A Comprehensive Strategy to Mitigate Institutional Maternal Mortality: Lessons From a Quality Improvement Initiative in Brazilian Maternity Hospitals

<b>Introduction:</b> Many pregnancy-related deaths can be avoided if health care workers reliably provide prompt, evidence-based care during the final stages of pregnancy and delivery. We report the impact on the institutional maternal mortality ratio (iMMR) of a quality improvement (QI) initiative that focused on establishing timely and reliable obstetric care in public Brazilian hospitals for the primary causes of maternal deaths.<b>Methods:</b> The QI initiative used a quasi-experimental time-series design implemented in 19 Brazilian maternity hospitals comparing 3 periods: baseline (January 2018 to November 2019), implementation (December 2019 to March 2021), and post-implementation (April 2021 to September 2021). We used a sequential approach, referred to as the "4Rs" (Recognize, Rescue, Reassess, and Refer), with the Modified Early Obstetric Warning Score (MEOWS) as a key tool, to identify clinical deterioration and implement care bundles directed at the management of 3 principal life-threatening conditions around the time of birth: postpartum hemorrhage, sepsis, and hypertensive disorders of pregnancy (HDPs). Clinical staff received tailored training and support within a structured learning system that brought multiple teams together to rapidly test and implement evidence-based changes.<b>Results:</b> Comparing the baseline with the implementation period, run charts detected an overall reduction of 34.2% in all causes of maternal deaths (from 83.7 to 55 deaths per 100,000 live births). Additionally, the iMMR due to the 3 analyzed life-threatening conditions decreased by 60.9% between baseline and the combined implementation and post-implementation periods (from 44.1 to 17.2 deaths per 100,000 live births); postpartum hemorrhage-related deaths were reduced by 72.9% (from 11.7 to 3.17 deaths per 100,000 live births) and sepsis-related deaths were reduced by 100% (from 20.4 to 0 deaths per 100,000 liv births). No changes by HDP-related deaths were observed. There was a 178% increase in iMMR by all causes (mainly attributed to COVID-19) during the post-implementation period, but deaths related to life-threatening conditions either remained stable or decreased even further.<b>Conclusion:</b> Optimization of the steps in a care sequence and reliable implementation of care bundles directed at the principal life-threatening conditions around the time of birth, using QI, tailored training, and the MEOWS tool, seem to be feasible and promising approaches for reducing iMMR.