Exercise Rehabilitation in Veterans With PAD: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Exercise Rehabilitation in Veterans With PAD (NCT02607046), a randomized na study. The primary outcome is Modified Physical Performance Test.

2. Clinical Question

3. Trial Design

The registry describes a completed, randomized, parallel-group trial titled “Exercise Rehabilitation in Veterans With PAD.” The official title identifies the focus as post-revascularization rehabilitation to improve function in veterans with PAD. The listed enrollment is 56. The trial compares exercise training with neuromuscular electrical stimulation, and the registry lists no masking. Its phase is recorded as NA, which is the registry entry rather than a clinical phase designation that can be interpreted from the posted record.

The registry dates the study start to 2015-10-01 and primary completion to 2019-12-31. It identifies the VA Office of Research and Development as sponsor and records the study status as completed. These registry entries describe the trial framework, but they do not by themselves establish how consistently the interventions were delivered or how completely participants were followed.

The planned clinical focus, as described in the registry summary, was to examine the effects of 3 months of exercise rehabilitation or neuromuscular stimulation on leg blood flow, physical function, quality of life, and general health in patients with PAD who require revascularization. The primary outcome posted is the Modified Physical Performance Test at 3 months. The registry also posts muscle perfusion measured by ultrasound scan at 3 months as a secondary outcome. Although the brief summary mentions quality of life and general health as areas of interest, the registry record does not report posted results for those measures.

The outcome summaries list 14 participants for Exercise and 10 for NMES for the primary measure. Those outcome counts do not match the enrollment figure, and the registry record does not report why or how the analysis counts differ from enrollment. Random allocation is recorded, but the method used to generate or conceal the allocation sequence is not reported in the registry record. The record also does not report details of treatment adherence, withdrawals, or the analysis population.

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: 56 participants enrolled; only 35 participants completed baseline testing and were randomized.

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 Modified Physical Performance Test, measured over 3 months.

The primary endpoint is the Modified Physical Performance Test, reported as a mean score with standard deviation at 3 months. The record describes it as a test of mobility function, with a minimum score of 0 and a maximum score of 36. Higher scores are associated with better outcome. These features make it a continuous measure: each participant contributes a score on a numerical scale, and group results can be summarized by a mean and the spread of scores around that mean.

A mobility measure is relevant to a study of rehabilitation in PAD because the registry summary identifies difficulty or pain with walking and activities using leg muscles as concerns associated with the condition. The test therefore addresses physical function rather than a direct measurement of blood flow. The record states that the study aimed to examine leg blood flow, physical function, quality of life, and general health, but only the specified test score is the primary endpoint posted for this analysis.

The mean offers a compact description of each group’s average score, while the standard deviation describes variation among participant scores. Neither measure describes the full distribution on its own. The reported group comparison is a difference in mean scores, with NMES contrasted against Exercise. A positive estimate indicates a higher average score in NMES in this coding; it should not be confused with proof of an important benefit. The registry record does not report a threshold for clinical importance or an interpretation rule for changes in the test score. Consequently, statistical comparison and clinical relevance are separate questions.

7. Results Posted in the Registry

• NMES: mean 32.30 versus 31.80 on control; difference 0.50 (95% CI -2.80 to 3.80), standardized effect d = 0.11; z-test p = 0.766.

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.

Muscle Perfusion Measured by Ultrasound Scan

• 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:
  • NMES: mean 4.59 versus 6.13 on control; difference -1.54 (95% CI -4.47 to 1.39), standardized effect d = -0.39; z-test p = 0.303.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between NMES and the control. The interval gives the range of effects that remain compatible with the data.

The posted secondary outcome concerns muscle perfusion measured by ultrasound scan at 3 months. Its group summaries and comparison provide a physiological measure alongside the primary measure of mobility function. The two outcomes address related but distinct aspects of the study’s stated aims: one is a mobility test score, and the other is a measure of muscle perfusion. A pattern across them can add descriptive context, but one outcome does not validate or replace the other.

The secondary result should be interpreted cautiously in relation to the primary result. A secondary outcome may help characterize what was measured, but it does not independently settle the primary clinical question. The registry record lists one secondary outcome and does not report a set of prespecified hypotheses, a hierarchy among outcomes, or an analysis plan describing how findings across outcomes were to be interpreted.

General: when several outcomes are examined, the chance of at least one apparently notable result arising by chance can increase if each is tested separately without accounting for multiplicity. An unadjusted secondary result therefore generally carries less confirmatory weight than a prespecified primary comparison, especially when the analysis strategy and outcome hierarchy are not reported. This is not a claim that a particular secondary result is false; it is a reason to avoid treating it as decisive evidence on its own.

The registry summary also mentions quality of life and general health among the areas the project aimed to examine, but posted results for those measures are not reported in the registry record. The available secondary outcome information is consequently limited to the ultrasound measure. There is no basis in the record to describe a broad pattern across quality of life, general health, or other unposted endpoints. Overall, the secondary measure offers complementary descriptive evidence, while the primary mobility comparison remains the principal posted result for the main study question.

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

• Muscle Perfusion Measured by Ultrasound Scan

9. Statistical Interpretation

10. Statistical Methods Explained

The posted primary outcome is continuous, so it is summarized by group means and standard deviations. General: a mean is the arithmetic center of a set of measurements, while a standard deviation describes how dispersed individual measurements are around that center. A standard deviation is not a confidence interval and does not indicate how certain the mean difference is. The standard errors listed for the group means describe sampling uncertainty in those means under the calculation used.

What does the mean difference represent?

The comparison is expressed as NMES minus Exercise. The reported estimate of 0.5 is therefore the difference between the two posted group means on the Modified Physical Performance Test scale. With this direction of subtraction, a positive estimate means the NMES group had the higher observed mean, while a negative estimate would mean the Exercise group had the higher observed mean. This sign convention is a description of the contrast, not a clinical judgment about which score difference matters.

The comparison is unadjusted: it uses the posted group summary statistics rather than a reported model controlling for other variables. General: an adjusted analysis can account for prespecified factors such as baseline measurements, if those factors and the model are appropriate and actually reported. The registry record does not report such an analysis, so the posted contrast should not be described as adjusted for baseline function or other characteristics.

How should the p-value be read?

The unadjusted p-value is 0.766. General: a p-value describes how unusual a result at least as far from the null value would be under a specified statistical model if the null hypothesis were true. It is not the probability that the null hypothesis is true, nor is it a measure of clinical importance. A large p-value does not prove that there is no difference. In a modest outcome dataset, estimates can remain imprecise even when the observed p-value is not small.

The registry record does not report the trial’s original statistical analysis plan, the significance threshold selected for the primary analysis, or any prespecified covariate adjustment. The posted sample-size illustrations use a two-sided significance level of 0.05, but those illustrations are planning calculations and do not establish the analysis plan used for the completed trial. The posted p-value is labeled unadjusted.

What information does the confidence interval add?

The confidence interval for the mean difference runs from -2.798 to 3.798. General: a confidence interval is a range of estimates generated by a method designed to have a stated coverage rate across repeated samples when its assumptions hold. It conveys the precision of an estimate and the range of values reasonably compatible with the data and calculation. It is not a guarantee that the true effect lies inside the particular interval, and it does not assign probabilities to individual effect values.

Because this interval crosses zero, it includes both a lower mean in NMES and a higher mean in NMES relative to Exercise. Its width also means that the posted data do not pin down the difference very closely. The registry does not report a clinical-importance threshold, so the interval cannot be divided into clinically meaningful and unimportant ranges using a registry-defined criterion.

Why can over-dispersion matter?

General: over-dispersion occurs when variability in observed data is greater than assumed by a statistical model. If it is not accounted for, standard errors may be too small and confidence intervals too narrow. The primary endpoint is a continuous score and the record reports standard deviations; it does not identify an over-dispersion issue or report a model for addressing one. Accordingly, over-dispersion is a general statistical consideration, not a finding about this trial.

The secondary ultrasound outcome is also continuous and is reported using means and standard deviations. Its comparison is unadjusted as posted. The same broad distinction applies: the observed group means describe the recorded values, while uncertainty and analysis assumptions govern what can be inferred about a wider population. No multiplicity adjustment or formal model is reported for the secondary outcome.

What does randomization contribute?

General: random assignment is intended to reduce systematic differences between treatment groups at the point of allocation, supporting a causal comparison when the trial is conducted and analyzed as planned. It does not guarantee balance in every characteristic, particularly in a relatively small outcome dataset, and it does not resolve missing outcome information or deviations after assignment. The registry records randomized allocation but does not report sequence generation, concealment, adherence, or the analysis population. Those details are therefore not available to assess from the record.

The record posts means, standard deviations, mean differences, confidence intervals, and unadjusted p-values for the reported outcomes. It lists no reported analyses beyond those summaries. The estimates should therefore be read as descriptive, unadjusted comparisons with uncertainty, rather than as results from a fully documented model or a complete account of the statistical plan.

11. Confidence Intervals and Precision

Precision concerns how narrowly the data estimate the group difference. For the primary outcome, the mean difference is 0.5 points, and its confidence interval ranges from -2.798 to 3.798. This interval includes zero and extends to differences in either direction. It therefore does not isolate a single direction or a narrow range of possible differences. The unadjusted p-value of 0.766 is consistent with the fact that the observed estimate is not clearly separated from zero under the reported comparison, but the confidence interval conveys the range of uncertainty more directly.

The outcome counts are 14 for Exercise and 10 for NMES. The standard deviations are 5.2 and 3.0. These group summaries help describe the observed data, but neither the group counts nor standard deviations alone determine whether the estimate is precise. The confidence interval incorporates the variability and the available outcome information under the calculation used. With the registry record not explaining the difference between enrollment and the outcome counts, the completeness of the analyzed data cannot be assessed from the posted record.

A confidence interval is not a range of guaranteed effects, and values inside it should not be treated as equally likely. General: narrower intervals indicate more precise estimates, while wider intervals leave more uncertainty about the underlying effect. Whether the values in an interval would be clinically important requires a meaningful-difference criterion. The registry record does not report such a criterion for this test. As a result, the statistical uncertainty can be described, but the clinical interpretation of particular values cannot be resolved using a threshold in the record.

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 participants per arm, not total events. The calculations are based on a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. They use the observed effect as a reference and show planning for 100%, 75%, and 50% of that observed effect. The trial’s own original sample-size assumptions are not reported in the registry record.

The 80% power column gives the estimated number of participants per arm for a hypothetical future study to have an 80% chance of detecting the stated effect under the listed planning assumptions, if that effect is the true one and the calculation’s assumptions hold. The 90% power column gives the corresponding estimate for a higher chance of detecting that effect, with other planning assumptions unchanged. These are prospective planning illustrations, not a description of how many participants were analyzed in this completed trial.

The rows use different fractions of the observed effect. Planning for 75% of the observed effect requires more participants per arm than planning for the full observed effect; planning for 50% requires still more. This reflects a general statistical principle: smaller effects are harder to distinguish from random variation, so detecting them with a chosen power generally requires a larger study. The figures should not be read as an enrollment recommendation without considering clinical importance, feasibility, the intended analysis, follow-up, and any expected loss of outcome information.

For the primary outcome, the estimates are 1,235 participants per arm for 80% power and 1,653 for 90% power when planning for 100% of the observed effect. For 75%, the figures are 2,195 and 2,938; for 50%, they are 4,938 and 6,611. These are planning illustrations based on the observed effect, not observed power. Because the observed effect is an uncertain estimate from the posted data, calculations anchored to it can yield highly uncertain planning numbers, particularly when that estimate is small or imprecise. The record does not report the original design rationale, target effect, or assumptions used when the trial was planned.

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 = 0.500; pooled standard deviation σ = 4.434.
2. n = 2 × (1.96 + 0.842)² × 4.434² / 0.500² = 1,234.5, rounded up to 1,235 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 <- 0.499999999999996                      # observed difference in means
sigma <- 4.43395985547907                     # 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(2607046)
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 = 1,235 (80% power), 1,653 (90% power)
 75% of observed effect: n per arm = 2,195 (80% power), 2,938 (90% power)
 50% of observed effect: n per arm = 4,938 (80% power), 6,611 (90% power)
power.t.test: 1235.4 
Simulated power at n = 1,235 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. Multiplicity, Missing Data and Analysis Populations

The registry identifies the Modified Physical Performance Test as the primary outcome and lists muscle perfusion measured by ultrasound scan as a secondary outcome. It does not report a multiplicity adjustment, a formal hierarchy for testing outcomes, or a complete list of prespecified analyses. The record also mentions quality of life and general health as study aims but does not report results for those measures. It is therefore not possible to determine whether these or other outcomes were included in the planned testing family.

General: multiplicity arises when multiple hypotheses or outcomes are examined. Without an appropriate strategy, interpreting each unadjusted p-value as if it were the only test can overstate the strength of evidence. A prespecified primary outcome helps focus interpretation, while secondary outcomes can be useful for description or hypothesis generation. The registry record does not state whether the secondary outcome was intended to support a confirmatory claim, and no adjustment should be assumed.

Missing-data handling is also not reported. Enrollment is listed as 56, while the primary outcome summaries include 14 participants in Exercise and 10 in NMES. The record does not explain the difference, state how many participants were randomized to each arm, or describe whether outcome data were unavailable, excluded, or analyzed under a particular rule. It does not report withdrawals, reasons for missing measurements, imputation, sensitivity analyses, or a comparison between participants with and without outcome data.

The registry record does not identify an intention-to-treat analysis, a per-protocol analysis, or another analysis population. It therefore cannot establish whether the posted summaries include all randomized participants according to assigned group or only participants with available measurements. General: missing outcomes can affect both precision and the validity of group comparisons, depending on why data are missing and how they are handled. Without those details, the direction or size of any such effect cannot be inferred.

These reporting gaps do not demonstrate that multiplicity or missing-data problems occurred. They limit what can be assessed from the record. The estimates and p-values are best understood as the posted unadjusted summaries, without attributing to them procedures that the registry does not describe.

15. 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

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: chest pain (4.5%).

The registry reports serious adverse event figures for participants at risk in each group. In Exercise, 22 were at risk and 1 was affected by a serious adverse event; in NMES, 13 were at risk and 0 were affected. The registry also lists other adverse event figures: 22 at risk and 4 affected in Exercise, and 13 at risk and 1 affected in NMES. Deaths are recorded as 0 in each group.

These counts describe the safety information posted in the registry. The serious-event entries are counts, not a comparison adjusted for differences in follow-up, participant characteristics, or exposure. The registry lists Chest Pain at 4.5% among the top serious terms, but does not provide further context in the record that would support attributing that term to a particular group or interpreting its clinical circumstances.

The counts should not be treated as proof that one intervention is safer. The registry record does not report event definitions, timing, severity beyond the serious/non-serious categorization, relatedness assessments, follow-up duration for safety, or whether events were counted per participant or by episode. Small counts also leave substantial uncertainty. These results are a descriptive account of the adverse event entries, not a complete comparative safety analysis.

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

The main result is a small observed difference in average mobility scores at 3 months, with the NMES mean higher than the Exercise mean under the reported contrast. The confidence interval includes differences in either direction and is broad enough that the data do not determine a precise group difference. This is a statement about the posted estimates and their uncertainty, not a conclusion that the treatments are equivalent or that either treatment does or does not work.

The result also does not compare either intervention with no treatment, usual care, or an untreated group. Both randomized groups received an intervention, and the registry record does not report another comparator. The results therefore address only the posted comparison between Exercise and NMES at the specified follow-up. They do not establish the effects of rehabilitation relative to another care pathway.

The outcome is a mobility test score, not a direct measure of every aspect of function or health mentioned in the study summary. Muscle perfusion was posted as a secondary measure, but quality of life and general health results are not reported in the registry record. Nor does the record report longer-term outcome results. Conclusions should remain limited to the measures and timing actually posted.

Whether a difference of a particular size would matter to patients or clinicians cannot be determined from the registry record because it does not report a threshold for clinical importance. A p-value alone cannot answer that question, and a confidence interval cannot supply a clinical threshold that has not been specified. Interpretation should distinguish statistical uncertainty, observed scale scores, and judgments about meaningful change.

Finally, the difference between the enrollment entry and the outcome counts is unexplained. Without details on allocation, follow-up, missing measurements, and analysis population, it is not possible to assess how those factors affect the comparison. The most defensible conclusion is descriptive: the posted group averages were close, but the available comparison does not precisely determine the underlying difference.

17. Why This Trial Matters Statistically

This trial is statistically informative because it illustrates the difference between a point estimate and the precision of that estimate. The primary comparison has a numerical estimate, but its confidence interval spans values on either side of no difference. Focusing only on the estimate or the p-value would miss how much uncertainty remains. For clinical researchers, reading the interval alongside the group means and outcome counts is essential to avoid treating an imprecise result as a definitive answer.

The study also connects a functional endpoint with a physiological secondary measure. Those outcomes can illuminate different parts of the research question, but they should not be conflated. A mobility score and an ultrasound measure do not have the same meaning, and evidence on a secondary measure cannot substitute for a clear interpretation of the primary outcome. The absence of posted results for some aims noted in the registry summary further illustrates why conclusions must track the outcomes actually reported.

The planning illustrations show how strongly required sample size can depend on the effect chosen for detection. Smaller target effects require larger planning numbers under the stated assumptions. That lesson is useful for prospective design, but calculations based on an observed estimate are not a substitute for a prespecified, clinically justified design target. The registry record does not report the original sample-size rationale, so the posted illustrations cannot explain the enrollment decision.

Finally, the gap between listed enrollment and reported outcome counts highlights the value of transparent reporting about participant flow, missing outcomes, and analysis populations. Such details are necessary to evaluate whether a randomized comparison retains the intended benefits of random assignment. The record supports a careful statistical reading, while leaving important design and analysis questions unresolved.

18. 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 intervention protocols. It names exercise training and neuromuscular electrical stimulation and describes a 3 months rehabilitation period, but does not report the specific content, intensity, frequency, supervision, adherence, or co-interventions. Differences in delivery cannot be assessed from the posted information.

The participant and outcome reporting is also incomplete for interpreting the comparison. Enrollment is listed as 56, whereas primary outcome summaries list 14 participants for Exercise and 10 for NMES. The registry record does not explain this difference, detail participant flow, or state the analysis population. Baseline characteristics, baseline primary scores, withdrawals, missingness reasons, and methods for handling missing outcomes are not reported.

The record indicates randomized allocation and no masking, but does not report how the allocation sequence was generated or concealed, whether outcome assessors were masked, or whether any departures from assigned treatment occurred. The absence of masking in the registry entry may be relevant to interpreting a performance-based test, but the record does not describe assessment procedures well enough to evaluate that possibility.

Only unadjusted outcome comparisons are posted, and no reported analyses are listed. The registry record does not report a prespecified statistical plan, covariate adjustment, multiplicity strategy, or clinical-importance threshold for the primary measure. These omissions limit assessment of whether the reported comparisons match the original analysis plan and how the score differences should be judged clinically.

Finally, the primary outcome is reported at 3 months, and the registry record does not report longer-term results. The secondary information is limited to muscle perfusion measured by ultrasound scan; results for quality of life and general health are not reported in the registry record. The findings therefore have a narrow evidential scope: the posted measures, groups, and follow-up only.

19. Related Tutorials

20. Related Statistical Calculators

21. Sources

• ClinicalTrials.gov: Exercise Rehabilitation in Veterans With PAD — NCT02607046.
• Registry record last updated 2024-01-18; status last verified 2023-12.
• Results first posted 2021-02-15.
• 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.

22. Record Summary

The registry describes a completed randomized parallel-group study of exercise training versus neuromuscular electrical stimulation in veterans with PAD requiring revascularization. The listed enrollment is 56, and the study is recorded as unmasked. The primary endpoint is the Modified Physical Performance Test at 3 months, a mobility score from 0 to 36 for which higher scores are associated with better outcome.

The posted primary outcome means are 31.8 for Exercise and 32.3 for NMES. The unadjusted mean difference, NMES minus Exercise, is 0.5, with a confidence interval from -2.798 to 3.798 and an unadjusted p-value of 0.766. The registry also posts muscle perfusion measured by ultrasound scan as a secondary outcome. The outcome counts are smaller than the enrollment entry, and the registry record does not explain the difference or report the analysis population.

The record does not report the original sample-size assumptions, a detailed analysis plan, missing-data methods, or a clinical-importance threshold. The available findings are therefore best read as posted, unadjusted comparisons with substantial uncertainty, rather than definitive evidence about the comparative effects of the interventions.