declining in popularity, laparo- relieve the obstruction. If, in a similar situ- scopic adjustable gastric banding has been a ation, the tubing for some reason cannot be commonly performed bariatric procedure. disconnected and reconnected to the port, Compared to other forms of bariatric we suggest that the tubing is cut as close to surgery it has very low 30-day morbidity the port as possible and, after repositioning, the cut end is sutured in the subcutaneous

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 90 www.nzma.org.nz/journal CLINICAL CORRESPONDENCE

fat close to the port in order to facilitate a patient presenting with SBO and LAGB in future reconnection. situ. As we have demonstrated, this can occur As LAGBs were once a popular procedure many years after the band has been placed. we will continue to see patients who have The published cases demonstrate the varying bands in situ present with symptoms of approaches to intraoperative management in obstruction. Therefore, one should include order to prevent recurrence. Where possible, internal herniation secondary to redundant patient preference on removal of the band band tubing in the differential diagnosis in entirely should be sought.

Figure 1: Coronal view computer tomography showing gastric band tubing at the root of the small bow- el mesentry and dilated loops of small bowel.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 91 www.nzma.org.nz/journal CLINICAL CORRESPONDENCE

Figure 2: Schematic representation of intraoperative fi ndings: gastric band tubing tightly wrapped around the root of the small bowel mesentry and dilated loops of congested small bowel.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 92 www.nzma.org.nz/journal CLINICAL CORRESPONDENCE

Figure 3: Intraoperative photo showing disconnection of the gastric band tubing from the subcutaneous access port.

Competing interests: Nil. Author information: Erika Fernandes, Surgical Registrar, Department of General Surgery, Taranaki Base Hospital, New Plymouth; James Tan, Surgical Registrar, Department of General Surgery, Taranaki Base Hospital, New Plymouth; Glenn Farrant, General Surgeon, Department of General Surgery, Taranaki Base Hospital, New Plymouth; Karl Kodeda, General Surgeon Department of General Surgery, Taranaki Base Hospital, New Plymouth. Corresponding author: Dr Erika Fernandes, Starship Children’s Health, Department of Paediatric Surgery, Auckland. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2018/vol-131-no-1480- 17-august-2018/7669

REFERENCES: 1. Ding TMS-A, Vernon 3. Strobos E, Antanavicius 5. Agahi A, Harle R. A serious AH, Goldfi ne AB. Bariat- G, Josloff R. Unusual but rare complication of ric Surgery. 2016. In: complication: small bowel laparoscopic adjustable Endocrinology: Adult obstruction caused by gastric banding: bowel and Pediatric [Internet]. tubing of gastric band. obstruction due to caecal Philadelphia: Elsevier Surg Obes Relat Dis. volvulus. Obes Surg. Saunders. 7th. [479–90]. 2009; 5(5):637–40. 2009; 19(8):1197–200. 2. Eid I, Birch DW, Sharma 4. Hamed OH, Simpson L, 6. Ng MK, Thompson G. AM, Sherman V, Karmali S. Lomenzo E, Kligman MD. Laparoscopic adjustable Complications associated Internal hernia due to gastric band tubing: with adjustable gastric adjustable gastric band unusual cause of an banding for morbid obesi- tubing: review of the internal hernia. Surg Obes ty: a surgeon’s guides. Can J literature and illustrative Relat Dis. 2009; 5(4):517–8. Surg. 2011; 54(1):61–6. case video. Surg Endosc. 2013; 27(11):4378–82.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 93 www.nzma.org.nz/journal LETTER

A new report on the effects of a policy used to restrict access to elective inguinal hernia surgery Philip Bagshaw

n article published in this journal in A new combined report from the British 2015, described the dangers of a poli- Hernia Society and the Royal College of Acy called ‘Watchful Waiting’, whereby Surgeons of England, released in June 2018, patients with inguinal hernias were refused described how similar Freedom of Infor- surgery because two studies indicated it mation requests were used to investigate might be safe to do so, unless they were the regrettable UK national consequences of having serious symptoms or complications. this fl awed policy.3 The authors of the report Using Offi cial Information Act requests, the should be congratulated for publishing the article showed the policy was being imple- report and facilitating its widespread uptake mented by some New Zealand district health by the UK media.4 boards (DHBs), even though there were clear It is to be hoped that we will all learn from concerns about its safety and cost-effective- this long salutary tale, not to put dubious 1 ness. short-term, cost-saving policies before the A letter was then published in this journal obvious long-term clinical interests of those in 2016, describing further research, which we are here to serve. seriously called the safety and cost-effec- tiveness of the policy into question.2

Competing interests: Nil. Author information: Philip Bagshaw, Clinical Associate Professor, University of Otago, Christchurch. Corresponding author: Philip Bagshaw, Clinical Associate Professor, University of Otago-Christchurch, 349–351 Harewood Road, Christchurch, PO Box 20409, New Zealand. [email protected] URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2018/vol-131-no-1480- 17-august-2018/7670

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 94 www.nzma.org.nz/journal LETTER

REFERENCES: 1. Bagshaw PF. What should to inguinal hernia surgery &utm_medium=dotmail- be the management in England. A report by er&utm_source=emailmar- policy for asymptomatic the British Hernia Society keting&dm_i=4D4N,BH- inguinal hernias? NZ and the Royal College 4G,1RMBGW,19I6R,1 Med J. 2015; 128;83–8. of Surgeons of England. on 23rd July 2018. 2. Bagshaw P. The clinical June 2018. Accessed at: 4. RCS Political Update 21st consequences of under- https://i.emlfi les4.com/ July 2018. Accessed at: funding elective healthcare: cmpdoc/9/3/6/3/0/2/ https://updates-rcseng. A second red fl ag warning. fi les/19222_rcs--bhs- co.uk/4D4N-BH4G-491RM- NZ Med J. 2016; 129:135–6. hernia-report-june-2018. BGW8C/cr.aspx on pdf?utm_campaign=535408_ 3. A dangerous waiting game? 23rd July 2018. Political%20update%20 A review of patient access 20%20July%20

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 95 www.nzma.org.nz/journal METHUSELAH

Five-year risk of stroke after TIA or minor ischemic stroke After a transient ischemic attach (TIA) or minor stroke, the long-term risk of stroke and other vascular events is not well known. In this paper the authors report on an international study that included 4,789 appropriate patients from a registry of TIA clinics in 21 countries. Among the 61 sites that participated, the researches selected 42 sites that had follow-up data on more than 50% of their enrolled patients at fi ve years. They analysed the data on these 3,847 patients. The rate of cardiovascular events including stroke in this cohort was 6.4% in the fi rst year and 6.4% in the second through fi fth years. N Engl J Med 2018; 378:2182–90

Prevalence of diabetes in Australia Accurate diabetes prevalence estimates are important for health service planning. In this study, data from Freemantle and its surrounding area has been evaluated. In a population of 157,000 the researchers identifi ed 6,307 diabetic subjects. The majority of residents with diabetes were born in Australia/New Zealand (56.4%) an just over a third originated from Europe. Based on their data, the researchers estimate that 4.8% or 1.1 million Australian have diabetes, 85.8% have type 2 diabetes, 7.9% have type 1 diabetes and 6.3% have other more rare types. It was concluded that approximately one in 20 Australians have diabetes. Although most have type two diabetes, one in seven have other types that may require more specialised diagnosis and/or management. Internal Medicine Journal 2018; 48:803–809

Comparative effi cacy and acceptability of 21 antidepressant drugs for the acute treatment of adults with major depressive disorder Major despressive disorder is one of the most common, burdensome, and costly psychi- atric disorders worldwide in adults. Pharmacological and non-pharmacological treatments are available; however, because of inadequate resources, antidepressants are used more frequently than psychological interventions. To throw light on this important issue the researchers examined data from 522 trials comprising over 100,00 participants. All of the 21 drugs were more effective than placebo and amitriptyline was the most effective and reboxetine the least effective. The more effective drugs including amitriptyline tended to be less acceptable and had more dropouts from treatment. The researchers concluded that all antidepressants were more effi cacious than placebo in adults with major depressive disorder. Smaller differences between active drugs were found when placebo-controlled trials were included in the analysis, whereas there was more variability in effi cacy and acceptability in head-to-head trials. Lancet 2018; 391:1357–66

URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2018/vol-131-no-1480- 17-august-2018/7671

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 96 www.nzma.org.nz/journal 100 YEARS AGO

A Case of Adherent Meckel’s Diverticulum Causing Fatal Strangulation By E. H. WILKINS, M.B.

LABOVE LAWS IN TBJS COUNTRY. N.Z. Citizen: Who are those fellows stretching themselves out about the farm? Oh, those are the farm hands. You see, they belong to the Farm Labourers' Union, and this is stretch-lime 6y the award. Tourist: And who is that broken-down, hard-worked looking old man? Why doesn't he do a stretch ? N.Z. Citizen: Oh, that's the farmer. You see, the farm life won't take a stretch, too, and so he hat to attend to it himself. (Observer, 26 October 1907). Alexander Turnbull Library, Wellington, New Zealand. /records/7050431

man, aged 22, farm labourer, suffered moist. I gave a turpentine enema with a with severe colicky pain in umbilical good result of faeces and fl atus, after which A region, accompanied by moderate he felt better. and inconstant vomiting, from 8 a.m. till 5 I considered the possibility of intestinal p.m. when I fi rst saw him. He said his bow- obstruction, but thought it unlikely as the els had acted that morning, that the pain vomiting was not frequent, continuous, came on suddenly, and that he was previous- distressing, or faecal, and his pulse, colour, ly in good health. The abdomen was slightly and general condition were so good. I gave tender: not rigid nor distended. Pulse quiet, instructions for another enema to be given full and strong; temperature normal. Face during the night, if necessary, and word to rosy and of good colour; tongue clean and be sent to me if symptoms continued.

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 97 www.nzma.org.nz/journal The patient lived ten miles away and I Post-mortem showed that portions of was not called to see him again till about 11 the strangulated bowel had not recovered, a.m. next day. He was then in very severe and that perforation had taken place. A pain, abdomen rather full, rigid, very Meckel’s diverticulum four inches long, tender, and not moving much with respi- three-quarters of an inch wide at base ration; temperature 101; pulse slightly above and tapering to a point was found, the end normal, but full and strong. He had not of which was rough and shaggy and had vomited since. evidently been adherent. Fully four feet of Immediate arrangements were made to bowel had been strangulated. bring the patient into hospital for operation, It was subsequently ascertained from which was done two hours later by Dr. H. T. his friends that he had been a very small Dawson. As he went under the anaesthetic eater for over two years, owing to fullness a large tumour, not so spherical as a full and discomfort after food, and that he had bladder nor so prominent as a six months eaten practically nothing for fi ve or six days gravid uterus, was evident in mid-abdomen. before I saw him. This probably explains The incision showed the subcutaneous fat to the very scanty subcutaneous fat. The be not more than an eighth of an inch thick. adhesion had evidently been causing partial On opening the peritoneum black blood obstruction for some time. ran out and two large coils of black small Unusual features of this case seem to be: bowel presented. A constricting band was the good colour and pulse of the patient felt across the root of the coils, which gave all through; he was never pale, pinched, or way under the fi nger and was not seen. The collapsed; the mildness of the vomiting in peritoneum covering these coils was of good the early stage, and its absence during the lustre and pulsation was now evident in the rest of the illness, although the obstruction vessels. No normal viscera were seen; every- was complete; that he drank water, retained thing that presented was black. There was it, and absorbed it, with nothing more than remarkably little distention and the wound a transient hiccough; and that there was was closed. comparative absence of distention—all Severe pain, tenderness, and rigidity this in the presence of four feet of disabled continued for four days, during which bowel. frequent injections of morphia were The only explanation I can give o these unavoidable. Enemas, including ox-gall, unusual features is that his alimentary canal and large doses of belladonna failed to give was so completely empty of food. Severe relief. All the time the patient drank much strangulation can evidently take place water, passed a normal quantity of urine, without distressing vomiting, without foecal and never vomited. He was of good colour vomiting, and with a good pulse. all through, pulse strong and not much above normal until the day before he died, and distention very slight.

URL: http://www.nzma.org.nz/journal/read-the-journal/all-issues/2010-2019/2018/vol-131-no-1480- 17-august-2018/7672

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 98 www.nzma.org.nz/journal also published by the New Zealand Medical Association

NZMJDigest http://www.nzma.org.nz/publications/nzmjdigest

The NZMA publishes the e-magazine NZMJDigest 10 times a year. It contains news and views from the profession and the NZMA, including the NZMA Chair's editorial, along with highlights from and links to the New Zealand Medical Journal. Click on the image above to view the latest issue. We welcome contributions from members and readers. To contribute to the NZMJDigest, please email [email protected]

NZMJ 17 August 2018, Vol 131 No 1480 ISSN 1175-8716 © NZMA 99 www.nzma.org.nz/journal