Can a quality improvement project impact maternal and child health outcomes at scale in northern Ghana?

Background: Quality improvement (QI) interventions are becoming more common in low- and middle-income countries, yet few studies have presented impact evaluations of these approaches. In this paper, we present an impact evaluation of a scale-up phase of 'Project Fives Alive!', a QI intervention in Ghana that aims to improve maternal and child health outcomes. 'Project Fives Alive!' employed a QI methodology to recognize barriers to care-seeking and care provision at the facility level and then to identify, test and implement simple and low-cost local solutions that address the barriers.

Methods: A quasi-experimental design, multivariable interrupted time series analysis, with data coming from 744 health facilities and controlling for potential confounding factors, was used to study the effect of the project. The key independent variables were the change categories (interventions implemented) and implementation phase - Wave 2a (early phase) versus Wave 2b (later phase). The outcomes studied were early antenatal care (ANC), skilled delivery, facility-level under-five mortality and attendance of underweight infants at child welfare clinics. We stratified the analysis by facility type, namely health posts, health centres and hospitals.

Results: Several of the specific change categories were significantly associated with improved outcomes. For example, three of five change categories (early ANC, four or more ANC visits and skilled delivery/immediate postnatal care (PNC)) for health posts and two of five change categories (health education and triage) for hospitals were associated with increased skilled delivery. These change categories were associated with increases in skilled delivery varying from 28% to 58%. PNC changes for health posts and health centres were associated with greater attendance of underweight infants at child welfare clinics. The triage change category was associated with increased early antenatal care in hospitals. Intensity, the number of change categories tested, was associated with increased skilled delivery in health centres and reduced under-five mortality in hospitals.

Conclusions: Using an innovative evaluation technique we determined that 'Project Fives Alive!' demonstrated impact at scale for the outcomes studied. The QI approach used by this project should be considered by other low- and middle-income countries in their efforts to improve maternal and child health.