Evaluating the Quality of Recommendations From Maternal Death Reviews—Fort Portal, Uganda [13R

Abstract

INTRODUCTION: Facility-based Maternal Death Reviews (MDRs) aim to identify contributors to maternal deaths, generate recommendations to prevent future deaths, and implement and follow-up those recommendations. Globally, MDR committees' recommendations lack clearly-defined achievable and measureable actions, and an internationally-recognized standard framework for evaluating recommendations does not exist. METHODS: We developed a scoring framework for MDR recommendations based on International Federation of Gynecology and Obstetrics recommendation structuring guidelines. We piloted the framework to evaluate the quality of 2014 MDR recommendations generated by three facility-based MDR committees in Fort Portal, Uganda. We developed a composite score for grading quality of recommendations (0–6) composed of: specificity (0-vague, 1-clear, 2-explicit), relevance to a committee identified “avoidable factor” (0-no, 1-yes), presence of implementation method (0-no, 1-yes), individual responsibility for implementation (0-no, 1-yes), and timeline (0-no, 1-yes). Two scorers coded recommendations independently and reviewed differences to reach consensus. RESULTS: Committees generated 83 recommendations during 42 MDRs; 61% prescribed facility-based changes. The mean composite score was poor (1.8, range 0–5). Recommendation specificity was “vague” in 20% of recommendations, “clear” in 58%, and “explicit” in 22%. Half (51%) of recommendations were relevant, although 81% of recommendations lacked a method of implementation, and 88% had no implementer assigned. All recommendations lacked a timeline. CONCLUSION: A framework for evaluating the quality of MDR recommendations was developed and identified specific areas for improvement. Clearly defining method, implementer and timeline, and improving recommendations' specificity and relevance should lead to better implementation which will help to reduce maternal mortality.