Cancer Medicine (Jan 2024)

Performance of potentially inappropriate medications assessment tools in older Indian patients with cancer

  • Sharath Kumar,
  • Renita Castelino,
  • Abhijith Rao,
  • Shreya Gattani,
  • Anita Kumar,
  • Anupa Pillai,
  • Arshiya Sehgal,
  • Pallavi Rane,
  • Anant Ramaswamy,
  • Ratan Dhekale,
  • Jyoti Krishnamurthy,
  • Shripad Banavali,
  • Rajendra Badwe,
  • Kumar Prabhash,
  • Vanita Noronha,
  • Vikram Gota

DOI
https://doi.org/10.1002/cam4.6797
Journal volume & issue
Vol. 13, no. 1
pp. n/a – n/a

Abstract

Read online

Abstract Background Polypharmacy and potentially inappropriate medication (PIM) use are common problems in older adults. Safe prescription practices are a necessity. The tools employed for the identification of PIM sometimes do not concur with each other. Methods A retrospective analysis of patients ≥60 years who visited the Geriatric Oncology Clinic of the Tata Memorial Hospital, Mumbai, India from 2018 to 2021 was performed. Beer's‐2015, STOPP/START criteria v2, PRISCUS‐2010, Fit fOR The Aged (FORTA)‐2018, and the EU(7)‐PIM list‐2015 were the tools used to assess PIM. Every patient was assigned a standardized PIM value (SPV) for each scale, which represented the ratio of the number of PIMs identified by a given scale to the total number of medications taken. The median SPV of all five tools was considered the reference standard for each patient. Bland–Altman plots were utilized to determine agreement between each scale and the reference. Association between baseline variables and PIM use was determined using multiple logistic regression analysis. Results Of the 467 patients included in this analysis, there were 372 (79.66%) males and 95 (20.34%) females with an average age of 70 ± 5.91 years. The EU(7)‐PIM list was found to have the highest level of agreement given by a bias estimate of 0.010, the lowest compared to any other scale. The 95% CI of the bias was in the narrow range of −0.001 to 0.022, demonstrating the precision of the estimate. In comparison, the bias (95%) CI of Beer's criteria, STOPP/START criteria, PRISCUS list, and FORTA list were −0.039 (−0.053 to −0.025), 0.076 (0.060 to 0.092), 0.035 (0.021 to 0.049), and −0.148 (−0.165 to −0.130), respectively. Patients on polypharmacy had significantly higher PIM use compared to those without (OR = 1.47 (1.33–1.63), p = <0.001). Conclusions The EU(7)‐PIM list was found to have the least bias and hence can be considered the most reliable among all other tools studied.