Variation between nursing homes in drug use and in drug-related problems

Residents at nursing homes (NHs) are at particular risk for drug related harm. Regular medication reviews using explicit criteria for pharmacological inappropriateness and classification of drug related problems (DRPs) have recently been introduced as measures to improve the quality of medication use and for making the treatment more uniform across different institutions. Knowledge about variation in DRPs between NHs is scarce. To explore if increased attention towards more appropriate drug treatment in NHs have led to more uniform treatment, we have analyzed variations between different nursing homes' drug use and DRPs.

Cross-sectional medication review study including 2465 long-term care residents at 41 NHs in Oslo, Norway. Regular drug use was retrieved from the patients' medical records. DRPs were identified by using STOPP/START and NORGEP criteria and a drug-drug interactions database. NHs were grouped in quartiles based on average levels of drug use. The upper and lower quartiles were compared using independent samples t-test and associations between drug use and DRPs were tested by logistic regression.

Patients' mean age was 85.9 years, 74.2% were women. Mean numbers of regular drugs per patient was 6.8 and varied between NHs from 4.8 to 9.3. The proportion of patients within each NH using psychotropic and analgesic drugs varied largely: antipsychotics from three to 50%, benzodiazepines from 24 to 99%, antidepressants from nine to 75%, anti-dementia drugs from no use to 42%, opioids from no use to 65% and paracetamol from 16 to 74%. Mean DRPs per patient was 2.0 and varied between NHs from 0.5 to 3.4. The quartiles of NHs with highest and lowest mean drugs per patient (7.7 vs. 5.7, p < 0.001) had comparable mean number of DRPs per patient (2.2 vs. 1.8, p = 0.2). Using more drugs and the use of opioids, antipsychotics, benzodiazepines and antidepressants were associated with more DRPs.

The use of psychotropic and analgesic drugs was high and varied substantially between different NHs. Even if the use of more drugs, opioids and psychotropic drugs was associated with DRPs, no difference was found in DRPs between the NHs with highest vs. lowest drug use.