Maternal mortality and severe maternal morbidity remain a major public health concern in low-resource settings. Although Ethiopia has reduced its maternal mortality ratio over recent decades, rates remain well above global targets, and many deaths and morbidities are attributable to preventable conditions such as hemorrhage, infection, hypertensive disorders, complications of delivery, and unsafe abortion. Delays in recognizing and managing clinical deterioration contribute substantially to adverse outcomes.
The Maternal Early Warning System (MEWS) is a structured, bedside monitoring tool that tracks physiological parameters and assigns color-coded escalation levels—Green, Yellow, Red—to prompt predefined clinical responses. Prior evidence from high-income and some middle-income countries indicates that MEWS can improve routine monitoring, reduce delays in response to abnormal findings, and lower the severity of maternal morbidity. However, effectiveness data from low-resource settings including Ethiopia have been limited, and implementation fidelity can be a challenge.
This study evaluated the impact of a statistically developed and validated MEWS chart on predicting and reducing severe maternal outcomes among obstetric inpatients in four public hospitals in North Shewa Zone, Ethiopia.
A parallel, quasi-experimental design was implemented. Recruitment occurred between 05/05/2025 and 31/08/2025. A total of 1,138 obstetric inpatients were enrolled and evenly allocated to intervention and control groups (569 per group). The intervention replaced the standard vital signs sheet with the validated MEWS chart in the intervention hospitals. Control hospitals continued with standard clinical monitoring.
The study took place in North Shewa Zone, Amhara Region. The zone includes multiple public hospitals serving a large population. Eligible participants were pregnant and postpartum women admitted with antepartum, intrapartum, or postpartum complications, abortion-related cases, those who had obstetric or gynecologic surgery, and those with high-risk conditions. Exclusions included women in normal labor planned for discharge within 24 hours, deaths from accidental causes, and those transferred directly to ICU without inpatient admission.
The MEWS chart used in the intervention arm is a simple observation-based tool incorporating twelve maternal clinical parameters. Parameters included routine vital signs, oxygen saturation, urine output, level of consciousness, pain assessment, and postpartum-specific items such as vaginal bleeding, uterine contraction, and perineal status. Each parameter is categorized into color-coded zones: Green (normal), Yellow (moderate abnormality; increased observation required), and Red (severe derangement; immediate attention required). The MEWS chart had been previously validated and is described in the study’s supplementary materials.
Primary and secondary outcomes included timing metrics (time from admission to first trigger, time from trigger to physician evaluation, time from trigger to clinical intervention), resource use (number of ultrasound scans, length of hospital stay), and clinical endpoints including occurrence of severe maternal outcomes and frequency of clinical triggers. The authors used multivariate generalized estimating equation (GEE) models with Poisson regression to compare outcomes between groups and to estimate adjusted risk ratios (aRR) with 95% confidence intervals.
Implementation of the MEWS chart was associated with measurable improvements in process and clinical outcomes compared with standard monitoring. The mean time from admission to the first trigger was reduced by 4.7 hours in the intervention group (5.61 hours) compared with the control group (10.27 hours). After a trigger, mean time to physician evaluation was shorter by 22.6 minutes in the MEWS group (49.3 vs. 71.9 minutes), and mean time from trigger to clinical intervention was reduced by 11.3 minutes (14.6 vs. 25.9 minutes).
Resource use differed between groups: women monitored with MEWS had fewer ultrasound scans on average (1.32 vs. 2.30) and a shorter mean hospital stay by approximately 0.5 days (4.83 vs. 5.29 days).
Clinically, monitoring with the MEWS chart was associated with a lower risk of severe maternal outcomes. The adjusted risk ratio for severe maternal outcomes in the MEWS group was 0.85 (95% CI: 0.73–0.99), indicating a 15% relative reduction after adjustment. In addition, women monitored with MEWS were more likely to be triggered for timely clinical response (aRR = 1.15, 95% CI: 1.03–1.28).
In this quasi-experimental evaluation, replacing routine monitoring sheets with a validated MEWS chart improved early detection of clinical deterioration and shortened the interval to physician assessment and intervention. The intervention coincided with reductions in diagnostic imaging use and length of stay, and with a modest but statistically significant reduction in severe maternal outcomes.
These findings align with international literature suggesting that structured early warning systems can standardize monitoring, prompt escalation, and reduce delays that contribute to preventable maternal morbidity and mortality. The study offers context-specific evidence from a low-resource Ethiopian setting, addressing a gap in the literature where implementation challenges have previously been noted.
Limitations noted in the report include the quasi-experimental design and the number of clusters, which may limit generalizability. The authors recommend further studies with a larger number of clusters and assessments across different maternal risk groups and settings to confirm effectiveness and to guide scale-up.
Implementation of the MEWS chart in the study hospitals was associated with earlier detection of deterioration, faster clinical response, reduced use of some investigations, shorter hospital stays, and a lower adjusted risk of severe maternal outcomes. The authors suggest that MEWS could be considered for broader evaluation and potential scale-up in similar low-resource settings, but they call for larger cluster studies to evaluate effectiveness across diverse contexts and risk groups.
The trial was registered with the Pan African Clinical Trial Registry (PACTR202506739780428). The study received funding from the International Institute for Primary Health Care–Ethiopia (IPHC-E), and every study phase was monitored by Debre Berhan University, Asrat Woldeyes Health Science Campus, and IPHC-E; the funder had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript. The authors report that all relevant data are incorporated within the paper and that original raw data were submitted as supplementary material.