Oral Nirmatrelvir-Ritonavir for Covid-19 in Higher-Risk Outpatients.
Butler CC, Pinto AD, Harris V, Holmes J, Rahman NM, Cureton L, Hayward G, Richards DB, Lowe DM, Standing JF, Breuer J, Hood K, Png ME, Petrou S, Dorward J, Patel MG, Thomas NPB, Evans P, Hart ND, Jani BD, Hosseini B, Murthy S, McBrien K, Condon A, McDonald EG, Daley P, Greiver M, da Costa BR, Selby P, Jüni P, Lee TC, Shi H, Detry MA, Saunders CT, Fitzgerald M, Berry NS, Saville BR, Khoo SH, Nguyen-Van-Tam JS, Hobbs FDR, Yu LM, Little P, PANORAMIC Trial and CanTreatCOVID Trial Collaborative Groups
- DOI
- 10.1056/NEJMoa2502457
- Record issued
- 2026-08-15
- Engine
- 7.39.0
- Exported
- 2026-09-21
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/09f9d74a-bd94-48ff-9bb9-49b1dceee81d is authoritative.
How this rating was calculated
- IntegrityIntegrity concern ×3−1.5★
- StatisticsStatistic did not reproduce−0.5★
- ClaimsTreatment effect not shown to be clinically meaningful−0.5★
- ReportingData & code availability partially met−0.25★
- The numeric-impossibility checks (GRIM/GRIMMER/DEBIT/SPRITE) did not run: 10 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.
- 01Printed percentage does not match its own count
32.1% does not match the reported count 112/358
“112/358 (32.1)”
Table 3Find in source - 02Treatment effect not shown to be clinically meaningful
The primary outcome showed no significant benefit (PANORAMIC: 0.8% vs 0.7%; CanTreatCOVID: 0.6% vs 1.2%), with odds ratios crossing 1 and low probability of superiority. The secondary recovery outcomes showed benefit but are not the primary efficacy claim and are not anchored to a minimal clinically important difference.
“Nirmatrelvir–ritonavir did not reduce hospitalization or death ((PANORAMIC: 14/1698 (0.8%) vs 11/1673 (0.7%); adjusted-odds ratio [OR], 1.18; 95% Bayesian credible interval [BCI], 0.55 to 2.62; probability of superiority 0.334, CanTreatCOVID: 2/343 (0.6%) vs…”
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.
The paper reports two well-conducted randomized trials with clear methodology, appropriate Bayesian statistical analysis, and transparent reporting of ethics, registration, and limitations. The main weakness is the lack of a concrete data availability statement and code sharing, which limits reproducibility.
Both reviewers classified the study as interventional, and no divergence was noted. The evaluation covered all eight dimensions; several sub-criteria were marked not applicable (e.g., animal-related items, blinding since open-label, exact p-values since Bayesian). The statistics verification covered only a limited subset of tests; the one flagged inconsistency could not be substantiated and is not treated as a demonstrable error.
Numerical inconsistencies
2 findings · worst highValues 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.
- Summary statistic impossible for the stated N (GRIM/GRIMMER)Recomputed
- Internal contradictions in the reported numbersAssessed
1 reported summary statistic mathematically impossible for the stated N (PERCENT).
- PERCENT32.1% does not match the reported count 112/358
“112/358 (32.1)”
Table 3Find in source
- lowinternal contradictionThe abstract reports 3516 participants in PANORAMIC, but the results section reports 1743+1773=3516, which is consistent. However, the text says '29,295 underwent randomization' overall, which includes other arms; this is not a contradiction but could be confusing.
“From December 8, 2021 to September 30, 2024, 3516 participants in PANORAMIC and 716 in CanTreatCOVID were randomized.”
AbstractFind in source - lowinternal contradictionIn Table 1, the CanTreatCOVID control group has 27 missing sex and 27 missing ethnicity, but the total N is 358. The sum of sex categories (231+100+0+27=358) is consistent, but the missing ethnicity is also 27, which is plausible but not explained.
“Missing | 0 | 0 | 2 (0.6%) | 27 (7.5%)”
Table 1Find in source - lowinternal contradictionIn Table 1, the sex percentages for CanTreatCOVID control group do not sum to 100% (64.5% female + 27.9% male + 7.5% missing = 99.9%), which is likely due to rounding.
Female | 237 (66.2%) | 231 (64.5%) | Male | 118 (33%) | 100 (27.9%) | Other | 1 (0.3%) | 0 | Missing | 2 (0.6%) | 27 (7.5%)
Table 1reviewer’s wording
Overstated conclusions
2 findings · worst highConclusions 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.
- Treatment effect not shown to be clinically meaningfulAssessed
- Conclusions only partially backed by the presented evidenceAssessed
4 major claims checked against the paper's own evidence: 1 only partially supported (evidence backs part of the claim; gaps or caveats remain); the rest adequately supported.
- partialReviewers 1, 2The open-label design is unlikely to bias the primary outcome of hospitalization or death.The authors argue that objective outcomes are unlikely to be biased, but acknowledge that self-reported outcomes may be affected by open-label design.Evidence: Discussion states open-label design is unlikely to lead to bias with primary outcomes such as hospital admission and/or mortality.
“Such a design facilitates trial conduct and is unlikely to lead to bias with primary outcomes such as hospital admission and/or mortality.”
Discussion ¶5Find in source - supportedReviewers 1, 2Nirmatrelvir-ritonavir did not reduce hospitalization or death within 28 days in vaccinated high-risk outpatients.The primary outcome results from both trials show no statistically significant difference, with credible intervals crossing 1.Evidence: Primary outcome results: PANORAMIC adjusted-OR 1.18 (95% BCI 0.55 to 2.62); CanTreatCOVID adjusted-OR 0.48 (95% BCI 0.08 to 2.23).
“Nirmatrelvir–ritonavir did not reduce hospitalization or death ((PANORAMIC: 14/1698 (0.8%) vs 11/1673 (0.7%); adjusted-odds ratio [OR], 1.18; 95% Bayesian credible interval [BCI], 0.55 to 2.62; probability of superiority 0.334, CanTreatCOVID: 2/343 (0.6%) vs 4/324 (1.2%); adjusted-OR, 0.48; 95% BCI, 0.08 to 2.23; probability of superiority 0.830).”
AbstractFind in source - supportedReviewers 1, 2Viral load was reduced by the end of treatment.The virology sub-study shows a significant reduction in viral load at day 5, supporting the claim.Evidence: Virology sub-study: viral load below detection at day 5 29.2% vs 16.5%, adjusted OR 2.15 (95% BCI 1.37 to 3.44).
“Viral load was reduced by the end of treatment.”
AbstractFind in source - supportedReviewers 1, 2Early sustained recovery was higher in the nirmatrelvir-ritonavir group.Both trials show higher early sustained recovery in the treatment group with credible intervals excluding 1.Evidence: PANORAMIC: 33.0% vs 22.1%, adjusted OR 1.74 (95% BCI 1.48 to 2.04); CanTreatCOVID: 69.0% vs 53.1%, adjusted OR 1.99 (95% BCI 1.40 to 2.87).
“In PANORAMIC, 33.0% reported early sustained recovery in the nirmatrelvir-ritonavir group compared to 22.1% in the usual care group (adjusted odds ratio, 95% BCI = 1.74 (1.48 to 2.04), and 69.0% vs 53.1% in CanTreatCOVID, respectively (adjusted odds ratio, 95% BCI = 1.99 (1.40 to 2.87)).”
ResultsFind in source
Premise concern: effect size not shown to be clinically meaningful.
- N/ASurrogate endpointThe primary efficacy claim is based on a hard clinical outcome (all-cause hospitalization or death), not a surrogate.
“Primary outcome was all-cause hospitalization or death within 28 days.”
- INADEQUATEEffect sizeThe primary outcome showed no significant benefit (PANORAMIC: 0.8% vs 0.7%; CanTreatCOVID: 0.6% vs 1.2%), with odds ratios crossing 1 and low probability of superiority. The secondary recovery outcomes showed benefit but are not the primary efficacy claim and are not anchored to a minimal clinically important difference.
“Nirmatrelvir–ritonavir did not reduce hospitalization or death ((PANORAMIC: 14/1698 (0.8%) vs 11/1673 (0.7%); adjusted-odds ratio [OR], 1.18; 95% Bayesian credible interval [BCI], 0.55 to 2.62; probability of superiority 0.334, CanTreatCOVID: 2/343 (0.6%) vs 4/324 (1.2%); adjusted-OR, 0.48; 95% BCI, 0.08 to 2.23; probability of superiority 0.830).”
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 EPIC-HR and EPIC-SR trials and notes the lack of evidence in vaccinated populations. It also acknowledges limitations of observational studies (residual confounding, confounding by indication, immortal time bias). The rationale directly follows from the cited evidence, and the study objectives are clearly stated.
“Observational studies following licensure have been and are being conducted, but all have issues with residual confounding, confounding by indication, and immortal time bias.”
“The national UK PANORAMIC and the pan-Canadian CanTreatCOVID trials assessed the effectiveness of nirmatrelvir-ritonavir in reducing hospital admissions and/or death, in mostly vaccinated adults in the community with risk factors for serious COVID-19 illness.”
“Observational studies following licensure have been and are being conducted, but all have issues with residual confounding, confounding by indication, and immortal time bias.”
“Since the EPIC-HR and EPIC-SR studies many more people have been multiply vaccinated and infected naturally so it is unclear whether nirmatrelvir-ritonavir still benefits those at higher risk.”
Randomization methods are described for both trials (web-based systems, stratified by age and vaccination status). Blinding is not applicable as the trials are open-label, but the authors provide a rationale. Power analysis is reported with revisions. Inclusion/exclusion criteria are specified. Outlier handling is addressed through sensitivity analyses for missing data. Controls are the usual care arm. Independent replication is not applicable for a single trial, but the two trials together provide some replication.
“A secure, web-based system (Spinnaker; version custom built for the PANORAMIC trial; Spiral Software, Wellington, New Zealand) was used for randomization, stratified by age (<50 years vs ≥50 years) and vaccination status (yes vs no).”
“In contrast to efficacy trials, we conducted an open-label trial, suited to answering pragmatic questions of the effectiveness during routine clinical care because placebos are generally not used in routine care.”
“With a control event rate of 1.3%, a relative risk of 0.23 (corresponding to an event rate of 0.3% in the intervention arm and which is smaller than the 88% relative risk reduction reported in the EPIC-HR trial ), 1438 participants per group would be required to achieve 80% power at a two-sided 5% significance level.”
“A secure, web-based system (Spinnaker; version custom built for the PANORAMIC trial; Spiral Software, Wellington, New Zealand) was used for randomization, stratified by age (<50 years vs ≥50 years) and vaccination status (yes vs no).”
“In contrast to efficacy trials, we conducted an open-label trial, suited to answering pragmatic questions of the effectiveness during routine clinical care because placebos are generally not used in routine care.”
“With a control event rate of 1.3%, a relative risk of 0.23 (corresponding to an event rate of 0.3% in the intervention arm and which is smaller than the 88% relative risk reduction reported in the EPIC-HR trial ), 1438 participants per group would be required to achieve 80% power at a two-sided 5% significance level.”
Sex is reported for both trials, and both sexes are enrolled, so sex_justified is not applicable. Age, weight (not reported but not required for this outpatient population), and health status (comorbidities) are reported. Demographics include age, sex, ethnicity, and socioeconomic measures. Species/strain and housing conditions are not applicable for human trials.
“Age, mean(SD) [min,max] | 54.7 (12.1) [18 to 96] | 54.8 (11.7) [18 to 93] | 54.7 (13.6) [19 to 88] | 55.0 (13.5) [18 to 89]”
“Comorbidity, n(%) | 1127 (64.9%) | 1185 (67.0%) | 173 (48.3%) | 160 (44.7%)”
“Age, mean(SD) [min,max] | 54.7 (12.1) [18 to 96] | 54.8 (11.7) [18 to 93] | 54.7 (13.6) [19 to 88] | 55.0 (13.5) [18 to 89]”
“Ethnicity category, n(%) | White | 1647 (94.9%) | 1661 (93.9%) | 290 (81%) | 272 (76%)”
PANORAMIC was approved by the South Central-Berkshire Research Ethics Committee, and CanTreatCOVID by Health Canada and provincial research ethics boards. Written informed consent was obtained from all participants. Regulatory compliance is implied through approvals and adherence to protocols.
“The UK Medicines and Healthcare products Regulatory Agency and the South Central-Berkshire Research Ethics Committee of the Health Research Authority approved the PANORAMIC trial protocol. The CanTreatCOVID trial was approved by Health Canada and research ethics boards in the participating provinces across Canada.”
“Written informed consent was obtained from all participants or a legal representative.”
“The UK Medicines and Healthcare products Regulatory Agency and the South Central-Berkshire Research Ethics Committee of the Health Research Authority approved the PANORAMIC trial protocol.”
“Written informed consent was obtained from all participants or a legal representative.”
The drug is identified by name and dosage (300mg/100mg BID x 5 days). No other biological or chemical resources are used. Statistical software (STATA and R) is identified. Since this is a drug trial, the investigational product is the key resource and is adequately described.
“participants in the nirmatrelvir–ritonavir group were asked to take nirmatrelvir at a dosage of 300 mg (two 150 mg tablets) with 100 mg ritonavir (one 100 mg tablet) orally twice daily for 5 days.”
“All analyses were performed in STATA (version 18.0) and R (version 4.2.1).”
“participants in the nirmatrelvir–ritonavir group were asked to take nirmatrelvir at a dosage of 300 mg (two 150 mg tablets) with 100 mg ritonavir (one 100 mg tablet) orally twice daily for 5 days.”
“All analyses were performed in STATA (version 18.0) and R (version 4.2.1).”
Tests are named (Bayesian logistic regression, time-varying piecewise exponential model). Assumptions are handled through pre-specified models and sensitivity analyses. Exact p-values are not reported; instead, Bayesian credible intervals and probabilities of superiority are used, which is appropriate. Effect sizes are reported with credible intervals. Software is identified. Data presentation includes per-group n and percentages. Mathematical plausibility checks were not possible for most outcomes due to continuous or model-derived statistics, but no obvious errors were found.
“Sensitivity analyses were also carried out to assess the robustness of the primary outcome analysis due to missing data.”
“Effect of treatment (and corresponding 95% Bayesian credible intervals (BCIs)) on the primary outcome was estimated using a Bayesian logistic regression model with weakly informative Cauchy priors, adjusting for comorbidity, age, and vaccination status”
“PANORAMIC: adjusted-odds ratio [OR], 1.18; 95% Bayesian credible interval [BCI], 0.55 to 2.62; probability of superiority 0.334”
The paper states that protocols and statistical analysis plans are available at NEJM.org, but does not provide a data availability statement for the raw data. No repository deposit or accession numbers are given. Code is not shared. For clinical trial data, managed access is acceptable, but the mechanism is not described.
“Full details and results are available at NEJM.Org (https://www.nejm.org/) .”
“Full details and results are available at NEJM.Org (https://www.nejm.org/) .”
Trial registration numbers are provided (EudraCT and ClinicalTrials.gov). Methods are detailed enough for replication. All outcomes are reported, including negative results. Limitations are discussed, including open-label design and potential bias. Conclusions are proportional to the evidence. Funding and COI are disclosed. No reporting guideline (e.g., CONSORT) is explicitly mentioned, but the manuscript appears to follow clinical trial reporting standards.
“EudraCT Number 2021-005748-31 (PANORAMIC); ClinicalTrials.gov (https://clinicaltrials.gov/) number NCT05614349 (CanTreatCOVID).”
“However, an open label design does not allow estimation of the contribution of either placebo or nocebo effects to any observed differences in self-reported outcomes such as time to recovery.”
“EudraCT Number 2021-005748-31 (PANORAMIC); ClinicalTrials.gov (https://clinicaltrials.gov/) number NCT05614349 (CanTreatCOVID)”
“However, an open label design does not allow estimation of the contribution of either placebo or nocebo effects to any observed differences in self-reported outcomes such as time to recovery.”
Registered (2 IDs: ClinicalTrials.gov, EudraCT). No reporting guideline cited.
Broken references and links
None foundReferences 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.
Checked — nothing surfaced.
Checked 26 references by DOI: 21 verified — 5 no DOI (shown, not verified).
- NO DOICOVID-19 epidemiological update – 17 September 2024No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOINirmatrelvir plus ritonavir, sotrovimab and tocilizumab for treating COVID-19No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOINirmatrelvir plus ritonavir, sotrovimab and tocilizumab for treating COVID-19: Technology appraisal guidance Reference number:TA878No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIIDSA Guidelines on the Treatment and Management of Patients with COVID-19No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIEQ-5D-5L User Guide, 2019No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
1 of 3 data/code links checked; 1 live; 2 not probed.
- datahttps://www.nejm.org/doi/full/10.1056/NEJMoa2502457UNVERIFIEDHTTP 403Liveness indeterminate — content not checked.
- datahttps://www.nejm.org/UNVERIFIEDHTTP 403Liveness indeterminate — content not checked.
- datahttps://clinicaltrials.gov/LIVEHTTP 200Resolves, but the content could not be matched to the paper.
Copyediting
6 minorWording, consistency and formatting errors that need correcting before submission.
No major wording or formatting errors. 6 minor suggestions below.
6 copyedit issues flagged: mostly consistency, grammar, typo.
- MINORtypoMethods, Outcome Measures“PANORMIC”→ PANORAMICTypo in trial name.
- MINORconsistencyResults, Participants“29,295 underwent randomization”→ 29,295 participants underwent randomizationMissing word for clarity.
- MINORgrammarMethods, Statistical Analysis“5300 per arm to ensure a would ensure a reduction”→ 5300 per arm to ensure a reductionRedundant phrase.
- MINORgrammarMethods, Statistical Analysis“5300 per arm to ensure a would ensure a reduction to 2%”→ 5300 per arm to ensure a reduction to 2%Repetitive phrase.
- MINORconsistencyResults, Participants“721 underwent randomization ... 358 participants received nirmatrelvir–ritonavir and 358 usual care alone, 5 were randomized to other arms”→ Ensure the sum of randomized participants matches the total (358+358+5=721).Numbers are consistent, but the phrasing could be clearer.
- MINORclarityMethods, Statistical Analysis“there were no interim performed in either trial”→ there were no interim analyses performed in either trialMissing word 'analyses'.
The published work is methodologically robust and transparent, with only minor reporting gaps. An informed reader should weigh the absence of a formal data availability statement and code sharing, and the open-label design's potential impact on subjective outcomes. No erratum is warranted based on the identified issues, but a data availability clarification would strengthen the record.
- 1.HIGHdata codeAdd a formal data availability statement in the Methods or a dedicated section specifying how de-identified participant data can be accessed (e.g., through a data access committee or repository like Vivli/YODA) with conditions and timeframe.The current statement only points to NEJM.org and lacks a concrete access mechanism, which is a reproducibility gap for a data-driven clinical trial.
- 2.HIGHdata codeDeposit statistical analysis code in a public repository (e.g., Zenodo, GitHub) with a DOI and link it in the data availability statement.Code sharing is not reported, and providing it would enhance reproducibility and allow independent verification of the Bayesian analyses.
- 3.MEDIUMreportingExplicitly mention adherence to CONSORT or other relevant reporting guidelines in the Methods or acknowledgments, and include the checklist as supplementary material.The manuscript follows CONSORT-like structure but does not state it, which is a transparency expectation for randomized trials.
- 4.MEDIUMreportingProvide a CONSORT flow diagram for each trial in the main text or appendix to improve transparency of participant flow.A flow diagram would clarify the numbers randomized, treated, and analyzed, especially given the multiple arms in the platform trials.
- 5.MEDIUMreportingState the version of the statistical analysis plan and where it can be accessed (e.g., NEJM.org) to allow verification of pre-specified analyses.This would help readers confirm that the reported analyses match the pre-specified plan.
- 6.MEDIUMreportingClarify the manufacturer/source of nirmatrelvir-ritonavir (e.g., Pfizer) in the Methods to fully identify the investigational product.Reviewer 2 noted the manufacturer is not stated; specifying it improves resource identification.
- 7.MEDIUMstatisticsConsider reporting exact p-values for secondary outcomes where frequentist analyses were performed, to complement the Bayesian credible intervals.This would provide additional inferential detail for readers accustomed to frequentist reporting.
- 8.MEDIUMreportingAdd a note on the generalizability of the findings to other healthcare systems, given the specific recruitment and care pathways in the UK and Canada.This would help readers interpret the applicability of the results beyond the study settings.
- 9.MEDIUMreportingDiscuss the potential impact of the open-label design on secondary outcomes more thoroughly, and suggest a blinded substudy for recovery outcomes.The open-label design is acknowledged but its effect on subjective outcomes could be explored further.
- 10.LOWcopyeditFix the typo 'PANORMIC' to 'PANORAMIC' in Methods, Outcome Measures.Correcting the trial name typo avoids confusion.
- 11.LOWcopyeditFix the redundant phrase '5300 per arm to ensure a would ensure a reduction' in Methods, Statistical Analysis.The duplicated phrase is a grammatical error that should be corrected.
- 12.LOWcopyeditAdd the missing word 'analyses' in 'there were no interim performed in either trial' in Methods, Statistical Analysis.The sentence is incomplete without 'analyses'.
- 13.LOWcopyeditClarify the phrasing in Results, Participants: '29,295 underwent randomization' to '29,295 participants underwent randomization'.Adding 'participants' improves clarity.
- 14.LOWcopyeditEnsure the sum of randomized participants in CanTreatCOVID is clearly presented (358+358+5=721) in Results, Participants.The numbers are consistent but the phrasing could be clearer to avoid confusion.
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.