Categorization and recommendations for outcome measures for lower limb absence by an expert panel.

Background: Understanding the psychometric strengths and limitations of outcome measures for use with people with lower limb absence (LLA) is important for selecting measures suited to evaluating patient outcomes, answering clinical and research questions, and informing health care policy. The aim of this project was to review the current psychometric evidence on outcome measures in people with LLA to determine which measures should be included in a stakeholder consensus process.

Methods: An expert panel was assembled, and a 3-stage review process was used to categorize outcome measures identified in a systematic literature review into 3 distinct categories (recommended for measures with better than adequate psychometric properties; recommended with qualification; and unable to recommend). Panelists were asked to individually categorize measures based on results of a systematic review of identified measures' psychometric properties. Each measure's final categorization was based on ≥70% agreement by all panelists.

Results: No outcome measure attained the ≥70% consensus threshold needed to achieve a rating of "recommend." Hence, panelists suggested combining "recommend" and "recommend with qualifications" into a single category of "recommend with qualifications." Using this approach, consensus was reached for 59 of 60 measures. Consensus could not be reached on 1 outcome measure (socket comfort score). Thirty-six outcome measures were categorized as "unable to recommend" based on available evidence; however, 23 (12 patient-reported measures and 11 performance-based measures) demonstrated adequate psychometric properties in LLA samples and were thus rated as "recommend with qualification" by the expert panel. The panel of experts were able to recommend 23 measures for inclusion in the subsequent stakeholder review. A key strength of this process was bringing together international researchers with extensive experience in developing and/or using LLA outcome measures who could assist in identifying psychometrically sound measures to include in a subsequent stakeholder consensus process.

Conclusion: The above categorizations represent the current state of psychometric evidence on outcome measures for people with LLA and hence may change over time as additional research becomes available. The results will be used to achieve wider consensus from clinicians, health policymakers, health clinic managers, researchers, and end users (i.e., individuals with LLA) on outcome measures for the International Society of Prosthetics and Orthotics lower limb Consensus Outcome Measures for Prosthetic and Amputation ServiceS.