Comparative Analysis Of Anticholinergics Prescribed To Elderly Patients At A Korean Long-Term Care Facility According To Beers Criteria 2003, 2012, And 2015 And Anticholinergic-Burden Rating Scales: A Cross-Sectional Retrospective Study

BACKGROUND: The Beers Criteria is used as a reference to identify potentially inappropriate medications (PIMs) prescribed to older people, and anticholinergic risk measurement scales (ARMSs) have been continuously made for measuring the anticholinergic burden. This study aimed to evaluate the concordance between any anticholinergics among PIMs identified by the Beers Criteria and those assessed by 9 different ARMSs. METHODS: This study was retrospectively conducted with Korean older patients hospitalized in the long-term care facility between March 2014 and August 2015. The data were collected through the chart review of electronic medical records of the patients. The Beers Criteria 2003, 2012, and 2015 were used to detect PIMs, and the following ARMSs were also employed to assess their potential anticholinergic effects: Anticholinergic Cognitive Burden Scale (2008), Anticholinergic Risk Scale (2008), Chew's Scale (2008), Anticholinergic Drug Scale (ADS; 2006), Anticholinergic Activity Scale (AAS; 2010), Anticholinergic Load Scale (2011), Clinician-Rated Anticholinergic Scale (2008), Duran's Scale (2013), and Anticholinergic Burden Classification (2006). RESULTS: The eligible patients who met inclusion and exclusion criteria were 216 during the study period. Most patients were females (70.4%), and the mean age was 81.0 ± 6.7 years. Approximately 70%, 86%, and 87% of the patients included were identified as using at least one PIM according to the Beers Criteria 2003, 2012, and 2015, respectively. Compared with the Beers Criteria 2003, the versions of 2012 and 2015 showed more improved concordance associated with the ARMSs. When the Beers Criteria 2015 was compared with the ARMSs, the lowest concordance was found for AAS (κ = 0.153; 95% CI, 0.079-0.227), whereas the highest concordance was observed for ADS (κ = 0.530; 95% CI, 0.406-0.654). CONCLUSION: The heterogeneity between the Beers Criteria and the ARMSs was observed. Compared with the Beers Criteria 2003, the versions of 2012 and 2015 showed more enhanced concordance associated with the ARMSs.