Screening tests for intended medication adherence among the elderly.

BACKGROUND: Medication nonadherence is increasingly recognized as a cause of preventable adverse events, hospitalizations, and poor healthcare outcomes. While comprehensive medication adherence assessment for the elderly is likely to identify and prevent drug-related problems, it is time consuming for patient and healthcare providers alike. OBJECTIVE: To identify screening tools to predict elderly patients' intended medication adherence that are suitable for primary-care settings and community pharmacies. METHODS: This study evaluated 57 English-speaking persons aged 65 years and older who were from diverse socioeconomic backgrounds. Intended adherence was quantified, and the relationships to demographic, medical history, socioeconomic, and literacy variables were determined. RESULTS: In a multivariate analysis with the composite MedTake Test (a quantitative measure of each subject's intent to adhere to prescribed oral medications) as the dependent variable, independent predictors of intended adherence included: age, car ownership in the last 10 years, receipt of food assistance in the last 10 years, number of over-the-counter (OTC) medicines, and REALM (Rapid Estimate of Adult Literacy in Medicine). The strongest predictor was the REALM word-recognition pronunciation test (beta = 0.666; R2 = 0.271; p < 0.001). CONCLUSIONS: We observed that the REALM word-recognition pronunciation test, along with age, number of OTC drugs, and 2 socioeconomic questions, predicted the intent of seniors to correctly take their own prescribed oral medications.

Individualized drug use assessment in the elderly.

STUDY OBJECTIVE: To quantify how seniors' ability to take oral prescription drugs safely may correlate with age, sex, socioeconomic status, education, cognitive impairment, depression, and drug self-management. DESIGN: Cross-sectional study SETTING: Three retirement communities and an adult day care center. PATIENTS: Fifty-seven elderly individuals (mean age 79.49 +/- 7.26 yrs; mean education 11.33 +/- 3.8 yrs; 72% women). INTERVENTION: After completing a comprehensive medical history, and with drug vials and pillboxes available for consultation, each subject described how he or she was taking prescribed oral drugs. MEASUREMENTS AND MAIN RESULTS: The MedTake test evaluated dosage, indication, food or water coingestion, and regimen. For each agent, the test was scored as percentage of correct actions, equally weighted, and compared with label directions or self-expressed physician changes. A composite MedTake test score (0-100%) summarized a subject's overall ability to take their drug(s) safely A follow-up qualitative assessment by a single pharmacist assigned each agent to one of four potential risk categories: correct use, partial correct use without potential clinical significance, partial correct use with potential clinical significance, or incorrect use with high potential of clinical significance. Most subjects (80%) managed their own drug therapy; 70% used reminder systems (calendar, pillbox). The number of medical conditions and prescription drugs was 6.11 +/- 4.2 and 5.88 +/- 3.44, respectively. Of 325 agents, correct dosage was reported for 94% (306), correct indication for 95% (309), correct coingestion with food or water for 97% (314), and correct regimen for 89% (288). The composite MedTake test score was 88.5 +/- 21.3%. The multivariate model, with that score as the dependent variable, adjusted for age and sex, used Mini-Mental State Examination (p = 0.002) and Medicaid assistance within 10 years (p = 0.021) as significant factors. The most frequent problem was underdosing of cardiovascular drugs. CONCLUSION: Seniors' ability to take oral prescription drugs safely was affected by cognitive function and socioeconomic status. Although the MedTake test helped identify some problems with therapy adherence, a pharmacist's follow-up evaluation of comprehensive medical and drug histories identified additional potentially clinically significant problems in 20% of subjects.