Low-dose digoxin in patients with heart failure with reduced or mildly reduced ejection fraction: a randomized controlled trial.
van Veldhuisen DJ, Rienstra M, Mosterd A, Alings M, Voors AA, Damman K, van Asselt ADI, Bouvy ML, Schaap J, van der Wall EE, Crijns HJGM, Touw DJ, Hoogslag PAM, van de Swaluw JEC, Schuurman RJ, van der Sluis A, Bondarenko O, Römer TJ, Oosterhof T, Bartels GL, Koudstaal S, Dijkmans PA, Linssen GCM, Aksoy I, Dorman HGR, Schut A, Hemels MEW, Tieleman RG, Lok DJA, Westendorp ICD, Vijver MAT, Voordes GHD, de Vos AH, Maas-Soer EL, Postmus D, Lunter G, Tijssen JGP, van der Meer P
- DOI
- 10.1038/s41591-026-04406-6
- 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/8d19ad32-b097-4d00-b759-b02055a343f8 is authoritative.
How this rating was calculated
- IntegrityIntegrity concern ×2−1★
- 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. These checks need the count the mean was averaged over, so none was performed.
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, rigorously reported randomized controlled trial. The main methodological strengths are the double-blind placebo-controlled design, pre-specified power analysis, clear reporting of demographics and ethics, and transparent trial registration. The primary weakness is the vague data availability statement and lack of code sharing, which limits reproducibility.
Both reviewers classified the study as interventional (randomized controlled trial), and I adopt that classification. The evaluation covered all eight dimensions; several sub-criteria were marked not applicable for a human clinical trial (e.g., species/strain, housing, IACUC, cell line authentication). The statistics verification recomputed only 4 tests (all consistent); the remaining statistics are unverified. Citation check found no retracted or unresolved 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 3 tests: 3 consistent, 0 inconsistent; 1 recomputed directly from the reported test statistics, 2 via agent-written checks.
- CONSISTENTreported p = .133 · recomputed p = .142Recomputed rate ratio 0.81 (95% CI 0.61–1.07), reported p=0.133
“rate ratio 0.81; 95% confidence interval (CI) 0.61–1.07, P = 0.133”
Taken as given: 0.61–1.07 is a two-sided 95% confidence interval for the rate ratio of 0.81, not a range, an IQR, or a different interval level; the rate ratio is a RATIO measure, so the interval is symmetric on the log scale; p=0.133 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(0.81, 0.61, 1.07, 1) - CONSISTENTreported p = .095 · recomputed p = .106Reviewer 1Total worsening heart failure events rate ratio: 0.76 (0.54–1.05). Recompute p from CI.
“The total number of worsening heart failure events was 155 and 203 in the digoxin and placebo groups, respectively (rate ratio 0.76, 95% CI 0.54–1.05)”
Taken as given: The rate ratio is on a log scale.; The CI is a 95% confidence interval.; The p-value is two-sided.Method: pCI function for a ratio estimate with 95% CI, two-sided.How we recomputed it: pCI(0.76, 0.54, 1.05, 1) - CONSISTENTreported p = .630 · recomputed p = .637Reviewer 1Cardiovascular mortality hazard ratio: 0.93 (0.69–1.26). Recompute p from CI.
“cardiovascular mortality occurred in 83 patients (17%) and 88 (18%) in the digoxin and placebo groups, respectively (hazard ratio 0.93, 95% CI 0.69–1.26)”
Taken as given: The hazard ratio is on a log scale.; The CI is a 95% confidence interval.; The p-value is two-sided.Method: pCI function for a ratio estimate with 95% CI, two-sided.How we recomputed it: pCI(0.93, 0.69, 1.26, 1) - UNCOMPUTABLEreported p = .133 · recomputed p = .637Reviewer 2Cardiovascular mortality hazard ratio p-value from reported HR and 95% CI
“cardiovascular mortality occurred in 83 patients (17%) and 88 (18%) in the digoxin and placebo groups, respectively (hazard ratio 0.93, 95% CI 0.69–1.26)”
Taken as given: The hazard ratio is 0.93 and the 95% CI is 0.69-1.26.; The CI is two-sided at 95%.; The p-value is two-sided.Method: Recomputed two-sided p-value from the hazard ratio and its 95% CI using the normal approximation for the log hazard ratio.How we recomputed it: pCI(0.93, 0.69, 1.26, 1)
- lowinternal contradictionThe text reports 28% women (n=284) and 72% men (n=717), which sums to 1001, consistent with the primary analysis population. However, the baseline table shows 136 (27%) women in digoxin and 148 (30%) in placebo, which sums to 284, but the percentages are rounded.
“Of the whole study population, 28% were women ( n = 284) and 72% were men ( n = 717)”
Table 1Find in source - lowinternal contradictionThe abstract states 1,001 patients were randomized, but the results section says 1,002 were randomly assigned and one was excluded. This is consistent, but the abstract's phrasing might be misread.
“1,001 patients with symptomatic chronic heart failure and a left ventricular ejection fraction of 50% or less were randomized to low-dose digoxin or placebo”
AbstractFind in source
Overstated conclusions
None foundConclusions 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.
Checked — nothing surfaced.
6 major claims checked against the paper's own evidence: all adequately supported.
- supportedReviewers 1, 2Low-dose digoxin did not significantly reduce the composite endpoint of total worsening heart failure events or cardiovascular mortality.The primary outcome analysis shows a rate ratio of 0.81 (95% CI 0.61–1.07, P=0.133), which is not statistically significant. The claim is directly supported by the presented data.Evidence: Primary outcome analysis: rate ratio 0.81, 95% CI 0.61–1.07, P=0.133.
“low-dose digoxin did not significantly reduce the composite endpoint of total worsening heart failure events or cardiovascular mortality.”
AbstractFind in source - supportedReviewers 1, 2Low-dose digoxin was generally well tolerated and safe.The paper reports similar rates of serious adverse events and no significant differences in side effects between groups, supporting the safety claim.Evidence: Serious adverse events: 19.9 vs 18.3 per 100 patient-years (rate ratio 1.09, 95% CI 0.86–1.37). Treatment-related SAEs: 40 vs 25 patients (rate ratio 1.53, 95% CI 0.91–2.55). No significant differences in patient-reported side effects.
“Low-dose digoxin was generally well tolerated and safe”
Results ¶5Find in source - supportedReviewers 1, 2Results were similar between men and women.The subgroup analysis shows a P-value for interaction of 0.61, indicating no significant difference in treatment effect between sexes.Evidence: Subgroup analysis: rate ratio digoxin versus placebo: 0.71 and 0.85 for women and men, respectively, P value for interaction 0.61.
“there were no differences with regard to the effect of digoxin on the primary endpoint between women and men (rate ratio digoxin versus placebo: 0.71 and 0.85, respectively, P value for interaction 0.61)”
Results ¶3Find in source - supportedReviewers 1, 2The effect of digoxin was not different in patients with atrial fibrillation compared to those in sinus rhythm.The subgroup analysis shows a P-value for interaction of 0.80, supporting the claim.Evidence: Subgroup analysis: rate ratio 0.76 and 0.82 for atrial fibrillation and sinus rhythm, P value for interaction 0.80.
“the effect of digoxin was not different in these patients as compared to those in sinus rhythm (rate ratio 0.76 and 0.82, respectively, for atrial fibrillation and sinus rhythm, P value for interaction 0.80)”
Results ¶3Find in source - supportedReviewer 1Low-dose digoxin was safe in the 284 women studied.The paper reports no significant interaction by sex and no safety signal specific to women, supporting the claim.Evidence: Subgroup analysis by sex shows no significant interaction (P=0.61). The paper also discusses the historical concern and states that low-dose digoxin was safe in women.
“In the present study, low-dose digoxin was safe in the 284 women studied”
Discussion ¶4Find in source - supportedReviewer 2The effect of low-dose digoxin was not different between patients with HFmrEF and HFrEF.Subgroup analysis shows no significant interaction by ejection fraction category (P for interaction 0.75).Evidence: Rate ratio 0.72 and 0.82 for HFmrEF and HFrEF, P for interaction 0.75.
“the effect of low-dose digoxin was not different between patients with HFmrEF and those with heart failure with reduced ejection fraction (HFrEF; rate ratio 0.72 and 0.82, respectively, P value for interaction 0.75)”
Results ¶4Find in source
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 the DIG trial, post-hoc analyses suggesting benefit of low SDC, the DIGIT-HF trial, and RATE-AF, establishing the premise. It acknowledges that prior evidence is from post-hoc analyses and that no larger study with low-dose digoxin has been conducted, addressing limitations of prior work. The rationale linking these findings to the study objectives is logical.
“Although these findings suggest that low-dose digoxin might be beneficial and guidelines advocated to avoid SDC >1.2 ng ml −1 (ref. ), no larger study with low-dose digoxin has been conducted in heart failure.”
“The present data are therefore important since the safety of digoxin has been questioned in the past, based on post hoc, retrospective, nonrandomized data.”
Randomization was 1:1 to double-blind treatment, though the specific randomization method (e.g., computer-generated sequence) is not detailed, but the trial is described as 'randomized' and 'double-blind'. Blinding is described: 'double-blind treatment of either (low-dose) digoxin or matched placebo' and dose adjustments were made centrally with dummy values for placebo. A power analysis is reported: 982 patients, 472 events, 80% power to detect a 22% reduction. Inclusion/exclusion criteria are provided in supplementary tables. The primary analysis is modified intention-to-treat. Outlier handling is not explicitly discussed, but the analysis uses a pre-specified model (LWYY) which is robust. Controls are the placebo group. Independent replication is not applicable for a single pivotal trial.
“DECISION was an investigator-initiated, double-blind, randomized, placebo-controlled trial”
“Under the assumptions of a placebo event rate of 18 events per 100 person-years and a median follow-up of 36 months, we estimated that the enrollment of 982 patients yielding 472 primary outcome events would provide 80% power to detect a 22% reduction”
“The full list of inclusion and exclusion criteria is depicted in Supplementary Tables and .”
“Patients were randomly assigned in a 1:1 ratio to double-blind treatment of either (low-dose) digoxin or matched placebo.”
“DECISION was an investigator-initiated, double-blind, randomized, placebo-controlled trial”
“Under the assumptions of a placebo event rate of 18 events per 100 person-years and a median follow-up of 36 months, we estimated that the enrollment of 982 patients yielding 472 primary outcome events would provide 80% power to detect a 22% reduction in the event rate”
Table 1 reports age (72±9 years), sex (28% women), BMI (28±5), LVEF, NYHA class, NTproBNP, eGFR, and comorbidities. Demographics are adequately reported. Sex is reported and both sexes are included, so no justification for single-sex is needed. Species/strain and housing conditions are not applicable for a human trial.
“The mean age was 72 years and 28% were women.”
“Age, in years | 72 ± 9 | 73 ± 8”
“Type 2 diabetes mellitus | 136 (27) | 119 (24)”
The methods state: 'the study was approved by the ethics committees affiliated with each of the investigational sites.' It also states: 'The trial was conducted according to the principles of Good Clinical Practice and the Declaration of Helsinki.' and 'All patients provided written informed consent.' This satisfies irb_ethics_statement, informed_consent, and regulatory_compliance. IACUC is not applicable for a human trial.
“the study was approved by the ethics committees affiliated with each of the investigational sites.”
“All patients provided written informed consent.”
“the study was approved by the ethics committees affiliated with each of the investigational sites.”
“All patients provided written informed consent.”
The investigational product is described: 'Disphar International B.V., Tiofarma B.V. and TEVA Nederland B.V. provided the investigational product'. The dosing algorithm is provided in Extended Data Fig. 5. Software is identified: 'Analyses were conducted using R (version 4.4.2) and STATA/SE 19.5.' Antibodies, cell lines, mycoplasma testing, and organisms are not applicable for this clinical trial. Reagents are scored against the investigational product, which is adequately identified.
“Disphar International B.V., Tiofarma B.V. and TEVA Nederland B.V. provided the investigational product and financially supported the trial.”
“Analyses were conducted using R (version 4.4.2) and STATA/SE 19.5.”
“patients received a starting dose of either 0.2 or 0.1 mg depending on age, kidney function, eGFR or interacting medication”
“Analyses were conducted using R (version 4.4.2) and STATA/SE 19.5.”
The primary analysis uses the LWYY semi-parametric proportional rates model, and time-to-first-event analyses use Cox proportional hazards. Effect sizes are reported as rate ratios or hazard ratios with 95% CIs. The primary outcome p-value is exact (P = 0.133). Software is identified. Data presentation includes cumulative-incidence plots and Kaplan-Meier curves. Assumptions are not explicitly verified, but for a large clinical trial using standard methods, this is acceptable. Mathematical plausibility is not applicable for continuous outcomes with large N.
“rate ratio 0.81, 95% confidence interval (CI) 0.61–1.07, P = 0.133”
“0.81 (0.61–1.07) | 0.133”
“rate ratio 0.81; 95% confidence interval (CI) 0.61–1.07, P = 0.133”
“rate ratio 0.81, 95% confidence interval (CI) 0.61–1.07, P = 0.133”
The data availability statement says: 'The datasets generated and/or analyzed during this study are not publicly available, but anonymized participant data can be made available upon requests directed to the corresponding author.' It describes a review process but does not name a managed-access platform or provide a clear timeframe, making it reported_but_inadequate. Code sharing is not mentioned. Repository deposit and accession numbers are not applicable for patient-level clinical data.
“The datasets generated and/or analyzed during this study are not publicly available, but anonymized participant data can be made available upon requests directed to the corresponding author.”
“The datasets generated and/or analyzed during this study are not publicly available, but anonymized participant data can be made available upon requests directed to the corresponding author.”
The trial is registered at ClinicalTrials.gov (NCT03783429). Methods are detailed enough for replication. A CONSORT diagram is provided. Limitations are discussed in the Discussion (e.g., mostly male population, high discontinuation rate). Conclusions are proportional: 'low-dose digoxin did not significantly reduce the composite endpoint'. Funding sources and conflicts of interest are disclosed. A reporting guideline is not explicitly named but a CONSORT diagram is included.
“ClinicalTrials.gov registration: NCT03783429”
“Our trial has several limitations. We included a mostly male population with NYHA functional class II heart failure, and extrapolation to other patient populations should be done with caution.”
“We acknowledge the funding provided by the Dutch Heart Foundation for the execution of the study (NHS:2018B024).”
“ClinicalTrials.gov registration: NCT03783429”
“Our trial has several limitations. We included a mostly male population with NYHA functional class II heart failure, and extrapolation to other patient populations should be done with caution.”
“We acknowledge the funding provided by the Dutch Heart Foundation for the execution of the study (NHS:2018B024).”
Registered (1 ID: ClinicalTrials.gov). 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 23 references by DOI: 22 verified — 1 no DOI (shown, not verified).
- NO DOIStaat van infectieziekten in Nederland, 2023No DOI in the reference — shown for manual review; not independently verifiable (not a fabrication signal).
1 data/code link checked; 1 live.
- datahttps://clinicaltrials.gov/study/NCT03783429LIVEHTTP 200Resolves, but the content could not be matched to the paper.
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 consistency, clarity, grammar.
- MINORconsistencyResults, paragraph 1“1,002 patients were randomly assigned... one patient who was randomized never received study medication and was excluded... Therefore, 1,001 patients were included in the primary analysis.”→ Consider rephrasing for clarity: '1,002 patients were randomized; one never received study medication and was excluded per protocol, leaving 1,001 in the primary analysis.'Minor clarity issue.
- MINORclarityResults, paragraph 2“The total number of cardiovascular hospitalizations and urgent cardiovascular visits was 191 in the digoxin group and 169 in the placebo group (rate ratio 1.12, 95% CI 0.85–1.49)”→ This result (more events in the digoxin group) is notable; consider adding a brief comment in the text.Not an error, but a point of interest.
- MINORconsistencyAbstract“1,001 patients with symptomatic chronic heart failure and a left ventricular ejection fraction of 50% or less were randomized”→ Consider specifying the randomization ratio (1:1) in the abstract for clarity.The abstract states the number randomized but not the ratio, which is mentioned later.
- MINORclarityResults, paragraph 2“The total number of cardiovascular hospitalizations and urgent cardiovascular visits was 191 in the digoxin group and 169 in the placebo group (rate ratio 1.12, 95% CI 0.85–1.49)”→ Clarify that this is a secondary outcome and not part of the primary composite.The sentence could be misread as a primary outcome component.
- MINORgrammarDiscussion, paragraph 1“In patients with heart failure and a reduced or mildly reduced ejection fraction, low-dose digoxin on top of current guideline-recommended treatment did not result in a significantly lower rate of a composite of total worsening heart failure events and death from cardiovascular causes (primary outcome) than placebo.”→ Consider rephrasing for readability: '...did not result in a significantly lower rate of the composite of total worsening heart failure events and cardiovascular death (primary outcome) compared with placebo.'The sentence is grammatically correct but could be clearer.
The published work is robust and well-reported; an informed reader should weigh the minor reporting gaps (data availability vagueness, lack of code sharing, unspecified randomization method) as limitations but not as validity threats. No erratum or re-analysis is warranted based on the evidence reviewed.
- 1.HIGHdata codeIn the Data availability section, specify a concrete access mechanism, such as a named data-access committee or platform (e.g., YODA, Vivli), and provide the conditions and timeframe for data requests.The current 'available upon request' statement is vague and does not meet best practices for data sharing, which reviewers and readers may flag.
- 2.HIGHdata codeAdd a statement about code sharing, including a link to a public repository (e.g., GitHub, Zenodo) for any custom analysis code, or state that no custom code was used.Code sharing is not mentioned, which limits reproducibility of the statistical analyses.
- 3.HIGHreportingIn the Methods, Trial procedures, describe the randomization method in detail (e.g., computer-generated random sequence, block size, stratification factors).The randomization method is currently reported_but_inadequate; specifying it strengthens the design description.
- 4.MEDIUMreportingIn the Methods, Statistical analysis, explicitly state that normality and proportional hazards assumptions were tested or handled (e.g., by using robust methods).Assumptions verification is currently reported_but_inadequate; adding this detail improves statistical transparency.
- 5.MEDIUMreportingIn the Methods, Trial design and oversight, name the specific ethics committees (e.g., 'the Medical Ethics Committee of the University Medical Center Groningen') and provide protocol approval numbers.Naming the ethics committees and approval numbers strengthens the ethics statement.
- 6.MEDIUMreportingIn the Methods, Statistical analysis, add a statement on how outliers were handled in the analysis, or note that no outliers were excluded.Outlier handling is not reported; clarifying this avoids ambiguity.
- 7.MEDIUMreportingIn the Reporting transparency section, explicitly name the reporting guideline followed (e.g., CONSORT) and include a statement that the checklist was submitted.The reporting guideline is only implied via the CONSORT diagram; naming it explicitly is clearer.
- 8.MEDIUMcopyeditIn the Results, paragraph 1, rephrase for clarity: '1,002 patients were randomized; one never received study medication and was excluded per protocol, leaving 1,001 in the primary analysis.'The current phrasing is slightly confusing about the number randomized vs. analyzed.
- 9.MEDIUMcopyeditIn the Results, paragraph 2, clarify that the total number of cardiovascular hospitalizations and urgent visits (191 vs 169) is a secondary outcome and not part of the primary composite.The sentence could be misread as a primary outcome component.
- 10.MEDIUMcopyeditIn the Abstract, specify the randomization ratio (1:1) for clarity.The abstract states the number randomized but not the ratio, which is mentioned later.
- 11.LOWcopyeditIn the Discussion, paragraph 1, rephrase for readability: '...did not result in a significantly lower rate of the composite of total worsening heart failure events and cardiovascular death (primary outcome) compared with placebo.'The sentence is grammatically correct but could be clearer.
- 12.LOWreportingIn the Discussion, consider adding a paragraph on the generalizability of the findings to women and underrepresented groups, given the mostly male population.Enhances transparency about the study's limitations.
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.