Determinants of polypharmacy in patients with behavioral and psychological symptoms of dementia

Background Polypharmacy is common in patients with dementia and is associated with several adverse effects. Factors associated with polypharmacy in patients with behavioral and psychological symptoms of dementia (BPSD) are unclear. This study examined the factors associated with general and psychotropic polypharmacy in patients with BPSD. Method Baseline data was obtained from the Standardizing Care for Neuropsychiatric Symptoms and Quality of Life in Dementia (StaN) study, a multisite trial currently underway in Canada at long‐term care and inpatient sites. The Cohen‐Mansfield Agitation Inventory (CMAI) was used to assess agitation and aggression and Cumulative Illness Rating Scale for Geriatrics (CIRS‐G) was used to assess medical burden. General polypharmacy was defined as concomitant use of five or more scheduled medications of any kind, and psychotropic polypharmacy was defined as concomitant use of two or more scheduled psychotropic medications. Correlation, and linear and logistic regressions were performed to investigate the associations of agitation/aggression and polypharmacy. Result 120 participants were enrolled [50.4% female; mean (SD) age= 80.25 (9.75) years]. 80.2% of participants had general polypharmacy and 56.2% had psychotropic polypharmacy. Total number of medications was positively correlated with CIRS_G (Spearman's r = 0.28, p = 0.002). In linear regression models, CMAI‐frequency (R2 = 0.137, p = 0.003, B = 0.269, p = 0.003) and CMAI‐disruptiveness (R2 = 0.136, p = 0.003, B = 0.268, p = 0.004) were significant predictors, whereas age, current psychiatric diagnosis and setting (inpatient versus long‐term) did not predict the number of psychotropics. Similarly in logistic regression models, CMAI‐frequency (R2 = 0.142, p = 0.002, Wald = 6.93, p = 0.008) and CMAI‐disruptiveness (R2 = 0.141, p = 0.002, Wald = 6.86, p = 0.009) were significant predictors, whereas age, current psychiatric diagnosis and setting did not predict psychotropic polypharmacy. Conclusion As expected, medical burden was positively correlated with general medication use. On the other hand, BPSD (agitation and aggression) were associated with total number of psychotropic medications, rather than psychiatric morbidity, age or setting. Improving the management of agitation/aggression might be an important factor in addressing polypharmacy in patients with BPSD.