New approaches to identifying and managing risk during pregnancy

No low-cost test for early prediction of preeclampsia (PE) risk exists. Yet, PE is a cause of 10-15% of maternal deaths globally. Early prediction would enable preventive treatment with low dose aspirin, timely triage and medical intervention, especially in low resource settings. Our goal is to develop a low-cost test for PE prediction at 10-12 weeks of gestation. In our previous study of pregnant women at Stanford we used machine learning to identify four highly predictive proteins. We validated the use of low-cost enzyme-linked immunosorbent assays (ELISA) technology for measuring these biomarkers with high accuracy (area under the receiver operating characteristic curve [AUC]: 0.94, 95% confidence interval [CI]: 0.93, 0.95). Further validation of these biomarkers using ELISAs is being performed in two large cohorts from Bangladesh and the United States. Using low-cost ELISA technology will facilitate a widely available test, drastically improving prenatal care globally.

National guidelines recommend the use of glycosuria and gestational diabetes mellitus (GDM) screening checklist for screening GDM during ANC. GDM screening is not routinely done in Tanzania. We introduced and screened women attending ANC in two district hospitals using a mixed methods approach. One in ten women had diabetes during pregnancy. Women liked the opportunity for GDM screening and wanted it to be routinely done for all pregnant women. They recommended quality education on GDM should be included during ANC and immediate linkage to specialized care for those who screen positive for GDM. Women were willing to do a fasting glucose test. GDM screening should be routinely done during ANC. Glycosuria should not be used as a screening test for GDM. Women's voices should be considered in the operationalization of GDM screening during ANC.

The Caribbean nation of Haiti experiences ongoing sociopolitical crisis. This situation disrupts HIV care access, mainly affecting vulnerable groups including pregnant women and lactating mothers and children and economically disadvantaged individuals. Zanmi Lasante launched a quality improvement strategy from February to September 2024 to enhance viral load coverage at their rural site in Haiti's Centre department from 67% to 95%. The project implemented two essential interventions. They consisted of community based viral load sample collection procedures and early viral load testing during standard clinic appointments; both approaches underwent two Plan Do Study Act (PDSA) cycles. The intervention period resulted in substantial growth of sample collection coverage rates which rose from 67% to 92%. We improved patient care continuity and lowered the chances of mother to child transmission. This also minimized service access disparities. The achievement of UNAIDS' 95 targets in fragile settings depends heavily on community-based intervention strategies.

Nigeria bears a high burden of maternal mortality rates, largely due to delays in seeking and accessing appropriate level of care, contributing to preventable deaths. To address this, we evaluated the effectiveness of the Antenatal Risk Stratification (ARS) tool, a digital solution designed to identify high-risk pregnancies and support timely referrals to appropriate health facilities. Implemented over 18 months, across selected LGAs in Kano and Lagos, the ARS tool was tailored to serve four distinct categories of users, based on their literacy, mobile phone access, and level of engagement with healthcare systems. Both the USSD and HeliumDoc platforms showed excellent performance, with sensitivity rates exceeding 99% in identifying individuals at risk. The tool successfully flagged critical conditions, such as pre-eclampsia, facilitating timely medical intervention. These results demonstrate the potential of digital solutions to enhance frontline care and strengthen maternal health systems, particularly in low-resource environments.