HMO Research Network CERT: Acute Myocardial Infarction: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for HMO Research Network CERT: Acute Myocardial Infarction (NCT00211172), a randomized na study. The primary outcome is Adjusted Mean Monthly Percent of Days Covered With B-blocker Following Intervention Date.

2. Clinical Question

The study focused on a behavior following AMI: filling prescriptions for beta-blockers. Its primary outcome was not a clinical event or a direct measure of medication-taking behavior. Rather, it used prescription records to estimate the proportion of days covered in each month. This approach gives a repeated measure of prescription coverage across the outcome period, but coverage is not identical to confirmed use of medication.

The registry describes a randomized, parallel-group comparison. It reports an adjusted mean of 70.62 for the Reminder Mailing group and 66.37 for Usual Care, with a mean difference of 4.25. The confidence interval for the difference extends from -0.249 to 8.749, and the unadjusted p-value is 0.064. These results describe the posted measure and its statistical uncertainty; they do not establish what happened to individual participants or whether any clinical outcome changed.

3. Trial Design

The registry describes a randomized, single-masked, parallel-group study with 836 participants. Its phase is recorded as NA, and its status is completed. The listed sponsor is Kaiser Permanente. The study start is recorded as 2003-10 and primary completion as 2005-12. The registry title is “HMO Research Network CERT: Acute Myocardial Infarction.”

Participants were assigned to Reminder Mailing or Usual Care. The intervention is classified as behavioral. The arm description says participants in the mailing group received two mailings about the importance of beta-blocker use, while the comparison group received usual care. The registry does not state who was masked, or what procedures were used to maintain masking, so the term single-masked cannot be made more specific from this record.

The study summary describes focus groups with people who had previously experienced AMI to identify issues and information to include in patient materials, as well as ways to present that information. It also says the materials were pre-tested with AMI patients for clarity, readability, and understanding of key messages. These descriptions give context for the intervention but do not specify the exact wording, delivery dates, or whether all participants received the mailings as intended.

The registry lists 410 participants in Usual Care and 426 in Reminder Mailing. It posts one primary outcome measure and no secondary outcomes. No analyses are listed in the record. The randomization procedure, allocation concealment, detailed eligibility rules, recruitment setting, and analysis population are not reported in the registry record. The original sample-size rationale and the assumptions used in the trial's design are also not reported.

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

Reasons for not completing this period:

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 Adjusted Mean Monthly Percent of Days Covered With B-blocker Following Intervention Date, measured over 9 months.

The primary endpoint is titled “Adjusted Mean Monthly Percent of Days Covered With B-blocker Following Intervention Date.” It measures prescription coverage after the intervention, using pharmacy records. For each month, the measure was constructed from prescription quantity dispensed and days posted to estimate the proportion of days covered. The registry gives a timeframe of 9 months and describes the outcome as a measure of adherence to beta-blocker therapy after the intervention.

This kind of endpoint is useful when a study aims to assess medication adherence over time using dispensing information. It can represent whether a person had medication available according to recorded prescriptions, and monthly summaries can describe patterns across the follow-up period. It does not directly establish whether a participant took each dose, and the registry does not describe how periods without a prescription record were handled.

The reported endpoint is an adjusted mean, and the registry gives group means of 66.37 for Usual Care and 70.62 for Reminder Mailing. The reported mean difference is 4.25. Because the measure is a monthly percentage of days covered, the difference is interpreted in percentage-point units of coverage, not as a relative percentage increase. The adjustment variables and the method used to form the adjusted means are not reported in the registry record.

7. Results Posted in the Registry

• Reminder Mailing: mean 70.62 versus 66.37 on control; difference 4.25 (95% CI -0.25 to 8.75), standardized effect d = 0.13; z-test p = 0.064.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is a continuous measure: an adjusted mean monthly percentage of days covered with a beta-blocker. A mean describes the average value in a group. For this endpoint, the average is based on monthly prescription coverage, constructed from pharmacy records using the quantity dispensed and days posted. The registry reports a standard deviation for each group as a description of variability. It does not report the adjustment variables or the procedure used to estimate the adjusted means.

What does a mean difference represent?

A mean difference compares the average outcome in one group with the average in the other. The posted estimate is 4.25 for Reminder Mailing relative to Usual Care. Since the outcome is expressed as a percentage of days covered, this is a difference in percentage points on the coverage measure. It is not a relative percent change. A positive estimate indicates a higher mean in the mailing group for the recorded outcome, while the confidence interval describes uncertainty around that estimated contrast.

The comparison includes group means, standard deviations, group sizes, a mean-difference estimate, a confidence interval, and an unadjusted p-value. These summaries provide a view of the difference and its precision. They do not reveal individual participant patterns, the distribution of the outcome, or how the adjustment was performed. The registry does not list the analysis model, covariates, or analysis population.

What does a confidence interval add?

A confidence interval gives a range of estimates compatible with the observed data under the statistical procedure and its assumptions. In repeated use of a method with its assumptions satisfied, intervals constructed in the same way would cover the underlying value at the method's stated rate. It is not a range containing a known proportion of individual outcomes, nor is it a direct probability statement about the particular true value. For this comparison, the interval extends from -0.249 to 8.749 and includes zero, indicating uncertainty about the direction and size of the difference.

How should the p-value be read?

The unadjusted p-value is 0.064. In general, a p-value quantifies how unusual data at least as incompatible with a specified null model would be if that model and the analysis assumptions were correct. It does not report the probability that the null model is true, and it does not measure practical importance. The registry does not report a multiplicity adjustment or a more complete account of the testing procedure, so the p-value should be understood as the unadjusted value posted for the comparison.

Why does variability matter?

Standard deviations describe variation among participant-level observations around a group mean. The posted standard deviations are 34.25 for Usual Care and 32.03 for Reminder Mailing. When outcomes vary substantially, a given difference in means can be harder to estimate precisely than when outcomes are less variable, all else being equal. The relationship also depends on sample size and the analysis method. This general principle does not identify the cause of variability in this study; the record does not report distributional details or sources of variation.

General statistical explanation: a standard error describes the expected sampling variability of an estimate under a statistical model. The record lists standard errors of 1.691 for Usual Care and 1.552 for Reminder Mailing, but does not describe how these were incorporated into the adjusted comparison. An estimate's standard error is not the same as the standard deviation of participant outcomes. Standard errors typically become smaller as information increases, while standard deviations describe the spread of observed outcomes.

What does randomization contribute?

In general, random assignment is intended to make treatment groups comparable in expectation with respect to measured and unmeasured characteristics. It supports causal interpretation when assignment and follow-up are conducted appropriately and the analysis respects the design. The registry identifies the study as randomized, but does not report the sequence-generation method, concealment procedures, adherence to assigned intervention, or losses to follow-up. Therefore, the design label alone cannot answer how these aspects were implemented.

The record describes a single-masked design but does not identify who was masked. Masking can reduce some forms of bias when participants, staff, or outcome assessors do not know group assignment, depending on who is masked and what outcomes are assessed. Without further detail, the role and effectiveness of masking cannot be evaluated from the registry record.

What can be learned from the posted analysis?

The reported group estimates and comparison describe the primary measure over the stated 9-month timeframe. There are no analyses listed in the registry record. The record also does not provide a model specification, missing-data strategy, prespecified covariates, or details of any sensitivity analyses. These omissions limit the ability to reproduce or fully evaluate the analysis. They do not invalidate the posted values, but they mean that interpretation must remain tied to the information documented in the registry.

General explanation: adjustment can account for selected variables in a statistical model and may change estimated group means or their uncertainty. The meaning of an adjusted estimate depends on which variables were included and how the model was specified. Since those details are not reported, the posted adjusted means should not be attributed to particular covariates or treated as a fully reproducible model result.

10. Confidence Intervals and Precision

The confidence interval for the mean difference extends from -0.249 to 8.749. Its width indicates that the estimate is not pinpointed at a single value: the data are compatible with differences spanning from a value just below zero to appreciably positive values on the days-covered scale. Because the interval crosses zero, it does not rule out no difference under the method used to form it. The record does not report the confidence level or the model details used for this interval.

The point estimate is 4.25, but the interval communicates more than the point estimate alone. It shows that uncertainty remains about both the direction and magnitude of the difference. The lower bound is close to zero, while the upper bound allows for a more substantial positive difference. That range should not be collapsed into a categorical conclusion based only on the point estimate or the p-value.

Precision is distinct from variability among participants. The group standard deviations, 34.25 and 32.03, describe dispersion in the outcome, while the confidence interval addresses uncertainty in the estimated group contrast. A confidence interval can be narrow or broad depending on the amount of information and the analysis, even when participant-level outcomes vary. The registry does not provide enough detail to assess how choices about adjustment, missing records, or analysis population affected precision.

The unadjusted p-value is 0.064. It is consistent with the interval's inclusion of zero, but the interval is often more informative because it displays a range of plausible effect estimates rather than a single threshold comparison. Neither statistic indicates the probability that a treatment effect exists. The registry does not report a separate multiplicity-adjusted interval or p-value.

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 sample-size figures are planning illustrations stated as participants per arm, not total event counts. They are based on the observed effect, rather than the trial's original design assumptions. The listed 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 the stated fraction of the observed effect. The original trial's own sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the planned participants per arm for a design with an 80% chance of detecting the specified effect under the planning assumptions, if that effect and assumptions were correct. The 90% power column gives the corresponding per-arm planning count for a higher chance of detection under those same assumptions. These are design calculations, not descriptions of what the completed study achieved.

The rows consider planning for 100%, 75%, and 50% of the observed effect. As the target effect becomes smaller, more participants are needed to distinguish it from random variation under otherwise similar assumptions. For example, the per-arm figures for 80% power are 955 when planning for 100% of the observed effect, 1,697 when planning for 75%, and 3,818 when planning for 50%. The corresponding 90% power figures are 1,278, 2,272, and 5,111. These illustrations are not observed power.

The planning figures use the posted observed effect as their reference, so they should not be mistaken for an independent estimate of a clinically important target. They also do not establish what sample size the investigators planned before the trial. The registry's original design rationale is not reported.

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 = 4.250; pooled standard deviation σ = 33.137.
2. n = 2 × (1.96 + 0.842)² × 33.137² / 4.250² = 954.3, rounded up to 955 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 <- 4.25                      # observed difference in means
sigma <- 33.1372942435585                     # 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(211172)
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 = 955 (80% power), 1,278 (90% power)
 75% of observed effect: n per arm = 1,697 (80% power), 2,272 (90% power)
 50% of observed effect: n per arm = 3,818 (80% power), 5,111 (90% power)
power.t.test: 955.3 
Simulated power at n = 955 per arm: 0.797 (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. Multiplicity, Missing Data and Analysis Populations

The registry lists one primary outcome measure and no secondary outcomes. It lists no analyses. The record does not state whether multiple hypotheses or time points were considered during the study, whether any multiplicity adjustment was planned, or whether there were interim analyses. These matters are not reported in the registry record and should not be inferred from the single posted primary endpoint.

The comparison's p-value is identified as unadjusted. No adjusted p-value is listed, and there is no reported description of a family of tests or a procedure for controlling the chance of false-positive findings across tests. With no secondary outcomes posted, the registry record does not permit an assessment of a broader set of outcome comparisons. The absence of listed analyses should not be taken as evidence that no other analyses were conducted.

Missing-data handling is not reported. The registry does not describe whether pharmacy records were complete for all participants throughout the 9-month timeframe, how gaps in records were treated, or whether participants with incomplete follow-up contributed to the endpoint. It also does not report attrition by group or the reason for any missing observations. Those details matter because missing outcome information can affect both estimated group means and the uncertainty around a difference.

The record does not identify the analysis population, such as whether participants were analyzed according to randomized assignment or according to another rule. It does not state how protocol deviations, nonreceipt of mailings, or changes in care were handled. The listed group counts are not enough to determine which participants contributed data at each month or to establish that the endpoint was available for everyone.

These reporting gaps limit appraisal of robustness and reproducibility. They do not justify assuming a particular missing-data method, multiplicity strategy, interim-monitoring plan, or analysis population. The posted result is best interpreted as the registry's reported primary comparison, with these aspects of the statistical plan left undocumented.

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

The result describes a difference in adjusted mean monthly prescription coverage after the intervention date. Reminder Mailing has a reported mean of 70.62 and Usual Care a reported mean of 66.37, with a mean difference of 4.25. The confidence interval extends from -0.249 to 8.749, and the unadjusted p-value is 0.064. These statistics convey a positive point estimate alongside uncertainty that includes a value close to zero on the negative side.

This outcome is a pharmacy-record measure of days covered, constructed from dispensed quantity and days posted. It is relevant to prescription adherence, but it does not directly show whether people took the medication as prescribed. Nor does it measure a cardiovascular event, survival, symptoms, or another clinical outcome. The registry does not post secondary outcomes that could address such outcomes.

The study summary explains that the mailings were intended to increase adherence to prescribed beta-blockers following AMI. The posted outcome comparison can be read in relation to that aim, but it cannot confirm how participants received, understood, or responded to the messages. It also cannot establish whether observed coverage reflected the mailing itself or other aspects of care without fuller information about implementation and analysis.

A statistically uncertain estimate is not evidence that the groups were identical, just as a positive estimate is not proof of a meaningful effect. The interval leaves room for different possible sizes and directions of the contrast. The record does not state what difference would be considered clinically important, and it does not provide a clinical threshold for interpreting the measure.

Accordingly, the posted data support a cautious description of the measured comparison, not a broad claim about patient health or the effectiveness of mailings in other settings. The registry does not report enough information to assess generalizability, participant characteristics, missing data, or the full analysis plan. No medical advice follows from these results.

15. Why This Trial Matters Statistically

This study is statistically interesting because it evaluates a behavioral intervention using a continuous pharmacy-record measure rather than a clinical event. A monthly days-covered outcome can capture prescription coverage over time, but interpretation depends on understanding what the record represents and what it cannot confirm. Separating medication availability from medication-taking is important when reading adherence measures.

The comparison also illustrates why a point estimate should be read together with its uncertainty. The reported difference is 4.25, but the confidence interval extends from -0.249 to 8.749. The interval includes zero and allows a range of possible differences. The p-value of 0.064 alone does not convey that range or indicate whether a difference matters clinically.

The registry's planning illustration highlights the relationship between target effect and required sample size. Smaller target effects need more participants under the stated assumptions. Because these figures are based on the observed effect and are not the original design rationale, they are useful as planning illustrations rather than a record of the investigators' initial calculation or the trial's achieved power.

Finally, the record provides only limited detail on the statistical plan. The absence of listed analyses, secondary outcomes, missing-data methods, and analysis-population information shows why registry results require careful reading. An estimate can be described precisely while important questions about model specification and data handling remain unanswered.

16. 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 does not report participant eligibility criteria, baseline characteristics, or the setting in sufficient detail to assess how closely the enrolled group represents other people with AMI. Enrollment is listed as 836, with 410 in Usual Care and 426 in Reminder Mailing. The record does not explain whether those group counts represent all randomized participants, participants with outcome data, or another analysis set.

The intervention is described as two mailings about the importance of beta-blocker use, but the exact content, timing, delivery success, and participant engagement are not reported. Usual care is not described beyond its label. The record therefore does not show how much the groups' experiences differed apart from the stated intervention description.

The primary outcome depends on pharmacy records and a monthly proportion-of-days-covered construction. The record does not explain how incomplete pharmacy information, overlapping prescriptions, treatment changes, or gaps in recorded supply were handled. It also does not report the adjustment variables behind the adjusted means. These omissions make it difficult to evaluate the measurement and reproduce the analysis.

The registry lists no secondary outcomes or analyses, and does not report multiplicity handling, missing-data methods, or the analysis population. It does not identify who was masked despite classifying the design as single-masked. These limitations concern reporting and do not establish that the underlying procedures were absent; they are simply not documented in the registry record.

Finally, the posted endpoint is an adherence-related measure rather than a direct clinical outcome. The record does not provide evidence about health events or longer-term outcomes beyond the stated 9-month timeframe.

17. Related Tutorials

18. Related Statistical Calculators

19. Sources

• ClinicalTrials.gov: HMO Research Network CERT: Acute Myocardial Infarction — NCT00211172.
• Registry record last updated 2014-10-09; status last verified 2011-07.
• Results first posted 2014-10-09.
• 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.

20. Record Summary

The registry describes a completed, randomized, single-masked, parallel-group behavioral study of reminder mailings about beta-blocker use after AMI. Enrollment is 836; 410 participants are listed in Usual Care and 426 in Reminder Mailing. The primary outcome is adjusted mean monthly percentage of days covered over 9 months, derived from pharmacy prescription records.

The posted adjusted means are 66.37 for Usual Care and 70.62 for Reminder Mailing. The reported mean difference is 4.25, with a confidence interval from -0.249 to 8.749 and an unadjusted p-value of 0.064. The registry lists no secondary outcomes and no analyses. Details of the adjustment model, missing-data handling, analysis population, and original sample-size rationale are not reported in the registry record.