Skip to main content
Tessa NewsLink
Paste a health news link, or browse

Source study found

Story checked

HIV services plunge at four providers in Uganda and Zimbabwe after U.S. aid cuts (opens in a new tab)

medicalxpress.com · 2026-10-05

Short answerEvidenceSource

Short answer

Mixed

Mixed.

One claim goes further than the study. 2 other points were not covered by the paper.

  • 2 supported
  • 1 overstated
  • 2 not covered

Checked against the study summary. The full text wasn't available, so some details couldn't be settled either way.

Share this check

Follow the evidence trail
1
2

NewsLink checks it

Mixed

One claim overstates the study. Two of five check out. Two claims the study doesn't address.

  • 2 supported
  • 1 overstated
  • 2 not covered
Open claim evidence
3
Source paper

Source layer

The 2 papers the story cites

Source study separated from background citations.

The research anchor for the report.

Then inspect each claim

Evidence layer

Claim by claim

Each claim gets a verdict. Expand it to see the evidence directly below.

5 claims in this story

Showing all 5 claimsChoose a verdict to focus the list.

Claim 1 of 5OverstatedHIV prevention, PrEP, HIV testing, and antiretroviral therapy all fell sharply after the 2025 PEPFAR and USAID stop-work orders and had not recovered by March 2026.View evidenceHide evidence

As statedfell by between a half and two-thirds; prevention roughly 37% to 63%, PrEP 43% to 72%, treatment 22% to 32%, testing 59% to 64%

Why this verdict

The direction, timing, and persistence are supported: prevention, PrEP, treatment, and testing indices all declined after the February 2025 interruption and remained substantially below 2024 levels through March 2026. However, the story’s broad magnitude framing that services fell “between a half and two-thirds” overgeneralizes the paper’s estimates: treatment declined by about 22.0–31.5 index points, and some prevention/PrEP estimates were below one-half. The more specific domain ranges are largely consistent with the profile, but the headline-style aggregate magnitude outruns the evidence.

1

Study evidence

Prevention service volumes fell substantially at the February 2025 interruption and remained below 2024 mean through March 2026.−36.8 to −62.7 index points

“In this interrupted time-series study, we used monthly and quarterly service-delivery data from NGOs in Uganda and Zimbabwe from January 2023 up to March 2026.”
2

Study evidence

Both the organisation-level unweighted ITS and the client-weighted (key-population-level) ITS estimated large, immediate declines in NGO service-domain indices in February 2025, sustained through March 2026.Prevention: -36.8 to -62.7 index points; PrEP: -43.0 to -71.6; Treatment: -22.0 to -31.5; Testing: -58.7 to -63.8 (index points = percentage change from mean 2024).

“...an unweighted panel interrupted time-series (the organisation level) and a client-weighted interrupted time-series (the key-population level) on these indices.”
Claim 2 of 5Not coveredThe paper focuses on nongovernmental organizations serving key populations in Uganda and Zimbabwe, including men who have sex with men, female sex workers, people who inject drugs, prisoners, and transgender people.View evidenceHide evidence

Why this verdict

The supplied abstract profile supports the broad scope—NGOs serving key populations in Uganda and Zimbabwe—but it does not enumerate the specific key-population categories listed in the story, such as men who have sex with men, female sex workers, people who inject drugs, prisoners, or transgender people. Those details may be in the full paper, but they are not verifiable from this abstract-depth profile.

1

Study evidence

Prevention service volumes fell substantially at the February 2025 interruption and remained below 2024 mean through March 2026.−36.8 to −62.7 index points

“In this interrupted time-series study, we used monthly and quarterly service-delivery data from NGOs in Uganda and Zimbabwe from January 2023 up to March 2026.”
2

Study evidence

Both the organisation-level unweighted ITS and the client-weighted (key-population-level) ITS estimated large, immediate declines in NGO service-domain indices in February 2025, sustained through March 2026.Prevention: -36.8 to -62.7 index points; PrEP: -43.0 to -71.6; Treatment: -22.0 to -31.5; Testing: -58.7 to -63.8 (index points = percentage change from mean 2024).

“...an unweighted panel interrupted time-series (the organisation level) and a client-weighted interrupted time-series (the key-population level) on these indices.”
Claim 3 of 5Not coveredThe article says the humanitarian waiver meant to protect lifesaving care had limited impact, and even treatment services declined after the waiver took effect because the funding and delivery apparatus had been dismantled.View evidenceHide evidence

As statedlimited impact

Why this verdict

The abstract profile supports the narrower point that the ART humanitarian waiver attenuated treatment-domain effects but did not prevent a substantial treatment-service decline. However, it does not provide the quantitative waiver effect, the analytic form, timing and implementation details, or evidence for the mechanism that services declined because the funding and delivery apparatus had been dismantled. That causal mechanism is not verifiable at abstract depth.

1

Study evidence

The ART humanitarian waiver attenuated the decline in treatment-domain service volumes after the February 2025 funding withdrawal, but treatment services nonetheless experienced a substantial, sustained reduction through March 2026.Treatment decline at interruption: 22.0–31.5 index points; attenuation stated qualitatively (magnitude not reported).

“The antiretroviral therapy humanitarian waiver attenuated treatment effects but did not prevent substantial decline.”
Claim 4 of 5SupportedA new study from the London School of Economics and Political Science and Duke University, published in The Lancet Global Health, documents severe and sustained declines in HIV service delivery at four nongovernmental organizations in Uganda and Zimbabwe after 2025 U.S. foreign aid cuts.View evidenceHide evidence

As statedsevere and sustained declines

Why this verdict

The abstract-profile evidence supports the core scientific claim: an interrupted time-series study of four included NGOs in Uganda and Zimbabwe estimated large, sustained declines in NGO-delivered HIV service volumes after the February 2025 PEPFAR/USAID stop-work orders and funding cuts. The supplied paper profile does not independently verify the institutional affiliations or journal venue, but the service-disruption claim itself is supported.

1

Study evidence

Prevention service volumes fell substantially at the February 2025 interruption and remained below 2024 mean through March 2026.−36.8 to −62.7 index points

“In this interrupted time-series study, we used monthly and quarterly service-delivery data from NGOs in Uganda and Zimbabwe from January 2023 up to March 2026.”
Claim 5 of 5SupportedThe study used routine monthly and quarterly service data from January 2023 to March 2026 and found large declines compared with 2024 averages across prevention, PrEP, treatment, and testing.View evidenceHide evidence

As statedroutine monthly and quarterly service data; large declines in every core service area

Why this verdict

Supported. The abstract profile states that the study used monthly and quarterly NGO service-delivery data from January 2023 through March 2026, normalized domain indices to the 2024 mean, and found large declines across prevention, PrEP, treatment, and testing.

1

Study evidence

Prevention service volumes fell substantially at the February 2025 interruption and remained below 2024 mean through March 2026.−36.8 to −62.7 index points

“In this interrupted time-series study, we used monthly and quarterly service-delivery data from NGOs in Uganda and Zimbabwe from January 2023 up to March 2026.”
2

Study evidence

Both the organisation-level unweighted ITS and the client-weighted (key-population-level) ITS estimated large, immediate declines in NGO service-domain indices in February 2025, sustained through March 2026.Prevention: -36.8 to -62.7 index points; PrEP: -43.0 to -71.6; Treatment: -22.0 to -31.5; Testing: -58.7 to -63.8 (index points = percentage change from mean 2024).

“...an unweighted panel interrupted time-series (the organisation level) and a client-weighted interrupted time-series (the key-population level) on these indices.”
Then look for missing context

Context layer

What the story left out

Important study details the story did not include.

  • Scope is four included NGOs in Uganda and Zimbabwe serving key populations; four other organisations were excluded for data quality, lack of funding impact, or location.

    The story reflects the four-NGO Uganda/Zimbabwe scope and key-population focus, but the supplied caveats do not indicate that four initially participating organisations were excluded or why. That exclusion information is material to generalisability and selection concerns.

    From Interrupted time-series (panel and client-weighted); Interrupted time-series (unweighted organisation-level vs client-we

  • Outcomes are service-volume domain indices normalized to mean 2024 = 100, not direct counts of population health outcomes.

    The story mentions declines compared with 2024 averages, but the caveats do not clearly state that the outcomes are normalized service-volume indices and do not measure downstream HIV incidence, morbidity, mortality, or other population-health outcomes.

    From Interrupted time-series (panel and client-weighted); Interrupted time-series (unweighted organisation-level vs client-we

  • Statistical uncertainty differed by specification; 95% confidence intervals excluded zero for all estimates except PrEP in the unweighted panel specification.

    The story’s caveats do not mention that the unweighted PrEP estimate had a confidence interval including zero. This is material because it qualifies certainty for one domain under one analytic specification.

    From Interrupted time-series (panel and client-weighted); Interrupted time-series (unweighted organisation-level vs client-we

  • Two analytic framings were used: an organisation-level unweighted panel ITS and a client-weighted key-population-level ITS.

    The story reports service declines but does not appear to explain the two specifications or that effect-size ranges reflect different weighting/estimand choices.

    From Interrupted time-series (panel and client-weighted); Interrupted time-series (unweighted organisation-level vs client-we

  • Causal interpretation depends on interrupted time-series assumptions, including that other concurrent shocks did not drive the observed level changes.

    The story includes a generic caveat that the study should not be stretched beyond what the data support, but it does not specifically acknowledge the key ITS causal assumption about concurrent shocks or alternative explanations.

    From Interrupted time-series (panel and client-weighted)

5 things the story did carry across
  • Interrupted time-series design with February 2025 specified as the interruption point for PEPFAR/USAID stop-work orders and funding freeze.
  • Data source and period: monthly and quarterly NGO service-delivery data from January 2023 through March 2026.
  • Estimated declines by domain: prevention −36.8 to −62.7, PrEP −43.0 to −71.6, treatment −22.0 to −31.5, and testing −58.7 to −63.8 index points, with services remaining below 2024 amounts through March 2026.
  • ART humanitarian waiver attenuated treatment-domain effects but did not eliminate substantial treatment-service declines; the abstract provides no numerical attenuation estimate or detailed model form.
  • Generalisability is limited to the studied NGOs whose U.S. funding was terminated or sharply cut, and may not extend to all HIV service providers or all countries.
Then read the study layer

Study layer

Study at a glance

Scan the study first. Expand only the parts you want to inspect.

Pieces of work

3

Evidence read

study summary

Lead result

secondary data

1Lead resultsecondary dataEstimate the causal effect of the February 2025 PEPFAR/USAID stop-work orders and subsequent funding cuts on NGO-delivered HIV service volumes for key populations in Uganda and Zimbabwe (prevention, PrEP, testing, and treatment), using interrupted time-series analyses.Interrupted time-series (panel and client-weighted)Expand

In plain English

Interrupted time‑series analysis of NGO service-delivery data from January 2023–March 2026 in Uganda and Zimbabwe estimating the effect of the February 2025 PEPFAR/USAID stop-work orders and funding cuts on four domain indices (prevention, PrEP, treatment, testing) normalized to mean 2024 = 100. Analyses used an unweighted panel ITS at the organisation level and a client-weighted ITS at the key-population level to estimate level changes at the interruption; all four domains showed large, sustained declines in service volumes after February 2025, with most 95% confidence intervals excluding zero.

Key findings

  • Prevention service volumes fell substantially at the February 2025 interruption and remained below 2024 mean through March 2026.−36.8 to −62.7 index points
  • PrEP service volumes showed large estimated declines at the interruption, with wider estimates under the two specifications; the unweighted panel estimate had a 95% CI that included zero.−43.0 to −71.6 index points
“In this interrupted time-series study, we used monthly and quarterly service-delivery data from NGOs in Uganda and Zimbabwe from January 2023 up to March 2026.”
What this piece can’t prove
  • Analysis is based on a small set of NGOs (four included; four excluded for data quality, lack of funding impact, or location) as reported in the paper.
  • Outcome measures are service-volume domain indices (relative to 2024 mean) and do not measure downstream population-level health outcomes.
  • Interrupted time-series inference assumes the February 2025 interruption is the primary cause of the observed level change and that other concurrent shocks are not driving results (assumption implicit in ITS design).
2secondary dataAssess robustness/contrast of estimated service disruptions using two analytic framings: an organisation-level unweighted panel interrupted time-series and a client-weighted (key-population-level) interrupted time-series.Interrupted time-series (unweighted organisation-level vs client-weighted key-population-level)Expand

In plain English

The study applied two interrupted time‑series specifications—an organisation-level unweighted panel ITS and a client-weighted (key-population-level) ITS—on the same NGO service-delivery indices (normalized to mean 2024 = 100) with February 2025 as the interruption. Both specifications estimated large, sustained declines across prevention, PrEP, treatment, and testing domains, though magnitudes varied by specification and the unweighted PrEP estimate had a 95% CI that included zero. Four organisations were included in the analysis after exclusions.

Key findings

  • Both the organisation-level unweighted ITS and the client-weighted (key-population-level) ITS estimated large, immediate declines in NGO service-domain indices in February 2025, sustained through March 2026.Prevention: -36.8 to -62.7 index points; PrEP: -43.0 to -71.6; Treatment: -22.0 to -31.5; Testing: -58.7 to -63.8 (index points = percentage change from mean 2024).
  • The statistical certainty of some domain estimates differed by specification: 95% confidence intervals excluded zero for all reported estimates except PrEP in the unweighted organisation-level panel, indicating sensitivity of inference for that domain to the weighting/estimand choice.
“...an unweighted panel interrupted time-series (the organisation level) and a client-weighted interrupted time-series (the key-population level) on these indices.”
What this piece can’t prove
  • Small analytic sample reported: four organisations were included after excluding four others for data quality, lack of funding impact, or location.
  • Findings are based on NGO service-delivery data for key populations in Uganda and Zimbabwe and may not generalize beyond this setting or to non-NGO providers.
  • Abstract does not report full model specifications or the exact confidence-interval bounds for each estimate.
3secondary dataEvaluate whether the antiretroviral therapy (ART) humanitarian waiver attenuated treatment-domain effects after the funding withdrawal.Interrupted time-series with policy modifier (ART humanitarian waiver)Expand

In plain English

The abstract reports that an antiretroviral therapy (ART) humanitarian waiver attenuated the decline in treatment-domain service volumes after the US funding withdrawal, but did not prevent a substantial sustained reduction in treatment services through March 2026.

Key findings

  • The ART humanitarian waiver attenuated the decline in treatment-domain service volumes after the February 2025 funding withdrawal, but treatment services nonetheless experienced a substantial, sustained reduction through March 2026.Treatment decline at interruption: 22.0–31.5 index points; attenuation stated qualitatively (magnitude not reported).
“The antiretroviral therapy humanitarian waiver attenuated treatment effects but did not prevent substantial decline.”
What this piece can’t prove
  • Abstract lacks methodological detail about how the waiver effect was modelled (no specification of model terms, timing of waiver relative to interruption, or covariate adjustment).
  • Analytic sample is small (four organisations included), with four others excluded for data quality, lack of impact, or location—potential selection and generalisability concerns.
  • Possible concurrent changes or confounders around the waiver period are not described, limiting causal interpretation of attenuation.

1 further detail could not be confirmed from the summary.

Finally, the search trail

Method layer

NewsLink found the paper. Tessa takes you deeper.

NewsLink checks the story. Tessa is where you inspect the paper, authors, evidence, and research context.

Papers considered

The selected paper, plus nearby candidates.

PubMed, Europe PMC, Crossref · 38 candidate papers

And 32 more candidates considered.