Polypharmacy, often defined as the concurrent use of 5 or more medications, is a growing concern in health care, affecting a significant proportion of adults [1]. The more medications a patient takes, the greater the likelihood that one or more may be potentially inappropriate medications (PIMs)—medications for which the potential harms outweigh the benefits [2]. The use of PIMs contributes to adverse drug events, reduced quality of life, and increased health care costs [3], highlighting an essential need for strategies to optimize appropriate medication use [4].

Despite rising rates of polypharmacy and adverse drug events among all age groups [5-7], existing PIM guidelines predominantly address older adults (aged ≥65 years) [8-10]. Guidelines such as the American Geriatrics Society Beers Criteria were initially developed for this population, as they are more likely to experience polypharmacy and drug-related harms [8]. For example, one in 4 community-dwelling older adults in Canada takes 10 or more different classes of medications [11], which is associated with adverse drug reactions, impaired balance and cognition, and increased hospitalizations [12]. However, the accumulation of medications leading to polypharmacy often begins earlier in life due to the increasing prevalence of chronic disease and the clinical management of patients with multiple comorbidities [1]. As a result, adults of all ages may face unnecessary risks from PIMs [13,14].

Existing PIM criteria for older adults cannot be generalized to younger populations. Many medications are considered inappropriate only under specific conditions—such as prolonged use, the presence of certain comorbidities, or interactions with other drugs—or become problematic primarily due to age-related physiological changes or cognitive vulnerability [13]. Conversely, some medications may pose inappropriate risks for adults when considering individual and clinical circumstances, regardless of age. For example, the prolonged use of proton pump inhibitors may increase the risk of bone fractures and nutrient deficiencies in adults of any age, particularly when used without clear indication or beyond recommended durations [15,16]. A consensus list adapted to the clinical realities and risk profiles of younger and middle-aged adults is therefore needed to ensure relevance and utility in practice. Addressing this gap could proactively improve medication safety by preventing PIM exposure earlier in life and enabling deprescribing before adverse events occur.

This study employs a modified Delphi method—adapted to better reflect the clinical complexity of medication appropriateness decisions—a structured, iterative process used to achieve consensus among a panel of experts [17,18]. This approach is particularly effective for addressing complex or emerging issues where evidence is limited. In a Delphi process, experts participate in multiple rounds of confidential surveys, providing input, reviewing group feedback, and refining their opinions until consensus is reached [17-20]. By engaging a diverse panel of international experts, this project aims to develop a consensus-based list of PIMs relevant to adults aged 18 to 65 years, thereby extending the scope of deprescribing efforts beyond geriatric populations and promoting safer and more appropriate medication use across adulthood.

The primary objective of this study is to develop a consensus-based list of PIMs for adults aged 18 to 65 years using a modified Delphi process. This list will inform the creation of a clinical guideline for this population. By focusing on common medications with the highest potential for harm or lack of utility, this study seeks to ensure that younger patients benefit from proactive deprescribing strategies, complementing existing recommendations for older adults and addressing a critical gap in the management of polypharmacy.

The initial list of PIMs was developed through a focused literature search and narrative synthesis, drawing on existing deprescribing guidelines, national and international PIM criteria (eg, Beers Criteria, STOPP/START [Screening Tool of Older Persons’ Prescriptions/Screening Tool to Alert to Right Treatment], FORTA [Fit for the Aged], Prescrire, etc) [8-10,28,29], and peer-reviewed publications focusing on medication safety in adults aged 18 to 65 years. To ensure clinical relevance and completeness, several experts with different professional backgrounds contributed sequentially to the review and refinement of the draft list, including steering committee members, 2 licensed pharmacists, and a hospital-based general internist. This process involved the addition of medications or classes not previously captured, clarification of subcategories (eg, formulation or indication), and identification of scenarios where appropriateness may vary depending on the clinical context.

The resulting list (see Multimedia Appendix 1) was then pilot-tested among 3 individuals (EF, NS, and EGM) and further adapted based on feedback from this pilot phase to ensure usability and clarity for the Delphi panel. It forms the basis for the round 1 questionnaire in the Delphi process.

To minimize participant attrition, participants were informed of the study timeline and expected commitment prior to enrollment. Surveys were designed with several features to reduce response burden: sections were divided to allow progress to be saved and resumed, items were grouped by therapeutic area, an estimated completion time was provided at the outset, and the survey was optimized for mobile access. Panelists also had the option to select “Insufficient knowledge to evaluate” for any item they did not feel comfortable assessing. Two reminder emails were sent during each round to encourage timely and continued participation.

To preserve the integrity of the Delphi process, only participants who complete each round will be allowed to proceed to subsequent rounds. No data imputation will be performed for participants who drop out in later rounds to avoid introducing bias or uncertainty from imputed values. This approach also simplifies analysis, reducing the risk of misinterpretation [18,31].

Ethics approval has been obtained from the Research Ethics Board of the Research Institute of the McGill University Health Centre in Montréal, Québec (approval number 2025‐11172). Prior to participation, all panelists will be provided with an informed consent form outlining the study’s purpose, procedures, risks, benefits, and confidentiality measures. Participants will be required to review and sign the informed consent form before receiving access to the first survey round.

All data collected during the study will be stored and managed using REDCap hosted at McGill University [30]. Identifying information (eg, names and email addresses) will be stored separately from survey response data and will only be accessible to authorized members of the research team. This information will be used exclusively for administrative purposes, such as sending reminders and sharing round-specific feedback.

Data will be coded prior to analysis and dissemination. The Delphi process will remain anonymous among participants; panelists will not be informed of the identities or individual responses of others. Feedback between rounds will be presented in aggregate form using deidentified data (eg, medians and IQRs). At no point will individual responses be linked to identifying information in any communications with participants or in any published reports or presentations.

All study data will be retained for a period of 7 years, following study completion, in accordance with institutional guidelines. Only authorized members of the study team will have access to the study key and identifiable data.

This study was reported in accordance with the ACCORD (Accurate Consensus Reporting Document) guideline for health care–based consensus studies [20]. An ACCORD guideline checklist for reporting consensus methods is provided in Checklist 1, which aided in the reporting of this study protocol.

The tentative timeline for this Delphi study spans 5 months (Figure 1).

Month 1: The initial list of PIMs will be generated, and recruitment for the Delphi panel will be conducted.

Months 2 to 4:

Month 5: Comprehensive data analysis will be completed, and the results will be compiled into a final report. The findings will be disseminated to stakeholders and will support subsequent guideline development.

This timeline ensures a structured and efficient progression through each phase of the Delphi process while allowing sufficient time for analysis and feedback integration. The exact timeline outlined for this study is subject to change based on participant availability, unforeseen delays, and the iterative nature of the Delphi process.

This study is anticipated to generate a consensus-based list of PIMs for adults aged 18 to 65 years, addressing a critical gap in polypharmacy research. While existing PIM criteria have historically focused on older adults [8-10,29], no prior effort has systematically identified inappropriate medications for younger populations, despite the common use of many potentially problematic medications in this age group [13]. By identifying medications that warrant reassessment in adults below 65 years, this research promotes a more comprehensive, proactive, and inclusive approach to safer medication use and deprescribing across the adult lifespan.

The resulting list is intended for use by clinicians, including physicians, nurse practitioners, and pharmacists, who may not have formal training in deprescribing outside of geriatric care. It aims to serve as a practical reference to help identify candidate medications for review or discontinuation in adult patients. The 4 evaluation criteria considered throughout the Delphi process (ie, importance to guideline inclusion, balance of benefits and harms, relevance to patient preferences and values, and availability of alternatives) will provide structured context for each medication’s inclusion in the final list, with more detailed evidence-based rationale to be developed during the subsequent guideline development phase.

The findings will serve as a foundation for developing an evidence-based clinical guideline, following established principles for guideline development, to support health care providers in assessing and deprescribing PIMs in younger and middle-aged adults [35-40]. The guideline development is anticipated to follow the WikiGuidelines methodology, a collaborative, iterative, and transparent approach that integrates expert consensus, current evidence, and stakeholder input to ensure clinical utility and adaptability over time [40]. For each prioritized PIM, a dedicated section will provide concise, evidence-informed recommendations, presented in a modular format to facilitate integration into diverse clinical settings.

In addition to the overarching guideline, individual deprescribing documents may be developed to further support implementation, tailored to specific clinical contexts or care providers. Ultimately, this work aims to enhance medication safety and improve health outcomes by enabling clinicians and patients to identify PIMs across the lifespan, thereby reducing the risks associated with polypharmacy, including adverse drug reactions, medication burden, and diminished quality of life among adults of all ages.

Several methodological considerations should be acknowledged. First, recruitment through established deprescribing networks may introduce self-selection bias, as participants may already be predisposed toward deprescribing. Second, although study materials were available in both English and French, the process was limited to these 2 languages, which may not fully capture the scope of perspectives across the international panel. Third, while consensus thresholds used in this study are consistent with published Delphi methodology, no universally accepted statistical cutoffs exist for Delphi consensus, and the chosen criteria involve an inherent degree of judgment. The detailed interpretation of findings and their limitations will be addressed in the forthcoming paper.