EDUCATE: The MEDTRONIC Endeavor Drug Eluting Stenting: Understanding Care, Antiplatelet Agents and Thrombotic Events: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for EDUCATE: The MEDTRONIC Endeavor Drug Eluting Stenting: Understanding Care, Antiplatelet Agents and Thrombotic Events (NCT01069003), a randomized phase 4 study. The primary outcome is Percentage of Participants With Composite of All Death, Target Vessel Myocardial Infarction (MI) and Stroke (Defined as MACCE) for Randomized Subjects.

2. Clinical Question

The central randomized comparison concerns whether the reported composite outcome differed between subjects assigned active thienopyridine and aspirin and those assigned placebo thienopyridine and aspirin. Both groups had received a stent and met the stated eligibility conditions before randomization. The registry also reports the composite outcome for a larger, nonrandomized surveillance group. That group offers a descriptive outcome profile, but its nonrandomized status means its results should not be read as if it had been assigned treatment by the same process as the two randomized groups.

The clinical question is therefore not a single three-arm randomized comparison. It includes a randomized comparison of the two antiplatelet regimens, alongside observational follow-up of the surveillance group. The different outcome timeframes also matter when interpreting contrasts: randomized subjects are reported over 12–30 months, whereas surveillance subjects are reported over 0–24 months. The registry record does not report additional details about the reason for these distinct timeframes or about how the surveillance group was selected.

3. Trial Design

EDUCATE is listed as a completed Phase 4, prospective, multicenter study of real-world safety and clinical outcomes among subjects receiving one or more Endeavor zotarolimus-eluting stents. The registry reports randomized allocation, triple masking, and a parallel model for the trial. It records enrollment of 2272 and lists coronary artery disease as the condition. The study start is recorded as 2010-04, with primary completion at 2014-04.

The randomized comparison was between placebo thienopyridine plus aspirin and active thienopyridine plus aspirin, with both assignments planned for 18 months. Eligibility for these randomized groups required absence of the specified clinical events in the first 12 months before randomization. A third group, called the surveillance arm, was nonrandomized and followed through 24 months. The registry does not report further information about how subjects entered that group or how its follow-up related to the randomized groups.

The primary outcome is reported as a percentage of participants with a composite of all death, target-vessel myocardial infarction, and stroke. The listed timeframe is 12–30 months for the randomized placebo and thienopyridine groups, and 0–24 months for the surveillance arm. The posted outcome group sizes are 392 for placebo, 393 for thienopyridine therapy, and 1451 for surveillance. Those are the group sizes attached to the outcome, and they should not be assumed to represent the number enrolled in every analysis or safety assessment.

The registry does not report the trial’s original sample-size justification, detailed allocation procedures, the identities of people masked, the handling of treatment changes, or the prespecified statistical analysis plan. Although the study is described as designed to collect real-world outcomes, that description does not by itself establish how the registry-defined randomized and surveillance populations were assembled or analyzed.

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: Enrolled 2272 but only 2262 subjects were analysis receiving the Endeavor stent only.

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.

Population: The intent to treat arm contains non randomized subjects followed through 24 months

6. Primary Endpoint

The registry lists the primary outcome as Percentage of Participants With Composite of All Death, Target Vessel Myocardial Infarction (MI) and Stroke (Defined as MACCE) for Randomized Subjects, measured over Placebo and Thienopyridine 12 - 30 months; Surveillance Arm (0 - 24 months).

The primary endpoint is a composite: the percentage of participants experiencing all-cause death, target-vessel myocardial infarction, or stroke. It is reported for randomized subjects and for the surveillance arm, with different stated follow-up windows. Combining several clinically important outcomes into one endpoint can provide a single summary of whether any component occurred during the specified period. It can also increase the number of participants who experience the composite compared with considering a less frequent component alone. These are general properties of composite endpoints, not additional claims about the trial’s design or results.

A composite percentage does not show which component accounts for the reported events. The registry title and endpoint description identify the components, but a breakdown of the primary composite by death, target-vessel myocardial infarction, and stroke is not reported in the registry record. The posted outcome percentages are 4.3% for placebo, 4.8% for thienopyridine therapy, and 16.9% for surveillance. These percentages describe participants with the composite, not the severity or timing of individual events.

The different timeframes are important to the endpoint’s meaning. The randomized groups are reported for 12–30 months, while the surveillance group is reported for 0–24 months. A percentage over one follow-up window is not automatically directly comparable to a percentage over another. The registry record does not report further endpoint-adjudication details, component-specific results, or how incomplete follow-up was handled.

7. Results Posted in the Registry

• Thienopyridine Therapy: 4.8% (19/393) versus 4.3% (17/392) on control; risk difference 0.5 percentage points (95% CI -2.4 to 3.4), risk ratio 1.11 (0.59–2.11); unadjusted two-proportion z-test p = 0.739.
• Surveillance Arm: 16.9% (245/1451) versus 4.3% (17/392) on control; risk difference 12.5 percentage points (95% CI 9.8 to 15.3), risk ratio 3.89 (2.41–6.28); unadjusted two-proportion z-test p = <0.001.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary at the participant level: a participant either experienced the composite during the stated timeframe or did not. The registry reports percentages of participants with an event. For the randomized arms, the comparison uses the posted event counts of 17 among 392 placebo participants and 19 among 393 thienopyridine participants. The surveillance comparison uses 17 among 392 placebo participants and 245 among 1451 surveillance participants. These counts accompany the posted comparisons; they do not supply missing details about follow-up completeness or the analysis population.

What does a risk difference describe?

A risk difference expresses the difference between the two event proportions on an absolute scale. A value of zero represents equal proportions. Positive or negative values indicate which group has the larger observed proportion according to the comparison’s group order. For thienopyridine therapy versus placebo, the posted risk difference estimate is 0.005 and its confidence interval includes zero. For surveillance versus placebo, the estimate is 0.126 and its interval does not include zero. The registry record does not report an adjusted risk-difference analysis; the posted comparison is unadjusted.

How does a risk ratio differ?

A risk ratio expresses one group’s event proportion relative to the other group’s proportion. A value of one represents equal proportions. In the randomized comparison, the reported risk ratio is 1.115 and its confidence interval ranges from 0.588 to 2.113. In the surveillance comparison, it is 3.893, with a confidence interval from 2.412 to 6.284. Relative and absolute measures answer different questions: the risk ratio describes proportional contrast, while the risk difference describes the absolute separation between event proportions. Neither measure alone explains why the outcomes differ.

What does the p-value tell the reader?

A p-value describes how compatible the observed data are with a specified null hypothesis under the assumptions of the statistical test. It is not the probability that the null hypothesis is true, and it does not measure the clinical importance of a difference. The unadjusted p-value for thienopyridine therapy versus placebo is 0.739; for surveillance versus placebo it is <0.001. The registry record does not report test details beyond the posted comparisons, and it does not report an adjustment for multiple comparisons. The p-values should therefore be read as unadjusted results rather than as a complete account of the trial’s analysis strategy.

Why does randomization matter?

In general, random assignment can make groups more comparable at baseline, on average, with respect to measured and unmeasured factors. That supports a causal interpretation of a comparison when the trial’s conduct and analysis also preserve the randomized assignment. The registry identifies the placebo and thienopyridine arms as randomized. It identifies surveillance as nonrandomized, so its contrast with placebo can reflect differences in who entered each group, in addition to any treatment or care differences. The registry record does not report detailed analysis-population definitions or methods for accounting for group differences.

What does a confidence interval add?

A confidence interval gives a range of estimates compatible with the observed data under the method’s assumptions. A wide interval signals imprecision: the data allow a broader range of possible effects. For the randomized comparison, the risk-ratio interval includes values below and above one, and the risk-difference interval includes zero. Thus, the posted results do not sharply distinguish among possible directions and sizes of the relative contrast. For surveillance versus placebo, the risk-ratio interval lies above one, but the nonrandomized design and different timeframes still limit causal interpretation. Confidence intervals quantify sampling uncertainty, not all sources of bias or measurement error.

These are general explanations of common statistical measures. The original analysis plan, model specifications, handling of missing outcome data, and any prespecified covariate adjustments are not reported in the registry record. The posted results therefore support interpretation of the stated unadjusted comparisons, not reconstruction of undocumented analyses.

10. Confidence Intervals and Precision

Precision concerns how narrowly the data estimate an outcome contrast. A confidence interval is useful because a point estimate alone can conceal substantial uncertainty. For the randomized thienopyridine-versus-placebo comparison, the reported risk ratio is 1.115 and its confidence interval runs from 0.588 to 2.113. Since the interval includes one, the posted data are compatible with relative outcome proportions below, near, or above equality. The risk-difference confidence interval also includes zero. The registry’s interval therefore does not pin down a single direction of difference for this randomized comparison.

The surveillance-versus-placebo risk ratio is 3.893, with a confidence interval from 2.412 to 6.284. The interval lies above one, indicating that the posted relative comparison is not centered around equal proportions. Even so, the interval spans a range of possible relative contrasts, and it does not address nonrandomized group selection or the distinct follow-up windows. Precision within a statistical model does not remove design limitations.

The reported percentages are rounded to one decimal place, while the comparison estimates and intervals use their own displayed precision. These values are not interchangeable: percentages describe each group, while the risk difference and risk ratio quantify a contrast. The registry does not report confidence intervals for the individual arm percentages or for the serious adverse-event counts. It also does not report component-specific confidence intervals for death, target-vessel myocardial infarction, or stroke.

In general, more observed information can narrow intervals, while variability, sparse outcomes, and incomplete data can limit precision. The registry record does not report enough information to assess all such influences, including missing-data handling or any adjustment strategy. Confidence intervals represent sampling uncertainty under the analysis assumptions; they do not encompass every possible source of error or bias.

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 based on the observed effect, not the trial’s original design calculation. For these binary-outcome comparisons, the entries refer to participants per arm, not total events. The 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. The original trial’s sample-size assumptions and justification are not reported in the registry record.

The 80% power column gives the illustrative number of participants per arm under those assumptions for a design that would have an 80% chance of detecting the specified effect if that effect were the true one. The 90% power column describes the corresponding planning target for a 90% chance. These are design calculations, not a measure of how much power the completed study actually had. They are not observed power.

Each row changes the assumed effect size relative to the observed comparison. Planning for 100% of the observed effect uses the full observed effect as the target; rows for 75% and 50% use smaller targets. Smaller effects generally require more participants to detect under otherwise unchanged assumptions. The estimates should be read separately for each comparison: the randomized thienopyridine-versus-placebo contrast and the placebo-versus-surveillance contrast have different observed effects and therefore different illustrative requirements.

The calculation assumes equal allocation, even though the posted group sizes are not equal in every comparison. It also uses a normal approximation and does not incorporate any unreported trial-specific design features. These figures illustrate how planning requirements depend on the target effect and desired power; they do not establish what the original investigators planned, nor do they replace the registry’s reported outcome results.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.043; treatment proportion p₁ = 0.048; average p̄ = 0.046.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.2958 = 0.5798.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.2958 = 0.2489.
4. n = (0.5798 + 0.2489)² / (0.0050)² = 27,707.8, rounded up to 27,708 participants per arm for 80% power.

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

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
p0 <- 0.0433673469387755                          # control proportion
p1 <- 0.0483460559796438                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

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

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(1069003)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 27,708 (80% power), 37,093 (90% power)
 75% of observed effect: n per arm = 48,623 (80% power), 65,092 (90% power)
 50% of observed effect: n per arm = 107,968 (80% power), 144,537 (90% power)
power.prop.test: 27707.8 
Simulated power at n = 27,708 per arm: 0.798 (20000 replicates)

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

13. Multiplicity, Missing Data and Analysis Populations

The registry record lists one primary outcome and three outcome measures overall, but no secondary outcomes are posted. It lists no analyses beyond the posted comparisons. The record does not report a prespecified statistical analysis plan, a hierarchy of endpoints, or a procedure for controlling multiplicity. The posted p-values are unadjusted. When several outcomes or contrasts are examined, the chance of at least one apparently notable result can increase; in general, multiplicity procedures can address that issue. Whether any such procedures were used for this study is not reported in the registry record.

Missing-data methods are not reported in the registry record. In particular, the record does not explain whether all randomized participants were included in the primary analysis, how participants with incomplete follow-up were handled, or whether imputation or sensitivity analyses were used. It also does not report definitions such as intention-to-treat, per-protocol, or as-treated populations. The outcome group sizes are reported as 392, 393, and 1451, while the safety reporting uses different numbers at risk. The reason for those differences is not stated, so the populations should not be assumed interchangeable.

The registry does not report interim analyses, stopping rules, covariate adjustments, subgroup analysis plans, or procedures for adjudicating endpoints. These omissions do not establish that such methods were absent from the study; they mean they are not reported in the registry record. Interpretation should therefore stay close to the posted unadjusted estimates and avoid attributing undocumented analysis choices to the investigators.

For the randomized contrast, random assignment is relevant to the interpretation of treatment groups, but the registry does not provide the detailed analysis-population information needed to assess whether the analysis preserved that assignment. For surveillance comparisons, nonrandomized group membership adds further potential for differences between groups. These considerations are separate from multiplicity and missingness, and the registry record does not report methods that would resolve them.

14. Safety Results

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

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 (6.7%), coronary artery disease (4.4%), acute myocardial infarction (3.8%), gastrointestinal haemorrhage (3.0%), angina pectoris (2.7%).

The safety record reports serious adverse events among 399 participants at risk in the placebo group, with 117 affected; among 412 at risk in the thienopyridine group, with 111 affected; and among 1451 at risk in the surveillance group, with 664 affected. Other adverse events are reported among the same respective numbers at risk, with 0 affected in placebo, 0 in thienopyridine therapy, and 90 in surveillance. These are counts of participants reported as affected, not a comparison adjusted for group differences or follow-up.

The listed serious adverse-event terms and percentages are Chest Pain, 6.7%; Coronary Artery Disease, 4.4%; Acute Myocardial Infarction, 3.8%; Gastrointestinal Haemorrhage, 3.0%; Angina Pectoris, 2.7%; and Cardiac Failure Congestive, 2.7%. The registry record does not specify in this listing which group each term percentage represents, so these terms should not be assigned to an individual arm. The death fields have no reported values.

The safety counts use different at-risk denominators from the primary outcome groups. The record does not explain the difference, and it does not report comparative uncertainty estimates or a detailed safety-analysis method. Serious adverse events are distinct from the primary composite endpoint and should not be treated as its component breakdown.

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

The randomized results describe the observed frequency of a composite outcome in two groups assigned different thienopyridine regimens alongside aspirin. The posted percentages are 4.3% for placebo plus aspirin and 4.8% for active thienopyridine plus aspirin. The unadjusted p-value is 0.739, and the confidence intervals for the reported comparative measures include the values corresponding to no difference. Statistically, this means the posted data do not estimate the randomized contrast with enough precision to rule out a range of possible differences. It does not establish that the regimens produce identical outcomes, and it does not demonstrate that one is better.

The surveillance group’s reported outcome percentage is 16.9%. Its contrast with placebo is statistically distinct in the posted unadjusted comparison, but surveillance was nonrandomized and had a different stated timeframe. The contrast may reflect differences in the groups, follow-up, or other factors that are not described in the registry record. It should not be interpreted as a causal effect of surveillance or as a randomized comparison of treatment strategies.

The composite combines all-cause death, target-vessel myocardial infarction, and stroke. The total percentage does not reveal how often each component occurred, and a difference in a composite could be influenced by one or more components. Component-specific results are not reported in the registry record. The safety summaries provide separate affected-participant counts and serious adverse-event terms, but do not supply a complete account of the primary endpoint components.

Overall, the posted figures describe outcome frequencies and unadjusted contrasts for the stated populations and follow-up periods. They do not settle the clinical interpretation of either regimen, identify the causes of the observed surveillance contrast, or answer questions beyond the endpoint as defined. The registry record does not report the additional analysis details needed to make broader claims about missing observations, adjusted comparisons, or component-specific outcomes.

16. Why This Trial Matters Statistically

This trial is statistically interesting because it places a randomized comparison alongside a nonrandomized surveillance group. Those designs answer different kinds of questions. Randomization supports interpretation of the placebo and active-thienopyridine contrast, subject to the study’s conduct and analysis. The surveillance contrast is descriptive because group membership was not randomized. Keeping the two comparisons distinct is essential: a large numerical separation in a nonrandomized comparison does not have the same causal meaning as a randomized treatment contrast.

The outcome is also a composite, so the reported percentage compresses several events into one measure. This can make a single summary easier to interpret, but it can conceal how the components contribute. Component-specific outcome results are not reported in the registry record. The differing follow-up windows add another interpretive issue: randomized participants are reported over 12–30 months, while surveillance participants are reported over 0–24 months. A reader must consider both the composition of the groups and the time period captured by the outcome.

The randomized estimates demonstrate why effect estimates and confidence intervals matter alongside p-values. The reported p-value of 0.739 is not proof of equivalence, and the confidence intervals include a range of possible effects. The surveillance comparison has an unadjusted p-value of <0.001, yet its design and timeframe limit what can be concluded from that statistical result. Statistical evidence and design-based interpretation must be considered together.

Finally, the registry does not report the original sample-size rationale, detailed missing-data procedures, or a full analysis plan. Those absences restrict assessment of how closely the posted analysis follows the intended design. The planning illustrations show how required sample size depends on the effect selected for detection and desired power, but they do not reveal the study’s original planning assumptions.

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.
• Participant counts were reconstructed from posted percentages and arm sizes, so small rounding differences are possible.

Several trial-specific limitations shape interpretation of the posted results. The randomized endpoint window is listed as 12–30 months, while the surveillance window is 0–24 months. These intervals are not the same, so the surveillance and randomized percentages do not describe identical periods of observation. The registry record does not explain the rationale for the different windows or provide a time-to-event analysis that would make the periods directly comparable.

The surveillance group was nonrandomized. The registry record does not report how participants entered that group, what baseline characteristics differed, or whether any adjusted analysis was used. Consequently, its contrast with placebo may reflect differences in group composition as well as differences in care or outcomes. It should not be given the causal interpretation of a randomized contrast.

The randomized population was restricted to subjects without specified events during the first 12 months before randomization. The results therefore concern this eligible population, not all people receiving the stent described in the study. The registry does not report details about treatment adherence, crossover, or the effects of changes in therapy after randomization.

The primary endpoint is composite, but component-specific results are not reported. Missing-data handling, analysis-population definitions, adjustment procedures, interim analyses, and the original sample-size justification are also not reported in the registry record. Safety denominators differ from the primary outcome group sizes, and the reason is not given. These reporting gaps limit independent assessment of the analyses and prevent the safety counts from being treated as directly interchangeable with the primary outcome results.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: EDUCATE: The MEDTRONIC Endeavor Drug Eluting Stenting: Understanding Care, Antiplatelet Agents and Thrombotic Events — NCT01069003.
• Registry record last updated 2016-01-20; status last verified 2015-12.
• Results first posted 2016-01-20.
• 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

EDUCATE (NCT01069003) is a completed Phase 4, prospective, multicenter study listed as randomized, triple-masked, and parallel. The registry records enrollment of 2272 and a study period from 2010-04 to primary completion at 2014-04. The study concerns subjects receiving one or more Endeavor zotarolimus-eluting stents and antiplatelet therapy.

The primary outcome is the percentage with all-cause death, target-vessel myocardial infarction, or stroke. Reported percentages are 4.3% for placebo plus aspirin, 4.8% for active thienopyridine plus aspirin, and 16.9% for the nonrandomized surveillance group. The randomized comparison has an unadjusted p-value of 0.739; the surveillance-versus-placebo comparison has an unadjusted p-value of <0.001. The surveillance contrast requires caution because assignment was nonrandomized and the endpoint windows differ.

The registry record does not report the original sample-size justification, detailed missing-data methods, or a full statistical analysis plan. Serious adverse-event counts and selected terms are posted separately from the primary endpoint.