To assess the MPAI-4 and related measures’ measurement properties and the quality
of evidence supporting these results; and identify the interpretability and feasibility
of the MPAI-4 and related measures.
We conducted a systematic review according to COnsensus-based Standards for the selection
of health Measurement Instruments (COSMIN) guidelines. We searched nine electronic
databases and registries, and hand searched reference lists of included articles.
Two independent reviewers screened and selected all articles. From 605 retrieved articles,
48 were included.
Two independent reviewers appraised the evidence quality and rated the extracted Classical
Test Theory and Rasch results from each study.
We used meta-analysis and COSMIN's approach to synthesize measurement properties evidence
(insufficient, sufficient), and the modified GRADE approach to synthesize evidence
quality (very-low, low, moderate, high) by diagnosis (traumatic brain injury (TBI),
stroke) and setting (inpatient, outpatient). The MPAI-4 and its subscales are sufficiently
comprehensible (GRADE: very-low), but there is currently no other content validity
evidence (relevance, comprehensiveness). The MPAI-4 and its participation index (M2PI)
have sufficient interrater reliability for stroke and TBI outpatients (GRADE: moderate),
whereas interrater reliability between TBI inpatients and clinicians is currently
insufficient (GRADE: moderate). There is no evidence for measurement error. For stroke
and TBI outpatients the MPAI-4 and M2PI have sufficient construct validity (GRADE:
high) and responsiveness (GRADE: moderate-high). For TBI inpatients the MPAI-4 and
M2PI have mixed indeterminant/sufficient construct validity and responsiveness evidence
(GRADE: moderate-high). There is one study with mixed insufficient/sufficient evidence
for each MPAI-4 adaptation (21- and 22-item MPAI, 9-item M2PI) (GRADE: low-high).
Users can be most confident in using the MPAI-4 and M2PI in TBI and stroke outpatient
settings. Future research is needed on reliability, measurement error, predictive
validity and content validity of the MPAI-4 and its related measures across populations