Inappropriate prescribing before and after nursing home admission

Objectives — To compare the prevalence of inappropriate prescribing before and after nursing home admission and to determine which patient and physician characteristics are associated with inappropriate prescribing in the nursing home setting.

Design — A pre/post retrospective, cohort study.

Setting — All licensed nursing homes in Ontario, Canada.

Participants — Nineteen thousand nine hundred eleven individuals aged 66 and older, newly admitted to nursing homes in Ontario between April 1, 1997, and March 31, 1999.

Measurements — For each patient in the cohort, a subset of the Beers criteria was used to characterize and compare the prevalence of inappropriate prescribing (as indicated by the prescription of one of 49 inappropriate drugs) before and after nursing home admission. A logistic regression model was used to study the association between inappropriate prescribing and patient and physician characteristics.

Results — The proportion of patients receiving a prescription for at least one inappropriate drug decreased from 25.4% before nursing home admission to 20.8% afterward (P <.001). Most patients who had been prescribed an inappropriate agent before nursing home entry had that agent discontinued after admission. The most commonly prescribed inappropriate drugs after nursing home admission were strongly anticholinergic antidepressants (6.4%) and long-half-life benzodiazepines (5.9%). Patients younger than 85 were more likely to receive inappropriate drug therapy (odds ratio (OR) = 1.25, 95% confidence interval (CI) = 1.15-1.35) than those aged 85 and older. Other significant predictors were having more than one prescriber (OR = 1.40, 95% CI = 1.29-1.51), having a physician aged 50 or older (OR = 1.14, 95% CI = 1.05-1.23), having a male physician (OR = 1.20, 95% CI = 1.05-1.37), having a nonspecialist physician (OR = 1.23, 95% CI = 1.01-1.49), having a nonurban physician (OR = 1.13, 95% CI = 1.03-1.24), and having a physician practicing outside the greater Ontario metropolitan area (OR = 1.31, 95% CI = 1.19-1.51).

Conclusions — Although a substantial number of nursing home residents receive inappropriate drug therapy, the prevalence of inappropriate prescriptions in our cohort declined after nursing home admission despite an overall increase in drug use. Patient and physician characteristics were associated with inappropriate prescribing. Targeted interventions such as regionally based education programs or drug use restrictions may reduce the prevalence of inappropriate prescribing.