| dc.description.abstract | Background: Interventions to reduce maternal deaths are urgently required, especially in low- and middle-income countries (LMICs). Timely maternal death surveillance and response (MDSR) has contributed to reductions in maternal deaths in LMIC settings, including Ethiopia, Rwanda, and South Africa. However, MDSR rates have been low in many parts of Uganda, including the National Referral Hospital where <50% of all maternal deaths underwent appropriate review at the start of this study. We hypothesized that implementation of an intervention consisting of training in standardized MDSR and stakeholder engagement would (i) increase the proportion of deaths subjected to timely MDSR and (ii) increase knowledge and confidence about performing MDSR in the study centre. The objectives were (i) to determine the proportion of maternal deaths notified and reviewed, causes of deaths and their preventability at the study centre, over a 3-year baseline period (1st January, 2016 to December 31st 2018); (ii) to explore the perspectives of various stakeholders on barriers and facilitators to implementation of maternal death surveillance and response (MDSR); (iii) to use these results to develop and implement an effective intervention to promote MDSR in the study centre, (iv) to evaluate the effect of the intervention on MDSR rates at the study centre; and (iv) to explore healthcare workers perspectives on changes in practice in MDSR implementation that occurred within 18-24 months of the intervention period. Methods: A cross-sectional study to review maternal death records and MDSR forms from 1st January 2016 to 31st December 2018 was conducted. We determined the proportion of deaths subjected to timely notification and review, established what proportion of deaths were scored preventable and described the contextual factors contributing to maternal deaths. In addition, we determined how different the maternal characteristics of those who were reviewed compared to non-reviewed deaths. Specific causes of deaths were subjected to further analysis (uterine rupture and abortion related cases) to understand preventability. Then, in-depth and key-informant interviews, and focus group discussions were conducted with key stakeholders to explore barriers and facilitators to the implementation of MDSR. Analysis done to identify emerging themes, drawing on the Theory of Planned Behaviour (TPB). Based on data emerging from both our quantitative and qualitative data, an intervention (training in MDSR with confidential enquiry using national MPDSR guidelines was conducted. The training was followed by stakeholder engagement activities for 19 months. The training in MDSR involved 176 health workers (midwives, SHOs, Specialists, records team, administrators, anaesthesia, and laboratory staff). Priority training targeted senior clinical staff (both doctors and midwives), residents (senior house officers) and those with leadership roles. The training was followed by regular stakeholder engagement for 19 months (March 2020 to September, 2021). Following training and stakeholder engagement, maternal death data were again reviewed by the independent assessors to investigate the institutional performance of MDSR and compared before and after the intervention period. Qualitative interviews were also conducted at the end of 19 months with 33 participants to explore their perspectives on positive changes and the barriers that persisted despite the intervention. Results Thirty-three percent (115/350) of maternal deaths underwent MDSR during the pre-intervention period. In 48% (167/350) of cases, notification to the Ministry of Health occurred, but only 11% of deaths (39/350) were notified within the recommended timeframe, and only 7% (25/350) of the maternal deaths were reviewed within the recommended timeframe. The major causes of deaths were obstetric haemorrhage (158/350; 45%), hypertensive disorders of pregnancy (87/350; 25%) and infection (95/350; 27%). Overall, 294/350 (84%) of maternal deaths were considered preventable. In 95% (332/350) of cases, delays within healthcare facilities were identified. Delays at the study centre occurred in (64%; 226/350) of cases. The results suggest that a well-supported, and timely maternal death review process with targeted interventions could be effective in reducing maternal deaths in this setting. Seventy-six participants were included in our baseline qualitative study to explore barriers and facilitators to MDSR implementation at the National Referral Hospital. We conducted 24 in-depth interviews with health workers and 4 focus-group discussions (with 28 participants). We also performed 24 key-informant interviews with health sector managers, implementing partners, and lawyers. The major barriers to implementation of MDSR were: inadequate knowledge and skills, fear of blame and litigation, failure to implement recommendations, burn out because of workload at the National Referral Hospital, and inadequate leadership to support health workers. Major facilitators that we identified included training and mentorship to enhance knowledge and skills in MDSR, involve all health workers as possible (including midwives), eliminate blame, and functionalize lower health care facilities. The participants strongly recommended training and mentorship in MDSR as an intervention to improve performance. We also identified a clear need to revise Uganda’s legal framework surrounding maternal death. After our intervention of MDSR training and ongoing stakeholder engagement, there was a significant increase in proportion of maternal deaths notified to the Ministry of Health within a 24 hr target (11.1% v. 85.3%, p<0.001) and proportion of maternal deaths reviewed within 7 days (7% v. 67%, p< 0.001) in the post-intervention compared to pre-intervention period. The mean interval from death to MDSR improved from 112 ±87.3 days to 12.3 ±21.4 days (p<0.001). After 19 months of regular stakeholder engagement, study participants reported improvement in MDSR implementation. In particular, they reported enhanced leadership and stronger commitment, stronger ownership of the MDSR process, and commitment to implementing recommendations. Conclusions: At baseline, MDSR implementation at the study centre in urban Uganda was sub-optimal. The majority of the mothers who died had experienced delays in their care (64%), however only 7% of deaths in our setting underwent review within 7 days. Our initial qualitative data generated new insights into the barriers to developing a robust system of MDSR, and were used to inform new multi-disciplinary training. After the training and stakeholder engagement intervention was applied, a higher percentage of maternal deaths (67%) underwent timely review and health workers expressed stronger ownership of the MDSR process in our setting. More sustained efforts of timely MDSR for a longer period could contribute to reducing institutional maternal mortality. Utility of the study: Findings from this work have been used to inform better practice at facility level. The study findings have also been disseminated to Ministry of Health, National Maternal and Perinatal Death Surveillance and Response Committee/ National Safe Motherhood Expert Committee (NASMEC) discussions, and Kampala City Council Authority (KCCA). Further dissemination and sharing experience within the East African region of our experience with improving the MDSR process has taken place through presentations at conferences and peer-reviewed publications. During the engagement with the Ministry of Health in the latter stages of our study, the budget of National Referral Hospital was increased from 1.5 billion Uganda shillings (416,666 USD) to 4.5 billion (1,250,000 USD), new equipment for maternal monitoring was purchased, and a new Intensive Care Unit was created within the maternity service. Key words: MDSR, maternal death, multi-disciplinary team training, legal aspect and policy issues | en_US |
