Inter-rater reliability of STOPPFrail [Screening Tool of Older Persons Prescriptions in Frail adults with limited life expectancy] criteria amongst 12 physicians

Amanda Hanora Lavan, Paul Gallagher, Denis O’Mahony

STOPPFrail is an explicit tool, developed by Delphi consensus, to assist physicians with deprescribing medications in frail older adults with poor survival prognosis. This study aimed to determine the inter-rater reliability (IRR), amongst physicians, of STOPPFrail application. Twenty clinical cases were collated to represent frail older patients. Eighteen cases met STOPPFrail inclusion criteria. They had a mean age of 79.5 (SD6) years and a median of 7 (IQR6-8.25) comorbidities and were prescribed a median of 9 (IQR7.75-11.25) medications. Two of the STOPPFrail originators reached complete agreement (gold standard) in determining 91 of 165 medications (55.2%) as inappropriate. Twelve physicians (6 geriatricians, 3 general practitioners and 3 palliative care physicians) independently applied STOPPFrail criteria. IRR between physicians and gold standard (GS) assessment was determined using Cohen's kappa statistic. Eighteen of the 20 cases that met STOPPFrail inclusion criteria were correctly identified by 9 of 12 physicians (75%). The average time taken per clinical case was 2.7 (SD0.94) minutes. The kappa co-efficient between physicians and GS assessment ranged from 0.71 (substantial) to 0.86 (good), with a mean kappa value of 0.758 (SD0.059). The Fleiss kappa coefficients between GS assessment and geriatricians, GPs and palliative care physicians were 0.80 (SD0.6), 0.77 (SD0.9) and 0.75 (SD0.1), respectively. No significant difference was noted, between groups or between participants within groups, as determined by one-way ANOVA, (df (2, 9) = 0.712, p = 0.516). IRR of STOPPFrail criteria between physicians, practising in different specialties, is substantial, despite no prior knowledge of the criteria.