Community-linked maternal death review (CLMDR) to measure and prevent maternal mortality: a pilot study in rural Malawi Olivia Bayley, Hilda Chapota, Esther Kainja, Tambosi Phiri, Chelmsford Gondwe, Carina King, Bejoy Nambiar, Charles Mwansambo, Peter Kazembe, Anthony Costello, Mikey Rosato, Tim Colbourn
Global MNH Conference, Mexico City, 19th October 2015 Open Access Research Community-linked maternal deathOpen Access To enable participation of the communityreview in (CLMDR) the identification to and measure review of deaths, and improves the quality process, we accepted a reduced level of confidentiality of MDR meetings, provides opportunities for education compared with traditional MDR. Whileprevent this was essential maternalon maternal mortality: health, and stimulates a pilot action in study communi- and no known adverse events occurred,in we rural recognise Malawities and health facilities. The potential of CLMDR has the potential risk of information being shared beyond been recognised by the Malawi Ministry of Health who the intended audience. We recognise that public discus- have begun a nationwide rollout. CLMDR is relevant to 1 2 2 2 sion of health workers’ actions could preventOlivia Bayley, disclosureHilda Chapota,similar settingsEsther Kainja, with highTambosi maternal Phiri, mortality and needs Chelmsford Gondwe,3 Carina King,1 Bejoy Nambiar,1 Charles Mwansambo,4,5 or result in recriminations. We believePeter that emphasising Kazembe,5,6 Anthonyfurther Costello, assessment.1 Mikey Rosato,7 Tim Colbourn1 the blame-free culture of the process was sufficient to avoid this in our pilot study, but these issues should be Author affiliations To cite: Bayley O, Chapota H, ABSTRACT 1University College London Institute for Global Health, London, UK carefully consideredKainja with E, et al. Community-reference to each cultural Strengths and limitations of this study Background: In Malawi, maternal2MaiMwana mortality remains Project, Mchinji, Malawi context where CLMDRlinked maternal is deathintroduced. review high. A Existing single maternal data- death reviews fail to adequately (CLMDR)BMJ to measure Open and 2015;5:e007753 3 ▪ This pilot study in Mchinji District, central region review most deaths, or capture thoseDepartment that occur of Safe Motherhood, Mchinji District Health Management Team, collection form travellingprevent maternal between mortality: the community and of Malawi, shows that a community-linked a pilot study in rural Malawi. outside the health system. We assessedMchinji, the Malawi value of maternal death review (CLMDR) process identi- fi 4 the health facility hadBMJ Open bene2015;5:e007753.ts for communicationcommunity involvement but to improveGovernment capture and of Malawi Ministryfied of twice Health, as many Lilongwe, maternal Malawi deaths as the exist- doi:10.1136/bmjopen-2015- response to communityfi maternal5 deaths. ing facility review process; yielded richer data; raised the risk of information007753 being lostMethods: or conWedential- designed and pilotedParent a community-linked and Child Health Initiative (PACHI), Lilongwe, Malawi 6 and led to more actions being taken after the ity being breached. The CLMDR managementmaternal death review team (CLMDR) processBaylor in College Mchinji of Medicine Childrenreview. ’s Foundation, Lilongwe, Malawi 7 ▸ Prepublication history District, Malawi, which partnered communityWomen and and Children health First,▪ London,Communities UK and health facility representatives making a copy of theand additional form material at each is stagefacility might stakeholders help to identify to and review maternal worked in partnership to investigate and respond mitigate this risk. available. To view please visit deaths and generate actions to prevent future deaths. to maternal deaths occurring in communities the journal (http://dx.doi.org/ The CLMDRfi process involved fiveTwitter stages:Follow community Timothy Colbournand health at @timcolbourn facilities. While CLMDR dramatically10.1136/bmjopen-2015- improvedverbal identi autopsy,cation community of and facility review meetings, ▪ Confidentiality of the death review was limited to deaths, the process007753). may strugglehttp://bmjopen.bmj.com/content/5/4/e007753.full.pdf to identifya public or meeting follow-up and bimonthly reviewsAcknowledgements involving both The authorsallow thank participation all the community of, gain information and health from, facility and community and facility representatives. spur action from the community. No adverse Received 22 January 2015 participants of the CLMDR process and hope they found it valuable. They also Results: The CLMDR process was found to be maternal deaths ofRevised transient 12 March workers 2015 due to the reduced thank all staff at MaiMwana Projecteffects who of this helped openness with were the logistics reported. and Accepted 27 March 2015 comparable to a previous research-driven surveillance ▪ Our pilot study delineated key issues to consider coverage of community teams in transientsystem communities at identifying deaths in Mchinjimanagement District of the project. for scale-up: the CLMDR process adds to exist- (eg, tobacco estates). Transient tenant farmers(population may456 500 have in 2008). 52 maternal deaths ing workload especially for community health were identified between July 2011Contributors and June 2012,OB and 27 HC conceivedworkers; the was study, not which started was for further some developed cases of little social support and may be more(52%) vulnerable of which would to not haveby been EK, identified TP, MR without and CG. OB, HC,death; EK, and TP andcan take CG were over 6involved months in for data each case maternal death, so efforts to improve identicommunityfication involvement. and Based oncollection. district estimates TC and OB of analysed(although the quantitative we believe data. this HC, can MR andbe beneficial). OB analysed population (500 000) and crude birth rate (35 births Raising the status of the community involved is follow-up of these deaths would be worthwhile.per 1000 population), In the at maternalthe mortality qualitative ratio data. was OB wroteessential the first draft to ensure of the paper the sustainability with significant of input the least three of the seven cases wherearound consent 300 maternal was deaths perfrom 100 000 MR live and births. TC. All Of authorsprocess. reviewed and revised the paper and approved the 41 cases that started the CLMDRthe final process, version 28 of the paper. declined or withdrawn, abortion or HIV(68%) contributed completed all five to stages. We found the CLMDR process to increase the quantityFunding of informationThis study was fundedIn 2013 by an a grant estimated from Engender 292 982 maternalHealth, no.: deaths the woman’s death, so it seems the CLMDRavailable process and to involve was a wider range of stakeholders 1 GMH-103-01. occurred worldwide, most preventable with not always able to facilitate the discussionin maternal of these deathsensi- review (MDR). The process resulted proven interventions. The UN Secretary in high rates of completion of community-planned General’s Commission on Information and tive topics. It is important that regardlessactions (82%), of and family district hospitalCompeting (67%) and interests health None declared. fi centre (65%) actions to prevent maternal deaths. Accountability recommends the introduction consent, all identi ed deaths should undergo at least a Ethics approval This studyof was better approved methods by the to National count Health maternal Sciences deaths Conclusions: CLMDR is an important addition to the 2 confidential health facility MDR. Womenestablished dying forms outside of MDR. It showsResearch potential Committee as a of Malawi,and to protocol review and 785. monitor progress. maternal death surveillance system, and may be Maternal death audit is an important tool the district prevented the process fromapplicable being completed to similar contexts withProvenance high maternal and peer reviewto preventNot commissioned; maternal deaths, externally and peer uses reviewed. knowl- during the pilot study, however, rollingmortality. out CLMDR edge of the circumstances of a death to help For numbered affiliations see Data sharing statement No additional unpublished data is available. end of article. prevent future deaths. Maternal death audit across neighbouring districts would potentially enable covers three approaches: confidential enquiry Open Access This is an Open Access article distributed in accordance with completion of the process regardless ofINTRODUCTION place of death. into maternal deaths, facility-based maternal Correspondence to the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, The CLMDR processDr Timothy attempts Colbourn; to challengeAround the the world, exist- many women continue to death review (MDR) and community-based [email protected] die as a result of pregnancywhich and permits childbirth. others to distribute,data-gathering remix, adapt, known build as verbal upon this autopsy. work non-WHO ing power imbalance between health workers and the commercially, and license their derivative works on different terms, provided communities they serve. Project staff notedBayley that O, et even al. BMJ Openthe2015; original5:e007753. work doi:10.1136/bmjopen-2015-007753 is properly cited and the use is non-commercial. See: http://1 though a community might be highly motivated to creativecommons.org/licenses/by-nc/4.0/ pursue the process, where the HSA failed or the health facility was slow to organise a meeting, the community had little recourse to push it forward. Any rollout of REFERENCES 1. Kassebaum NJ, Bertozzi-Villa A, Coggeshall MS, et al. Global, CLMDR should try to elevate the status of the commu- regional, and national levels and causes of maternal mortality during nity and hold all stakeholders to account, perhaps 1990–2013: a systematic analysis for the Global Burden of Disease Study 2013. Lancet 2014;384:980–1004. through written agreements or parallel advocacy activ- 2. WHO. Keeping promises, measuring results. Commission on ities. The traditional authority (or chief) rarely attended information and accountability for women’s and children’s health. community feedback meetings, instead being repre- Geneva: World Health Organization, 2011. 3. Lewis G. Beyond the numbers: reviewing maternal deaths and sented by a Group Village Headman. Raising traditional complications to make pregnancy safer. Br Med Bull 2003;67:27–37. authority involvement from the beginning might be an 4. Lewis G. Reviewing maternal deaths to make pregnancy safer. Best Pract Res Clin Obstet Gynaecol 2008;22:447–63. effective strategy to improve the sustainability of the 5. Mataya R, Malawi National Confidential Committee on enquiry into process in the long term. maternal death. Report on the Confidential Enquiry into Maternal Deaths in Malawi (2008–2012). Lilongwe, Malawi: Reproductive Health Unit, Ministry of Health, 2013. Conclusion 6. Kongnyuy EJ, van den Broek N. The difficulties of conducting CLMDR is a new and effective method of maternal maternal death reviews in Malawi. BMC Pregnancy Childbirth 2008;8:42. death audit. By harnessing the motivation of communi- 7. Danel I, Graham WJ, Boerma T. Maternal death surveillance and ties to prevent maternal deaths, CLMDR improves response. Bull World Health Organ 2011;89:779–9A.
Bayley O, et al. BMJ Open 2015;5:e007753. doi:10.1136/bmjopen-2015-007753 9 Key messages
CLMDR identified twice as many maternal deaths
More actions were taken
Communities and facilities can work together
Confidentiality is an issue but we found no adverse effects
Background
Maternal mortality in Malawi is still high
Existing maternal death reviews are insufficient: - inadequate review and action at facility - community deaths missed
Weaknesses of the current MDR system
1. Maternal death identification Only hospital deaths are notifiable
2. Review of maternal deaths 29% of facilities in Malawi in 2010 conducted MDRs these included only 26% of recorded deaths [1] Barriers: poor record keeping; staff/resource/skills shortages [2-4]
Weaknesses of the current MDR system cont’d
3. Quality & Quantity of Information Available Patient history, examination, monitoring, management not well recorded [4] Culture of blame prevent sharing of info
4. Stakeholder Involvement Needs non-clinical staff too e.g. pharmacists, lab workers, transport co-ordinators Weaknesses of the current MDR system cont’d
5. The Potential of the Community is Overlooked
6. Accountability of Health Workers No check on whether actions are taken
Design of CLMDR
MaiMwana, UCL, MoH
Based on social autopsy studies in Indonesia, India [5-7]
Input from Local Leaders, national Safe-Motherhood Task Force
Design of CLMDR
350 community teams formed + trained (GVH, HSA, volunteered)
12 Health Centre teams formed + trained
Exisiting MDR team at Mchinji DHO expanded to also include drivers, pharmacy, laboratory, support staff + further training on roles and responsibilities
! STAGE 1: A!WOMAN!DIES! ! Community!Verbal!AutopsyMonthly(Strategy(Evaluation(Meeting(using(Form(5A! ( ( the(team(meets(woman’s(family(and(seeks(consent(for(community(feed(back( Volunteer!and!HSA(interview(family(using(Section!1(and(collect(Health!Passport!( ( ! The ( STAGE 2:Community!CLMDR!Meeting ( CLMDR ( Community!CLMDR!team(meet,(read(information,(discuss(and(complete(Section!2! ( GVH(notifies(TA( HSA(notifies(CLMDR(management(team(and(arranges(CLMDR(meeting! Process (
STAGE 3:!Health!Facility!CLMDR!Meeting! ( Health!facility!CLMDR!team!meet,(read( Monthly(Strategy(Evaluation(( ( information,(discuss,(plan(strategies,( agree(action(points(for(health(centre(and( district(hospital(and(complete(Section!3!! Using(evaluation(form(
STAGE 4: STAGE 5: Community!Feedback!Meeting Bimonthly!Progress!Meeting HSA!meets(woman’s(family(to(seek(consent(for( ( District!hospital!CLMDR! community(feedback(meeting(and(agree( team,!chairpersons!of! information(to(be(shared( health!centre!CLMDR! teams!and!community! Community!CLMDR! HSAs(discuss(progress(on( team,!TA!and!health! community,(health( facility!representative( centre(and(district( Monthly(Strategy(Evaluation( hold(public(meeting,( hospital(strategies(and( discuss,(plan(strategies,( action(points.(Suggest( modifications(and(agree.( agree(action(points(and( Complete(Section!5! Using(evaluation(form( complete(Section!4( Sample
1 year implementation: July 2011 to June 2012
Whole of Mchinji district (around 500,000 people)
Maternal deaths of women resident in Mchinji Results 1. Maternal Death Identification
52 in total Only 25 (48%) identified by existing system
43 deaths (83%) identified by community CLMDR teams, including 4 that happened at DH but were overlooked by existing system
Estimated MMR: 300 per 100,000 similar to MaiMwana trial [8] CLMDR accurate method of measuring MMR Results 2. Review of Maternal Deaths
45 / 52 (86%) subject to review 37 (71%) at community CLMDR meeting 44 (85%) at health facility CLMDR meeting 32 (62%) at community feedback meeting 35 (67%) at bimonthly review meeting
28 cases completed all 5 stages Non-completion due to: family declining community feedback meeting (5 cases); community meeting not happening (3); HSA failing to organise meeting (2); form lost (1); death outside of Mchinji (2) Results 3. Quantity of Information
Verbal Autopsy form available in 39 of 44 cases discussed at health facility CLMDR meeting
Open-ended free text questions à more information on delays: disrespectful treatment by health workers, being turned away from health centres, misdiagnoses, slow referral pathways, lack of hospital transport and unavailability of life-saving treatments
“Using information from the deceased family together with hospital records during reviews assists to come up with a root cause of the problem which enables us to come up with real contributing factor and good strategies” [Midwife, HC]
Results 4. Stakeholder Involvement Lots of people involved! Community members Health Facility staff
Results
5. Community Mobilisation & Action Action points: • Community meetings to explore traditional beliefs • Bye-laws to prevent traditions posing a risk to pregnant women • Educating men on their roles and responsibilities • Lobbying health facility for more respectful treatment of women • Establishing mobile antenatal clinic • Mobilising community funds for bicycle ambulance maintenance • Establishing youth club • Organising female counsellors to support pregnant women
Ave. 2.2 action points (range 1– 4) made per community feedback meeting à1.8 action points (range 0–4) reported completed 82% of all proposed community action points were reported completed
Results 6. Accountability of Health Workers
Action points: • Designing new antenatal form to better capture risk factors • Improving drug supplies (antihypertensive drugs) • Training for clinicians following maternal deaths • Health education events for communities • Improved emergency transport, incl. motorcycle ambulance • Increased fuel allowance • Changing protocols to improve access to rural hospitals HC: Ave. 2.4 action points (range 1– 4) made per meeting (2.2 for DH)
à1.5 action points (range 0–3) reported completed (HC & DH) 65% of all proposed HC action points reported completed (67% for DH)
Discussion
Community elevated from passive ‘data collectors’ to active partners in Maternal Death Surveillance and Response (MDSR)
CLMDR doubled number of deaths reviewed
Valuable Discussions & Action to address the Three Delays Discussion
Improved Data à Evidence-based Decision Making
Lots of people involved à creative solutions & more action points completed
Publicised actions à increased motivation to fulfil commitments; increased trust in system
CLMDR challenges power hierarchy à positive change to patient-provider relationships Issues to consider for scale-up CLMDR adds to existing workload (but in long term: actions from CLMDR à less deaths à less work)
CLMDR not started for some cases (HIV? Abortion?)
Raising the status of the community is essential for sustainability of CLMDR Communities can transform maternal health challenges when given an opportunity to do so
Zikomo Kwambiri!
Tim Colbourn: [email protected]
References
1. Republic of Malawi Ministry of Health. Malawi 2010 EmONC needs assessment final report. 2010. 2. Kongnyuy EJ, van den Broek N. The difficulties of conducting maternal death reviews in Malawi. BMC Pregnancy Childbirth 2008;8:42. 3. Kongnyuy EJ, Mlava G, van den BN. Facility-based maternal death review in three districts in the central region of Malawi an analysis of causes and characteristics of maternal deaths. Womens Health Issues 2009;19:14–20. 4. Combs Thorsen V, Sundby J, Meguid T, et al. Easier said than done!: methodological challenges with conducting maternal death review research in Malawi. BMC Med Res Methodol 2014;14:29. 5. Supratikto G, Wirth ME, Achadi E, et al. A district-based audit of the causes and circumstances of maternal deaths in South Kalimantan, Indonesia. Bull World Health Organ 2002;80:228–35. 6. UNICEF. Maternal and perinatal death inquiry and response: empowering communities to avert maternal deaths in India. New Delhi: UNICEF, 2008. http://www.unicef.org/india/MAPEDIR- Maternal_and_Perinatal_Death_Inquiry_and_Response-India.pdf (accessed 17 Dec 2014). 7. Kalter HD, Salgado R, Babille M, et al. Social autopsy for maternal and child deaths: a comprehensive literature review to examine the concept and the development of the method. Popul Health Metr 2011;9:45. 8. Lewycka S, Mwansambo C, Rosato M, et al. Effect of women’s groups and volunteer peer counsellors on rates of mortality, morbidity and health behaviours in mothers and children in rural Malawi (MaiMwana): a factorial, cluster-randomised controlled trial. Lancet 2013;381:1721–35