Measurement Properties of the Multiple Errands Test: A Systematic Review.

Abstract

To systematically review, summarize, and evaluate published evidence on measurement properties of real-world versions of the Multiple Errands Test (MET) following Consensus-based Standards for the Selection of Health Measurement Instruments guidelines. Four databases were searched in May 2019 using multiple variants of the name of the MET from 1991 onward following the publication of the original MET. We included peer-reviewed original research articles in English that provided data on measurement properties (reliability, validity, and responsiveness to change) on real-world versions of the MET in any clinical population. Data on the MET characteristics, study population, and evidence for each measurement property were extracted using predefined criteria. The review team critically appraised the methodological quality and rated the results from each study as sufficient (+), insufficient (-), or indeterminate (?). Data on each measurement property were pooled. Pooled results were rated as sufficient (+), insufficient (-), mixed (±), or indeterminate (?). The overall quality of evidence per measurement property was graded based on risk of bias, sample size, and consistency of results. The overall evidence for each measurement property was determined as high, moderate, low, or very low. We found 33 studies that provided data on measurement properties of real-world versions of the MET. Pooled results revealed high-quality evidence for interrater reliability and moderate-quality evidence for known-group validity. Limited support for other kinds of reliability and validity was found. This review suggests the MET should be used cautiously. Reasons for the limited psychometric support are discussed, the value of generic forms of the MET that do not require site specific adaptations is noted, and areas for further psychometric work are highlighted.