A Study to Evaluate the Use of Mobile-phone Based Patient Support in Patients Diagnosed With Myocardial Infarction: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for A Study to Evaluate the Use of Mobile-phone Based Patient Support in Patients Diagnosed With Myocardial Infarction (NCT01874262), a randomized na study. The primary outcome is Non-adherence Score.

2. Clinical Question

The study description frames the question around whether mobile-phone-based support affects adherence and persistence with ticagrelor or lifestyle changes related to cardiovascular risk factors. The single posted primary outcome is the non-adherence score, not a direct measure of all those aims. It is based on missed-dose patterns and patient-reported treatment gaps, as defined in the endpoint description.

The reported group means were 16.6 for e-diary plus support and 22.8 for e-diary alone. The registry also records a sponsor-reported Wilcoxon (Mann-Whitney) p-value of 0.025. A separate unadjusted comparison of means gives a mean difference of 6.2, with a confidence interval that includes values on either side of zero. Because the sponsor's analysis and the unadjusted re-calculation differ, the sponsor's own analysis takes precedence; the re-calculation is useful for describing the posted summary statistics, not as a replacement for the reported analysis.

3. Trial Design

SUPPORT (NCT01874262) was a randomized, parallel-group study with no masking reported. The registry lists its phase as NA and its status as completed. Enrollment was 174. The study began in 2013-06, and primary completion was recorded in 2014-12. AstraZeneca is listed as the sponsor.

The active group was assigned mobile-phone-based patient support combined with an e-diary. The control group had access to the e-diary alone and recorded daily ticagrelor use without receiving feedback. The active-group description says that its application contained both the diary and support functions, and that patients received feedback. The rest of that description is truncated in the registry record, so the exact content and delivery of the feedback are not fully reported.

The primary outcome was the non-adherence score at 6 months. The registry lists one outcome measure and one analysis. It does not report whether the analysis population included all randomized participants, only participants with available outcome data, or another defined group. Details such as the randomization procedure, allocation concealment, planned sample-size justification, and the trial's original statistical analysis plan are not reported in the registry record.

The trial design therefore supports a comparison between two assigned approaches to recording and supporting treatment use, but the available record does not answer every design question. In particular, the lack of masking is recorded, while any steps taken to limit bias in outcome recording or analysis are not reported in the registry record. These gaps matter when interpreting a patient-reported, behavior-related endpoint.

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: Before discharge from the hospital patients diagnosed with myocardial infarction and treated with ticagrelor were offered participation in the study and those that accepted and signed informed consent were randomized into one of the two study groups.

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 Non-adherence Score, measured over 6 months.

The primary endpoint was the non-adherence score at 6 months, reported as a mean with a standard deviation. It combined adherence failure events and treatment gaps. An adherence failure event was defined as two missed doses during an observation cycle of up to 7 days. The first missed dose recorded in the e-diary started a one-week observation cycle. A second missed dose during that week counted as a failure event; a third missed dose started a new cycle. If the second missed dose occurred after more than one week, it did not count as a failure event and instead started a new cycle. A treatment gap was a patient-reported gap of four consecutive doses.

Combining these features into a score provides a way to represent more than one pattern of non-adherence in a single endpoint. It can capture repeated missed doses within a defined observation window as well as longer consecutive gaps. The registry does not report the scoring formula, weighting of the two components, possible score range, or how incomplete diary entries were handled. Without those details, the clinical interpretation of a particular score value and the contribution of each component cannot be fully assessed. The reported means describe the score as analyzed, rather than showing separate results for failure events and treatment gaps.

7. Results Posted in the Registry

• E-diary: mean 22.80 versus 16.60 on control; difference 6.20 (95% CI -6.77 to 19.17), standardized effect d = 0.15; z-test p = 0.349.

8. Statistical Interpretation

9. Statistical Methods Explained

The registry reports a Wilcoxon (Mann-Whitney) analysis for the two groups and gives p = 0.025. This is a rank-based method for comparing the distributions of observations in two independent groups. Rather than requiring the comparison to be expressed only as a difference between arithmetic means, the method works with the ordering of observations. Its interpretation is not automatically a test of a difference in means; a location-shift interpretation requires additional conditions about the shape of the distributions. The registry does not report the test statistic, the exact alternative hypothesis, or further details of the analysis.

Why is there also a mean-based comparison?

The posted results include a mean and standard deviation for each group. Those summaries allow an unadjusted re-calculation of the difference in means and an approximate confidence interval. This is a separate analysis from the sponsor's rank-based test. It does not reproduce the sponsor analysis, and its p-value should not be substituted for the sponsor-reported p-value. The re-calculation estimates the mean difference as 6.2, with a 95% confidence interval from -6.772 to 19.17 and an unadjusted p-value of 0.349.

The comparison uses the group summaries as posted: the support-plus-diary group has a mean of 16.6 and standard deviation of 42.9, while the e-diary group has a mean of 22.8 and standard deviation of 41.3. The recalculation uses a pooled standard deviation of 42.15 and standard errors of 4.653 and 4.707 for the group means. These are summary-statistic-based quantities, not adjusted estimates. No covariate adjustment is reported in the registry record.

What does the confidence interval represent?

A confidence interval describes the range of values reasonably compatible with the data and the assumptions of the method. In repeated use of the same procedure under its assumptions, the stated confidence level refers to the long-run coverage of the interval procedure. It is not a probability statement that a particular interval contains the true effect. Here, the recalculated interval includes zero and also includes values in either direction. Its width indicates that the summary-statistic re-calculation does not pin down a narrow range for the mean difference.

Why might the methods give different results?

The Wilcoxon (Mann-Whitney) method and a comparison of arithmetic means answer different statistical questions and use the data differently. A rank-based result depends on the relative ordering of individual observations, while the re-calculation of means uses group averages and variability. The registry does not report individual observations, distribution plots, the rank statistic, or an explanation of the discrepancy, so the source of the difference cannot be identified from the record. The record explicitly indicates a discrepancy; the sponsor's own analysis takes precedence.

Standard deviations of 42.9 and 41.3 are large in relation to the reported means. General statistical context: substantial variability can make estimates less precise, and unusual distributional shapes or extreme observations can affect means and standard deviations. Over-dispersion, in general, means that variability is greater than a model would expect; it can make standard errors and tests unreliable if the model does not account for that variability. The registry does not establish that over-dispersion occurred in this trial, and no claim about the outcome distribution can be made from the summary statistics alone.

The outcome is a composite score, but the registry does not report its scale range, component weights, or the analysis model used beyond naming the rank-based test. It also does not report whether the analysis was prespecified, whether there were interim analyses, how outliers were handled, or whether any sensitivity analyses were conducted. These details cannot be inferred from the reported p-value.

What can a p-value tell the reader?

A p-value is calculated under a null hypothesis and the assumptions of a statistical test. Smaller p-values indicate that data at least as incompatible with that null model would be less common under those assumptions. They do not measure the probability that chance alone produced the result, the probability that the intervention is effective, or the magnitude of an effect. The sponsor-reported p-value of 0.025 belongs to its rank-based analysis. The recalculated p-value of 0.349 belongs to the unadjusted mean comparison. Their difference is a reason to keep the analyses distinct, not to select whichever value seems preferable.

10. Confidence Intervals and Precision

The mean-based re-calculation estimates a difference of 6.2 between the e-diary-only mean and the support-plus-diary mean. Its 95% confidence interval extends from -6.772 to 19.17. Because zero lies within this interval, the recalculation is compatible with no mean difference as well as with values favoring either direction. The interval is also broad relative to the point estimate, so the mean difference is not estimated with high precision by this summary-statistic analysis.

A confidence interval is not a range that assigns a probability to the true value after the data have been observed. In general, it is an interval generated by a procedure that would cover the target value at its stated rate over repeated studies, if the model and sampling assumptions hold. Its interpretation depends on the analysis being used. The interval applies to the unadjusted mean difference, not to the Wilcoxon (Mann-Whitney) result. The registry does not report a confidence interval for the sponsor's rank-based analysis.

The group standard deviations, 42.9 and 41.3, indicate considerable spread in scores relative to the reported means. The summary statistics do not show the shape of either distribution or how observations are arranged, so they cannot reveal whether the mean is influenced by a small number of extreme scores. Nor do they show the score's clinically important threshold, which is not reported in the registry record. Precision of the estimate and importance of a difference are separate questions: the record provides no threshold with which to judge the latter.

The confidence interval and p-value from the recalculation should be read alongside, not in place of, the sponsor's reported p-value of 0.025. The methods differ, and the registry marks a discrepancy. Accordingly, the sponsor's own analysis takes precedence for the reported trial analysis; the recalculated interval is a separate description of uncertainty around the unadjusted difference in means.

11. 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 participants or total events. This is a two-group continuous-outcome illustration based on the observed effect, rather than the original trial design. The assumptions listed for the illustration 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. The registry's original sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the number of participants per arm that the illustration estimates would be needed to have an 80% chance of detecting the specified effect under those assumptions, if the assumed effect and model were correct. The 90% power column gives the corresponding number for a 90% chance. These are design-planning quantities, not results observed in the trial. They are not observed power.

The rows consider planning for 100%, 75%, and 50% of the observed effect. As the effect specified for planning becomes smaller, more participants are needed to distinguish it from random variation at the chosen significance level and power. For example, planning for 75% of the observed effect requires 1,290 participants per arm for 80% power and 1,727 participants per arm for 90% power; planning for 50% requires 2,902 and 3,885 participants per arm, respectively. The 100% row gives 726 participants per arm for 80% power and 972 for 90% power.

These numbers are illustrations rather than a reconstruction of the SUPPORT design. They rely on an effect estimated from this trial and on the stated approximation and allocation assumptions. They do not establish what sample size would be appropriate for another study without considering its target effect, outcome distribution, design, and analysis plan. The planning curve should be read in the same way: it depicts the relationship between the assumed effect size and the planning sample size under the listed assumptions, rather than evidence about the trial's observed outcome.

12. 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 = 6.200; pooled standard deviation σ = 42.148.
2. n = 2 × (1.96 + 0.842)² × 42.148² / 6.200² = 725.4, rounded up to 726 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 <- 6.2                      # observed difference in means
sigma <- 42.1475740701645                     # 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(1874262)
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 = 726 (80% power), 972 (90% power)
 75% of observed effect: n per arm = 1,290 (80% power), 1,727 (90% power)
 50% of observed effect: n per arm = 2,902 (80% power), 3,885 (90% power)
power.t.test: 726.4 
Simulated power at n = 726 per arm: 0.796 (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.

13. Analyses Reported by the Sponsor

The registry reports the sponsor's analysis as a Wilcoxon (Mann-Whitney) comparison, with p = 0.025. The reported-analysis fields do not include an effect estimate or confidence interval for that test. The record separately provides the group means and standard deviations, from which an unadjusted mean-based re-calculation gives a difference of 6.2, a 95% confidence interval from -6.772 to 19.17, and p = 0.349.

These are not two versions of the same test: the sponsor analysis is rank-based, while the re-calculation compares arithmetic means using posted summary statistics. They can differ because they use different features of the data and address different statistical quantities. The registry record marks the difference as a discrepancy. In keeping with that record, the sponsor's own analysis takes precedence. The recalculation should be treated as supplementary context about the posted means, not as a correction or replacement for the sponsor's result.

The registry does not explain why the results differ, whether the analyses target the same estimand, or whether the sponsor analysis followed a prespecified plan. It does not report a confidence interval for the rank-based result. Those details are not reported in the registry record, so no further reconciliation can be made from the posted information.

14. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome measure and one analysis, and no secondary outcomes are posted. It does not report whether the trial assessed additional outcomes that were not entered in the registry, whether there were multiple comparisons within the primary analysis, or whether any adjustment for multiplicity was planned. The analysis plan and any interim analyses are not reported in the registry record. Without those details, it is not possible to assess how the sponsor-reported p-value relates to a broader family of planned tests.

Missing-data handling is also not reported in the registry record. The enrolled total is 174, while the primary outcome summaries list 85 participants in the support-plus-diary group and 77 in the e-diary group. The record does not explain the difference between these counts and overall enrollment. It does not state how many participants lacked a 6-month score, whether diary records were incomplete, how missing observations were treated, or whether any imputation or sensitivity analysis was used. These omissions limit assessment of whether the posted summaries could be affected by missingness.

The analysis population is not described. The registry does not say whether the analysis followed assignment regardless of adherence to the assigned approach, included only participants with observed scores, or used another population definition. Nor does it report exclusions after randomization or their reasons. Randomized allocation is recorded, but the available information is not enough to establish how allocation was carried through in the analysis.

These are reporting gaps, not proof that particular procedures were or were not used. The correct description is that multiplicity plans, missing-data methods, and analysis-population definitions are not reported in the registry record. They should not be reconstructed from the single posted p-value or the group summaries.

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

The posted results describe a comparison of the non-adherence score at 6 months between participants assigned to mobile-phone support plus an e-diary and those assigned to an e-diary alone. The group means were 16.6 and 22.8, respectively. The sponsor's reported rank-based analysis gives p = 0.025. Separately, the unadjusted re-calculation of means gives a difference of 6.2 and a confidence interval from -6.772 to 19.17. Since those analyses differ, the sponsor's own analysis takes precedence, while the mean-based result should be understood as a distinct supplementary calculation.

The findings do not establish the size or importance of a difference on the score beyond what the reported summaries describe. The registry does not provide a threshold for a meaningful score difference, the score range, or separate results for its adherence-failure and treatment-gap components. It also does not report the distribution of individual scores or a confidence interval for the sponsor's rank-based analysis.

The record does not support conclusions about all aspects of adherence, treatment persistence, lifestyle change, or cardiovascular risk factors. Although these aims are described in the study summary, only the non-adherence score is posted as a primary outcome, and no secondary outcomes are posted. The result should not be generalized to outcomes that were not reported.

Nor does the record establish how missing scores, incomplete diary reporting, or post-randomization exclusions were handled. The analysis population is not reported, and the outcome group counts do not match the enrolled total. These limits affect how fully the posted results can be interpreted. The statistics describe the registered endpoint and analyses; they do not, by themselves, provide medical advice or establish a treatment recommendation.

16. Why This Trial Matters Statistically

This trial raises a useful statistical issue: the result depends on which feature of the outcome is being compared. The sponsor reported a Wilcoxon (Mann-Whitney) analysis, while the posted means and standard deviations allow an unadjusted comparison of means. The sponsor's p-value is 0.025; the re-calculated mean comparison has p = 0.349 and a confidence interval spanning zero. Keeping these analyses distinct is essential. A rank-based p-value cannot be read as if it were the p-value for a difference in means.

The endpoint also combines two patterns of non-adherence: repeated missed doses within a defined observation cycle and patient-reported gaps of consecutive doses. Composite outcomes can make a complex behavior easier to summarize, but their interpretation depends on the scoring rule and the contributions of each component. The score formula, range, and component-specific results are not reported in the registry record. That makes it difficult to understand what a difference in the overall score represents in practical terms.

Finally, precision and reporting completeness matter alongside the p-value. The mean-based confidence interval is broad, and the record does not provide an interval for the sponsor's analysis. Missing-data handling, the analysis population, and the original design assumptions are not reported. These gaps make transparent distinction between reported evidence, recalculation, and unknown details especially important for clinical trial interpretation.

17. 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.

Several trial-specific limitations arise from what the registry record reports and omits. The active-group description is incomplete, so the content and delivery of the mobile-phone feedback cannot be fully characterized. The record says that the control group had an e-diary without feedback, but does not detail other differences in contact or support between groups. Masking is listed as none, and the record does not describe procedures to reduce potential influence on diary reporting or outcome assessment.

The primary endpoint is a composite score, but its formula, possible range, weighting, and component-level results are not reported in the registry record. The group summaries do not reveal the individual score distributions. The analysis population and handling of missing diary entries or missing 6-month outcomes are also not reported. Outcome group counts total fewer participants than the overall enrollment, but the reason is not given.

The registry reports a sponsor rank-based p-value without an effect estimate or confidence interval for that analysis. The separate mean-based re-calculation produces a different p-value and an interval spanning zero; the registry marks a discrepancy, and the sponsor's own analysis takes precedence. The reason for the discrepancy is not reported. No secondary outcomes are posted, and details of the original sample-size justification, multiplicity plan, interim analyses, and analysis plan are not reported in the registry record. These limits constrain interpretation without establishing that any particular unreported method was absent.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: A Study to Evaluate the Use of Mobile-phone Based Patient Support in Patients Diagnosed With Myocardial Infarction — NCT01874262.
• Registry record last updated 2016-02-22; status last verified 2016-01.
• Results first posted 2016-02-22.
• 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.

21. Record Summary

NCT01874262, SUPPORT, was a completed randomized, unmasked, parallel-group study sponsored by AstraZeneca. The registry lists enrollment of 174 and compares mobile-phone-based patient support plus an e-diary with an e-diary alone among patients diagnosed with myocardial infarction. The primary endpoint was the non-adherence score at 6 months.

Group means were 16.6 for support plus diary and 22.8 for diary alone. The sponsor's reported Wilcoxon (Mann-Whitney) analysis gave p = 0.025. An unadjusted re-calculation of the mean difference gave 6.2, with a 95% confidence interval from -6.772 to 19.17 and p = 0.349. The registry identifies a discrepancy between these analyses, so the sponsor's own analysis takes precedence. Details on missing data, analysis population, original sample-size assumptions, and secondary outcomes are not reported in the registry record.