Randomized Controlled Trial of Group Prevention Coaching: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Randomized Controlled Trial of Group Prevention Coaching (NCT01838226), a randomized na study. The primary outcome is Risk of Fatal Coronary Event or Non-fatal MI.

2. Clinical Question

The intervention was intended to help Veterans work toward a health behavior goal of their choice that might improve heart disease risk. The study therefore asked whether a group-based, problem-solving form of prevention coaching was associated with a different cardiovascular risk estimate than usual care. The primary measure was a risk score rather than the occurrence of cardiac events during the study. A lower score represents lower estimated risk, while a higher score represents higher estimated risk.

The registry also listed patient activation and a later measurement of the cardiovascular risk score as secondary outcomes. These provide additional views of participant-reported activation and risk over time, but they do not replace the primary comparison at 6 months. The registry record does not report further details of the enrolled population or the specific behavior goals chosen by participants.

3. Trial Design

This was a randomized, parallel-group trial with single masking. The registry record lists an enrollment of 401 and describes the study as completed. The study start date was 2014-08-29, and the primary completion date was 2018-10-31. The sponsor was the VA Office of Research and Development. The registry record does not identify who was masked.

Participants were assigned to Group Prevention Clinics or Treatment as Usual Control. The posted outcome groups included 202 participants in the Group Prevention Clinics arm and 199 in the usual-care arm. The comparator is described as usual VA care. The coaching intervention combined group problem-solving sessions with interval phone calls to check goal progress and reinforce group learning. The groups were described as meeting monthly for 6 months, with each patient called between group sessions. The registry description says each group would consist of 10 patients; it does not report how many sessions or calls participants actually attended.

The primary outcome was the mean Framingham Risk Score for 10-year risk of fatal coronary event or non-fatal myocardial infarction at 6 months. Two additional outcomes were posted: Patient Activation Measure at 6 months and the cardiovascular risk score at 12 months. The registry lists three outcome measures and three analyses. It does not report the sequence-generation details, allocation concealment, study locations, identity of the masked party, or the trial's original sample-size justification.

Although the design was randomized, the posted results are group summaries and analysis estimates. They do not establish how the intervention was delivered in practice or whether participants followed the intended coaching schedule. Those implementation details are not reported in the registry record.

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.

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 Risk of Fatal Coronary Event or Non-fatal MI, measured over 6 months.

The primary endpoint was the mean 10-year risk of a fatal coronary event or non-fatal myocardial infarction, measured 6 months after enrollment using the Framingham Risk Score. The score is expressed as a percentage of estimated risk and ranges from 0 to 100. A higher score represents higher estimated risk. The registry describes the primary outcome as change in 10-year risk, while the posted analysis is labeled as a mean difference in final values.

A risk score is a continuous measure: each participant contributes a value on a numerical scale, and the group mean summarizes the values for that arm. The standard deviation describes how spread out individual scores are around the arm's mean. Comparing the group means gives a direct estimate of the difference in average scores, but does not show how many participants experienced an actual cardiac event.

The endpoint was used to assess cardiovascular risk at the end of the intervention period, rather than to count clinical events occurring during follow-up. The distinction matters: the posted result concerns a risk estimate, not an observed reduction or increase in fatal coronary events or myocardial infarctions. The registry record does not report additional detail about how the score was calculated beyond naming the Framingham Risk Score.

7. Results Posted in the Registry

• Group Prevention Clinics: mean 27.00 versus 26.30 on control; difference 0.70 (95% CI -2.50 to 3.90), standardized effect d = 0.04; z-test p = 0.668.

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.

Patient Activation Measure

• 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:
  • Group Prevention Clinics: mean 62.10 versus 62.60 on control; difference -0.50 (95% CI -2.82 to 1.82), standardized effect d = -0.04; z-test p = 0.673.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Group Prevention Clinics and the control. The interval gives the range of effects that remain compatible with the data.

Risk of Fatal Coronary Event or Non-fatal MI

• 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:
  • Group Prevention Clinics: mean 25.70 versus 28.20 on control; difference -2.50 (95% CI -5.66 to 0.66), standardized effect d = -0.16; z-test p = 0.120.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Group Prevention Clinics and the control. The interval gives the range of effects that remain compatible with the data.

The secondary outcomes give a broader but still limited view alongside the primary risk-score comparison. Patient Activation Measure was assessed at 6 months, while the cardiovascular risk score was also reported at 12 months. Across these outcomes, the posted group contrasts do not form a uniform pattern of clearly separated results: the activation comparison is close to zero, while the later risk-score comparison is more favorable in direction to the intervention arm. These results complement the primary finding but do not change what the primary endpoint estimates.

The secondary analyses have less evidential weight than the designated primary analysis, particularly because several outcomes and follow-up times create multiple opportunities to observe a result that appears noteworthy by chance. The registry record does not report a multiplicity adjustment or a prespecified hierarchy for interpreting the secondary outcomes. Their p-values and confidence intervals, where reported, should therefore be understood as individual analyses rather than as a family-wise controlled set of tests.

Some secondary analyses have model-based results, while the available record also includes unadjusted calculations from group summaries. The summaries and model estimates should not be treated as interchangeable. The 12-month risk score is a later measurement of the same type of risk estimate, not a count of cardiac events. Patient Activation Measure captures a separate construct and cannot be treated as a direct clinical outcome. The registry record does not report how these secondary measures were prioritized or whether additional secondary outcomes were assessed but not posted.

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

• Patient Activation Measure
• Risk of Fatal Coronary Event or Non-fatal MI

9. Statistical Interpretation

10. Statistical Methods Explained

What does the mean difference describe?

The primary outcome is continuous, so the posted arm results report a mean and standard deviation for each group. The mean is the arithmetic average of the scores in that arm. The standard deviation describes variation among participant scores; it is not the uncertainty around the mean difference. The primary arm summaries were 27.0 with a standard deviation of 17.2 for Group Prevention Clinics, and 26.3 with a standard deviation of 15.5 for Treatment as Usual Control.

The unadjusted re-calculation subtracts the control-arm mean from the intervention-arm mean. The resulting mean difference is 0.7. With this direction of subtraction, positive values mean a higher average score in the intervention arm. Since higher scores represent higher risk, the sign has a straightforward interpretation for this measure. This is a summary-statistic comparison and does not incorporate any covariate adjustment unless such adjustment is explicitly part of the analysis.

Why are there two estimates?

The registry reports a mixed-model analysis of mean difference in final values. Its estimate is 0.7, its 95% confidence interval is -1.3 to 2.8, and its p-value is 0.50. Separately, an unadjusted re-calculation based on the posted means and standard deviations gives the same central estimate, but a wider confidence interval, from -2.504 to 3.904, and p = 0.668. These estimates answer closely related questions but arise from different procedures. A mixed model can incorporate the structure and terms specified in its analysis; the registry record does not report enough detail to identify all of those choices. The re-calculation is not a substitute for the sponsor's reported analysis.

The discrepancy indicator is false, so the record does not flag a discrepancy between the reported and recalculated central estimates. The estimates nevertheless have different confidence intervals and p-values. That is consistent with the analyses using different methods. The record does not provide the model specification or enough information to reproduce the reported model result.

What does a confidence interval express?

A confidence interval gives a range of estimates compatible with the data under the analysis method. Its width reflects precision: a narrower interval indicates a more precise estimate, while a wider interval leaves a broader range of plausible differences. The primary reported interval crosses zero, the value that represents no mean difference. That means the data do not precisely distinguish a positive from a negative difference using this interval. It does not prove that the groups are equal.

A confidence interval is not a range containing a fixed proportion of individual participant scores. Nor is it the probability that the true difference falls inside the particular interval. It is a repeated-sampling measure of the performance of the interval procedure under its assumptions. Interpretation also depends on how participants were selected, how outcomes were measured, and whether the analysis assumptions are reasonable.

How should the p-value be read?

The reported primary p-value is 0.50. The unadjusted re-calculation gives p = 0.668. A p-value measures how unusual data at least as incompatible with the null model would be if the null model and the test assumptions held. It is not the probability that the null hypothesis is true, and it does not measure the size or practical importance of a difference. It should be considered alongside the estimated difference and its confidence interval.

The analysis is labeled superiority. That label describes the type of comparison, not a conclusion about the result. The registry record does not report an equivalence margin or non-inferiority framework. A non-significant superiority test cannot be interpreted as evidence that two approaches are equivalent.

What does variability add?

The standard deviations show that participant scores varied within each arm. The group means compress that variation into a single summary, while the confidence interval communicates uncertainty around the between-arm estimate. Both are useful: variability helps describe the observed participants, and the interval helps assess how precisely the group contrast was estimated. The registry record does not report distributions, medians, or graphical diagnostics for the participant-level scores, so the degree to which the means represent any skewed or unusual score patterns cannot be assessed from these summaries alone.

The statistical results concern the endpoint as recorded and analyzed. They do not by themselves establish whether the intervention was implemented consistently, whether the assumptions of the analysis were met, or how representative the enrolled participants were. Those issues require information not reported in the registry record.

11. Confidence Intervals and Precision

The reported primary estimate is a mean difference of 0.7, with a 95% confidence interval from -1.3 to 2.8. The range crosses zero and includes both negative and positive differences. Accordingly, the posted analysis leaves uncertainty about the direction as well as the magnitude of the group contrast. The interval is not narrow enough to support a precise estimate of a small difference, and it does not establish that the groups are equivalent.

The unadjusted re-calculation gives a mean difference of 0.7, with a 95% confidence interval from -2.504 to 3.904. This range is wider than the interval from the mixed-model analysis. The central estimates are the same, but the amount of uncertainty differs between analyses. The registry record does not report sufficient model details to identify why the intervals differ, and the re-calculation should be understood as a separate, unadjusted summary-statistic analysis.

Precision is affected by the amount of information available and the variation in the measured scores, among other factors. Here the arm summaries include 202 participants in Group Prevention Clinics and 199 in usual care, with standard deviations of 17.2 and 15.5. These summaries describe the posted groups, but they do not reveal whether outcome data were available for all enrolled participants or how missing observations were handled.

Neither confidence interval should be read as a statement about individual outcomes. It describes uncertainty in a group-level mean difference under the corresponding statistical method. It also does not address possible selection, measurement, or implementation limitations. The numerical estimates are most informative when considered together with those limitations and with the endpoint's meaning as a risk score rather than a count of observed cardiac events.

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 for the primary endpoint are participants per arm, not total participants and not total events. They are illustrations based on the observed effect, not the trial's original design calculation. The stated planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is set to the stated fraction of the observed effect.

For planning at 100% of the observed effect, the illustration gives 8,594 participants per arm for 80% power and 11,505 participants per arm for 90% power. For planning at 75% of the observed effect, it gives 15,278 participants per arm for 80% power and 20,453 participants per arm for 90% power. For planning at 50% of the observed effect, it gives 34,376 participants per arm for 80% power and 46,019 participants per arm for 90% power.

The 80% power column indicates the number of participants per arm that the stated approximation associates with an 80% chance of detecting the specified effect under its assumptions. The 90% column uses a higher target chance of detection and therefore requires more participants. Smaller effects are harder to distinguish from random variation, so the planning figures rise as the assumed effect becomes a smaller fraction of the observed effect.

These figures are planning illustrations, not observed power. They do not show what the completed trial's power was, and they do not establish that the trial was planned using these assumptions. The original design's sample-size assumptions are not reported in the registry record.

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, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 0.700; pooled standard deviation σ = 16.378.
2. n = 2 × (1.96 + 0.842)² × 16.378² / 0.700² = 8,593.9, rounded up to 8,594 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
delta <- 0.699999999999999                      # observed difference in means
sigma <- 16.37846100714                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^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 (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(1838226)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 8,594 (80% power), 11,505 (90% power)
 75% of observed effect: n per arm = 15,278 (80% power), 20,453 (90% power)
 50% of observed effect: n per arm = 34,376 (80% power), 46,019 (90% power)
power.t.test: 8594.8 
Simulated power at n = 8,594 per arm: 0.800 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-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 sponsor's primary result as a mixed-model mean difference in final values of 0.7, with a 95% confidence interval from -1.3 to 2.8 and p = 0.50. This is the study's reported analysis. The re-calculation from the posted arm means and standard deviations also gives a mean difference of 0.7, but its 95% confidence interval is -2.504 to 3.904 and its p-value is 0.668.

The central estimate is the same in both analyses, while their uncertainty estimates differ. The mixed-model analysis and the simple unadjusted re-calculation need not produce identical confidence intervals or p-values because they use different statistical procedures. The registry record does not provide the full mixed-model specification, so the distinction cannot be examined in detail. The record does not flag a discrepancy in the estimates.

The sponsor's result should be read as the reported model-based analysis, while the re-calculation is a transparent comparison based only on the posted group summaries. Both intervals include zero and neither gives a precise estimate of the direction of the difference. The re-calculation is useful as a check on the arm summaries, but it does not replace or reproduce the sponsor's model.

15. Multiplicity, Missing Data and Analysis Populations

The registry record lists three outcome measures and three analyses, including the primary cardiovascular risk score and two posted secondary outcomes. It does not report a multiplicity strategy, such as an adjustment to the significance threshold or a prespecified order for testing outcomes. When several outcomes or time points are examined, the chance of at least one apparently unusual result can increase. Without a reported plan, secondary results are best viewed as supportive or exploratory rather than as a set of independently confirmed findings.

Missing-data handling is not reported in the registry record. The arm summaries list 202 participants for Group Prevention Clinics and 199 for usual care, but the record does not explain whether these are all randomized participants with available outcome measurements, whether any observations were imputed, or how incomplete follow-up was handled. The numbers alone do not establish that all randomized participants contributed to every analysis.

The analysis population is also not reported in sufficient detail. The registry record does not specify whether the primary analysis followed intention-to-treat principles, used a per-protocol population, or applied another rule for classifying participants. It does not report exclusions after randomization or how adherence to coaching affected analysis inclusion.

The record does not report interim analyses, stopping rules, or the original sample-size justification. These details should not be inferred from the completed status or from the planning illustrations based on the observed effect. Without the statistical analysis plan and fuller reporting of missingness and populations, readers cannot fully assess whether the posted analyses followed a prespecified approach or how sensitive the results might be to incomplete data and analytic choices.

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 5% reporting threshold.

Most frequent serious adverse event terms

Serious events with the highest arm-level rate: death (0.5%).

The serious adverse event figures list one affected participant in the Group Prevention Clinics arm and none in the usual-care arm. The serious event term shown in the registry is death, reported as 0.5%. The death counts are one in Group Prevention Clinics and zero in Treatment as Usual Control.

The figures also list zero other adverse events in both arms. These are counts as posted in the registry record; they do not describe the circumstances, timing, cause, or relationship of the death to the intervention. The registry record does not provide a narrative assessment of causality or additional detail about the event.

These counts should be interpreted cautiously. A small number of observed events cannot establish whether there is a difference in safety between groups, and the absence of other listed events does not rule out unreported or unobserved events. The record does not report a broader safety analysis or detailed ascertainment procedures. The counts are descriptive, not evidence that the intervention caused or prevented an event.

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

The posted primary analysis estimates a mean difference of 0.7 in the 6-month Framingham risk score, with a 95% confidence interval from -1.3 to 2.8. The interval includes zero and allows for differences on either side of zero. The result therefore does not give a precise estimate of the direction or size of the average difference. It also does not establish that the two groups are equivalent.

The endpoint is an estimate of 10-year cardiovascular risk, not a record of whether participants experienced a fatal coronary event or non-fatal myocardial infarction during the trial. A difference in this score should not be described as an observed difference in cardiac events. The outcome was measured at 6 months, and the registry also posted a later risk-score outcome at 12 months. The registry record does not describe the score calculation beyond identifying it as the Framingham Risk Score.

The randomized comparison addresses the groups as analyzed, but the registry record does not provide enough detail to determine the analysis population, missing-data approach, or all model specifications. Those omissions limit how fully the result can be evaluated. The same caution applies to the secondary outcomes, for which a multiplicity plan is not reported.

The findings do not establish that group coaching does or does not work, nor do they establish that usual care is equivalent to coaching. They describe the posted estimates and their precision for specified measures and times. The clinical meaning of a difference in a risk score, its relation to actual event rates, and the relevance to people outside the enrolled study population are not resolved by these statistical results alone.

18. Why This Trial Matters Statistically

This trial illustrates why an estimate should be interpreted together with its uncertainty rather than through a single p-value. The primary mean difference is 0.7, and the 95% interval spans values below and above zero. The reported analysis therefore leaves uncertainty about both the direction and the magnitude of the average contrast, even though a numerical estimate is available.

It also demonstrates the distinction between a model-based reported analysis and an unadjusted re-calculation from arm summaries. The central estimates match, but the intervals and p-values differ. Understanding what each analysis does—and what information is missing from the record—is important for making a fair statistical reading without treating the simpler calculation as a replacement for the reported model.

The endpoint itself matters. A Framingham risk score is a continuous estimate of future risk, whereas an observed cardiac event is a clinical occurrence. The primary result concerns the former. Keeping that distinction clear prevents a score comparison from being recast as evidence about event prevention.

Finally, the study offers a useful example of how randomization, a defined primary outcome, and additional secondary measures fit together, while also showing the importance of reporting missing-data methods, analysis populations, multiplicity plans, and model specifications. The registry record does not provide those details, so readers must separate what the posted estimates establish from what remains uncertain.

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 does not report detailed eligibility criteria, participant demographics, study locations, or the number of sites. This limits assessment of how broadly the results might apply beyond the enrolled Veterans in the VA care setting. The condition is listed as cardiovascular disease, but the record does not give enough detail to characterize participants' baseline clinical status.

The intervention description outlines intended group sessions and interval phone calls, but actual attendance, completion, fidelity, and goal selection are not reported. It is therefore not possible to determine from these results how consistently the coaching model was delivered or how much exposure participants received. The identity of the party masked in this single-masked trial is also not reported.

The primary outcome is a calculated risk score rather than observed cardiovascular events. The registry record does not provide further details on score inputs, measurement procedures, or whether the score was calculated in the same way across arms. The posted primary analysis is described as mean difference in final values, while the study summary describes change in risk; additional analysis detail is not reported.

Important statistical details are absent, including the original sample-size justification, the full mixed-model specification, the analysis population, missing-data procedures, and any multiplicity plan. Interim analyses and stopping rules are not reported. The secondary outcomes have no reported multiplicity approach, and the results should be interpreted with that limitation in mind. The registry record also provides limited detail about the serious event, so the posted safety counts cannot establish event causality.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Randomized Controlled Trial of Group Prevention Coaching — NCT01838226.
• Registry record last updated 2023-07-27; status last verified 2023-07.
• Results first posted 2020-05-19.
• 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

This completed randomized, parallel-group trial compared Group Prevention Clinics with usual VA care and enrolled 401 participants. The primary endpoint was the mean Framingham estimate of 10-year risk of fatal coronary event or non-fatal myocardial infarction at 6 months.

The sponsor-reported mixed-model estimate was a mean difference of 0.7, with a 95% confidence interval from -1.3 to 2.8 and p = 0.50. The interval includes zero and does not provide a precise direction of difference. The unadjusted re-calculation produced the same central estimate with a wider interval. Two secondary outcomes were posted, but the registry record does not report a multiplicity plan, missing-data method, detailed analysis populations, or full model specifications.