Key Informant Interviews (KII) were conducted in 2021 using a standardized guideline as part of the primary data collection to complement the findings of the document review. ²⁵ The KII guidelines were provided by the author. It was pilot-tested before the interview. Two trained anthropologists collected the primary data through face-to-face interviews with key respondents. The interviewers were MBBS with MPH backgrounds. They were experts in taking KIIs of health and family planning managers at the national and district levels. First, a communication letter was circulated from the Line Director, MNC&AH, DGHS, to conduct these KIIs with respondents. Then, the interviewers communicated with the respondents over the phone and booked their time for the interviews. A guideline for KII and a consent form in Bangla was developed by the research team for this data collection. The respondents who were highly engaged with the MPDSR program in Bangladesh were selected purposively. A permission letter was collected from the government for data collection. The data collectors communicated with the selected key informants over the cell phone and booked their time for an interview. Then, they visited the respondents’ work and conducted the interview. Before starting the interview, consent was received for the interview and audio recording. The duration of each interview (KII) was 30-45 minutes. Audio records were included in each interview. The transcription was developed from audio recordings, and the translation was conducted from Bangla to English. A thematic analysis was then conducted using qualitative methods. KIIs were performed to understand critical stakeholders’ beliefs about MPDSR data use and how the implementation of MPDSR could be improved by identifying barriers. The respondents ( Table 1) were specifically asked about the gaps in the MPDSR implementation and process and their recommendations for improving the process for a higher impact.

This study also collected district MPDSR action plans developed by health managers through workshops and funded for implementation. The plans were implemented in the Sylhet, Maulvibazar, Bandarban, Cox’sbazar, Jamalpur, Netrokona, and Gazipur districts.

Secondary data were collected from public governmental health websites. There was a specific dashboard on the MPDSR program within the health dashboard. The MPDSR dashboard produced tables, pie charts, and trends. To provide a quick overview of the performance at the district level, a color-coded tool/table can be generated, with green indicating on-track, yellow indicating progress, and red indicating not on-track. Consent and permission were obtained from the Ethical Review Committee and the Government to access all data used in this study; thus, ethical clearance was ensured. Reviewing the Reproductive, Maternal, Neonatal and Child Health ( RMNCH) scorecard was another method applied in this study to observe progress in the health indicators in the MPDSR data. The baseline was 2019 and compared with 2021. The RMNCH scorecard data was collected from 27 districts from the government website ( Figure 1). The data from the website was extracted in 2022.

The causes of death data were also entered into DHIS2 after an expert’s review at the district level. The districts organize a one-day-long workshop with all their sub-district health managers to develop a plan based on the available data in DHIS2. Frequency analysis was used for MPDSR key actions in the seven districts and presented in a tabulated form. Quantitative data on death notifications were also analyzed in a Microsoft Excel (Version 2308) spreadsheet, shown in the Figures 2, 3 and 4.

It is worth noting that Bangladesh is one of the few countries that uses real-time MPDSR data through a dashboard, which the MOH creates for monthly monitoring. The study team examined data to measure performance in 27 districts and identify gaps in information systems. This dashboard is used for various meetings, including MPDSR committee meetings, monthly health coordination meetings, MPDSR action plan workshops, video conferences, and national MPDSR core committee meetings. It is also important to mention that the WHO provides projection-based population data to estimate the number of children born to each administrative unit.

In 2019, only 63% of the estimated 2056 maternal deaths were reported, and health workers reviewed only 32% of notified cases. According to the national MPDSR guidelines, the target for maternal death review was 100% of notified cases. This indicates the need for health workers to visit the deceased mother’s home to increase notifications and review the cause of death. Facility maternal deaths are notified and reviewed, with much better performance compared to community MPDSR. 76% of the cases were reviewed at the facility ( Figure 2).

According to the national MPDSR guidelines, it is recommended that 10% of neonatal deaths be reviewed to determine the causes of death and to identify social barriers. However, the performance of neonatal death surveillance in capturing community data is lacking, with only 29% of cases being recorded. Additionally, health workers reviewed 46% of the cases, even though this was not required ( Figure 3). However, the performance in reviewing facility deaths was satisfactory. Identifying the right target can help health workers reduce their workload by targeting appropriate number of cases.

We attempted to examine how the performance of MPDSR could be evaluated using data from the dashboard. We observed that the death notification system is unique in all 64 districts and captures data from around 5000 unions (lowest administrative units) through health assistant reports. The data show that the system captures 45% of maternal and 26% of neonatal deaths nationally ( Figure 4). This is to mention that private hospitals and urban health facilities are not fully reporting through DHIS2. In this case, the selected districts performed better than the national level in terms of death notification, which was common for all districts.

This study found that the current information system can capture important qualitative data to aid health managers in evidence-based planning. According to the three-delay model, maternal mortality is linked to delays in three areas: (1) deciding to seek care, (2) reaching the healthcare facility, and (3) receiving care. By analyzing the MPDSR dashboard, health managers could identify the major causes of maternal and neonatal deaths at different levels, which helps in planning. The maternal death review contains information about the cause of death and the three delays: decision-making, transport, and delays in receiving services. The MPDSR dashboard displays the percentage of different causes of maternal and neonatal deaths, which helps health managers prioritize their efforts. Figure 5 below shows that hemorrhage is the leading cause of maternal death, whereas birth asphyxia is the primary cause of neonatal death.

We collected detailed action plans from seven districts to determine whether the actions were linked to the information systems. Health managers developed MPDSR action plans annually through a 1-2 day-long workshop. All team members analyze the data and develop a 6-to 12-month plan for the district. The senior gynecologist and neonatologist led the technical expertise for interventions, and health managers develop key actions to reduce mortality. We observed that the key actions were being developed in different thematic groups by the cause of death for maternal and neonatal deaths. Some cross-sectoral health-system barriers have been identified. After reviewing the documents, we summarized the key actions to reduce maternal and neonatal deaths in the seven districts ( Table 2). During the workshops, health managers identified the major causes of maternal and neonatal deaths along with barriers. The experts recommend solutions to these problems and develop action plans that were costed for implementation. The actions in Table 2 were linked to the causes of death from the dashboard. The persons responsible for implementing these action plans at the community level were health education officers, community healthcare providers, health inspectors, and local elite persons in the community. At the facility level and in the MPDRS committee, the Residential Medical Officer, Medical Officer, Civil Surgeon, Upazila Health and Family Planning Officer, and Upazila Family Planning Officer were responsible for implementing actions to reduce death. The quality of MPDSR action plans varied from district to district. Significant gaps in data completeness have been addressed in the action plan by authorities to improve data entry.

This study analyzed maternal and neonatal indicators to evaluate the impact of information systems from 2019 to 2021 ( Table 3). Based on the scorecard system developed by MOH&FW for key indicators, it was revealed that all seven districts in 2019 had inadequate coverage of specific services, including the registration of pregnant mothers, antenatal care (ANC) and postnatal care (PNC) services, and delivery by skilled birth attendants. This highlights the need to improve these areas to ensure better health care services. The RMNCAH dashboard generates a color-coded table that provides a quick overview of performance at the district level. Green indicates on track, yellow indicates progress, and red indicates not on track ( Table 3). The information system can track the progress of health indicators in a district for comparison and trend analysis among health workers. This information system can track the progress of health indicators in a district for comparison and trend analysis. Despite the COVID pandemic, the seven districts that developed MPDSR action plans could measure progress using many indicators. This is to remind you that the COVID pandemic has heavily impacted routine essential health services such as ANC, facility delivery, routine immunization services, and PNC. In 2021, many efforts have been made to restore the services to the baseline year (2019). The impact was impressive in most cases, except in a few districts, which require further investigation.

This study reviewed the scorecard of the MPDSR program and found that its key performance indicators were linked with maternal and neonatal mortality rates. The program’s performance dashboard can visualize its impact based on surveys and routine data. Most indicators showed positive changes, except in the Bandarban and Gazipur districts, where mortality and service coverage decreased. Pregnancy registration, delivery antenatal 4th visits coverage, and postnatal care increased in Jamalpur and Maulvibazar, where the impact of reducing MMR was the highest. The data were subject to verification but could be discussed during the MPDSR meetings.

The interviews informed recommendations for improving MPDSR in developing countries.

Strengthening the health workforce through capacity building and incentives

MPDSR involves capacity-building opportunities, such as training and logistics, as well as motivating healthcare workers. Training helps workers understand the importance of death reviews and enables them to use this information during planning. Engaging professional associations and providing incentives can motivate workers to ensure their success. One of the interviewees mentioned that “There was remuneration for death review before, and now there is no incentive for death review and social autopsy in the field. The staff needs to be accountable for this work.”

Linking MPDSR information systems with quality improvement programs

MPDSR plays a crucial role in enhancing the quality of healthcare services. The results derived from the facility death review process should be connected with the quality improvement cycle at the national, subnational, and facility levels through QI/MPDSR committees. Every intervention at the facility level must be linked to the Plan Do Check and Act (PDCA) cycle, ensuring continuous improvement and monitoring through information systems.

Providing a sustainable and trouble-free information system

It is essential to provide technical assistance to ensure accurate data entry at all levels, especially at the district and upazila levels. With proper guidance and support from an IT cell/department, technical issues related to data entry, data accuracy, and server issues may be minimized or resolved.

Better coordination between policymakers and field-level staff

Coordination at all levels needs to be strengthened to ensure better implementation and participation of field-level health workers and their supervisors. During the national MPDSR core committee meetings, policymakers need more involvement to change the national strategy and address the findings from lower tiers, such as incentivizing success and appreciating champions at the national level. The managers mentioned that the MPDSR initiative aims to improve maternal and perinatal health outcomes by fostering a collaborative and comprehensive approach. To achieve this, the organogram must include UNICEF, WHO, CIPRB, and the ministry for death notification, data analysis, DHIS2 updates, coordination, and collaboration. By bringing together various stakeholders, the initiative can significantly impact reducing mortality rates and improving healthcare systems, aligning with global efforts to improve maternal and perinatal health outcomes. One of the managers said that:

Revision of National MPDSR guideline

Most key stakeholders believe that there is a need to revise the current MPDSR guidelines developed in 2014. The new guidelines should clarify certain functions and add improvements to the existing framework/process, as per the new WHO guidelines, including stillbirth.

Promote standardized response processes.

The actions taken during the MPDSR workshops were mostly similar and broad. Some solutions, such as HR and funds, are not manageable by local health managers. This includes the systematic follow-up and implementation of recommendations across system levels using a tracking system. The MPDSR committee can review the actions and define a set of common actions for all districts as a recommendation.

Legal enforcement of death reporting

In many countries, death reporting is required by law, which makes it essential for healthcare providers to report all types of deaths, including maternal and neonatal deaths and stillbirths. This will further strengthen the information system for the MPDSR. One respondent stated that:

Monthly district-level cause-analysis

The stakeholders recommend the provision of monthly cause analysis workshops at the district level instead of conducting the workshop at a higher level. At the district level, gynae-and-pediatric consultants are present, so conducting cause analysis workshops monthly using the local information system is possible. This will improve the accuracy and speed of local action plan development.

MPDSR-monitoring cell

Policymakers suggest the formation of a separate MPDSR-focused cell to concentrate on the successful implementation of the program, including partners supporting it in different districts. The capacity of MPDSR data analysis needs to be further strengthened by local health managers.