Using routine health data and intermittent community surveys to assess the impact of maternal and neonatal health interventions in low-income countries: A systematic review.

BACKGROUND

There is a need to provide increased evidence on effective interventions to reduce maternal and neonatal mortality in low- and middle-income countries (LMICs).

OBJECTIVES

To summarize the breadth of knowledge on using routine data (Routine Health Information Systems [RHIS] and Intermittent Community Surveys [ICS]) for well-designed maternal and neonatal health evaluations in LMICs.

SEARCH STRATEGY

We searched reports and articles published in Embase, Medline, and Google scholar. Selection criteria Studies were considered for inclusion if they were carried out in LMICs, using RHIS or ICS data with experimental or quasi-experimental design.

DATA COLLECTION AND ANALYSIS

A form was used to collect information on indicators used for interventions' impact assessment. Descriptive statistics and multiple correspondence analyses were then performed.

MAIN RESULTS

Of the 1201 publications identified, 46 studies met the inclusion criteria. Most of these were using RHIS data (n=40), mainly extracted from health facility registers (n=34), and non-controlled before and after design (n=30). The indicators, which were mostly reported, were related to the use of healthcare services (n=36) and maternal/neonatal health outcomes (n=31). Few studies used ICS data (n=6) or indicators of severity (n=2).

CONCLUSION

RHIS and ICS data should be increasingly used for impact studies on maternal and neonatal health in LMICs.