Mobile Health Apps for Systemic Lupus Erythematosus and Lupus Nephritis

A Critical Appraisal

Akshara Ramasamy; Poojitha Dugyala; Chandra Mohan

Disclosures

Arthritis Res Ther. 2022;24(110) 

In This Article

Methods

Two reviewers (AR and PD) independently searched the App Store (iOS) and USA Google Play Store (Android) to screen, review, and abstract eligible mHealth apps. Our search terms for the systemic lupus erythematosus screening included lupus and SLE. For the lupus nephritis screening, our search terms included nephritis, kidney, and CKD. Originally, "kidney" and "CKD" were not part of our search strategy. However, the search term "nephritis" did not yield any apps designed for LN patients, so we increased the scope of search by adding "kidney" and "CKD." We then reviewed the abstracted data to remove any duplicates. An Apple device (iPhone X), an Android device (Xiaomi Redmi Note 9 Pro), and an Android emulator (Pixel 4 API 30) were used to complete the screening. The search was done within the following time period: 04/05/2021–07/08/2021. The numbers of apps that were initially identified and subsequently assessed are summarized in the PRISMA diagrams in Figures 1 and 2.

Figure 1.

PRISMA diagram outlining all SLE apps that were initially identified and subsequently excluded or included for analysis. This figure follows the flow of the SLE app identification and selection process. The initial database search yielded 841 applications, with 495 apps from the Google Play Store and 346 from the App Store. After all 841 apps underwent a title and app description screening, 805 apps were excluded due to their irrelevance to SLE patients. This process left 36 apps eligible for a full app screening. For the following reasons, 19 applications were eliminated from the second phase of screening: there were 6 duplicates (the same app appeared in more than one search), 5 apps did not function, and 8 apps were unrelated to the purpose of this review. Following the full app screening, 17 apps were ultimately included in the final analysis

Figure 2.

PRISMA diagram outlining all LN apps that were initially identified and subsequently excluded or included for analysis. This figure follows the flow of the LN app identification and selection process. Our initial database search yielded 1152 applications, with 742 apps from the Google Play Store and 410 from the App Store. After all 1152 apps underwent a title and app description screening, 1148 apps were excluded due to their irrelevance to LN patients and the purpose of this review. This process left 4 apps eligible for a full app screening. Due to duplicates, 2 applications were eliminated from the second phase of screening. Following the full app screening, 2 apps were ultimately included in the final analysis

We included apps in English that were designed specifically for SLE and LN patients. There were no restrictions on app cost. Apps were excluded if they had a target audience different from SLE/LN patients or lacked function. For apps that satisfied our inclusion criteria but had functional problems, attempts were made to contact the developer to ensure that we were not excluding any eligible apps. Other reasons for exclusion are detailed in Figures 1 and 2.

Apps with vague functions were first discussed between AR and PD and later adjudicated by a third reviewer (CM) to determine inclusion. Conflicts were also resolved by the third reviewer. A Google Sheets spreadsheet was used to produce a final list of the included apps, with information about the developer, price, app version, and app content retrieved from each app.

The Mobile App Rating Scale (MARS), an objective and reliable technique for assessing app quality, was used to evaluate the included mHealth apps.[13] The MARS scale is categorized into five major areas of criteria, including four objective quality scales: engagement, functionality, aesthetics, and information, as well as one subjective quality scale, which were combined into a 23-item questionnaire. A 5-point scale (1, inadequate; 2, poor; 3, acceptable; 4, good; 5, excellent) is used to rate each of the 23 items. An option of "not applicable" was provided in instances where an item might not be applicable to all apps. Individual mean ratings of engagement, functionality, aesthetics, and information quality were combined to provide an overall mean app quality score. The subjective quality portion of the MARS is optional; therefore, it was excluded to ensure objectivity. In any circumstances where there was a disagreement (a difference of more than 2 points in any of the individual MARS subscale mean scores), a third reviewer was to be consulted (CM). However, the two reviewers had a high level of agreement, with ratings never deviating by more than the 2-point threshold. The reviewers also watched the MARS training video provided by the creators of the scale. If the reviewers needed clarification for any MARS elements, they first referred to the training video, and any lingering questions were discussed among the reviewers.

processing....