Predicting Statin Adherence Following Hospital Discharge for Acute Coronary Syndromes: What’s the Score?

The use of statins following acute coronary syndromes (ACS) is well documented to improve mortality and reduce hospitalisation, revascularisation and risk of stroke [1], and has become a frequently used benchmark to audit in-hospital care of such patients. Much less is known about adherence to statins following discharge—and the factors influencing this—which is the subject of the paper by Liao et al. in this edition of Heart, Lung and Circulation [2].

Liao et al. examined the statin medication possession ratio (MPR) at 16 months–18 months after hospitalisation for ACS using a national pharmaceutical dispensing database, for all patients enrolled in the All New Zealand (NZ) ACS Quality Improvement (ANZACS-QI) programme between 2013–2017 [2]. Medication adherence was defined as an MPR>0.8—a commonly accepted definition—and meaning the proportion of time (or number of doses) that a medication is used when a patient is not in hospital [3,4]. The authors then established a risk model for associations with statin non-adherence (an MPR<0.8) following an ACS hospital presentation, with the findings of particular interest.

Unsurprisingly, the most significant risk factor for statin non-adherence was previously not taking a statin; this was seen particularly in those who had had previous cardiovascular disease. Non-adherence was also associated with no revascularisation performed in hospital, younger age, being female, smoking, being of lower socio-economic background and in those of Māori or Pacific Islander ethnicity, though interestingly the relative risk in these latter groups was only modest, from 1.21–1.56 [2]. These findings were consistent with prior meta-analyses which identified those of female sex, low education, non-Caucasian background and low-income groups as being of greatest risk of medication non-adherence [5]. The authors then established a score, from 1 to 23, indicating a graded increase in the proportion of patients who were non-adherent to statin use. Rates of non-adherence in the lowest quartile, a score of less than 5, was 12%, compared to the highest quartile, with a score greater than 11, at 45%, though the relatively low C statistic of 0.67 (CI 0.66–0.68) suggests only modest applicability.

Risk scores have been well-established in assessing outcomes in ACS, such as the Global Registry of Acute Coronary Events (GRACE) ACS risk and mortality score [6,7]. This work by Liao et al. suggests similar techniques might in future also be applied to ACS management after discharge from hospital, and importantly, potentially allow for more targeted multidisciplinary approaches to approve adherence in populations identified as higher risk.

The work also suggests there is much more to be learnt about reasons for adherence and non-adherence to statin therapy, given the variation in relative risk demonstrated [2]. While the reasons for these differences were not explored in this work, they are raised as key further questions for study. Also important in future work will be to establish how generalisable are the findings, given the patients were drawn from a specific programme using an invasive management strategy and the New Zealand health care system where medication cost subsidies likely differ from those found elsewhere. Lastly, this novel and interesting work highlights the importance of education initiatives, shared decision making between patients and health professionals, and involvement of pharmacy-based programs pre- and post-discharge [8]; and, emphasises the importance of ongoing evaluation of adherence with patients and the need for further investigation into this important area.