Development and preliminary evaluation of OPTI-3S: A deprescribing support tool for hospitalised frail older adults.

Abstract

Introduction: In frail older people, polypharmacy (5 or more regular prescribed medicines) and the prescription of potentially inappropriate medicines (PIMs) are known to be prevalent and linked to negative clinical outcomes. There is insufficient but consistent evidence that deprescribing is feasible, beneficial and safe in frail older adults. The aim of this research was the development and preliminary evaluation of the OPTI-3S list: new criteria for OPTImising medicines by Stopping, Stepping down or Switching to safer alternatives in hospitalised frail older patients. Methods: A scoping review was conducted to provide a comprehensive overview of the published explicit or implicit criteria that aid assessment of potentially inappropriate medicines (PIMs) in older adults (1991-2020). Additional literature (2010-2021) was extensively reviewed to develop preliminary statements for the OPTI-3S criteria. Findings on the safe and/or effective use of medicines in older adults and frailty-related subgroups (e.g. patients with advanced age, multimorbidity, physical/cognitive limitations, or polypharmacy) were retrieved to draft the preliminary statements. A pilot study was conducted on the initial statements to explore their suitability and ease of use. The final preliminary list was split to yield core and full lists and circulated to a multidisciplinary panel of eleven experts in clinical practice for content validation via a three-round modified Delphi method. The interrater reliability of the OPTI-3S core list was then assessed by five practising professionals using case studies based upon publicly archived data from ten participants in The Irish Longitudinal Study on Ageing (TILDA). Finally, the retrospective applicability of the OPTI-3S core list to all frail older (65 years or more) participants with polypharmacy in the third wave of TILDA, and the ability of this novel tool to identify potentially inappropriate prescribing and reduce polypharmacy within this population, were compared to three published PIM screening tools. Results: The scoping review of tools to identify PIMs yielded 73 criteria sets. Fifty-two tools (71%) were developed or validated using Delphi methods and their modifications. Almost half of the criteria (35, 47%) targeted older adults in multiple or all healthcare settings, with the remainder targeting specific cohorts (e.g. patients in long-term residential care). However, no consensus-based criteria had been specifically developed for hospitalised frail older adults. All tools were designed to screen for inappropriate medication use and overprescribing, with less than one-third of all tools addressing potential prescribing omissions (PPOs). Two tools with explicit statements suggested regimens for the safe withdrawal of potentially inappropriate medications. Further literature reviews led to analysis of 1500 published articles and the drafting of 109 OPTI-3S preliminary statements, applicable to patients at designated Clinical Frailty Score (CFS) levels. The pilot study resulted in the refinement and division of this initial list into 130 less complex statements (29 in the core list and an additional 101 in the full list). Nine experts completed all three rounds of the Delphi study, after which 45 statements were accepted for inclusion in the final OPTI-3S core list. The core statements related to deprescribing or optimising medicines acting on the central nervous system in medical and surgical patients (n=7), genito-urinary medications (n=7), hypoglycaemics (n=7), antithrombotics (n=5), antihypertensives (n=5), statins (n=4), heart failure medications (n=2), perioperative analgesia (n=2), osteoporosis medications (n=3), proton-pump inhibitors (n=1), vitamins and supplements (n=1), and anticholinergics (n=1). Cohen's kappa coefficients ranged from 0.65 to 0.86, indicating substantial to perfect interrater reliability. 583 frail older patients with polypharmacy were identified within TILDA wave 3. When applied to this cohort, the OPTI-3S core list identified at least one PIM in 62.6% of the cohort and could be used to reduce the incidence of polypharmacy by 20.6%: a higher PIM detection rate and polypharmacy reduction rate than any of the comparator PIM tools tested. The OPTI-3S core list detected 53.5 % of all PIMs identified by any tool and was the sole tool to detect 543 (36.4%) PIMs. The most common PIMs identified were the inappropriate use of aspirin, benzodiazepines and related drugs, and antihypertensive medications. Conclusion: This research identified that no PIM tool had previously been specifically developed for hospitalised frail older adults or served as a comprehensive guide. The range of inappropriate prescribing detectable using the OPTI-3S core list and its linkage to an easy and inpatient-validated frailty assessment (CFS) are significant strengths of this new consensus-based tool. Although the OPTI-3S core list proved to be a reliable and valuable tool to support prescribing optimisation in a cohort of frail community-dwelling older adults, further research should be conducted on a representative sample of hospitalised patients.