The implementation setting included three English-speaking East-African countries (Malawi, Tanzania and Uganda) and one French-speaking West-African country (Benin). These countries were purposively selected as they reflect different health system contexts and leveraged pre-existing research collaborations. ALERT was implemented in 16 district and referral hospitals in Benin, Malawi, Tanzania, and Uganda, including three public and one private-for-non-profit hospitals in each country.
We developed and successfully implemented an intervention composed of co-design, competency-based training, quality improvement and mentoring.
A total of 134,630 women were included in our stepped-wedge trial, with 60,403 during the comparison and 74,227 during the intervention period. More than 4,000 women were interviewed on their experiences with the care received in the hospitals after giving birth. Our preliminary results indicate that the intervention successfully reduced perinatal mortality by one-quarter (0.76 95% CI 0.63-0.91). Secondary indicators of maternal and perinatal morbidity also showed a reduction in morbidity, such as hypoxic-ischemic events (Apgar score <7). The intervention led to an increase in Caesarean sections. We observed no impact on the reduction of mistreatment during labour and birth, although the overall prevalence was low.
The study found that the total economic cost of implementing the intervention across four countries—Benin, Malawi, Tanzania, and Uganda—was approximately 130,000 USD, with significant variability due to local conditions such as travel distances and the frequency of the hospital visits by the in-country implementation teams. The Incremental Cost-Effectiveness Ratio (ICER) was calculated to compare the cost per perinatal death averted, showing costs around $800 per perinatal death averted in Benin and Uganda, while Malawi (around $2000) and Tanzania ($3200) had a higher ICER. In Tanzania, the ICER was especially high when based on pre-intervention perinatal mortality rates ($20,000). Preliminary results estimating a 24% reduction in perinatal mortality indicate that the intervention could improve health outcomes for newborns born in hospitals at a reasonable cost in typical high mortality and low resource settings in Sub-Saharan Africa, though further research is needed to explore country-specific dynamics.
The realist evaluation indicated improved hospital-level decision-making through the clinical training sessions in the training programme. Interviews supported the usefulness of the training as a refresher of previously obtained knowledge and a source of new knowledge, such as on ambulation during labour and birth position. The QI component contributed to improving care in all hospitals, with a focus on the improvement of the quality of care, and this contributed to better interactions between the different services related to emergency Caesarean sections. Departments such as the laboratory, pharmacy, and operating room improved their responsiveness, which facilitated emergency Caesarean sections. Tracking the QI component’s work helped facility managers in charge lobby for cleaning materials to improve infection prevention, foetal heart rate monitoring devices and drugs for emergency management of women, thus increased negotiating power.
The dissemination meetings in the four countries suggest that the ALERT team created a major momentum beyond the hospitals, with the intervention receiving traction by district health authorities and national policy makers. The two key innovations receiving the most interest was i) the perinatal e-registry and ii) the co-design component to better engage health care workers. The co-design process was emphasised as an empowering bottom-up approach that helped to identify and understand i) the critical problems as well as opportunities in the maternity wards, ii) the training needs to define the key priority areas for the competency-based training, and iii) prioritise the improvement work.
