A Multicomponent Intervention to Improve Maternal Infection Outcomes.
Lissauer D, Gadama L, Waitt C, Whyte S, Burnside G, Anilkumar A, Makuluni R, Okwaro P, Yang L, Waitt P, Musopole O, Bilesi R, Maseko B, Lwasa J, Mugahi R, Olaro C, Lamorde M, Makuta M, Kachiwaya C, Mkandawire T, Malunga A, Chitsulo N, Abitimo P, Ayabo T, Weeks A, Martin J, Hemming K, Gallos I, Monk EJM, Riches J, Chapuma C, Nanyondo S J, Lorencatto F, Monahan M, Allegranzi B, Dunlop C, Atkins L, Rosala-Hallas A, Roberts T, Gamble C, Malata A, Desmond N, Kommwa E, Merriel A, Parry-Smith W, Smith R, Ndumu I, Williams E, Faque B, Banda G, Nyondo-Mipando AL, Twimukye A, Chater T, Diplas A, Brizuela V, Souza JP, Rylance J, Cheshire J, Hawker L, Coomarasamy A, Bonet M
- DOI
- 10.1056/NEJMoa2512698
- Record issued
- 2026-08-15
- Engine
- 7.39.0
- Exported
- 2026-09-20
Prepared by Alpha1. This document is confidential: it is intended for the recipient it was shared with and must not be redistributed. The live record at alpha1science.com/verify/a4a7423b-4feb-43c7-acc6-4e1cb91fc92c is authoritative.
How this rating was calculated
- IntegrityIntegrity concern−0.5★
- ReportingData & code availability partially met−0.25★
- The numeric-impossibility checks (GRIM/GRIMMER/DEBIT/SPRITE) did not run: 12 reported means were read, and their group size is not stated where the values are printed (this source has no machine-readable table structure). These checks need the count the mean was averaged over, so none was performed.
- No data or code availability links were detected to verify.
This Kaimen Rigor review uses Kaimen Rigor reviewers trained on a curated corpus of high-fidelity and retracted papers, with expert supervision and curation. It can still make mistakes; verify each finding against the source before relying on it.
This is a well-conducted cluster-randomized trial with rigorous design, clear ethical approvals, and appropriate statistical methods. The main weaknesses are the lack of a clear data availability statement and minor reporting gaps (statistical software, demographics, reporting guideline).
Both reviewers independently scored all eight dimensions and agreed on all statuses; no divergence to reconcile. The statistics verification component checked only 2 tests (both consistent); the remaining statistics are unverified. The citation check found no retracted or non-existent references.
Numerical inconsistencies
1 finding · worst lowValues that contradict each other or are impossible for the stated sample: recomputed p-values and test statistics, GRIM/GRIMMER checks on summary numbers, percentages against their own counts, totals against their parts, and estimates against their own confidence intervals.
- Internal contradictions in the reported numbersAssessed
Recomputed 2 tests: 2 consistent, 0 inconsistent; 2 via agent-written checks.
- CONSISTENTreported p < .001 · recomputed p = <.001Reviewers 1, 2Primary outcome risk ratio p-value
“risk ratio, 0.68; 95% CI, 0.55 to 0.83; P<0.001”
Taken as given: The risk ratio is 0.68 with 95% CI 0.55 to 0.83.; The CI is two-sided at 95%.; The p-value is two-tailed.Method: Recomputed p-value from the reported risk ratio and 95% CI using the pCI function for a ratio.How we recomputed it: pCI(0.68, 0.55, 0.83, 1) - CONSISTENTreported p < .001 · recomputed p = <.001Reviewers 1, 2Severe infection-related morbidity risk ratio p-value
“risk ratio, 0.68; 95% CI, 0.55 to 0.84”
Taken as given: The risk ratio is 0.68 with 95% CI 0.55 to 0.84.; The CI is two-sided at 95%.; The p-value is two-tailed.Method: Recomputed p-value from the reported risk ratio and 95% CI using the pCI function for a ratio.How we recomputed it: pCI(0.68, 0.55, 0.84, 1)
- lowinternal contradictionThe number of live births in Table 1 (94,730 intervention, 95,770 control) sums to 190,500, which matches the baseline phase total. However, the intervention phase totals (124,298 and 116,596) sum to 240,894, which matches the stated total. This is consistent.
“There were a total of 431,394 women with live births during the study (190,500 in the baseline phase and 240,894 in the intervention phase, .)”
Table 1Find in source
Overstated conclusions
1 finding · worst lowConclusions that reach past what the paper's own results support — including a significance claim that no longer holds when the statistic is recomputed, and efficacy resting on an unvalidated surrogate endpoint.
- Conclusions only partially backed by the presented evidenceAssessed
8 major claims checked against the paper's own evidence: 2 only partially supported (evidence backs part of the claim; gaps or caveats remain); the rest adequately supported.
- partialReviewer 1The intervention improved early detection of sepsis and use of the FAST-M bundle.Some implementation outcomes improved (e.g., complete observations at admission), but the difference in antibiotics within 1 hour was not statistically significant (mean difference 8%, 95% CI -3% to 19%).Evidence: Complete observations at admission: 48% vs 15%, mean difference 32% (95% CI 25-40%); antibiotics within 1 hour: 44% vs 38%, mean difference 8% (95% CI -3% to 19%).
“Several measures related to Goal 3 also supported primary outcomes findings; for example, complete observations at admission in 48% vs 15%, respectively (mean difference 32%; 95% CI 25% to 40%) and, among patients with suspected sepsis, antibiotics administered within 1 hour in 44% vs 38%, respectively (mean difference 8%; 95% CI -3% to 19%).”
ResultsFind in source - partialReviewer 2The intervention is feasible beyond the trial setting.The paper argues feasibility based on minimal additional resources, but this is an inference from the trial context and not directly tested.Evidence: Discussion paragraph 3.
“The intervention involved the provision of relatively few additional resources beyond what was generally available within the hospital systems, suggesting feasibility beyond the trial setting.”
DiscussionFind in source - supportedReviewers 1, 2The APT-Sepsis intervention reduced the composite of maternal infection-related mortality and severe morbidity compared to usual care.The primary outcome analysis shows a statistically significant reduction with a risk ratio of 0.68 (95% CI 0.55-0.83, P<0.001), directly supporting the claim.Evidence: Primary outcome result: 1752/124,298 (1.4%) vs 2208/116,596 (1.9%), RR 0.68 (0.55-0.83), P<0.001.
“The incidence of maternal infection–related mortality or severe morbidity was 1.4% in the intervention group and 1.9% in the usual care group (risk ratio, 0.68; 95% confidence interval, 0.55 to 0.83; P<0.001).”
AbstractFind in source - supportedReviewers 1, 2The effect was consistent across countries and facility size.The paper states the effect was generally consistent across subgroups, and forest plots are referenced, though not all subgroup results are shown in the text.Evidence: Subgroup analyses by country and facility size are mentioned, with forest plots in Figure 3.
“This effect appeared generally consistent between countries ( and ) and across small, medium and large facilities.”
ResultsFind in source - supportedReviewers 1, 2The effect was sustained over time.The paper shows a progressive decrease in incidence and increasing effect size over the intervention period, supporting the claim of sustainability.Evidence: Incidence decreased from 2.4% at baseline to 0.9% in the final month; risk ratio improved from 0.92 to 0.53.
“The incidence of the primary outcome in the intervention group progressively decreased following randomization, with a mean rate of 2.4% at baseline, 2% in the first month following implementation and completion of the transition phase, and 0.9% in the final month of the study.”
ResultsFind in source - supportedReviewer 1The intervention improved hand hygiene compliance.Implementation outcome shows a significant improvement in hand hygiene compliance (mean difference 14%, 95% CI 10-19%).Evidence: Hand hygiene compliance: 33% vs 15%, mean difference 14% (95% CI 10-19%).
“Mean hand hygiene compliance (Goal 1) was 33% in the intervention sites compared to 15% in the usual care sites (mean difference 14%; 95% CI 10% to 19%).”
ResultsFind in source - supportedReviewer 1The intervention improved appropriate antibiotic prophylaxis at cesarean section.Implementation outcome shows a significant improvement in correct antibiotic prophylaxis (mean difference 15%, 95% CI 4-26%).Evidence: Correct cesarean section antibiotic prophylaxis: 74% vs 58%, mean difference 15% (95% CI 4-26%).
“Appropriate antibiotic prophylaxis was appropriately administered prior to cesarean section (Goal 2) in 74% and 58% of patients, respectively (mean difference 15%; 95% CI 4% to 26%).”
ResultsFind in source - supportedReviewer 2The intervention improved hand hygiene compliance, antibiotic prophylaxis, and sepsis management.Implementation outcomes show significant improvements in hand hygiene, antibiotic prophylaxis, and vital sign recording, though some measures like antibiotics within 1 hour did not reach statistical significance.Evidence: Table 2 implementation outcomes.
“Mean hand hygiene compliance (Goal 1) was 33% in the intervention sites compared to 15% in the usual care sites (mean difference 14%; 95% CI 10% to 19%).”
ResultsFind in source
Efficacy claim is anchored to an adequate endpoint and a meaningful effect.
- ADEQUATESurrogate endpointThe primary outcome is a composite of hard clinical outcomes: infection-related maternal mortality, infection-related maternal near-miss, and severe infection (deep surgical/perineal site or body-cavity infection). These are clinical events, not surrogate biomarkers. The intervention also shows target engagement through implementation outcomes (e.g., hand hygiene compliance, antibiotic prophylaxis, vital sign recording) that are directly linked to the intervention's mechanisms.
“The primary outcome was infection-related maternal death or severe morbidity. This was defined as a composite of infection-related maternal mortality, infection-related maternal “near-miss” events, and severe infection (deep surgical-site, deep perineal or body-cavity infection) during pregnancy, childbirth or within 42 days of pregnancy ending, at any time up to 28 days of discharge (whichever occurred first).”
- ADEQUATEEffect sizeThe primary outcome showed a risk ratio of 0.68 (95% CI 0.55-0.83) with absolute rates of 1.4% vs 1.9%, representing a 32% relative reduction. This is statistically significant and clinically meaningful, as it reflects a reduction in serious maternal morbidity and mortality. The effect was consistent across subgroups and increased over time.
“The incidence of maternal infection–related mortality or severe morbidity was 1.4% in the intervention group and 1.9% in the usual care group (risk ratio, 0.68; 95% confidence interval, 0.55 to 0.83; P<0.001).”
Data authenticity concerns
None foundAn adversarial read for patterns associated with data that may not be genuine: results that look too clean, implausibly large effects, duplicated data or images, and methods that do not match the results reported.
Checked — nothing surfaced.
Reporting gaps
1 finding · worst mediumRequired detail the manuscript never states — study design, biological variables, ethics approval and consent, key resources, statistical reporting, data and code availability, and overall transparency.
- Data/code availability incompleteAssessed
The introduction cites multiple studies and WHO recommendations on maternal infection and sepsis, highlighting the burden and gaps in care. It logically links these to the need for a multicomponent intervention and states the study objective. Limitations of prior work (e.g., lack of established effect of maternal sepsis bundles) are acknowledged.
“The World Health Organization (WHO) has issued recommendations on adherence to hand-hygiene standards and evidence-based practices to prevent and treat maternal infections.”
“A maternal sepsis bundle has been developed specifically for low resource settings, but its effect on maternal outcomes has not been established.”
“We conducted a large cluster-randomized trial in Malawi and Uganda to evaluate whether implementation of the APT-Sepsis program in health facilities would reduce infection-related maternal mortality and severe morbidity.”
“The World Health Organization (WHO) has issued recommendations on adherence to hand-hygiene standards and evidence-based practices to prevent and treat maternal infections.”
“A maternal sepsis bundle has been developed specifically for low resource settings, but its effect on maternal outcomes has not been established.”
Randomization used a minimization algorithm with a random element, and the unit of randomization (facility) is stated. Blinding is described for outcome adjudication, though outcome assessors were not blinded. A detailed power analysis is provided, including a pre-specified re-estimation. Inclusion/exclusion criteria for hospitals are specified. The analysis population (ITT) and handling of missing data (none reported) are adequate.
“A minimization algorithm generated by an independent statistician was used to ensure balance between facilities allocated to the intervention and control groups within each country.”
“If there was discordance in assessments, or if uncertainty in the causation was recorded by either group, the case was adjudicated by a separate case classification committee blinded to site allocation.”
“We calculated that at least 60 clusters (a minimum of 30 in Malawi and 30 in Uganda) would be required for the trial to have 95% power, with 2 sided p < 0.05, to detect a 25% relative reduction in the composite primary outcome from 3% to 2.25%.”
“A minimization algorithm generated by an independent statistician was used to ensure balance between facilities allocated to the intervention and control groups within each country.”
“If there was discordance in assessments, or if uncertainty in the causation was recorded by either group, the case was adjudicated by a separate case classification committee blinded to site allocation.”
“We calculated that at least 60 clusters (a minimum of 30 in Malawi and 30 in Uganda) would be required for the trial to have 95% power, with 2 sided p < 0.05, to detect a 25% relative reduction in the composite primary outcome from 3% to 2.25%.”
Sex is inherently female (pregnant women), and demographics such as age are not explicitly reported but the population is defined. Health status is captured through obstetric outcomes. Species/strain and housing are not applicable. Demographics are partially reported via facility characteristics and outcomes.
“Postpartum hemorrhage (>1 liter) – no. (%) | 1,680 (1.8) | 1,355 (1.4)”
“Live births – no. | 94,730 | 95,770”
The paper lists specific ethics committees that approved the trial, including the University of Liverpool, WHO Ethics Review Committee, and national committees in Malawi and Uganda. It explains that individual patient consent was not obtained because the intervention targeted healthcare providers and components were considered best practice, with agreement obtained at national and facility levels. This is adequate for a cluster trial of a provider-level intervention.
“The trial was approved by the University of Liverpool, the WHO Ethics Review Committee, the College of Medicine Research Ethics Committee in Malawi, the Infectious Diseases Institute Research Ethics Committee and Uganda National Council for Science and Technology in Uganda.”
“Patients were not individually consented as the intervention was delivered to healthcare providers, and the components were considered best practice.”
“The trial was approved by the University of Liverpool, the WHO Ethics Review Committee, the College of Medicine Research Ethics Committee in Malawi, the Infectious Diseases Institute Research Ethics Committee and Uganda National Council for Science and Technology in Uganda.”
“Patients were not individually consented as the intervention was delivered to healthcare providers, and the components were considered best practice.”
The paper describes the intervention components, including the FAST-M bundle with specific antibiotic regimens, and mentions that key resources such as antibiotics were obtained through usual procurement. It also describes the provision of soap, alcohol hand rub, thermometers, and blood pressure machines. However, specific product identifiers (e.g., manufacturer, catalog numbers) are not provided, which is typical for a health systems intervention.
“Antibiotics - According to source or if unknown ceftriaxone 2g IV OD and metronidazole 500mg IV TDS (or 400mg PO TDS) with additional singe dose gentamicin 5mg/kg IV if haemodynamically unstable.”
“Implementation strategies were developed from a behavior change perspective and include the following key components: hospital leadership engagement; program champions selected from existing facility staff; multi-disciplinary training with comprehensive training materials; implementation tools (eg FAST-M checklist); and performance feedback with dashboards and quarterly visits”
“FAST-M bundle: To be completed within 1 hour. Fluids - 500ml crystalloid bolus, repeated if hypotension persists. Antibiotics - According to source or if unknown ceftriaxone 2g IV OD and metronidazole 500mg IV TDS (or 400mg PO TDS) with additional singe dose gentamicin 5mg/kg IV if haemodynamically unstable.”
The primary analysis uses generalized linear mixed effects models with robust standard errors and marginal standardization. Effect sizes are reported as risk ratios with 95% CIs. Exact p-values are reported for the primary outcome. Software is not explicitly identified, but the methods are standard. Data presentation includes per-group Ns and CIs.
“In the primary analysis, we used generalized linear mixed effects models incorporating a constrained baseline approach.”
“The primary outcome occurred in 1752 of 124,298 women (1.4%) in the intervention group and in 2208 of 116,596 (1.9%) in the usual care group (risk ratio, 0.68; 95% CI, 0.55 to 0.83; P<0.001)”
“The primary outcome occurred in 1752 of 124,298 women (1.4%) in the intervention group and in 2208 of 116,596 (1.9%) in the usual care group (risk ratio, 0.68; 95% CI, 0.55 to 0.83; P<0.001)”
“In the primary analysis, we used generalized linear mixed effects models incorporating a constrained baseline approach.”
The paper states that the protocol is available with the full text, but does not provide a clear data availability statement. No repository deposits or accession numbers are mentioned. Code sharing is not applicable as no custom code is described.
“The first and last author, and the trial statisticians from the Liverpool Clinical Trials Centre vouch for the accuracy and completeness of the data and for the fidelity of the trial to the protocol, available with the full text of this article at NEJM.org.”
The trial is registered (ISRCTN42347014). Methods are comprehensive. Limitations are explicitly discussed, including lack of blinding and potential underreporting. Conclusions are proportional to the evidence. Funding and COI are disclosed.
“APT-Sepsis ISRCTN number, ISRCTN42347014.”
“The trial has several limitations. The multicomponent nature of the intervention precludes attribution of the effect to individual elements.”
“This project is supported by the Joint Global Health Scheme with funding from the UK Foreign, Commonwealth and Development Office, the UK Medical Research Council (MRC), The UK Department of Health and Social Care through the National Institute of Health Research (NIHR) and Wellcome (Grant ref: MRV005782/1)”
“APT-Sepsis ISRCTN number, ISRCTN42347014.”
“The trial has several limitations. The multicomponent nature of the intervention precludes attribution of the effect to individual elements.”
“This project is supported by the Joint Global Health Scheme with funding from the UK Foreign, Commonwealth and Development Office, the UK Medical Research Council (MRC), The UK Department of Health and Social Care through the National Institute of Health Research (NIHR) and Wellcome (Grant ref: MRV005782/1)”
Registered (1 ID: ISRCTN). No reporting guideline cited.
Broken references and links
None found · partly checkedReferences checked against Crossref, OpenAlex and Retraction Watch for retractions and resolvability, plus declared data and code links probed for whether they resolve to content matching the paper.
Nothing surfaced — but not everything feeding this category ran (missing: data/code link verification), so read this as a partial clean bill.
Checked 30 references by DOI: 21 verified — 9 no DOI (shown, not verified).
- NO DOIStatement on maternal sepsisNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOI2030 global agenda for sepsisNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIImproving the prevention, diagnosis, and clinical management of sepsisNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIAccelerate progress towards reducing maternal, newborn, and child mortality in order to achieve Sustainable Development Goal targets 3.1 and 3.2No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIToo little, too late: the recurrent theme in maternal deaths due to sepsisNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIWHO guidelines on hand hygiene in health careNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIWHO recommendations for the prevention and treatment of maternal peripartum infectionsNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIAbortion care guidelineNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIWHO recommendation on antibiotic prophylaxis during labour for vaginal birthNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
Copyediting
5 minorWording, consistency and formatting errors that need correcting before submission.
No major wording or formatting errors. 5 minor suggestions below.
5 copyedit issues flagged: mostly typo, clarity, consistency.
- MINORtypoFigure 1 (in paper text)“singe dose gentamicin”→ single dose gentamicinTypographical error.
- MINORconsistencyTable 2“116596”→ 116,596Missing comma for consistency with other numbers.
- MINORclarityMethods, Statistical analysis“The second minimization factor (proportion of births with the composite primary outcome) was not included as it was already in the model as the outcome variable.”→ Clarify that the outcome variable is the composite primary outcome, not the proportion.Potential ambiguity.
- MINORtypoFigure 1 legend“singe dose gentamicin”→ single dose gentamicinTypographical error in the FAST-M bundle description.
- MINORclarityMethods, Statistical analysis“The second minimization factor (proportion of births with the composite primary outcome) was not included as it was already in the model as the outcome variable.”→ The second minimization factor (proportion of births with the composite primary outcome) was not included as a covariate because it was already the outcome variable in the model.Clarify the wording to avoid ambiguity.
The published work is robust and well-reported; an informed reader should weigh the minor reporting gaps (data availability, software identification, demographics) as transparency issues, not validity threats. No erratum is warranted for the statistical results, but a data availability statement would improve reproducibility.
- 1.HIGHdata codeAdd a clear data availability statement in the Methods or a dedicated section, specifying that de-identified individual participant data may be available upon reasonable request to the trial steering committee, with conditions and timeframe.The current statement only mentions protocol availability, not data access, which is a transparency gap for a data-driven trial.
- 2.HIGHstatisticsMention the statistical software (e.g., Stata, R) and version used for analyses in the Statistical analysis section.Identifying the software is standard for reproducibility and was flagged by both reviewers.
- 3.MEDIUMreportingExplicitly state adherence to CONSORT guidelines for cluster-randomized trials, and include the CONSORT checklist as supplementary material.Reporting guidelines improve transparency and are expected for RCTs; currently not mentioned.
- 4.MEDIUMreportingProvide more detailed demographic characteristics of the study population (e.g., maternal age, parity, comorbidities) in Table 1 or supplementary tables.Demographics are only partially reported; age and other key variables are missing from the main text.
- 5.MEDIUMstatisticsReport exact p-values for secondary outcomes or clearly state that they are not adjusted for multiplicity and should be interpreted with caution.Some secondary outcomes lack exact p-values, which limits interpretability.
- 6.MEDIUMreportingInclude a statement about the availability of the trial protocol and statistical analysis plan, possibly with a link or DOI.The protocol is mentioned but not linked; a DOI would improve accessibility.
- 7.LOWcopyeditFix the typo 'singe dose' to 'single dose' in Figure 1 and its legend.Typographical error in the FAST-M bundle description.
- 8.LOWcopyeditAdd a comma to '116596' in Table 2 for consistency with other numbers (e.g., '116,596').Inconsistent formatting in the table.
- 9.LOWcopyeditClarify the sentence in Methods, Statistical analysis about the second minimization factor to avoid ambiguity.The current wording could be misread; suggest rephrasing to 'was not included as a covariate because it was already the outcome variable in the model.'
The star rating is the report’s one-glance summary. Every paper starts at 5★ and loses stars for the concrete problems the review finds — so a rating is never a vague average, it’s a running total you can read line by line under “How this rating was calculated.”
- Reporting — 8 dimensionseach dimension that fully fails−½★
- each dimension partially met−¼★
- Statistics · Integrity · Claimseach serious problem−1★
- each medium problem−½★
- Citationseach retracted or unverifiable reference−¼★
- Copyeditonly when the manuscript needs a full edit−½★
The rating never drops below 1★, and a demonstrable critical failure (an impossible statistic, a proven ethics violation) caps it at 1★ on its own — so the stars can never look healthy when the verdict is CRITICAL.
The rating draws on a panel of agents. Three independent Kaimen Rigor reviewers grade the eight dimensions below across several independent passes (the shown verdict is their majority vote — steadier than any single run), isolate the paper’s major claims and check its own evidence backs them, and flag integrity concerns. Alongside them, a citation agent resolves every reference against Crossref, OpenAlex, and Retraction Watch; a statistics agent recomputes reported tests; and rule-based checks verify that declared data/code links actually resolve. Full text is required — an abstract-only submission is not analyzed.
Graded against NIH, MDAR, ARRIVE 2.0, CONSORT, EQUATOR, and RRID guidelines. A dimension that doesn’t apply to the study type is skipped, never penalized.
This Kaimen Rigor review is model-assisted and is not a substitute for formal expert review. It complements human evaluation by surfacing potential methodological concerns — verify each finding against the source.