Explicit, evidence-based criteria to assess the quality of prescribing to elderly nursing home residents

Abstract

Background: prescribing in nursing homes is frequently suboptimal. Indicators to measure prescribing quality, including appropriateness of prescribing certain drugs or combinations of drugs, to hospital inpatients have been developed previously. Objective: to modify prescribing indicators, including appropriateness of prescribing algorithms developed in the hospital setting, for use in nursing homes. Design: an audit of prescribing to patients resident in a random sample of nursing homes on a single day. Setting, subjects: 22 nursing homes in the former South Thames Region selected from lists of nursing homes with more than 35 residents. All residents aged 65 years or over were eligible. Methods: prescribing indicators, including evidence-based indicators of appropriateness of prescribing benzodiazepines, steroids with β2 agonists, antithrombotics with digoxin and aspirin with nitrates were adapted: to reflect where prophylaxis was not justified in terms of quality of life; and for use with primary care clinical records. Indicators were used to evaluate drugs prescribed to each resident to determine whether prescribing was appropriate. Results: 13 indicators were successfully modified and applied. The 934 residents included were prescribed a mean of 5.1 regular items. Only 496/934 (55%) drug sensitivity statements were completed. Although 24% residents received benzodiazepines, clinical data indicated that only 7% received benzodiazepines appropriately. Over three-quarters of residents with ischaemic heart disease received appropriate aspirin therapy, but fewer than half residents with atrial fibrillation received appropriate antithrombotic therapy. It was not possible to derive reference ranges of observed prescribing that included homes demonstrating appropriate prescribing whilst excluding those with inappropriate prescribing. Intra-cluster correlations ranged from 0.027 to 0.335. Conclusion: quality of prescribing indicators were successfully modified for the nursing home setting. Application identified suboptimal prescribing to nursing home residents.