Measurement properties of the Mental Health Literacy Scale (MHLS): A systematic review
Rouwida ElKhalil,Mohamad AlMekkawi,Matt O'Connor,Emad Masuadi,Moustafa Sherif,Messaouda Belfakir,Luai A Ahmed,Rami H Al-Rifai,Rasha Bayoumi,Iffat Elbarazi
DOI: https://doi.org/10.1016/j.ajp.2024.104214
Abstract:Background: Since its creation, the Mental Health Literacy Scale (MHLS) has been used worldwide in mental health literacy studies. Objective: This study aimed to systematically evaluate, summarize, and compare the measurement properties of MHLS validation studies. Methods: PsycINFO, CINAHL, ERIC, Scopus, Embase, MEDLINE, and PubMed databases were searched from May 30, 2015, to December 31, 2023. Peer-reviewed studies validating the MHLS and its measurement properties were included, irrespective of language, study population, and setting. Studies using the MHLS as an outcome measure, as a comparative instrument to validate another instrument, or using other MHL measures and grey literature was excluded. Results: Of the 685 search results, 16 studies were deemed eligible. The COnsensus-based Standards for the selection of health status Measurement INstruments (COSMIN) RoB criteria showed 15/15 studies exhibited 'Very Good' or 'Adequate' internal consistency, 3/6 reliability, 1/8 content validity, 14/14 structural validity, 6/7 hypothesis testing for convergent validity, 2/7 hypothesis testing for known-group validity, and 0/1 error measurement. The Cronbach's alpha ranged from 0.720 to 0.890, and the Intra-class Correlation Coefficient ranged from 0.741 to 0.99, while content validity was limited regarding the quality of evidence rating. The four-factor and unidimensional structures were 35.7 % and 28.6 %, respectively, the most common models. Conclusion: The MHLS exhibited strong evidence of construct validity and reliability, ensuring consistent and accurate evaluation of MHL and improving research credibility and generalizability. However, the low number of identical language versions of MHLS studies prohibited statistical pooling and quantitative summaries.