The use of psychotropic medication among persons with dementia (PWD) both in nursing home (Wetzels et al., Reference Wetzels, Zuidema, de Jonghe, Verhey and Koopmans2011) and community settings (Maust et al., Reference Maust, Langa, Blow and Kales2016) far exceeds what might be expected based on their limited evidence for benefit (Kales et al., Reference Kales, Gitlin and Lyketsos2015). This relatively high use persists despite years of evidence about the potential harms associated with their use in older adults generally and PWD in particular (e.g. Wang et al., Reference Wang, Bohn, Glynn, Mogun and Avorn2001; Schneider et al., Reference Schneider, Dagerman and Insel2005). However, the solution to relatively high psychotropic use is not to end all psychotropic use, as there are individual patients for whom use of such medication is appropriate. For example, a policy that defines all antipsychotic use as inappropriate may simply lead to increased use of alternatives with even less evidence of benefit, as suggested by the response to antipsychotics’ black box warning in the United States (Kales et al., Reference Kales2011).

The paper in this month's issue by van der Spek and colleagues (van der Spek et al., Reference van der Spek2016) presents an important step forward in our ability to help define the “right” amount of psychotropic medication use among PWD. They use the Appropriate Psychotropic drug use In Dementia (APID) index to review medication use on the following seven domains of appropriateness: indication, evaluation, dosage, drug–drug interactions, drug–disease interactions, duplications, and therapy duration. Among 559 PWD living in 44 different dementia special care units, just 9.6% of psychotropic use was fully appropriate. With 14.4% considered fully appropriate, antidepressants were the class most appropriately prescribed, while antiepileptics were most inappropriately used (0% appropriate). The domains of indication, evaluation, and duration accounted for 89.5% of overall inappropriateness: 35.8% of medications had no indication, and over one half had no evidence of evaluation.

The APID index provides a potentially useful approach for facilities and providers to address medication use among PWD. Rather than taking a blanket approach where all psychotropic use or all use of a particular class among PWD is deemed inappropriate, using APID could help a facility identify either particular medications or particular domains as problematic. For example, while van der Spek and colleagues identified antidepressant use as the most appropriate, antidepressants benefited from scoring 100% appropriate on therapy duration, as there is not a maximum therapy duration according to drug formularies. This artificially high score for antidepressants on this domain masked the fact that antidepressants scored next to last for appropriateness of indication, consistent with prior findings that antidepressant use among older adults frequently lacks a clear indication (Maust et al., Reference Maust2011; Wiechers et al., Reference Wiechers, Kirwin and Rosenheck2014). A facility may find that interventions that specifically limit prescribing to specific documented indications for several classes may be the most meaningful way to reduce inappropriate medication use overall.

While potential application of the authors’ approach in nursing home settings is relatively straightforward, use in community settings may be more challenging. However, this is a pressing need, as the majority of psychotropic use occurs among PWD in community settings (Maust et al., Reference Maust, Langa, Blow and Kales2016). Finally, any intervention to reduce psychotropic use among PWD will be far more likely to succeed where there are alternative environmental and psychosocial interventions in place, with a structured approach such as DICE that guides their use (Kales et al., Reference Kales, Gitlin and Lyketsos2014).