Credibility of Minimal Clinically Important Differences for Otolaryngology Patient-Reported Outcome Measurements: A Scoping Review.

Abstract

Anchor-based approaches for determining a minimal clinically important difference (MCID) for patient-reported outcome measures (PROMs) are critical for detecting meaningful changes in clinical status. However, the methodologies for MCID determination vary across the literature and have been found to be inconsistent. Our objective is to evaluate the credibility of anchor-based MCIDs in otolaryngology. PubMed, Embase, and Web of Science. Studies identifying novel anchor-based MCIDs for PROMs in otolaryngology from inception to November 2, 2021, in a previous review were included. A systematic search of PubMed, Embase, and Web of Science was conducted for original articles on the same topic from November 3, 2021, to January 4, 2024. The credibility of MCIDs was assessed using a previously developed instrument with five main criteria, including anchor interpretability and relevance to patients, correlation between the anchor and the patient-reported outcome, sample size, and determination if the selected anchor threshold is small and meaningful. In total, 43 studies reported anchor-based MCIDs in otolaryngology. A total of 74 MCIDs were identified across subspecialties, most commonly in rhinology (38%) and otology (26%). Only 25% (18/74) were deemed credible, whereas 48% (35/74) had low credibility. In 34% (24/74) of cases, correlations between the anchor and PROMs were not reported. Most MCIDs (70%) were calculated using differences in their anchors that were considered sufficiently small and clinically relevant to patients. All MCIDs were calculated using at least an anchor-based method, with 50 (58%) relying on the mean change anchor-based method. A small subset of MCIDs for PROMs in otolaryngology were found to be credible. These findings highlight the need for more rigorous methods and standardized guidelines for defining MCIDs, ensuring that criteria, such as sufficient sample size and appropriate anchor thresholds, are met.