Aim: As part of a multi-country implementation trial, we tested a regionally specific model of kangaroo mother care (KMC). Effective KMC was defined as ≥8 h of newborn-caregiver skin-to-skin contact daily plus exclusive breast feeding. The study was designed to achieve ≥80+% effective KMC coverage at the population level. Methods:The Amhara KMC model was designed using global evidence, formative research in the region and input from government officials, clinicians, newborn families and global scientists. We optimised the initial model using continuous quality improvement with process feedback, outcome measurement and collaborative redesign. Outcomes from the evaluation period are reported.Results: At discharge, the final model resulted in a median of 16 h per day of skin-toskin contact with 63% effective KMC coverage. Fifty-three percent sustained effective KMC to 7 days post-discharge.Conclusions: It is possible to achieve high coverage (63%), high-quality KMC at public hospitals without prior KMC services using government-owned, multisectoral collaborative design. Targeted co-design, real-time data and customisation of KMC interventions with input from impacted stakeholders was critical in achieving high coverage and sustained quality.

Background Measuring facility readiness to manage basic obstetric emergencies is a critical step toward reducing persistently elevated maternal mortality ratios (MMR). Currently, the Signal Functions (SF) is the gold standard for measuring facility readiness globally and endorsed by the WHO. The presence of tracer items classifies facilities’ readiness to manage basic emergencies. However, research suggests the signal functions may be an incomplete indicator. The Clinical Cascades (CC) have emerged as a clinically-oriented alternative to measuring readiness. The purpose of this study is to determine Amhara's clinical readiness and quantify the relationship between SF and cascade estimates of readiness. Methods Data were collected in May 2021via Open Data Kit (ODK) and KoBo Toolbox. We surveyed 20 hospitals across three levels of the health system. Commodities were used to create measures of SF-readiness (e.g., % tracers) and cascade-readiness. We calculated differences in SF and cascade estimates and calculated readiness loss across five emergencies and 3 stages of care in the cascades. Results The overall Signal Function estimate for all 6 obstetric emergencies was 17.5% greater than the estimates using the cascades. Consistent with global patterns, hospitals were more prepared to provide medical management (70.0% ready) compared to manual procedures (56.7% ready). The SF overestimate was greater for manual procedures (26.7% overall for retained placenta and incomplete abortion) and less for medical treatments (8.3%). Hospitals were least prepared to manage retained placentas (30.0% of facilities were ready at treatment and 0.0% were ready at monitor and modify) and most able to manage hypertensive emergencies (85.0% of facilities were ready at the treatment stage). When including protocols in the analysis, no facilities were ready to monitor and modify the initial therapy when clinically indicated for 3 common emergencies—sepsis, post-partum hemorrhage and retained placentas. Conclusions We identified a significant discrepancy between SF and CC readiness classifications. Those facilities that fall within this discrepancy are unprepared to manage common obstetric emergencies, but regional health planners are unable to identify the need. Future research should explore the possibility of modifying the SF or replacing it with a new readiness measurement.

Measuring facility readiness to manage basic obstetric emergencies is a critical step toward reducing persistently elevated maternal mortality ratios (MMR). Currently, the Signal Functions (SF) is the gold standard for measuring facility readiness globally and endorsed by the World Health Organization. The presence of tracer items classifies facilities’ readiness to manage basic emergencies. However, research suggests the SF may be an incomplete indicator. The Clinical Cascades (CC) have emerged as a clinically-oriented alternative to measuring readiness. The purpose of this study is to determine Amhara’s clinical readiness and quantify the relationship between SF and CC estimates of readiness. Data were collected in May 2021via Open Data Kit (ODK) and KoBo Toolbox. We surveyed 20 hospitals across three levels of the health system. Commodities were used to create measures of SF-readiness (e.g., % tracers) and CC-readiness. We calculated differences in SF and CC estimates and calculated readiness loss across six emergencies and 3 stages of care in the cascades. The overall SF estimate for all six obstetric emergencies was 29.6% greater than the estimates using the CC. Consistent with global patterns, hospitals were more prepared to provide medical management (70.0% ready) compared to manual procedures (56.7% ready). The SF overestimate was greater for manual procedures 33.8% overall for retained placenta and incomplete abortion) and less for medical treatments (25.3%). Hospitals were least prepared to manage retained placentas (30.0% of facilities were ready at treatment and 0.0% were ready at monitor and modify) and most prepared to manage hypertensive emergencies (85.0% of facilities were ready at the treatment stage). When including protocols in the analysis, no facilities were ready to monitor and modify the initial therapy when clinically indicated for 3 common emergencies—sepsis, post-partum hemorrhage and retained placentas. We identified a significant discrepancy between SF and CC readiness classifications. Those facilities that fall within this discrepancy are unprepared to manage common obstetric emergencies, and employees in supply management may have difficulty identify the need. Future research should explore the possibility of modifying the SF or replacing it with a new readiness measurement.