Metformin as an Adjunctive Therapy to Catheter Ablation in Atrial Fibrillation: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Metformin as an Adjunctive Therapy to Catheter Ablation in Atrial Fibrillation (NCT04625946), a randomized phase 4 study. The primary outcome is Freedom From Recurrent Atrial Arrhythmias Between 3 to 12 Months After a Single Ablation to Eliminate Atrial Fibrillation (AF).

2. Clinical Question

The trial’s central comparison is the addition of metformin to an ablation and lifestyle-recommendation approach, rather than a comparison of ablation with no ablation. Both arms received standard-of-care ablation and lifestyle recommendations. The primary endpoint excludes recurrences during the 3-month blanking period and concerns freedom from recurrent atrial arrhythmias thereafter, through 12 months after ablation.

The registry summary describes the trial as investigating whether metformin can reduce atrial fibrillation in people having ablation. It also gives an anticipated benefit, but that expectation is not an observed result. The statistical question is whether the recorded endpoint differs between randomized arms, and how precisely that difference is estimated. The posted outcome counts and comparison estimates address that question; they do not by themselves establish why any difference occurred or whether it would be reproduced in another study.

3. Trial Design

This was a phase 4, randomized, parallel-group trial with no masking. The registry record identifies the University of Michigan as sponsor and gives enrollment as 117. The trial status is completed. Its recorded start date is 2021-01-19 and its primary completion date is 2025-07-01.

Participants were assigned to metformin or standard of care. Both arms included standard-of-care ablation and recommendations for lifestyle modification. The metformin arm also received metformin; the control arm did not. The summary says follow-up visits were planned up to approximately 1 year after ablation. It also describes lifestyle and exercise education as standard of care for all participants. An AliveCor device is listed among the interventions, but further details on its role are not reported in the registry record.

The primary outcome is reported for 49 participants in the metformin arm and 50 in the standard-of-care arm. Those outcome denominators are smaller than the registry’s enrollment figure. The reasons for that difference, and whether it reflects exclusions, unavailable measurements, or another analysis-population definition, are not reported in the registry record. Randomization helps make treatment groups comparable on average, but it does not remove the need to understand who contributed outcome data and how missing observations were handled.

The registry record does not report the original sample-size justification, allocation ratio details beyond the randomized parallel-group structure, or the planned statistical analysis in sufficient detail to identify adjustments, analysis populations, or handling of missing data. These details should not be inferred from the posted results or from planning illustrations based on the observed effect.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Before assignment: 18 of 117 randomized withdrew consent before ablation, did not undergo ablation, or had surgical ablation, and were not assigned to a study arm (modified ITT). All 99 assigned participants completed the 12-month visit.

Overall Study

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Freedom From Recurrent Atrial Arrhythmias Between 3 to 12 Months After a Single Ablation to Eliminate Atrial Fibrillation (AF), measured over 12 months after ablation.

The primary endpoint counts participants who remained free from recurrent atrial arrhythmias between 3 and 12 months after a single ablation. Its recorded timeframe is 12 months after ablation. The 3-month period immediately following ablation is treated as a blanking period: recurrences during that interval do not contribute to this endpoint.

This endpoint is useful because it expresses the result in terms of participants meeting a clinically interpretable event-free criterion over a defined follow-up period. It asks whether a participant had a recurrence during the specified interval, rather than measuring the number or duration of all rhythm episodes. The resulting count can be compared between randomized arms, with a risk difference expressing the difference in the proportions meeting the endpoint.

The endpoint is not the same as time until recurrence. A participant who recurs early in the evaluated interval and one who recurs later may both fail the event-free criterion, even though their timing differs. The registry lists median time to recurrence as a secondary outcome, which addresses timing separately. The primary outcome also does not describe episodes during the blanking period, recurrent events after the first recurrence, or outcomes beyond the recorded follow-up interval. Those distinctions matter when interpreting what “freedom from recurrence” captures.

7. Results Posted in the Registry

• Metformin: 77.6% (38/49) versus 58.0% (29/50) on control; risk difference 19.6 percentage points (95% CI 1.6 to 37.5), risk ratio 1.34 (1.01–1.77); unadjusted two-proportion z-test p = 0.038.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Freedom From Recurrent Atrial Arrhythmias at 6 Months After Ablation After the Blanking Period of 3 Months

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Metformin: 85.7% (42/49) versus 80.0% (40/50) on control; risk difference 5.7 percentage points (95% CI -9.1 to 20.5), risk ratio 1.07 (0.90–1.28); unadjusted two-proportion z-test p = 0.451.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Metformin and the control. The interval gives the range of effects that remain compatible with the data.

Atrial Fibrillation Severity Score (AFSS) - Part C

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Metformin: mean 2.70 versus 5.00 on control; difference -2.30 (95% CI -5.10 to 0.50), standardized effect d = -0.32; z-test p = 0.107.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Metformin and the control. The interval gives the range of effects that remain compatible with the data.

Change in Weight at 3 Months After Ablation

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Metformin: mean -5.10 versus -7.00 on control; difference 1.90 (95% CI -0.00 to 3.80), standardized effect d = 0.53; z-test p = 0.050.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Metformin and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes cover several distinct questions: recurrence timing and freedom from recurrence at a shorter follow-up interval, AFSS symptom severity, and changes in weight and hemoglobin A1c. Taken together, these outcomes do not provide one uniform pattern that can be reduced to a single conclusion. The measures differ in what they capture, their timeframes, and the number of participants contributing data. A result on one measure need not track a result on another.

Secondary outcomes can help place the primary endpoint in context, but they do not answer the primary question in the same way. The recurrence outcomes are related to the primary endpoint while using different follow-up windows or measures. Symptom scores and changes in weight or hemoglobin A1c assess other dimensions. Their findings may suggest questions for future work, but they should not be treated as independent confirmation of the primary result.

The registry record identifies 13 secondary outcomes, while detailed numerical results for only selected secondary measures are included in the posted results. Values and analyses for the other listed outcomes are not reported in the registry record. Even where results are posted, the record does not describe a multiplicity adjustment or a prespecified hierarchy for interpreting multiple secondary tests. When many outcomes are examined, some may appear notable by chance even when there is no underlying difference. Unadjusted secondary p-values therefore carry less weight than a clearly prespecified primary analysis, particularly when the analysis plan and missing-data procedures are not available.

Secondary findings are best read as exploratory context unless the registry record documents how they were prioritized and controlled for multiplicity. They do not establish a broader pattern beyond the outcomes and participants actually reported.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Median Time to Recurrence of Atrial Fibrillation After a 3 Month Blanking Period of Ablation
• Freedom From Recurrent Atrial Arrhythmias at 6 Months After Ablation After the Blanking Period of 3 Months
• Freedom From Recurrent Atrial Arrhythmias at 6 Months After Repeat Ablation
• Atrial Fibrillation Severity Score (AFSS) - Part C
• Change in Weight at 3 Months After Ablation
• Change in Weight at 6 Months After Ablation
• Change in Hemoglobin A1c at 6 Months After Ablation
• Change in Hemoglobin A1C at 12 Months After Ablation

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is binary at the participant level: each participant is counted according to whether they were free from recurrent atrial arrhythmias in the specified post-blanking interval. For this kind of outcome, the arm-level event-free counts and denominators describe the observed data. A risk difference subtracts the comparator arm’s event-free proportion from the intervention arm’s event-free proportion. Positive values therefore indicate a higher event-free proportion in the metformin arm. The reported primary risk difference is 0.196.

What does the confidence interval represent?

A confidence interval describes the range of effect values reasonably compatible with the observed data under the calculation’s assumptions. It conveys uncertainty around an estimate, rather than identifying a guaranteed range for the true effect. The primary 95% interval runs from 0.016 to 0.375. Its width reflects that the estimate is based on a relatively limited number of outcome observations. The interval also helps distinguish the point estimate from its precision: a single estimate cannot show how much uncertainty surrounds it.

How should the p-value be read?

The unadjusted p-value for the primary risk-difference comparison is 0.038. In general, a p-value quantifies how unusual data at least this discrepant from a specified null model would be if that model and the analysis assumptions were true. It does not measure the probability that the null hypothesis is true, and it does not indicate the size or practical importance of an effect. The registry record does not report enough detail to confirm whether the posted unadjusted test was the prespecified primary test or how it relates to the sponsor’s reported hazard-ratio analysis.

Why are different effect measures reported?

The risk difference compares the proportions of participants meeting the event-free criterion by the end of the specified interval. A hazard ratio, by contrast, summarizes relative event rates over follow-up while taking event timing into account, subject to its modeling assumptions. The registry reports a hazard ratio of 0.5 and a 95% confidence interval from 0.2 to 0.9, but does not identify the method used for that analysis. Without details about event coding, censoring, model specification, and analysis population, the hazard-ratio result cannot be reconstructed from the endpoint counts alone. Neither effect measure should be treated as a conversion of the other.

What does an unadjusted comparison leave out?

An unadjusted comparison uses the posted arm-level summary data without accounting for additional variables in a model. This can be a transparent way to describe the observed groups, but it may not match a prespecified model-based analysis. Randomization supports a causal comparison when the trial is properly conducted and analyzed, but it does not make every later summary equivalent to the planned analysis. The registry record does not report whether the primary analysis adjusted for baseline factors or used a particular analysis population.

Why does over-dispersion matter in general?

For count outcomes, over-dispersion means that variability is greater than would be expected under a simple model that assumes a fixed relationship between the mean and variance. If it is ignored when present, standard errors and confidence intervals can be too small, making estimates appear more precise than they are. This is a general statistical consideration, not a reported feature of the primary endpoint. The primary outcome is a participant count with a binary event-free classification, and no over-dispersion assessment is reported in the registry record.

How should continuous secondary outcomes be understood?

For a continuous outcome, a mean difference compares average values between arms. The accompanying standard deviation describes spread among participants, while the confidence interval for the mean difference conveys uncertainty in the between-arm estimate. The registry reports a standard deviation for the AFSS score and for weight change. A mean difference is not a statement that every participant changed by that amount, and group averages can conceal individual variation. Details such as adjustment for baseline score, repeated-measures structure, and missing-data handling are necessary to understand a model-based estimate fully; these are not reported for every outcome.

What can the registry’s analysis labels establish?

The registry includes a superiority label for the reported primary hazard-ratio analysis and identifies a mixed models analysis for the AFSS outcome. A method label gives some information about the analysis type, but it does not specify the full model, estimand, or assumptions. The hazard-ratio method is not named. For the AFSS mixed models analysis, model details are not reported, and the registry’s p-value differs from the simple unadjusted recalculation listed for the arm-level summaries. Without further specification, the two results cannot be reconciled by assuming they test precisely the same quantity under the same assumptions.

11. Confidence Intervals and Precision

The primary risk-difference estimate is 0.196, and its 95% confidence interval extends from 0.016 to 0.375. This interval is compatible with a relatively small positive difference as well as with a materially larger one. The point estimate is therefore not precise enough to settle the size of the between-arm difference. Its lower limit is near zero, so the data are also compatible with a result close to no difference, although the reported interval does not include zero.

Confidence intervals should be interpreted as a measure of estimation uncertainty, not as a range containing the true value with certainty. They depend on the statistical method and its assumptions. The registry record does not report the detailed method used for the hazard-ratio analysis, so the reported hazard-ratio interval cannot be evaluated against a fully specified model from the record alone. The HR estimate is 0.5, with a 95% confidence interval from 0.2 to 0.9.

In general, limited outcome counts tend to yield less precise estimates than larger amounts of information, although precision also depends on the outcome’s variability and the method of analysis. The primary arm denominators are 49 and 50. The registry enrollment is 117, and the reason the endpoint denominators are smaller is not reported in the registry record. That uncertainty about the analysis population is relevant to judging how broadly the estimate applies.

The interval does not express uncertainty about every possible source of bias. It does not account for unreported choices in outcome handling, missing-data procedures, or selective reporting. Precision is therefore only one part of interpretation: the endpoint definition, analysis method, and completeness of follow-up also matter.

12. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are illustrations based on the observed effect, not the trial’s original sample-size design. The record’s assumptions for those illustrations are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. Each planning row uses the stated fraction of the observed effect as the effect to detect. For the primary endpoint, the figures are participants per arm. The time-to-event planning figures, where relevant, are total events rather than participants per arm.

The 80% power column indicates the approximate number of participants per arm that the illustration says would be needed for an 80% chance of detecting the specified effect under those assumptions, if that effect were the true effect and the analysis followed the assumed method. The 90% power column gives the corresponding planning target for a higher chance of detection. These are design concepts, not guarantees about the outcome of a future trial. Smaller effects need more participants to detect with the same target power, which is why the planning figures increase as the assumed fraction of the observed effect becomes smaller.

For the primary endpoint, the planning display gives 89 participants per arm for planning at 100% of the observed effect, 165 per arm for planning at 75% of the observed effect, and 383 per arm for planning at 50% of the observed effect, each for 80% power. The corresponding figures for 90% power are 118, 220, and 512 participants per arm. These values illustrate how strongly a planning target depends on the assumed effect size; the observed estimate itself is uncertain, so treating it as a fixed design truth would be misleading.

This is not observed power. The trial’s own original sample-size assumptions and justification are not reported in the registry record, and the illustrations cannot establish what the trial was designed to detect.

13. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.580; treatment proportion p₁ = 0.776; average p̄ = 0.678.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.6609 = 1.2954.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.6463 = 0.5439.
4. n = (1.2954 + 0.5439)² / (0.1955)² = 88.5, rounded up to 89 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.58                          # control proportion
p1 <- 0.775510204081633                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(4625946)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 89 (80% power), 118 (90% power)
 75% of observed effect: n per arm = 165 (80% power), 220 (90% power)
 50% of observed effect: n per arm = 383 (80% power), 512 (90% power)
power.prop.test: 88.5 
Simulated power at n = 89 per arm: 0.804 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

14. Analyses Reported by the Sponsor

The registry reports the primary sponsor analysis as a hazard ratio of 0.5, with a 95% confidence interval from 0.2 to 0.9, and labels it as a superiority analysis. The method and p-value for that analysis are not reported in the registry record. The separate unadjusted recalculation uses the arm-level counts for freedom from recurrence and expresses the comparison as a risk difference, with an unadjusted p-value. These analyses use different effect measures, and one cannot be substituted for the other.

The registry also reports a mixed models analysis for the AFSS outcome with p = 0.8. The unadjusted recalculation from the posted arm summaries gives a p-value of 0.107 for the mean difference. The mixed-model specification, estimand, and handling of repeated measurements or missing observations are not reported, so the difference between these p-values cannot be resolved from the record. They should not be treated as competing calculations of an identical analysis unless those details are known.

In general, a recalculation from posted summary statistics is useful as a transparent check of a simple, unadjusted comparison. It does not recreate a model-based sponsor analysis. The registry record does not report sufficient analysis detail to determine exactly how the primary sponsor estimate was obtained or how it relates to the unadjusted risk-difference estimate.

15. Multiplicity, Missing Data and Analysis Populations

Multiplicity arises when a study evaluates several outcomes, timepoints, or comparisons. Each additional statistical test creates another opportunity for a result to appear notable by chance. This does not make secondary outcomes uninformative, but it changes how strongly an isolated unadjusted result should be interpreted. A prespecified hierarchy, adjustment procedure, or clear exploratory designation can help readers understand how multiple findings were intended to be assessed.

The registry record lists one primary outcome and 13 secondary outcomes. It does not report a multiplicity adjustment, testing hierarchy, or detailed statistical analysis plan. It is therefore not possible to determine from the registry record whether the reported p-values were intended as confirmatory tests or how the study planned to control error across outcomes. The unadjusted p-values should not be read as though they account for all the outcomes considered.

The record also does not report the analysis populations used for each endpoint, the number of participants with missing outcome data, reasons for missingness, imputation procedures, or sensitivity analyses. The primary endpoint denominators are 49 and 50, whereas enrollment is 117; the relationship between those figures is not explained. The secondary outcomes have varying denominators in the posted numerical results, but the record does not provide a complete account of why data are unavailable for some participants or how that availability may affect comparisons.

Other details that are not reported in the registry record include the original sample-size justification, the prespecified handling of protocol deviations, and whether interim analyses were planned or conducted. These omissions limit assessment of how closely the posted analyses follow an intended confirmatory plan. They do not invalidate the reported counts or estimates, but they narrow what can be concluded about the study’s analysis process.

16. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: atrial fibrillation (hospital admission) (10.2%).

The registry reports serious adverse events affecting 5 participants in the metformin arm and 3 in the standard-of-care arm. It lists atrial fibrillation requiring hospital admission as the top serious-event term, at 10.2%. Deaths are recorded as 0 in both arms. Other adverse events are reported for 12 participants in the metformin arm and 0 in the standard-of-care arm.

These are event counts within the reported safety groups, not estimates of comparative risk adjusted for participant characteristics or follow-up. The serious-event figures do not establish that a particular treatment caused an event. The registry record does not provide a detailed account of event timing, severity grading beyond the serious-event classification, exposure duration, or attribution to treatment. It also does not report a statistical comparison of these safety counts.

Safety counts should be interpreted alongside the number of participants at risk in each arm and the way events were collected. The relevant reported at-risk groups are 49 and 50. A zero count in one arm means no event was recorded in that group in the posted results; it does not prove that the event is impossible or that risk is absent. Differences in follow-up or reporting can also matter, and those details are not reported in the registry record.

17. What the Result Does, and Does Not, Mean

The primary results show a higher observed count of participants free from recurrent atrial arrhythmias in the metformin arm than in the standard-of-care arm during the specified interval after the blanking period. The unadjusted risk-difference estimate is positive, and its reported confidence interval does not include zero. The registry’s sponsor-reported hazard-ratio analysis also carries a superiority label. These are descriptions of the posted trial results, not medical advice or a guarantee about an individual outcome.

The result applies to the trial’s comparison: metformin added to standard-of-care ablation and lifestyle recommendations versus standard-of-care ablation and lifestyle recommendations alone. It does not compare metformin with ablation, and it does not isolate the effects of lifestyle recommendations because those were part of both arms. It also does not establish a biological mechanism for any observed difference.

The endpoint concerns freedom from recurrence between 3 and 12 months after a single ablation. It does not describe arrhythmias during the blanking period, all aspects of symptoms, recurrence timing in full, or longer-term outcomes beyond the recorded interval. Secondary outcomes provide different information and should not be substituted for the primary outcome.

The estimate remains uncertain, and the registry record does not explain why the primary endpoint denominators are smaller than enrollment. Details on missing data, analysis populations, multiplicity, and the primary analysis method are not fully reported. Those limits matter when judging how robustly the posted estimates represent all randomized participants and how well they might generalize beyond this trial. The findings are evidence about this trial’s reported comparison; they do not settle every question about metformin use alongside ablation.

18. Why This Trial Matters Statistically

This trial illustrates why the choice of estimand matters. A count of participants free from recurrence at a defined time captures a different aspect of outcome than a hazard ratio that incorporates event timing. Both can be useful, but they answer related, not identical, statistical questions. Readers need the endpoint definition and analysis details to understand what a numerical estimate represents.

It also shows why the denominator is part of the result. Enrollment is 117, while the primary outcome is reported for 49 and 50 participants in the two arms. Without an explanation of this difference, interpretation depends on information the registry record does not report, including how outcome availability and analysis populations were handled. A clear account of participant flow and missing data is important for judging the reach of randomized comparisons.

The planning illustrations underline the influence of assumed effect size on study size. Planning around the full observed effect requires fewer participants than planning for a smaller effect, but an observed estimate is itself uncertain and may not be a reliable design target. The illustrations are distinct from the trial’s original design assumptions, which are not reported in the registry record.

Finally, the presence of multiple secondary outcomes makes transparent analysis planning important. An unadjusted secondary result has a different evidential role from a prespecified primary comparison. Taken together, these statistical issues show how endpoint definition, precision, analysis transparency, and multiplicity shape interpretation—not merely whether a single p-value crosses a conventional threshold.

19. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The registry record gives limited detail about the trial’s operational and statistical methods. The primary outcome denominators do not account for all reported enrollment, and the reason is not reported. Without participant-flow details, it is difficult to assess which randomized participants contributed to the primary comparison or whether exclusions differed between arms.

The record does not report the original sample-size calculation, the planned primary analysis in full, or a detailed method for the sponsor’s hazard-ratio estimate. It also does not give a complete account of missing-data handling, analysis populations, protocol deviations, interim analyses, or multiplicity control. These gaps limit assessment of whether the posted analyses follow a prespecified plan and how sensitive the results are to analytic choices.

The trial is described as open-label, so masking was absent. That matters particularly for outcomes involving participant reporting, follow-up behavior, or assessment decisions, although the registry record does not specify how such risks were addressed. Both arms received ablation and lifestyle recommendations, which helps define the comparison but means the result concerns adding metformin to that care context rather than the separate effects of each component.

The primary endpoint is limited to freedom from recurrence over a defined interval following a blanking period. The record does not establish longer-term outcomes, and the available secondary results cover varied measures and follow-up times. Some listed secondary outcomes do not have detailed numerical results in the registry record. The reported safety counts likewise lack detailed event timing, attribution, and comparative analyses.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Metformin as an Adjunctive Therapy to Catheter Ablation in Atrial Fibrillation — NCT04625946.
• Registry record last updated 2026-10-06; status last verified 2026-07.
• Results first posted 2026-10-06.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

NCT04625946 is a completed, randomized, open-label, parallel-group phase 4 trial sponsored by the University of Michigan. It enrolled 117 participants with atrial fibrillation who were having ablation. The comparison was metformin plus standard-of-care ablation and lifestyle recommendations versus standard-of-care ablation and lifestyle recommendations alone.

The primary outcome counts participants free from recurrent atrial arrhythmias between 3 and 12 months after a single ablation, excluding the 3-month blanking period. The posted arm counts and unadjusted comparison favor a higher event-free proportion in the metformin arm, with uncertainty around the size of the difference. The registry also reports a hazard-ratio analysis labeled as superiority.

Interpretation is limited by incomplete reporting of analysis methods, missing-data handling, analysis populations, and multiplicity procedures. The original sample-size justification is not reported in the registry record.