Low-dose spironolactone and cardiovascular outcomes in moderate stage chronic kidney disease: a randomized controlled trial.
Hobbs FDR, McManus RJ, Taylor CJ, Jones NR, Rahman JK, Wolstenholme J, Kim S, Kwon J, Jones L, Hirst JA, Yu LM, Mort S, BARACK-D Investigators, BARACK-D Investigators, Regional coordinating centre teams
- DOI
- 10.1038/s41591-024-03263-5
- Record issued
- 2026-08-16
- 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/383861bf-d691-4c43-9766-5ac05b91c789 is authoritative.
How this rating was calculated
- IntegrityIntegrity concern ×4−2★
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 and transparently reported phase 3 randomized trial. The methodology is rigorous, with clear randomization, blinding, power calculation, and comprehensive reporting of demographics, ethics, resources, statistics, data/code availability, and limitations. The main weaknesses are minor reporting inconsistencies and the significant under-recruitment, which is transparently acknowledged.
Both reviewers classified the study as interventional, and this was adopted. The evaluation covered all eight dimensions; several sub-criteria were marked not applicable (e.g., animal-related, cell line, and replication criteria) due to the human trial design. The statistics verification covered only a subset of tests (those with test statistics/df or effect estimates with CIs); the remaining tests were not machine-verifiable and are not confirmed.
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 19 tests: 19 consistent, 0 inconsistent; 4 recomputed directly from the reported test statistics, 15 via agent-written checks.
- CONSISTENTreported p = .013 · recomputed p = .013Recomputed odds ratio 1.37 (95% CI 1.07–1.76), reported p=0.013
“odds ratio 1.37, 95% confidence interval (CI): 1.07–1.76, P = 0.013”
Taken as given: 1.07–1.76 is a two-sided 95% confidence interval for the odds ratio of 1.37, not a range, an IQR, or a different interval level; the odds ratio is a RATIO measure, so the interval is symmetric on the log scale; p=0.013 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(1.37, 1.07, 1.76, 1) - CONSISTENTreported p = .970 · recomputed p = .964Recomputed HR 0.99 (95% CI 0.64–1.53), reported p=0.970
“HR = 0.99, 95% CI: 0.64–1.53, P = 0.970”
Taken as given: 0.64–1.53 is a two-sided 95% confidence interval for the HR of 0.99, not a range, an IQR, or a different interval level; the HR is a RATIO measure, so the interval is symmetric on the log scale; p=0.970 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(0.99, 0.64, 1.53, 1) - CONSISTENTreported p = .478 · recomputed p = .463Recomputed HR 1.14 (95% CI 0.80–1.61), reported p=0.478
“HR = 1.14, 95% CI: 0.80–1.61, P = 0.478”
Taken as given: 0.80–1.61 is a two-sided 95% confidence interval for the HR of 1.14, not a range, an IQR, or a different interval level; the HR is a RATIO measure, so the interval is symmetric on the log scale; p=0.478 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(1.14, 0.8, 1.61, 1) - CONSISTENTreported p = .699 · recomputed p = .703Recomputed HR 1.09 (95% CI 0.70–1.70), reported p=0.699
“HR = 1.09, 95% CI: 0.70–1.70, P = 0.699”
Taken as given: 0.70–1.70 is a two-sided 95% confidence interval for the HR of 1.09, not a range, an IQR, or a different interval level; the HR is a RATIO measure, so the interval is symmetric on the log scale; p=0.699 is the p for THIS estimate, not for another comparison in the same sentenceMethod: back the two-tailed p out of the log-scale CI width and compare it against the printed pHow we recomputed it: pCI(1.09, 0.7, 1.7, 1) - CONSISTENTreported p = .702 · recomputed p = .716Reviewers 1, 2Primary outcome HR p-value from CI
“hazard ratio (HR) = 1.05, 95% CI: 0.81–1.37, P = 0.702”
Taken as given: The HR is 1.05 with 95% CI 0.81-1.37.; The CI is two-sided at 95%.; The HR is on a log scale.Method: Two-tailed p-value derived from the confidence interval for a hazard ratio.How we recomputed it: pCI(1.05, 0.81, 1.37, 1) - CONSISTENTreported p = .478 · recomputed p = .463Reviewer 1CVD component HR p-value from CI
“CVD (HR = 1.14, 95% CI: 0.80–1.61, P = 0.478)”
Taken as given: The HR is 1.14 with 95% CI 0.80-1.61.; The CI is two-sided at 95%.; The HR is on a log scale.Method: Two-tailed p-value derived from the confidence interval for a hazard ratio.How we recomputed it: pCI(1.14, 0.80, 1.61, 1) - CONSISTENTreported p = .699 · recomputed p = .703Reviewer 1Death component HR p-value from CI
“death (HR = 1.09, 95% CI: 0.70–1.70, P = 0.699)”
Taken as given: The HR is 1.09 with 95% CI 0.70-1.70.; The CI is two-sided at 95%.; The HR is on a log scale.Method: Two-tailed p-value derived from the confidence interval for a hazard ratio.How we recomputed it: pCI(1.09, 0.70, 1.70, 1) - CONSISTENTreported p = .004 · recomputed p = .004Reviewer 1eGFR 3-year adjusted treatment effect p-value from CI
“adjusted treatment effect −1.14 ml min −1 1.73 m −2 , 95% CI: −1.92 to −0.37, P = 0.004”
Taken as given: The treatment effect is -1.14 with 95% CI -1.92 to -0.37.; The CI is two-sided at 95%.; The effect is on a linear scale.Method: Two-tailed p-value derived from the confidence interval for a linear effect.How we recomputed it: pCI(-1.14, -1.92, -0.37, 0) - CONSISTENTreported p < .001 · recomputed p = <.001Reviewer 1Systolic BP 6-month adjusted treatment effect p-value from CI
“−3.32 (−5.05 to −1.59) | <0.001”
Taken as given: The treatment effect is -3.32 with 95% CI -5.05 to -1.59.; The CI is two-sided at 95%.; The effect is on a linear scale.Method: Two-tailed p-value derived from the confidence interval for a linear effect.How we recomputed it: pCI(-3.32, -5.05, -1.59, 0) - CONSISTENTreported p = .003 · recomputed p = .003Reviewer 1Systolic BP 1-year adjusted treatment effect p-value from CI
“−2.66 (−4.43 to −0.90) | 0.003”
Taken as given: The treatment effect is -2.66 with 95% CI -4.43 to -0.90.; The CI is two-sided at 95%.; The effect is on a linear scale.Method: Two-tailed p-value derived from the confidence interval for a linear effect.How we recomputed it: pCI(-2.66, -4.43, -0.90, 0) - CONSISTENTreported p = .169 · recomputed p = .168Reviewer 1Systolic BP 2-year adjusted treatment effect p-value from CI
“−1.33 (−3.22 to 0.56) | 0.169”
Taken as given: The treatment effect is -1.33 with 95% CI -3.22 to 0.56.; The CI is two-sided at 95%.; The effect is on a linear scale.Method: Two-tailed p-value derived from the confidence interval for a linear effect.How we recomputed it: pCI(-1.33, -3.22, 0.56, 0) - CONSISTENTreported p = .074 · recomputed p = .074Reviewer 1Systolic BP 3-year adjusted treatment effect p-value from CI
“−1.69 (−3.55 to 0.16) | 0.074”
Taken as given: The treatment effect is -1.69 with 95% CI -3.55 to 0.16.; The CI is two-sided at 95%.; The effect is on a linear scale.Method: Two-tailed p-value derived from the confidence interval for a linear effect.How we recomputed it: pCI(-1.69, -3.55, 0.16, 0) - CONSISTENTreported p = .034 · recomputed p = .032Reviewer 1Hypotension adjusted treatment effect p-value from CI
“1.60 (1.04–2.46) | 0.034”
Taken as given: The treatment effect is 1.60 with 95% CI 1.04-2.46.; The CI is two-sided at 95%.; The effect is on a log scale (rate ratio).Method: Two-tailed p-value derived from the confidence interval for a rate ratio.How we recomputed it: pCI(1.60, 1.04, 2.46, 1) - UNCOMPUTABLEreported p = .487 · recomputed p = .095Reviewer 1Natriuretic peptide 2-year adjusted treatment effect p-value from CI
“−1.09 (−1.39 to 1.17) | 0.487”
Taken as given: The treatment effect is -1.09 with 95% CI -1.39 to 1.17.; The CI is two-sided at 95%.; The effect is on a linear scale.Method: Two-tailed p-value derived from the confidence interval for a linear effect.How we recomputed it: pCI(-1.09, -1.39, 1.17, 0) - UNCOMPUTABLEreported p = .146 · recomputed p = .069Reviewer 1Natriuretic peptide 3-year adjusted treatment effect p-value from CI
“−1.20 (−1.53 to 1.06) | 0.146”
Taken as given: The treatment effect is -1.20 with 95% CI -1.53 to 1.06.; The CI is two-sided at 95%.; The effect is on a linear scale.Method: Two-tailed p-value derived from the confidence interval for a linear effect.How we recomputed it: pCI(-1.20, -1.53, 1.06, 0) - UNCOMPUTABLEreported p = .897 · recomputed p = .191Reviewer 1ACR 3-year adjusted treatment effect p-value from CI
“1.03 (−1.50 to 1.59) | 0.897”
Taken as given: The treatment effect is 1.03 with 95% CI -1.50 to 1.59.; The CI is two-sided at 95%.; The effect is on a linear scale.Method: Two-tailed p-value derived from the confidence interval for a linear effect.How we recomputed it: pCI(1.03, -1.50, 1.59, 0) - CONSISTENTreported p < .001 · recomputed p = <.001Reviewer 1eGFR 6-month adjusted treatment effect p-value from CI
“−1.68 (−2.41 to −0.94) | <0.001”
Taken as given: The treatment effect is -1.68 with 95% CI -2.41 to -0.94.; The CI is two-sided at 95%.; The effect is on a linear scale.Method: Two-tailed p-value derived from the confidence interval for a linear effect.How we recomputed it: pCI(-1.68, -2.41, -0.94, 0) - CONSISTENTreported p < .001 · recomputed p = <.001Reviewer 1eGFR 1-year adjusted treatment effect p-value from CI
“−2.33 (−3.08 to −1.58) | <0.001”
Taken as given: The treatment effect is -2.33 with 95% CI -3.08 to -1.58.; The CI is two-sided at 95%.; The effect is on a linear scale.Method: Two-tailed p-value derived from the confidence interval for a linear effect.How we recomputed it: pCI(-2.33, -3.08, -1.58, 0) - CONSISTENTreported p = .029 · recomputed p = .029Reviewer 1eGFR 2-year adjusted treatment effect p-value from CI
“−0.89 (−1.69 to −0.09) | 0.029”
Taken as given: The treatment effect is -0.89 with 95% CI -1.69 to -0.09.; The CI is two-sided at 95%.; The effect is on a linear scale.Method: Two-tailed p-value derived from the confidence interval for a linear effect.How we recomputed it: pCI(-0.89, -1.69, -0.09, 0) - CONSISTENTreported p = .037 · recomputed p = .033Reviewer 1EQ-5D-5L 3-year mean difference p-value from CI
“mean difference 0.032, 95% CI: 0.001–0.060, P = 0.037”
Taken as given: The mean difference is 0.032 with 95% CI 0.001-0.060.; The CI is two-sided at 95%.; The effect is on a linear scale.Method: Two-tailed p-value derived from the confidence interval for a linear effect.How we recomputed it: pCI(0.032, 0.001, 0.060, 0) - CONSISTENTreported p = .796 · recomputed p = .794Reviewer 1Total cost mean difference p-value from CI
“a mean difference of −£56 (bootstrapped 95% CI: −£463 to £378, P = 0.796)”
Taken as given: The mean difference is -56 with 95% CI -463 to 378.; The CI is two-sided at 95%.; The effect is on a linear scale.Method: Two-tailed p-value derived from the confidence interval for a linear effect.How we recomputed it: pCI(-56, -463, 378, 0) - CONSISTENTreported p = .004 · recomputed p = .004Reviewer 2Secondary eGFR treatment effect p-value from CI
“adjusted treatment effect −1.14 ml min −1 1.73 m −2 , 95% CI: −1.92 to −0.37, P = 0.004”
Taken as given: The effect is a difference (not a ratio).; The CI is a 95% confidence interval.Method: Two-tailed p-value derived from the estimate and its 95% CI using the normal approximation.How we recomputed it: pCI(-1.14, -1.92, -0.37, 0)
- lowinternal contradictionThe safety analysis table reports N=757 for usual care, while the primary analysis uses N=695. The paper explains this is due to including 33 non-treated spironolactone participants in the usual care arm for safety analysis, but this could be confusing.
“Usual care ( N = 757)”
Table 4Find in source - lowinternal contradictionThe abstract states 'Two-thirds of participants randomized to spironolactone stopped treatment within 6 months', but the results state 'only a third of patients continued taking spironolactone beyond 6 months'. These are consistent.
“Two-thirds of participants randomized to spironolactone stopped treatment within 6 months”
AbstractFind in source - lowinternal contradictionThe abstract states '1,372 (96%) were included in the primary analysis' but the results section says 62 were ineligible, leaving 1,372. The percentage is correct (1434-62=1372, 1372/1434=95.7%), but the abstract rounds to 96%.
“We recruited 1,434 adults from English primary care, of whom 1,372 (96%) were included in the primary analysis.”
AbstractFind 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: 1 only partially supported (evidence backs part of the claim; gaps or caveats remain); the rest adequately supported.
- partialReviewer 1The trial was underpowered but this is unlikely to have influenced the outcomes.The conditional power calculation suggests low probability of finding a significant result if fully recruited, but this is based on assumptions.Evidence: Post hoc conditional power calculation: 11.8% probability of significant result if recruited to target.
“The conditional power calculation suggests that there is an 11.8% probability that a statistically significant result would have been found if the trial had reached its recruited target and the observed event rate remained constant with the participants who were recruited.”
DiscussionFind in source - supportedReviewers 1, 2Low-dose spironolactone did not reduce cardiovascular outcomes in stage 3b CKD.The primary outcome analysis shows no significant difference between groups, supporting the claim.Evidence: Primary endpoint occurred in 113/677 (16.7%) spironolactone vs 111/695 (16.0%) usual care, HR=1.05, 95% CI 0.81-1.37, P=0.702.
“we found that spironolactone was frequently discontinued due to safety concerns, with no evidence that it reduced cardiovascular outcomes in people with stage 3b CKD.”
AbstractFind in source - supportedReviewers 1, 2Spironolactone was frequently discontinued due to safety concerns.The high discontinuation rate and reasons are documented in the results.Evidence: Two-thirds stopped within 6 months; 455 participants had treatment withdrawn due to safety concerns; common reasons include eGFR decrease (35.4%), side effects (18.9%), hyperkalemia (8.0%).
“Two-thirds of participants randomized to spironolactone stopped treatment within 6 months, predominantly because they met prespecified safety stop criteria.”
AbstractFind in source - supportedReviewers 1, 2Spironolactone should not be used for people with stage 3b CKD without another explicit treatment indication.The conclusion follows from the lack of benefit and high discontinuation due to safety.Evidence: No significant difference in primary outcome, high discontinuation, increased hyperkalemia and hypotension.
“Spironolactone should not be used for people with stage 3b CKD without another explicit treatment indication.”
AbstractFind in source - supportedReviewer 1Spironolactone was associated with a small but statistically significant greater deterioration in eGFR.The adjusted treatment effects show a small but significant decline in eGFR at all time points.Evidence: Adjusted treatment effects for eGFR at 6 months, 1 year, 2 years, 3 years were -1.68, -2.33, -0.89, -1.14 ml/min/1.73m2, all with p<0.05.
“Among participants randomized to spironolactone compared to usual care, there was a small reduction in mean eGFR over follow-up at 3 years (adjusted treatment effect −1.14 ml min −1 1.73 m −2 , 95% CI: −1.92 to −0.37, P = 0.004; Table ).”
ResultsFind in source - supportedReviewer 1Spironolactone was not cost-effective compared to usual care.The within-trial economic evaluation indicates low probability of cost-effectiveness.Evidence: At £20,000/QALY threshold, 12.6% probability of cost-effectiveness; no significant difference in total costs.
“This further indicated that the added treatment of spironolactone to usual care was not cost-effective when compared with usual care.”
ResultsFind in source - supportedReviewer 2The trial was underpowered due to recruitment difficulties.The paper acknowledges under-recruitment and provides conditional power calculations.Evidence: Recruited 45% of planned sample size; conditional power 11.8% based on observed event rate.
Limitations of the study include that we recruited only around half the number of participants we originally planned (>3,000 participants) due to difficulties with recruitment.
Discussionreviewer’s wording - supportedReviewer 2Spironolactone was associated with a small but significant eGFR decline.The adjusted treatment effects show a statistically significant decline at each time point.Evidence: Adjusted treatment effects at 6 months, 1, 2, 3 years: -1.68, -2.33, -0.89, -1.14 ml/min/1.73m2, all with p<0.05.
“Among those randomized to spironolactone, the adjusted treatment effects for eGFR at 6 months, 1 year, 2 years and 3 years, respectively, were −1.68 ml min −1 1.73 m −2 (95% CI: −2.41 to −0.94), −2.33 ml min −1 1.73 m −2 (95% CI: −3.08 to −1.58), −0.89 ml min −1 1.73 m −2 (95% CI: −1.69 to −0.09) and −1.14 ml min −1 1.73 m −2 (95% CI: −1.92 to −0.37; Table ).”
ResultsFind in source
Data authenticity concerns
1 finding · worst lowAn 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.
- Other integrity concernAssessed
4 integrity concerns flagged (0 high).
- lowotherThe trial failed to recruit to target (45% of planned sample size), which is a significant limitation but transparently reported.
“The trial failed to recruit to the target, achieving 45% of the planned sample size.”
ResultsFind in source
Reporting gaps
None foundRequired detail the manuscript never states — study design, biological variables, ethics approval and consent, key resources, statistical reporting, data and code availability, and overall transparency.
Checked — nothing surfaced.
The introduction cites prior research on CKD, cardiovascular risk, and MRAs, including finerenone trials, and acknowledges uncertainties regarding spironolactone. The rationale for the trial is clearly linked to the need to establish whether spironolactone provides similar benefits to finerenone. Limitations of prior work (e.g., finerenone trials excluding non-diabetic and non-albuminuric patients) are explicitly addressed.
“The nonsteroidal mineralocorticoid receptor antagonist (MRA), finerenone, offers cardiorenal protection for people with CKD and diabetes, but there is uncertainty if the steroidal MRA, spironolactone, provides the same protection.”
“It is therefore important to establish if the beneficial effects of the nonsteroidal MRA, finerenone, reported in FIDELIO-DKD and FIGARO-DKD could be reproduced through treatment with the steroidal MRA, spironolactone, given that it is a cheaper medication that is already widely used in clinical practice.”
“It is therefore important to establish if the beneficial effects of the nonsteroidal MRA, finerenone, reported in FIDELIO-DKD and FIGARO-DKD could be reproduced through treatment with the steroidal MRA, spironolactone, given that it is a cheaper medication that is already widely used in clinical practice.”
Randomization was 1:1 using a validated system (Sortition) stratified by GP practice. The PROBE design (open-label with blinded endpoint adjudication) is described. A power calculation was performed (80% power to detect 20% relative risk reduction, requiring 3,022 participants). Inclusion/exclusion criteria are detailed. The analysis population (ITT excluding post-randomization ineligible) and safety population are defined. Outlier handling is addressed through prespecified stop criteria and missing-data methods.
“Randomization was carried out using Sortition, a validated randomization system developed within our Primary Care Clinical Trials Unit.”
“We determined that a sample size of 1,511 participants per group (3,022 total) would be required at 80% power to detect a 20% relative risk reduction in death or cardiovascular events over 3 years of follow-up, accounting for an anticipated treatment withdrawal rate of 13%.”
“Randomization was carried out using Sortition, a validated randomization system developed within our Primary Care Clinical Trials Unit. Participants were enrolled from 329 general practice sites across the UK. We stratified by GP practice to ensure a balance of the two arms within each practice.”
“We determined that a sample size of 1,511 participants per group (3,022 total) would be required at 80% power to detect a 20% relative risk reduction in death or cardiovascular events over 3 years of follow-up, accounting for an anticipated treatment withdrawal rate of 13%.”
Sex, age, weight, health status, and demographics (including ethnicity) are reported in Table 1. Both sexes are enrolled, so sex justification is not applicable. Species/strain and housing conditions are not applicable for a human trial.
“At the time of randomization, the mean age of participants was 74.8 years (s.d. = 8.1) with 54.5% women (Table ).”
The study was approved by a named NHS Research Ethics Committee with a protocol number, and all participants provided written informed consent. Regulatory compliance is stated through MHRA approval and adherence to the Declaration of Helsinki.
“The study was approved by a National Health Service (NHS) Research Ethics Committee (REC-13/SC/0114) as well as the Medicines and Healthcare Regulatory Authority (MHRA), relevant NHS Research and Development departments and the host institutions.”
“All participants provided written informed consent.”
“The study was approved by a National Health Service (NHS) Research Ethics Committee (REC-13/SC/0114) as well as the Medicines and Healthcare Regulatory Authority (MHRA), relevant NHS Research and Development departments and the host institutions.”
“All participants provided written informed consent.”
Spironolactone is named with dose and regimen. The trial is a drug trial, so bench reagents are not applicable. Statistical software is identified (Stata, version not specified but implied).
“Eligible participants were randomized in a 1:1 ratio to usual care or usual care plus treatment with spironolactone 25 mg once daily by their usual doctor.”
“Eligible participants were randomized in a 1:1 ratio to usual care or usual care plus treatment with spironolactone 25 mg once daily by their usual doctor.”
The primary analysis uses Cox proportional hazards model, with Schoenfeld residuals to verify assumptions. Secondary analyses use linear mixed effects models with missingness adjustments. Exact p-values are reported (e.g., P = 0.702). Effect sizes with 95% CIs are reported throughout. Statistical software is identified (Stata). Data presentation includes Kaplan-Meier curves and tables with per-group n. Mathematical plausibility checks were not performed due to large N and continuous outcomes.
“For the primary outcome, we report the HR and 95% CI for the time to the first occurrence of a primary endpoint event using a Cox proportional hazards model adjusted for randomized treatment allocation.”
“Schoenfeld residuals were conducted for the primary outcome and confirmed that the proportional hazards assumption was not violated.”
“hazard ratio (HR) = 1.05, 95% CI: 0.81–1.37, P = 0.702”
“Schoenfeld residuals were conducted for the primary outcome and confirmed that the proportional hazards assumption was not violated.”
“hazard ratio (HR) = 1.05, 95% CI: 0.81–1.37, P = 0.702”
“adjusted treatment effect −1.14 ml min −1 1.73 m −2 , 95% CI: −1.92 to −0.37, P = 0.004”
The data availability statement provides a clear access route via the chief investigator with a timeframe. Code is available on GitHub. Repository deposit and accession numbers are not applicable for patient-level data.
“All available data can be obtained by contacting the chief investigator (F.D.R.H.). Individual patient data will be shared in datasets in a de-identified and anonymized format, following our data-sharing process. We will aim to make data available within 6–9 weeks.”
“The scripts for the statistical analysis are freely available at https://github.com/samjamesmort/BARACK-D_Statistical_Analysis .”
“All available data can be obtained by contacting the chief investigator (F.D.R.H.). Individual patient data will be shared in datasets in a de-identified and anonymized format, following our data-sharing process. We will aim to make data available within 6–9 weeks.”
“The scripts for the statistical analysis are freely available at https://github.com/samjamesmort/BARACK-D_Statistical_Analysis .”
The trial is registered (ISRCTN44522369). The paper follows CONSORT guidelines (flow diagram). All prespecified outcomes are reported, including negative results. Limitations are thoroughly discussed, including under-recruitment and lack of placebo. Conclusions are proportional to the evidence. Funding and COI statements are provided.
“ClinicalTrials.gov registration: ISRCTN44522369 (https://www.isrctn.com/ISRCTN44522369) .”
“Fig. 1 CONSORT flow diagram.”
“Limitations of the study include that we recruited only around half the number of participants we originally planned (>3,000 participants) due to difficulties with recruitment, despite over 300 GP surgeries acting as recruitment sites.”
“ClinicalTrials.gov registration: ISRCTN44522369 (https://www.isrctn.com/ISRCTN44522369) .”
“Limitations of the study include that we recruited only around half the number of participants we originally planned (>3,000 participants) due to difficulties with recruitment, despite over 300 GP surgeries acting as recruitment sites.”
“This article presents independent research commissioned and funded by the National Institute for Health and Care Research (NIHR) under a Health Technology Assessment grant (12/01.52) and pilot support from the School for Primary Care Research.”
Registered (1 ID: ISRCTN). Reporting guideline cited: CONSORT.
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 38 references by DOI: 35 verified — 3 no DOI (shown, not verified).
- NO DOIDapagliflozin for treating chronic kidney disease. Technology appraisal guidance (TA775)No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOIBenefits of Aldosterone Receptor Antagonism in Chronic Kidney Disease (BARACK-D) trial: a potential new treatment for kidney diseaseNo DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
- NO DOINICE health technology evaluations: the manual (PMG36)No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
1 data/code link checked; 1 live.
- codeGitHubLIVEHTTP 200https://github.com/samjamesmort/BARACK-D_Statistical_AnalysisResolves to GitHub (code repository).
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, clarity, typo.
- MINORconsistencyAbstract“ClinicalTrials.gov registration: ISRCTN44522369”→ ISRCTN is not ClinicalTrials.gov; consider rephrasing to 'Trial registration: ISRCTN44522369'.The registration number is from ISRCTN, not ClinicalTrials.gov.
- MINORtypoTable 1“1.358”→ 1,358Inconsistent use of comma as thousands separator.
- MINORclarityResults, Safety“An additional 33 participants who were randomized to spironolactone but who did not take any medication were included in the comparator usual care arm in this analysis.”→ Clarify that these participants were analyzed as part of the usual care arm for safety analysis.This sentence could be misinterpreted as a protocol violation.
- MINORconsistencyTable 1, Systolic BP on the left arm“137.5 (18.1; 1.358)”→ Change '1.358' to '1,358' for consistency with other numbers.Decimal point used instead of comma in the n value.
- MINORclarityResults, Safety“An additional 33 participants who were randomized to spironolactone but who did not take any medication were included in the comparator usual care arm in this analysis.”→ Clarify that this is a safety analysis population and not the primary analysis.This sentence could be misinterpreted as changing the randomization groups.
- MINORconsistencyTable 4, Safety analysis“Usual care ( N = 757)”→ Ensure the N=757 is clearly explained as the safety population including the 33 non-treated spironolactone participants.The N differs from the primary analysis N=695; a footnote explains this but could be more prominent.
The published work is robust and well-reported. An informed reader should weigh the significant under-recruitment (45% of planned sample) and the open-label design as limitations, but these are transparently disclosed. Minor copyedit issues (e.g., mislabeling ISRCTN as ClinicalTrials.gov, inconsistent number formatting) do not affect the scientific validity but could warrant a correction for clarity.
- 1.MEDIUMcopyeditIn the Abstract, change 'ClinicalTrials.gov registration: ISRCTN44522369' to 'Trial registration: ISRCTN44522369' to correctly identify the registry.ISRCTN is not ClinicalTrials.gov; this is a factual mislabeling that could confuse readers.
- 2.MEDIUMcopyeditIn Table 1, change '1.358' to '1,358' for the systolic BP n value to maintain consistent thousands-separator formatting.Inconsistent number formatting is a minor but visible copyedit issue.
- 3.MEDIUMreportingIn the Results/Safety section, clarify that the 33 non-treated spironolactone participants were included in the usual care arm only for the safety analysis, not the primary analysis.The current wording could be misinterpreted as a protocol violation or a change in randomization groups.
- 4.MEDIUMreportingIn Table 4, add a prominent footnote explaining that the usual care N=757 includes the 33 non-treated spironolactone participants, to reconcile with the primary analysis N=695.The differing N between tables could confuse readers without a clear explanation.
- 5.LOWreportingIn the Methods/Statistical analysis, specify the Stata version used (e.g., Stata 17) to enhance reproducibility.Specifying software versions is a standard reproducibility practice.
- 6.LOWreportingIn the Data availability section, provide more detail on the data-sharing process, including exact conditions and timeline for access.A more concrete data-sharing process enhances transparency and trust.
- 7.LOWreportingIn the Discussion, add a note on the generalizability of findings to non-UK primary care settings and non-White populations.The predominantly White, older UK cohort limits generalizability; acknowledging this is important.
- 8.LOWreportingIn the Results, report the number of participants screened but not eligible to complete the CONSORT flow.A complete CONSORT flow improves transparency about the recruitment process.
- 9.LOWreportingIn the Results, provide exact p-values for secondary outcomes where currently reported as '<0.001'.Exact p-values are more informative than threshold values.
- 10.LOWreportingIn the manuscript, add a statement about adherence to the CONSORT checklist, possibly as a supplementary file.Explicitly stating CONSORT adherence confirms reporting completeness.
- 11.LOWreportingIn the Discussion, expand on the potential impact of the COVID-19 pandemic on trial results, and consider mentioning it in the abstract.The pandemic may have affected recruitment and outcomes; more prominent discussion is warranted.
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.