Association of Statin Adherence with Mortality in Patients with Atherosclerotic Cardiovascular Disease
By Heather Nelson Cortes, PhD and Kevin C Maki, PhD
There is little doubt of the role that statins play in the reduction of mortality risk, mainly attributable to a benefit on death from cardiovascular causes. A meta-analysis of statin clinical trials reported a 12% proportional reduction in all-cause mortality per mmol/L reduction in low-density lipoprotein cholesterol (LDL-C) with the use of statins (rate ratio [RR] 0.88, 95% confidence interval [CI] 0.84-0.91; p < 0.0001).1 In a systematic review, De Vera et al. reported an increased risk of adverse outcomes with poor statin therapy adherence.2 With such a strong link between statin adherence and decreased mortality, it is unclear why so many patients stop taking their statins or what the long-term effects on health and healthcare costs will be. Results from surveys of statin adherence suggest that as many as 50% of patients for whom statin therapy is prescribed stop taking the medication within 12 months.3,4 Rates of discontinuation and poor adherence are high even among those with known atherosclerotic cardiovascular disease (ASCVD).3,4
To further assess the association between statin adherence and all-cause mortality, Rodriguez et al. conducted a retrospective cohort analysis of patients between the ages of 21-85 years with one or more International Classification of Diseases, Ninth Revision, Clinical Modification codes for ASCVD on two or more dates in the previous two years without intensity changes to their statin prescription.5 All patients were treated within the Veterans Affairs Health System between January 1, 2013 and April 2014.
The primary outcome was death from all causes adjusted for demographic and clinical characteristics, and adherence to other cardiac medications. Secondary outcomes included 1-year mortality, 1-year hospitalization for ischemic heart disease or ischemic stroke. A sensitivity analysis was also conducted to investigate an association between statin adherence and hospitalization for gastrointestinal bleeding and pneumonia. Finally, the researchers sought to determine if the association between statin adherence and mortality was modified by statin intensity (low, medium, high) or by patient-level or system-level characteristics.
The medication possession ratio (MPR) was used to measure patient medication adherence. The MPR is the number of days of outpatient statin supplied during a 12- month period divided by the number of days the patient was not hospitalized and alive in the same 12-month time frame. Medication adherence was categorized as <50% MPR, 50-69% MPR, 70-89% MPR and ≥90% MPR.
The study included 347,104 patients with ASCVD on stable statin prescriptions. The overall mean statin adherence in this population was ~88%; ~6% had a MPR of <50% and ~64% had a MPR of ≥90%. Overall, women were less adherent than men (odds ratio, 0.89; 95% CI, 0.84-0.94), as were minority groups, while younger and older patients were less likely to be adherent compared with those aged 65-74 years. During a mean (standard deviation) follow up of 2.9 (0.8) years there were 85,930 deaths (24.8%). Compared to the most adherent patients (MPR ≥90%), patients with a MPR <50% had a hazard ratio (HR) adjusted for clinical characteristics and adherence to other cardiac medications of 1.30 (95% CI, 1.27-1.34), while those with a MPR of 50-69% had a HR of 1.21 (95% CI, 1.18-1.24), and those with an MPR of 70-89% had a HR of 1.08 (95% CI, 1.06-1.09).
After one year, hospitalizations for ischemic heart disease and stroke were more frequent in patients who were less adherent to their statin therapy. The proportion of patients with a hospitalization for ischemic heart disease or ischemic stroke was 13.4% (n = 2653) for an MPR<50%, 13.1% (n = 4018) for an MPR of 50-69%, 11.5% (n = 8729) for an MPR of 70-89%, and 11.5% (n = 25434) for an MPR of ≥90% (p < 0.001). This association remained even after adjusting for baseline characteristics. There was no association between MPR and hospitalization for gastrointestinal bleeding or pneumonia.
Finally, in this cohort 42,010 (12%) patients were on low-intensity therapy, 217,570 (63%) were on moderate-intensity therapy, and 87,524 (25%) were on high-intensity treatment. Patients on moderate-intensity statin therapy were more likely to adhere to statin therapy compared to patients in the low- and high-intensity therapy groups. Patients with the highest MPR had lower LDL-C values (77.2 mg/dL for MPR ≥90% compared with 92.1 mg/dL for MPR <50%).
Comment. The role that statins play in the reduction of mortality is not surprising. When a patient consistently takes the statin as prescribed, their risk of cardiovascular mortality will likely decrease. What is surprising is the lack of adherence by patients, especially over time, given the evidence supporting statin effectiveness. Further research should focus on how to improve patient adherence to statin therapy.6
Another important consideration that is illustrated by the present study is that it is important to consider the effects of healthy and unhealthy user bias in observational studies. Ann Marie Navar makes this point in an editorial accompanying the paper.6 Those with the poorest adherence to statin therapy (MPR <50%) had a 30% increase in mortality. Adjustment for follow-up LDL-C levels reduced the mortality hazard by 10%. Thus, only one third of the effect appears to be attributable to the main pathway through which statins alter risk. This suggests the presence of residual confounding by other factors. People who are adherent to therapy recommendations differ in numerous ways relevant to health outcomes from those who do not. It is difficult, if not impossible, to fully account for differences in potential confounders through statistical modeling. Thus, while a portion of the higher mortality risk among those with poor adherence is likely due to less impact of the drug itself, other factors also likely contribute to a similar or even larger degree.
This concept was recently illustrated in an analysis from the Prostate, Lung, Colorectal and Ovarian (PLCO) Cancer Screening Trial.7,8 Healthy 55- to74-year old participants were randomly assigned to receive usual care or a more intensive screening program. Among those randomized to more intensive screening, 10.8% did not complete any of the recommended screening tests. After 10 years, those non-adherent subjects had 50% higher mortality compared to those with full adherence who completed all recommended tests. However, the increased mortality was not attributable only to cancers, but to a wide range of causes. Thus, non-adherence to recommended screening (or to prescribed medication) is likely a marker for an array of behaviors associated with increased mortality risk. This unhealthy or healthy user bias should be kept in mind when evaluating the results from observational studies of behaviors associated with health outcomes. Those who choose to engage in a behavior they view as health-promoting, such as taking prescribed medication, undergoing recommended screening tests, following a diet or exercise program, or taking a dietary supplement, may differ in important ways from those who choose not to engage in the behavior, resulting in healthy user bias. Conversely, those who engage in behaviors they know are unhealthy, such as cigarette smoking, may also engage in other unhealthy behaviors (unhealthy user bias). Thus, estimates of the effects of behavioral exposures from observational studies should be interpreted with caution and should ideally be verified in prospective, randomized, controlled trials.
- Baigent C, Keech A, Kearney PM, et al. Cholesterol Treatment Trialists’ (CTT) Collaborators. Efficacy and safety of cholesterol-lowering treatment: prospective meta-analysis of data from 90,056 participants in 14 randomised trials of statins. Lancet. 2005;366:1267-1278.
- DeVera MA, Bhole V, Burns LC, Lacaille D. Impact of statin adherence on cardiovascular disease and mortality outcomes: a systematic review. Br J Clin Pharmacol. 2014;78:684-698.
- Maddox TM, Chan PS, Spertus JA, et al. Variations in coronary artery disease secondary prevention prescriptions among outpatient cardiology practices: insights from the NCDR (National Cardiovascular Data Registry). J Am Coll Cardiol. 2014;63:539-546.
- Hirsh BJ, Smilowitz NR, Rosenson RS, et al. Utilization of and adherence to guideline-recommended lipid-lowering therapy after acute coronary syndrome: opportunities for improvement. J Am Coll Cardiol. 2015;66:184-192.
- Rodriguez F, Maron DJ, Knowles JW, et al. Association of statin adherence with mortality in patients with atherosclerotic cardiovascular disease. JAMA Cardiol. 2019; Epub ahead of print.
- Navar AM. Statins work, but only in people who take them. JAMA Cardiol. 2019; Epub ahead of print.
- Pierre-Victor D, Pinsky PF. Association of nonadherence to cancer screening examinations with mortality from unrelated causes: a secondary analysis of the PLCO Cancer Screening trial. JAMA Intern Med. 2019;179:196-203.
- Grady D. Why is nonadherence to cancer screening associated with increased mortality? JAMA Intern Med. 2018; Epub ahead of print.