Abstract
Due to the complexity of medication use in older adult patients living with cancer, the risk of potentially inappropriate medications (PIMs) was significantly increased, yet evaluation criteria vary. Therefore, this study aimed to detect the PIMs use by the 2023 American Geriatrics Society (AGS) Beers criteria, the 2023 Screening Tool of Older Persons' Prescriptions (STOPP) criteria, the geriatric oncology potentially inappropriate medication (GO-PIM) scale, and the 2024 Chinese criteria in older adult patients living with cancer and compare the prevalence of PIMs and the accordance between the 4 PIM criteria; and further to explore related risk factors for PIMs according to the 4 criteria. This retrospective study included 484 older adult patients living with cancer. PIMs were analyzed based on the 4 criteria. The consistency of the evaluation results was compared between the 4 criteria, and influencing factors for PIMs were analyzed using multivariate logistic regression. The incidence of PIMs in older adult patients living with cancer was high in our study, with certain differences in consistency among the 4 criteria. The 2024 Chinese criteria were the most sensitive for identifying PIMs, and polypharmacy was the main influencing factor for the occurrence of PIMs in all criteria. It is necessary to strengthen medication therapy management for older adult patients living with cancer. The prevalence of at least one PIM identified by the 4 criteria ranged from 42.15% to 71.07%. The drugs with the highest incidence were metoclopramide, medroxyprogesterone, and cimetidine. The kappa statistics for the 2023 AGS/Beers criteria with the 2024 Chinese criteria, the GO-PIM scale and 2024 Chinese criteria indicated a good concordance (κ = 0.788 and 0810). The consistency between the 2023 AGS/Beers criteria and the 2023 STOPP criteria, the GO-PIM scale were moderate (κ = 0.459 and 0732). The consistency between the 2023 STOPP criteria and the 2024 Chinese criteria was moderate (κ = 0.448), but with the GO-PIM scale was poor (κ = 0.345). The results of multiple logistic regression analysis showed that polypharmacy was the main influencing factor for the occurrence of PIMs in all criteria (P<.001).