Inappropriate drug prescription at nursing home admission

Abstract

Background: Inappropriate prescriptions are common in older people admitted to nursing homes. Commonly used instruments to detect potential inappropriate prescriptions have limitations that have precluded wide use, and new instruments are needed.

Objective: The goal of this study was to determine the value of the Screening Tool of Older Person's potentially inappropriate Prescriptions/Screening Tool to Alert doctors to the Right, ie appropriate, indicated Treatment (STOPP-START) criteria and the Australian criteria to detect potentially inappropriate drug prescriptions in older people on admission to nursing home care.

Methods: Cross-sectional study of 100 consecutive patients (mean age 84.7 ± 7.5 years, 80% women) admitted to 6 assisted living nursing homes, with systematic review of prescriptions used at the time of nursing home admission using the STOPP-START and the Australian criteria looking for potentially inappropriate drug treatments.

Results: Using the STOPP criteria, 79% of the subjects showed at least one potentially inappropriate prescription. Omissions of potentially appropriate drugs were found by the START criteria in 74% of them. The Australian criteria detected at least one potential problem in 95% of the sample. The number of subjects with 2 or more problems detected was highest using the Australian criteria (72%). The most frequent potentially inappropriately used drugs detected were proton-pump inhibitors, benzodiazepines, antipsychotic drugs, and anticholinergic drugs; many cases of duplicate medications and drug interactions were also detected. Underuse of statins and aspirin in patients with high cardiovascular risk, and of calcium and vitamin D in osteoporosis was also frequent.

Conclusions: A high number of potentially inappropriate drug prescriptions can be detected at the time of admission to nursing home care by the use of systematic instruments. Both STOPP-START criteria and the Australian criteria performed well in this setting. The impact of this detection on health outcomes and costs should be assessed before they can be widely recommended.