Author:

Abstract:

Introduction Mobile technology is an increasing, influential source of information, resulting in a steadily growth of health apps. Health professionals are concerned about the quality of apps. A generally accepted, reliable tool for evaluating health apps is the Mobile Application Rating Scale (MARS). However, there is no standardized Dutch tool. This pilot study investigated the usefulness of a Dutch MARS for assessing the quality of pregnancy apps and determined its psychometric properties. Materials and methods The original 23-item MARS, consisting of four objective subscales (engagement, functionality, aesthetics, information) and one subjective dimension, was translated and adapted according to the WHO-guideline ‘Translation and adaptation of instruments’. Nineteen most downloaded pregnancy apps were evaluated independently by two raters at two time points, after following an online training module and pilot testing on another five pregnancy apps. MARS total/subscale scores were calculated. Standard error of measurement values (SEM) was used to assess absolute realibility. Inter- and intrarater reliability (Wilcoxon signed-rank tests, intraclass correlation coefficients) and internal consistency (Cronbach’s alpha coefficients) were determined. Results Moderate to excellent inter- (0.672–0.765) and intrarater reliability (0.794-0.921) were found for the total score and the subjective part (0.601-0.761 resp. 0.729-0.882). The absolute reliability showed low %SEMs, indicating good reliability. The reliability for the subscale scores varied from poor to good. Intraclass correlation coefficients (ICCs) of the interrater reliability ranged from 0.176 to 0.796 and from 0.276 to 0.921 for the intrarater reliability, with poor ICCs in the functionality subscale. However, the absolute reliability of this subscale showed good %SEM. Internal consistency was acceptable for all subscales and the subjective part (0.786–0.943), except for the functionality subscale (0.010–0.596). Conclusions Our results were similar to those in the original study, except for the functionality subscale. This may indicate that the Dutch MARS is a reliably tool for rating the quality of pregnancy apps. However, further research is needed to optimize the instrument (e.g. functionality subscale) and to evaluate its psychometric properties (e.g. validity) on a larger number of apps. Also the usability of the MARS by health professionals and the implementation in daily practice should be further explored.