Link → Alcohol misuse does seem to predict poor medication adherence
Ann Intern Med 2008; 149: 795-903
Excerpt: Analysis of data from a controlled trial provides objective evidence that alcohol misuse is associated with poor adherence to medications.The authors of this study used data from a pre-existing cohort to examine the hypothesis that alcohol misuse would be associated with a dose-related effect on medication adherence. They used data from large randomised trial that looked at the effects of an audit and feedback intervention (ACQUIP). As part of the study, prospective data on alcohol use and medication adherence was obtained for a large cohort of primary care patients. At entry to the study, patients completed a questionnaire that assessed a range of health behaviours, including alcohol misuse using the AUDIT-C measure (Alcohol Use Disorder Identification Test–Consumption - score ranges from 0 to 12, 0 = no alcohol use). Data on medication use was obtained from pharmacy records, and medication adherence for medications used to treat hypertension, hyperlipidaemia, and diabetes; as in other work, having medication available for at least 80% of a study period was considered to be adherent. Outcome was the proportion of adherent patients according to AUDIT-C score.
There were 35,725 patients who participated in ACQUIP and returned questionnaires with completed AUDIT-C. Of these, 5,473 used a statin, 3,468 used an oral anti-diabetic drug, and 13,729 used an antihypertensive medication. Overall, about 20% had AUDIT-C scores of 4 or more, signifying possible alcohol misuse.
For lipid lowering and antihypertensive drugs, higher AUDIT-C scores were associated with reduced adherence: at one year, and after adjustment for confounding factors, in the statin group, 58% (95% CI, 52% to 65%) and 55% (95% CI, 47% to 63%) of those in the two highest alcohol misuse groups (scores 6-7 and 8-12) were adherent compared to 66% (95% CI, 64% to 68%) of non-drinkers. There was a similar difference for antihypertensive medications, however there was no significant difference between the groups for anti-diabetic drugs.
