Background Older people with dementia are at risk of adverse drug-related events associated with the prescribing of potentially inappropriate medications.

Aim To describe how tools designed to identify potentially inappropriate prescribing have been used in studies of older people with dementia and to determine the prevalence of potentially inappropriate prescribing in this cohort.

Methods Systematic review with narrative synthesis designed and conducted according to the recommendations set out in Preferred Reporting Items for Systematic Reviews and Meta-Analyses-Protocol (PRISMA-P) and PRISMA statements. Electronic databases, including MEDLINE, EMBASE, PsychINFO, Cochrane Library, Social Science Citation Index, OpenGrey and Grey Literature Report were searched, along with hand-searches, for studies using tools/criteria to identify potentially inappropriate prescribing in older people with dementia. Screening, extraction and quality assessment (Hawker score) were performed in duplicate.

Results A total of 3,626 unique records were identified; 26 studies met the inclusion criteria (which included 26,534 participants, of which 21,285 (80%) had dementia or cognitive impairment). The mean participant age in these studies ranged from 72.5 to 86.8. The Hawker score ranged from 22/36 to 36/36, indicating all studies were fair to good quality. The Beer's criteria was the most commonly used tool (15/26 studies). There were variations in how the tools were applied, with 13 out of 15 studies using the Beer's criteria not using the full tool. Eight studies used more than one tool to identify potentially inappropriate prescribing. The prevalence of potentially inappropriate prescribing ranged from 14% to 74% in older people with dementia. The most commonly prescribed potentially inappropriate medications were benzodiazepines and anticholinergics.

Conclusions Variations in tool application may partly explain variations in potentially inappropriate prescribing rates across studies. Recommendations include a more standardised tool usage, ensuring the tools comprehensively identify all potentially inappropriate medications and that the tools are kept up to date.