The SECURE trial is a phase 3, multinational, randomized, controlled trial. A total of 2499 older patients with myocardial infarction within six months were randomised to either the polipill or the usual care group. The polypill containing aspirin (100 mg), ramipril (2.5mg, 5mg or 10 mg), and atorvastatin (20 mg or 40 mg). The primary outcome was a composite of cardiogenic death, nonfatal myocardial infarction, nonfatal ischemic stroke, or urgent coronary revascularization. Medication adherence was measured with the use of the eight-item Morisky Medication Adherence Scale . During a median follow-up of 3 years among 2466 evaluable patients, the incidence of primary outcome events was significantly lower in the polypill group than in the usual-care group. Treatment adherence, as reported by the patients at 6 and 24 months, was higher with the polypill.

Zhang You and her colleagues are very interested in SECURE trial. After careful study, they communicate and talk about the design of this study with their international counterparts.

In the comment, Zhang et al. thought that this positive result was attributed to increased adherence, which was determined on the basis of data from patient questionnaires. Although patient questionnaires are widely used to measure adherence, this method is susceptible to recall bias with increases in time between visits. In the SECURE trial, data on patient adherence were available only at 6 and 24 months, but longer-term data such as at 36 and 48 months were not provided, and only 77.5% of surviving patients provided data on adherence at 24 months. Although events were adjudicated by persons who were unaware of the treatment assignments, it is possible that subjective factors related to the patients may have influenced the results, because data on medication adherence were collected by means of questionnaires and some outcomes were assessed by telephone follow-up at 36 and 48 months.

The authors reply: Zhang et al. raise the question about bias in reporting of patient adherence. Adherence was a secondary outcome in our trial, and we consciously sought to avoid overexposing participants to questionnaires during follow-up given that adherence is a behavior, and people tend to modify their behavior simply because they are being observed (in what is classically known as the Hawthorne effect). Because this effect will probably be more marked the more frequently participants are asked about the target behavior, a randomized clinical trial is probably not the most suitable setting in which to measure detected patterns of adherence. With additional questionnaires, the usual-care group would probably have had better adherence than they would had they not been enrolled in the SECURE trial.

Zhang Y, Zhu Z, Gao C. Polypill Strategy in Secondary Cardiovascular Prevention. N Engl J Med. 2022;387(23):2197. doi: 10.1056/NEJMc2213446.