Improving medication use in newly admitted home healthcare patients: a randomized controlled trial.

Objectives: To test the efficacy of a medication use improvement program developed specifically for home health agencies. The program addressed four medication problems identified by an expert panel: unnecessary therapeutic duplication, cardiovascular medication problems, use of psychotropic drugs in patients with possible adverse psychomotor or adrenergic effects, and use of nonsteroidal antiinflammatory drugs (NSAIDs) in patients at high risk of peptic ulcer complications. It used a structured collaboration between a specially trained clinical pharmacist and the patients' home-care nurses to improve medication use.

Design: Parallel-group, randomized controlled trial.

Setting: Two of the largest home health agencies in the United States.

Participants: Study subjects were consenting Medicare patients aged 65 and older admitted to participating agency offices from October 1996 through September 1998, with a projected home healthcare duration of at least 4 weeks and at least one study medication problem.

Intervention: Qualifying patients were randomized to usual care or usual care with the medication improvement program.

Measurements: Medication use was measured during an in-home interview, with container inspection at baseline and at follow-up (between 6 and 12 weeks) by interviewers unaware of treatment assignment. The trial endpoint was the proportion of patients with medication use improvement according to predefined criteria at follow-up.

Results: There were 259 randomized patients with completed follow-up interviews: 130 in the intervention group and 129 with usual care. Medication use improved for 50% of intervention patients and 38% of control patients, an attributable improvement of 12 patients per 100 (95% confidence interval (CI) = 0.0-24.0, P =.051). The intervention effect was greatest for therapeutic duplication, with improvement for 71% of intervention and 24% of control patients, an attributable improvement of 47 patients per 100 (95% CI = 20-74, P =.003). Use of cardiovascular medications also improved more frequently in intervention patients: 55% vs 18%, attributable improvement 37 patients per 100 (95% CI = 9-66, P =.017). There were no significant improvements for the psychotropic medication or NSAID problems. There was no evidence of adverse intervention effects: new medication problems, more agency nurse visits, or increased duration of home health care.

Conclusions: A program congruent with existing personnel and practices of home health agencies improved medication use in a vulnerable population and was particularly effective in reducing therapeutic duplication.