A hand-hygiene and sepsis-bundle programme cut severe maternal infection by a third

Maternal infection and sepsis remain major causes of maternal death and severe illness worldwide, particularly in low- and middle-income countries, and the reasons are less about missing knowledge than about missing execution: inconsistent implementation of evidence-based recommendations, and delays in detecting and treating sepsis once it starts [s1].

A cluster-randomised trial published in the New England Journal of Medicine on 19 November tested whether a programme designed around those two failures could shift outcomes at scale [s1].

What was tested

The Active Prevention and Treatment of Maternal Sepsis programme — APT-Sepsis — was built to support health care providers in achieving three goals: adherence to WHO hand-hygiene standards; adoption of evidence-based practices for maternal infection prevention and management; and early detection of sepsis with use of the FAST-M treatment bundle, which stands for fluids, antibiotics, source control, transfer if required, and monitoring [s1].

The control group received usual care along with dissemination of guidelines [s1]. That is a meaningful choice: the comparator was not "nothing," it was the guidelines themselves, which is the closest available test of whether implementation support adds anything beyond publishing recommendations.

Fifty-nine health facilities in Malawi and Uganda were randomly assigned — 30 clusters to the intervention and 29 to usual care [s1]. During the trial, 431,394 women gave birth across those facilities [s1].

The result

The primary outcome was a composite of infection-related maternal death, infection-related near-miss event (in which a woman survived a life-threatening complication), or severe infection-related illness, defined as deep surgical-site, deep perineal, or body-cavity infection, among women who were pregnant or had recently been pregnant [s1].

A primary-outcome event occurred in 1.4% of patients in the intervention group and 1.9% in the usual-care group — a risk ratio of 0.68 (95% CI 0.55 to 0.83; P<0.001) [s1].

The effect was generally consistent between the two countries and across facilities of different sizes, and was sustained over time [s1]. That last point is not a footnote. Implementation programmes commonly show an early effect that decays as attention moves elsewhere; a sustained effect is the harder result.

The trial was funded by the Joint Global Health Trials scheme and others, and is registered as ISRCTN42347014 [s1].

Where the bundle came from

The FAST-M bundle at the centre of the intervention was not new in 2025. A before-and-after evaluation published in BJOG in 2021 tested it across 15 government healthcare facilities in Malawi, comparing a two-month baseline phase with a six-month intervention phase [s2].

Its findings were about process rather than outcomes. After implementation, women were more likely to have a complete set of vital signs taken on admission to the wards — 169 of 252 (67.1%) versus 0 of 163 (0%), P<0.001 [s2]. Recognition of suspected maternal sepsis improved, with 107 of 166 cases (64.5%) identified after the intervention against 12 of 106 (11.3%) before, P<0.001 [s2]. Women were more likely to receive antibiotics within one hour of recognition — 72 of 107 (67.3%) versus 3 of 12 (25.0%), P=0.004 [s2]. Receipt of all bundle components within one hour improved from 0 of 12 (0%) to 21 of 107 (19.6%), but that comparison did not reach significance (P=0.091) [s2].

The 2021 study concluded that implementation was feasible and improved recognition and management [s2]. What it could not show, by design, was whether better recognition translated into fewer deaths and near-misses. That is the gap the 2025 trial fills.

What the trial does not establish

Because APT-Sepsis is a multicomponent programme, the trial cannot say which component did the work. Hand hygiene, infection-prevention practice, and sepsis detection were bundled together by design, and the result is a verdict on the package [s1].

Cluster-randomised designs in health facilities also cannot blind staff to allocation. Some of the measured improvement in a composite that includes clinician-diagnosed severe illness could reflect changed documentation rather than changed biology — though infection-related death is harder to document differently.

The absolute difference is half a percentage point [s1]. Across 431,394 births, that is a large number of events; for an individual facility, it is a change that would be invisible without systematic measurement.

Why it matters

Two of the three goals in the intervention — hand hygiene to WHO standards, and prompt antibiotics for suspected sepsis — have been recommended for decades. The trial's contribution is evidence that a structured programme to make those recommendations happen produces a measurable reduction in severe maternal infection outcomes, in two African health systems, at scale [s1].

This article is informational and does not constitute medical advice.