PRESCRIPTION DRUG USE AND DRUG-DRUG INTERACTIONS IN NURSING HOMES: PREVALENCE, PREDICTORS, AND HEALTH OUTCOMES (DRUG USE).

Description

The purpose of this study was to examine (1) the prevalence of prescription drug use and drug-drug interactions (DDI) in a sample of nursing home residents, (2) the demographic and health predictors of drug use and DDI, and (3) the relationship between drug use, DDI, and health outcomes over time. The sample consisted of 174 nursing home residents who resided in four nursing homes over a one-year period. Data were collected in conjunction with a larger study of interventions in nursing homes during August 1988 and August 1989. Drug use was assessed via Medication Administration Records (MARS) kept by each facility. Drug-drug interactions were detected using two separate methods, including a computerized system and a manual reference. Residents' health status was assessed by nursing staff using both the Multidimensional Observation Scale for Elderly Subjects (MOSES) and more global ratings. Subsequent factor analyses were used to reduce health and behavior data to three parsimonious dimensions (Mental Health, Physical Health, and Behavioral Health). The results indicated that residents were prescribed a mean of about 8 different drugs during the study interval, but they received only about 5 of those drugs, on average. The discrepancy between drugs ordered and drugs received was particularly pronounced for prn (pro re nata, or as needed) drugs, since over 90% of subjects were prescribed at least one prn medication, but only 40% actually received one during the study interval. In addition, about 30% of the residents in this sample were exposed to at least one clinically significant drug-drug interaction, though estimates differed markedly depending on the DDI detection system used. A subsequent series of regression, cross-lagged correlation, and path analyses was conducted to test the relationship between drugs and residents' physical, mental, and behavioral health over time. These analyses indicated few significant drug-health relationships. In particular, no significant health outcomes of central nervous system agents (including psychotropic drugs) were noted. Overall, the results suggest that the best predictor of Time Two drug use or health status is the residents' drug use or health status, respectively, at Time One. Implications for research, clinical practice for nurses and physicians, and public policy (especially recent legislative reform for nursing homes) were discussed.

Subjects