Evaluation of the Perinatal Death Surveillance and Response System in Eastern Province, Rwanda

Technical Summary: Evaluation of the Perinatal Death Surveillance and Response System in Eastern Province, Rwanda

Problem Statement
Perinatal mortality remains a critical public health challenge, particularly in low- and middle-income countries. While Rwanda established the Maternal, Perinatal, and Child Death Surveillance and Response (MPCDSR) system to improve the identification, reporting, review, and response to perinatal deaths, the performance of this system in specific settings has not been comprehensively assessed. Despite the existence of national guidelines and committees, perinatal mortality rates remain unacceptably high and largely preventable. There is a lack of evidence regarding the operational effectiveness of the Perinatal Death Surveillance and Response (PDSR) system in the Eastern Province, specifically regarding its ability to detect, notify, analyze, and act upon perinatal deaths.

Methodology
This study employed a descriptive cross-sectional evaluation design, adhering to the updated Centers for Disease Control and Prevention (CDC) guidelines for evaluating public health surveillance systems.

• Setting: The evaluation was conducted in the Eastern Province of Rwanda, covering six randomly selected district hospitals and twelve purposively selected health centers (two attached to each hospital, with one selected for having a high perinatal mortality rate and one for having a low rate).
• Data Sources: The study utilized a mixed-methods approach combining:
  • Retrospective review of routinely collected surveillance data from 2021 to 2025 (4,414 recorded perinatal deaths).
  • Document reviews of registers, notification forms, and HMIS reports.
  • Structured electronic questionnaires administered to 136 healthcare workers (members of MPCDSR committees) to assess system attributes and core functions.
• Evaluation Framework: The system was assessed based on eight attributes (usefulness, data quality, simplicity, acceptability, stability, timeliness, flexibility, representativeness) and five core functions (case detection, notification, data collection/management, analysis/interpretation, response/action).
• Scoring: Performance was categorized as poor (≤50%), fair (51–79%), or good (≥80%) based on composite scores derived from indicator percentages.

Key Results
The evaluation revealed a system with significant strengths in documentation but critical weaknesses in the analytical and responsive phases of the surveillance cycle.

• Epidemiology: A total of 4,414 perinatal deaths were recorded between 2021 and 2025. The majority (97%) occurred in hospitals, with minimal occurrences in health centers (3%) and none recorded in the community or private clinics. The data showed a 44% reduction in perinatal deaths over the study period, declining from 1,586 in 2021–2022 to 890 in 2024–2025.
• Surveillance Attributes:
  • Good Performance: Data quality (90.5%), timeliness (92.0%), usefulness (84.2%), simplicity (80.1%), and flexibility (80.3%) were rated as "good."
  • Fair Performance: Acceptability (77.5%), representativeness (74.2%), and stability (59.8%) were rated as "fair." Stability was notably low due to technical interruptions and inadequate support.
• Core Functions:
  • Strengths: Data collection and management achieved a perfect score (100%), indicating that all notified cases were fully documented.
  • Weaknesses:
    • Notification: Only 47% of detected cases were formally notified through the surveillance system.
    • Case Detection: 69% of deaths in registers were detected by the system.
    • Analysis and Interpretation: This function scored 0%, indicating that no perinatal deaths were analyzed for causes or avoidable factors during the evaluation period.
    • Response and Action: Only 58% of reviewed cases resulted in documented actions or feedback.

Key Contributions
This study provides one of the first comprehensive evaluations of the PDSR system in Rwanda's Eastern Province, assessing both system attributes and core functions simultaneously. It highlights a specific operational disconnect: while the system is effective at recording data once a case is notified (high data quality and timeliness), it fails at the upstream process of ensuring all detected deaths are formally entered into the system (low notification) and the downstream process of utilizing data for learning (zero analysis). The findings underscore that high data completeness does not equate to a functional surveillance system if the critical steps of notification and analysis are missing.

Significance and Claims
The authors claim that the PDSR system in the Eastern Province serves as a functional foundation for reducing perinatal mortality, evidenced by the observed 44% decline in deaths and high scores in data quality and timeliness. However, the study concludes that the system's effectiveness is currently compromised by gaps in notification, analysis, and the implementation of response actions.

The paper asserts that strengthening the system requires:

1. Improving the completeness of notification to ensure all detected deaths are captured.
2. Establishing routine mechanisms for the analysis and interpretation of surveillance data.
3. Enhancing feedback loops to ensure that audit findings lead to actionable changes.
4. Addressing system stability issues, such as technical support and resource availability.

The authors emphasize that without addressing these specific gaps, the potential of the MPCDSR framework to prevent avoidable perinatal deaths remains unrealized. The study does not claim that the PDSR system alone caused the reduction in mortality, noting that the observed trend may also reflect concurrent improvements in broader maternal and newborn health services.