Critical appraisal of a
simulated randomized trial
Your institution's pharmacy and therapeutics committee is considering a new cardiovascular drug. You have one randomized trial and a decision deadline.
First, inspect the simulated journal report. Then make five appraisal decisions about effect framing, clinical importance, uncertainty, applicability and formulary action.
CAREVA-1, Veralozin and every patient-level value in this experience are fictional and created solely for learning.
Veralozin 10 mg versus Placebo for Secondary Prevention of Major Adverse Cardiovascular Events in Post-Myocardial Infarction Patients: Results of the CAREVA-1 Randomized Controlled Trial
Department of Cardiology, University Medical Center Rotterdam; National Heart Institute, Lagos; Institut Cardiovasculaire de Paris; full list of co-author affiliations for the CAREVA-1 Study Group in the online supplement.
Journal of Cardiovascular Medicine, Vol. 18 Issue 4 · Simulated report ID: CAREVA-1-2026 · Learning sample
Reduction
Reduction
to Treat
(HR, 5 yr)
1.1 Study design and participants
CAREVA-1 was a phase III, multicenter, randomized, double-blind, placebo-controlled trial conducted at 148 sites in 22 countries across Europe, North America, and Latin America. Eligible patients were adults aged 40 to 80 years with a documented prior myocardial infarction, LDL-C ≥70 mg/dL, and stable statin therapy for at least 6 weeks before enrollment. Patients were excluded if they had an eGFR <30 mL/min/1.73m², severe hepatic impairment, active malignancy, pregnancy, or prior intolerance to agents with a similar mechanism.
1.2 Randomization and blinding
Patients were randomly assigned in a 1:1 ratio using a computer-generated sequence, stratified by geographic region and history of heart failure. Both patients and investigators were masked to treatment assignment. An independent clinical events committee, blinded to treatment allocation, adjudicated all primary and secondary endpoints.
1.3 Design specifications
| Element | Detail |
|---|---|
| Primary endpoint | Composite MACE: cardiovascular death, non-fatal MI, or non-fatal stroke at 5 years |
| Analysis | Intention-to-treat (ITT) as pre-specified primary; per-protocol as sensitivity |
| Statistical model | Cox proportional hazards; two-sided α = 0.05 |
| Power | 80% to detect ≥20% RRR assuming 12% placebo event rate over 5 years |
| Funding | Veral Therapeutics AG; independent academic steering committee; no publication restrictions |
1.4 Baseline characteristics
| Characteristic | Veralozin (n=2,100) | Placebo (n=2,100) |
|---|---|---|
| Age, years — mean (SD) | 63.2 (9.4) | 63.5 (9.1) |
| Female sex — no. (%) | 664 (31.6%) | 672 (32.0%) |
| White — no. (%) | 1,724 (82.1%) | 1,718 (81.8%) |
| LDL-C, mg/dL — mean (SD) | 88.4 (16.2) | 88.1 (15.9) |
| History of heart failure — no. (%) | 378 (18.0%) | 382 (18.2%) |
| Current smoker — no. (%) | 294 (14.0%) | 288 (13.7%) |
| eGFR <60 mL/min — no. (%) | 316 (15.0%) | 309 (14.7%) |
Table 1. Baseline characteristics. SD = standard deviation; LDL-C = low-density lipoprotein cholesterol; eGFR = estimated glomerular filtration rate.
2.1 Primary outcome
The primary composite outcome occurred in 164 of 2,100 patients (7.8%) in the Veralozin group and 210 of 2,100 patients (10.0%) in the placebo group over the 5-year follow-up period. The hazard ratio was 0.78 (95% confidence interval, 0.65 to 0.94; p = 0.008). The cumulative event risks correspond to an approximately 22% relative risk reduction and a 2.2-percentage-point absolute risk reduction. Using the unrounded absolute risk difference, the number needed to treat to prevent one primary outcome event was 46 patients over 5 years.
2.2 Secondary outcomes and safety
Each component of the composite endpoint occurred numerically less frequently in the Veralozin group: cardiovascular death (3.2% vs. 4.0%), non-fatal MI (3.6% vs. 4.7%), and non-fatal stroke (1.8% vs. 2.2%). The subgroup analysis was consistent with the overall finding across age, sex, region, LDL-C tertile, and history of heart failure, with no statistically significant interaction detected. Serious adverse events occurred in 18.4% and 17.9% of the Veralozin and placebo groups, respectively. Liver enzyme elevations greater than 3 times ULN occurred in 1.4% of Veralozin patients versus 0.8% placebo. Adherence at 5 years was 74.2% in the Veralozin group and 78.1% in the placebo group.
In this large international randomized trial, Veralozin 10 mg daily added to standard care lowered the hazard of the primary composite cardiovascular outcome compared with placebo in patients with prior MI receiving stable statin therapy. The estimated hazard ratio was 0.78 (95% CI 0.65–0.94). No statistically significant treatment-by-subgroup interactions were detected in the prespecified analyses.
The approximately 22% relative risk reduction corresponds to a 2.2-percentage-point absolute risk reduction and a number needed to treat of 46 over 5 years. The clinical relevance of this magnitude of benefit must be considered in the context of the patient population, which was high-risk by design (prior MI, elevated LDL-C on statin therapy). If the relative treatment effect is similar in patients with higher baseline event rates, their absolute benefit would be greater.
The trial has important limitations. The enrolled population was predominantly male (68%) and White (82%), with a mean age of 63 years; this reduces certainty when applying the estimate to populations that were underrepresented in the trial. Patients with eGFR <30 mL/min were excluded, leaving the benefit-risk profile in advanced kidney disease uncharacterized. Adherence declined to 74% in the treatment group by year 5. The per-protocol analysis yielded a somewhat lower hazard ratio (0.74; 95% CI 0.62–0.89), but it remains susceptible to adherence-related selection bias and does not replace the prespecified intention-to-treat analysis. Cost-effectiveness was not evaluated.
Veralozin 10 mg daily lowered the hazard of the composite major adverse cardiovascular event outcome over 5 years in patients with prior MI receiving stable statin therapy. Rates of serious adverse events were similar between groups, although treatment discontinuation and liver-enzyme elevations should be considered when interpreting the benefit-risk profile. These data support consideration of Veralozin as adjunctive cardiovascular risk-reduction therapy in appropriately selected patients. Formulary bodies should also consider local costs, applicability to their patient population, and available alternatives.
Trial review complete.
Begin critical appraisal.
At five appraisal stations, assess how the reported evidence should inform interpretation and formulary decision-making.
Relative and absolute
treatment effects
The abstract reports: "Veralozin significantly reduced the composite risk of major adverse cardiovascular events (HR 0.78; p = 0.008)." Relative and absolute effect measures describe different aspects of the same result and should be interpreted with their time horizon and baseline risk.
Statistical evidence and
clinical importance
Under the trial's prespecified null model, a result at least this incompatible with the null would be uncommon (p = 0.008). The p value does not give the probability that the null hypothesis is true, and it does not establish whether the benefit is large enough to justify use.
Precision of the
treatment-effect estimate
The point estimate is 0.78, but the confidence interval runs from 0.65 to 0.94. The entire interval lies below 1.0, yet it spans 29 percentage points of possible effect size.
Who was actually in this trial?
The discussion flags it directly: 68% male, 82% White, mean age 63, eGFR ≥30. Think about how the patients in your institution compare to that profile.
Formulary
recommendation
Integrate the trial findings, uncertainty, applicability, safety information and resource considerations into a recommendation for the committee.
Your critical appraisal of CAREVA-1
What the trial shows: Veralozin 10 mg daily was associated with a 2.2-percentage-point absolute reduction in the composite MACE outcome over 5 years in the enrolled post-MI population receiving stable statin therapy (HR 0.78; 95% CI 0.65–0.94).
What it does not resolve: Long-term safety beyond 5 years; the precision and applicability of treatment-effect estimates in underrepresented populations; outcomes in patients with advanced kidney disease; and cost-effectiveness in health systems with constrained formularies.
The appraisal skill: Statistical compatibility is only one part of the judgment. Effect magnitude, uncertainty, population fit, harms, alternatives and resource implications all affect the decision.
Appraisal references
- CONSORT 2025: reporting and interpretation of randomized trials.
- American Statistical Association statement on p-values: what a p value does and does not indicate.