Outcomes Study to Assess STatin Residual Risk Reduction With EpaNova in HiGh CV Risk PatienTs With Hypertriglyceridemia: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Outcomes Study to Assess STatin Residual Risk Reduction With EpaNova in HiGh CV Risk PatienTs With Hypertriglyceridemia (NCT02104817), a randomized phase 3 study. The primary outcome is The Composite of Major Adverse Cardiovascular Events (MACE).

2. Clinical Question

3. Trial Design

STRENGTH was a phase 3, randomized, triple-masked, parallel-group study sponsored by AstraZeneca. The registry record describes a double-blind, placebo-controlled design in its brief summary, while its design fields classify masking as triple. It planned to compare Epanova plus statin with corn oil plus statin. Participants were assigned in a 1:1 ratio according to the brief study description. The posted primary-outcome arm sizes are 6539 in each group, within a total enrollment of 13078.

The intended population was people at high risk for cardiovascular disease with hypertriglyceridemia and low HDL. The registry condition describes eligible men or women considered at high risk for atherosclerotic cardiovascular disease. The detailed eligibility requirements, including specific laboratory thresholds and the distribution of baseline cardiovascular history, are not reported in the registry record.

The primary endpoint was a composite of cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, emergent or elective coronary revascularization, and hospitalization for unstable angina. The endpoint timeframe ran from randomization through the end-of-treatment visit at Month 60 or study closure. The brief study description says treatment was planned for approximately 3-5 years and that duration depended on reaching the number of MACE outcomes. Actual participant-level follow-up and the date on which the event target was reached are not reported in the registry record.

The status is completed. The record includes the primary result and several secondary outcome results. It does not report the original protocol’s detailed statistical analysis plan, the basis for the planned enrollment, or additional operational details of treatment adherence and follow-up.

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:

Analysis population for the primary outcome, as defined in the registry: Full Analysis Set

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 The Composite of Major Adverse Cardiovascular Events (MACE), measured over From the date of randomization and up to completion of the end-of-treatment visit (Month 60) or at study closure.

The primary endpoint counts participants with a composite cardiovascular event. The composite includes cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, emergent or elective coronary revascularization, and hospitalization for unstable angina. Combining events allows a study to assess a broader set of clinically relevant outcomes under one primary definition. It also means that the endpoint is not a measure of any single component alone.

The registry describes follow-up from randomization until completion of the end-of-treatment visit at Month 60 or study closure. Participants without an observed event were censored at the earliest of withdrawal of consent and last study contact. The record defines last study contact using the latest date of assessments that could contribute to determining whether a participant had experienced every component under analysis. This is relevant because the sponsor’s reported method is a Cox regression analysis, which uses the timing of events and censoring rather than treating all follow-up as identical.

The posted outcome counts are 785 participants in the Epanova group and 795 in the placebo group, out of 6539 in each group. Those counts describe the number recorded with an event, but do not by themselves account for when events occurred or how long participants were observed. The sponsor’s time-to-event estimate is therefore the primary inferential result to consult alongside the event counts. The registry record does not give separate primary-endpoint estimates for each component.

7. Results Posted in the Registry

• Epanova: 12.0% (785/6539) versus 12.2% (795/6539) on control; risk difference -0.2 percentage points (95% CI -1.3 to 1.0), risk ratio 0.99 (0.90–1.08); unadjusted two-proportion z-test p = 0.788.

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.

The Composite of CV Events

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Epanova: 8.3% (541/6539) versus 7.9% (517/6539) on control; risk difference 0.4 percentage points (95% CI -0.6 to 1.3), risk ratio 1.05 (0.93–1.17); unadjusted two-proportion z-test p = 0.442.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Epanova and the control. The interval gives the range of effects that remain compatible with the data.

The Composite of Coronary Events

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Epanova: 8.5% (556/6539) versus 9.4% (616/6539) on control; risk difference -0.9 percentage points (95% CI -1.9 to 0.1), risk ratio 0.90 (0.81–1.01); unadjusted two-proportion z-test p = 0.066.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Epanova and the control. The interval gives the range of effects that remain compatible with the data.

CV Death

• Endpoint type: Binary: the number of participants with the event, compared as proportions.
• Test: Unadjusted two-proportion z-test, with a Wald 95% confidence interval for the risk difference.
• Result:
  • Epanova: 3.5% (228/6539) versus 3.2% (211/6539) on control; risk difference 0.3 percentage points (95% CI -0.4 to 0.9), risk ratio 1.08 (0.90–1.30); unadjusted two-proportion z-test p = 0.409.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Epanova and the control. The interval gives the range of effects that remain compatible with the data.

The registry lists multiple secondary cardiovascular outcomes, including composites of cardiovascular events and coronary events, cardiovascular death, all-cause death, and outcomes in the subgroup with established cardiovascular disease at baseline. The record reports 9 secondary outcomes. The posted results shown for selected outcomes do not all point in the same numerical direction: some count comparisons are higher in the Epanova group and some are lower. Across the reported analyses, the pattern does not provide a consistent basis for overturning the primary result.

The secondary outcomes complement the primary composite by describing other cardiovascular groupings and specific outcomes. They may help characterize the broader set of results, but they are not substitutes for the prespecified primary question. In particular, a result for a secondary composite cannot establish a difference in the primary MACE outcome, and it does not identify which individual component accounts for any composite pattern.

The posted secondary analyses include unadjusted comparisons and sponsor-reported time-to-event analyses. The registry record does not report a multiplicity adjustment across the secondary outcomes or a hierarchy that would control the interpretation of multiple tests. When many outcomes are examined, the chance of seeing an apparently notable result by chance can increase. This general concern means isolated unadjusted secondary p-values carry less weight than a primary analysis specified in advance, especially when interpreted without a documented testing strategy.

For that reason, the secondary results are most useful as descriptive context. They can indicate where the observed estimates lie and which outcomes were recorded, but they should not be treated as a separate confirmatory conclusion from the primary comparison. The full secondary-outcome analysis plan and any prespecified order of testing are not reported in the registry record.

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

• The Composite of MACE in the Subgroup of Participants With Established CV Disease(CVD) at Baseline
• The Composite of CV Events
• The Composite of CV Events in the Subgroup of Participants With Established CV Disease (CVD) at Baseline
• The Composite of Coronary Events
• The Composite of Coronary Events in the Subgroup of Participants With Established CVD at Baseline
• CV Death
• CV Death in the Subgroup of Participants With Established CVD at Baseline
• All-cause Death

9. Statistical Interpretation

10. Statistical Methods Explained

The primary endpoint is binary at the participant level: each participant is counted as having experienced an event or not having an observed event during the specified follow-up. The posted arm counts show 785 of 6539 participants in the Epanova group and 795 of 6539 in the placebo group with an event. The unadjusted calculation compares the observed proportions. It does not adjust for baseline variables, incorporate event timing, or account for censoring through a survival model.

What does the risk difference summarize?

General statistical explanation: A risk difference subtracts the event proportion in one group from the event proportion in the other, using a stated direction of comparison. A value of zero corresponds to equal observed proportions. Its confidence interval describes uncertainty about the difference under the calculation’s assumptions. In this analysis, the reported estimate is -0.002, and its 95% confidence interval runs from -0.013 to 0.010. Because that interval includes zero, the posted unadjusted calculation does not distinguish the estimated difference from no difference at the conventional interpretation of its interval. The registry’s method description does not state that this was the prespecified primary inferential method.

Why is the sponsor’s analysis time-to-event?

The registry reports a Cox regression analysis and a hazard ratio for the primary endpoint. The endpoint definition includes time from randomization and specifies how participants without an observed event are censored. A time-to-event method can use information about when events occurred and can retain participants who have no observed event through their last relevant contact, rather than simply classifying everyone by a common fixed observation period. The sponsor reported a hazard ratio of 0.99, a 95% confidence interval from 0.90 to 1.09, and p = 0.837.

General statistical explanation: A hazard ratio compares event rates over time as represented by the fitted model. It is not a risk difference and should not be read as a direct subtraction of the final proportions. The interpretation of a Cox model depends on its assumptions and implementation. The registry record does not provide enough detail to assess the model diagnostics, covariate specification, or other implementation choices.

What does the p-value say?

General statistical explanation: A p-value describes how incompatible the observed data are with a specified null model, under the assumptions of the analysis. It is not the probability that the null hypothesis is true, nor does it measure the size or clinical importance of an effect. For the sponsor’s primary analysis, p = 0.837. For the unadjusted recalculation, p = 0.788. These values arise from different analyses, so they should be interpreted in their respective contexts rather than treated as competing versions of one identical calculation.

Why are both estimates shown?

The sponsor’s result is the registry’s reported analysis and reflects the stated Cox regression approach. The unadjusted recalculation offers a transparent comparison based on the participant counts. The two methods use different information and answer slightly different statistical questions. The record marks no discrepancy between the posted results and the recalculated summary, but that does not make the estimands identical.

The registry’s primary population is the Full Analysis Set. Beyond that label, its record does not specify the detailed rules for inclusion in that set, the exact handling of intercurrent treatment changes, or a fuller description of censoring implementation. The endpoint description does state that participants without observed events were censored at the earliest of withdrawal of consent and last study contact. The record also does not report proportional-hazards diagnostics or a detailed missing-data sensitivity analysis. Those omissions limit independent assessment of the analysis, rather than changing the estimates that are posted.

11. Confidence Intervals and Precision

Precision is best read from the confidence interval alongside the point estimate. The sponsor’s primary hazard ratio is 0.99, with a 95% confidence interval from 0.90 to 1.09. The interval includes 1, the value corresponding to equal hazards under this measure. It spans estimates on either side of 1, so the reported result does not pin down the direction of a relative difference with high certainty.

The unadjusted risk difference is -0.002, with a 95% confidence interval from -0.013 to 0.010. Zero lies inside this interval. The interval therefore includes no difference in observed proportions, while also allowing differences in either direction. It is important not to turn the interval into a claim of equivalence: an interval that includes no difference does not show that the groups are exactly the same. No equivalence margin is reported in the registry record.

General statistical explanation: A confidence interval is a range of estimates compatible with the data and the statistical procedure at a stated confidence level. It conveys uncertainty around an estimate, not the probability that a particular true value lies within this specific calculated interval. A narrower interval generally indicates greater precision, while a wider interval leaves more values compatible with the data.

The event counts are close, but counts alone are not a measure of precision and do not incorporate event timing. The sponsor’s interval belongs to the time-to-event estimate; the recalculated interval belongs to the unadjusted risk difference. Their widths cannot be compared as if they described precisely the same quantity. The registry record does not report a prespecified equivalence or noninferiority framework for this primary result.

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 sample-size display is a planning illustration based on the observed effect, not the study’s original design calculation. For this binary primary endpoint, its entries are participants per arm, not total events. The planning 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 assumed effect to detect is the stated fraction of the observed effect. The trial’s original sample-size justification and assumptions are not reported in the registry record.

Read the 80% power column as the number of participants per arm that the illustration estimates would be needed to detect the specified effect with an 80% chance under those assumptions, if that effect were the true effect. The 90% power column describes a higher chance of detecting that same assumed effect, and consequently gives a larger planning requirement. For the row using 100% of the observed effect, the display gives 712,946 participants per arm for 80% power and 954,432 participants per arm for 90% power. For 75% of the observed effect, it gives 1,269,190 and 1,699,084 participants per arm, respectively. For 50% of the observed effect, it gives 2,859,567 and 3,828,148 participants per arm, respectively.

Smaller effects are harder to distinguish from random variation, so the illustrated requirement grows as the assumed effect becomes smaller. These calculations are not observed power and do not describe the chance that this completed trial succeeded in finding an effect after its results were known. The associated curve should be read as a visual display of how the planning requirement changes with the assumed effect, not as new evidence about the treatment comparison.

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

Worked calculation for this trial

1. Control proportion p₀ = 0.122; treatment proportion p₁ = 0.120; average p̄ = 0.121.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.4609 = 0.9034.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.4609 = 0.3879.
4. n = (0.9034 + 0.3879)² / (-0.0015)² = 712,945.5, rounded up to 712,946 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 <- 0.121578222969873                          # control proportion
p1 <- 0.120048937146353                          # 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(2104817)
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 = 712,946 (80% power), 954,432 (90% power)
 75% of observed effect: n per arm = 1,269,190 (80% power), 1,699,084 (90% power)
 50% of observed effect: n per arm = 2,859,567 (80% power), 3,828,148 (90% power)
power.prop.test: 712945.5 
Simulated power at n = 712,946 per arm: 0.800 (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.

14. Analyses Reported by the Sponsor

The sponsor’s reported primary result is a Cox regression hazard ratio of 0.99, with a 95% confidence interval from 0.90 to 1.09 and p = 0.837. It is the registry’s time-to-event analysis and incorporates the endpoint’s event timing and censoring framework. The separate recalculation from the posted counts gives an unadjusted risk difference of -0.002, with a 95% confidence interval from -0.013 to 0.010 and p = 0.788.

These results are not direct duplicates. A hazard ratio compares event rates over time, while a risk difference compares observed event proportions. Their confidence intervals and p-values arise from different methods, and exact agreement is not expected. Both show estimates close to their respective no-difference values and intervals that include those values. The registry record indicates no discrepancy, and there is no basis to substitute the recalculated result for the sponsor’s analysis. The recalculation is best read as a supplementary, unadjusted description of the posted counts.

The sponsor’s detailed statistical analysis plan, including any model specifications beyond the recorded method, is not reported in the registry record. The reported Cox result should therefore be interpreted as the sponsor analysis documented there, without assuming additional adjustment or implementation details.

15. Multiplicity, Missing Data and Analysis Populations

The registry record does not report how multiplicity across secondary outcomes was handled. It lists 9 secondary outcomes and 14 analyses, but does not state whether p-values were adjusted, whether outcomes followed a prespecified hierarchy, or whether any analyses were intended to be confirmatory. Without that information, secondary p-values should be read as individual results rather than as a family-wise controlled set of tests. This matters because repeated testing can make a chance finding more likely across a collection of outcomes.

The primary analysis is reported as a Cox regression hazard ratio for superiority. The detailed protocol and statistical analysis plan are not reported in the registry record. In particular, the record does not describe planned interim analyses, any alpha-spending approach, covariate adjustment, or model diagnostics. These items cannot be inferred from the posted estimates.

For the endpoint, participants without observed events were censored at the earliest of withdrawal of consent and last study contact, with last contact defined in relation to assessments relevant to endpoint ascertainment. This provides a stated rule for some censoring, but the record does not give a broader account of missing-data sensitivity analyses, assumptions about censoring, or imputation. The number of participants with incomplete follow-up for each component is not reported in the registry record.

The primary outcome specifies the Full Analysis Set. The exact operational definition of that set, including handling of participants who did not follow assigned treatment or had missing outcome information, is not reported in the registry record. The record therefore identifies the analysis population by name but does not allow a full independent reconstruction of inclusion and exclusion decisions. These gaps do not invalidate the posted estimates; they define what can and cannot be assessed from the registry results alone.

16. Safety Results

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

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 1% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: atrial fibrillation (2.3%), angina unstable (2.2%), pneumonia (1.9%), angina pectoris (1.6%), cardiac failure (1.4%).

The adverse-event summary reports serious events affecting 2259 of 6532 participants at risk in the Epanova group and 2195 of 6535 participants at risk in the placebo group. Other adverse events are listed for 1405 of 6532 participants at risk in the Epanova group and 826 of 6535 participants at risk in the placebo group. The record also lists 373 deaths in the Epanova group and 333 in the placebo group.

These are descriptive counts and denominators as recorded; they are not estimates adjusted for follow-up time or baseline differences. The serious-event terms listed most often include atrial fibrillation, unstable angina, pneumonia, angina pectoris, cardiac failure, and coronary artery disease, with the percentages shown in the posted safety display. The record does not state whether each event was considered related to treatment, nor does it supply a comparative statistical analysis for these summaries. The death counts should not be interpreted as a causal treatment effect from these data alone.

Safety summaries are distinct from the primary MACE analysis. Some events may overlap with endpoint components, but the registry record does not provide an event-level linkage. The arm denominators for safety differ slightly from the primary endpoint denominators, so the counts should be read with their own stated populations rather than combined or compared as if all participants contributed to every summary in exactly the same way.

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

The primary analysis does not show a clear separation between Epanova plus statin and corn oil plus statin for the composite MACE outcome in this study. The sponsor’s hazard ratio is 0.99, and its 95% confidence interval from 0.90 to 1.09 includes 1. The unadjusted risk-difference interval also includes zero. These results describe uncertainty around the observed comparison; they do not demonstrate that the treatments are exactly equivalent.

The endpoint is a composite. It combines cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, coronary revascularization, and hospitalization for unstable angina. The primary result therefore does not, on its own, answer whether any individual component differed. The registry record does not report separate estimates for the individual components within the primary endpoint.

The findings apply to the population, comparison, follow-up, and analyses recorded for STRENGTH. The trial enrolled a high cardiovascular risk population with hypertriglyceridemia and low HDL according to the study description. Results cannot be extended automatically to people outside that described population or to different treatment regimens. The registry record also does not provide enough detail to assess every feature of eligibility, adherence, or follow-up.

Finally, a nonsignificant result is not a probability statement about whether treatment has an effect, and it is not proof of no effect. The confidence intervals show that estimates on either side of no difference remain compatible with the analyses. The appropriate reading is bounded: the posted results do not establish superiority on the primary endpoint, while retaining uncertainty about the range of effects consistent with the data. This is a statistical interpretation of the record, not medical advice.

18. Why This Trial Matters Statistically

This trial is statistically informative because it tested a defined composite cardiovascular outcome in a randomized, parallel-group comparison and posted both event counts and a time-to-event analysis. Random allocation is intended, in general, to support a fair comparison by distributing participant characteristics across groups. The registry record identifies randomized allocation, although it does not give enough detail to evaluate every aspect of implementation.

The result also illustrates why effect estimates and confidence intervals matter more than a binary reading of whether a p-value crosses a threshold. The sponsor’s hazard ratio is close to 1, and its interval includes 1. The unadjusted risk difference is similarly accompanied by an interval that includes zero. These summaries communicate both the central estimates and the remaining uncertainty, while avoiding a claim that a nonsignificant test proves equality.

The comparison highlights a methodological distinction: event counts and unadjusted proportions provide an accessible summary, while a Cox model uses event timing and censoring. Neither should be mistaken for the other. The sponsor’s result is the reported time-to-event analysis; the recalculation offers a complementary view from the posted counts. Keeping those estimands separate helps prevent overinterpretation.

The sample-size illustration is also useful as a planning lesson. It shows how planning requirements depend strongly on the size of the assumed effect, while not representing the original study rationale or observed power. Finally, the range of secondary outcomes emphasizes why multiplicity and prespecified analysis plans matter. The registry record’s limits on those details are part of the statistical context in which the results should be read.

19. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

The primary endpoint combines several different cardiovascular events. A composite can summarize a broad clinical outcome, but the registry results do not show how much each component contributed to the overall count. Without component-specific estimates, the primary comparison cannot establish whether the observed pattern was similar across cardiovascular death, myocardial infarction, stroke, revascularization, and unstable-angina hospitalization.

The registry’s primary analysis is a Cox regression result, but the record does not report model diagnostics, detailed covariate handling, or a full account of the censoring assumptions. The endpoint definition states how participants without an observed event were censored, but there is no posted sensitivity analysis to show how alternative missing-data or censoring assumptions might affect the estimate. Detailed participant flow and the exact operational rules for the Full Analysis Set are also not reported in the registry record.

The secondary outcome set includes multiple analyses, while the registry record does not report a multiplicity strategy or testing hierarchy. Consequently, isolated secondary findings should be treated cautiously and are not a substitute for the primary result. The detailed statistical analysis plan, interim-analysis approach, and original sample-size justification are not reported in the registry record.

The study description identifies high-risk participants with hypertriglyceridemia and low HDL, but the registry summary does not provide the full eligibility criteria or enough baseline detail to judge how closely the enrolled participants match other populations. The actual follow-up distribution, treatment adherence, and reasons for withdrawal are not reported. These limitations restrict what can be inferred from the registry record; they do not change the estimates it posts.

20. Related Tutorials

21. Related Statistical Calculators

22. Sources

• ClinicalTrials.gov: Outcomes Study to Assess STatin Residual Risk Reduction With EpaNova in HiGh CV Risk PatienTs With Hypertriglyceridemia — NCT02104817.
• Registry record last updated 2021-08-17; status last verified 2021-08.
• Results first posted 2021-07-28.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

23. Record Summary

STRENGTH was a completed phase 3 randomized, triple-masked, parallel-group study sponsored by AstraZeneca, with 13078 participants enrolled. It compared Epanova plus statin with corn oil plus statin in a high cardiovascular risk population described as having hypertriglyceridemia and low HDL.

The primary outcome was a composite of cardiovascular death, nonfatal myocardial infarction, nonfatal stroke, coronary revascularization, or hospitalization for unstable angina. The posted counts were 785 of 6539 participants in the Epanova group and 795 of 6539 in the placebo group. The sponsor-reported Cox analysis gave a hazard ratio of 0.99, a 95% confidence interval from 0.90 to 1.09, and p = 0.837. The unadjusted recalculation gives a risk difference of -0.002, with a 95% confidence interval from -0.013 to 0.010.

Both summaries are compatible with no clear difference on the primary endpoint, while leaving uncertainty about the possible effect range. Details of the original sample-size justification, multiplicity handling, and several analysis-plan features are not reported in the registry record.