Using a ResearchKit Smartphone App to Collect Rheumatoid Arthritis Symptoms From Real-World Participants: Feasibility Study

Abstract

Background: Using smartphones to enroll, obtain consent, and gather self-reported data from patients has the potential to enhance our understanding of disease burden and quantify physiological impact in the real world. It may also be possible to harness integral smartphone sensors to facilitate remote collection of clinically relevant data.

Objective: We conducted the Patient Rheumatoid Arthritis Data From the Real World (PARADE) observational study using a customized ResearchKit app with a bring-your-own-device approach. Our objective was to assess the feasibility of using an entirely digital approach (social media and smartphone app) to conduct a real-world observational study of patients with rheumatoid arthritis.

Methods: We conducted this observational study using a customized ResearchKit app with a bring-your-own-device approach. To recruit patients, the PARADE app, designed to guide patients through a series of tasks, was publicized via social media platforms and made available for patients in the United States to download from the Apple App Store. We collected patient-reported data, such as medical history, rheumatoid arthritis-related medications (past and present), and a range of patient-reported outcome measures. We included in the assessment a joint-pain map and a novel objective assessment of wrist range of movement, measured by the smartphone-embedded gyroscope and accelerometer.

Results: Within 1 month of recruitment via social media campaigns, 399 participants self-enrolled, self-consented, and provided complete demographic data. Joint pain was the most frequently reported rheumatoid arthritis symptom to bother study participants (344/393, 87.5%). Severe patient-reported wrist pain appeared to be inversely linked with the range of wrist movement measured objectively by the app. At study entry, 292 of 399 participants (73.2%) indicated a preference for participating in a mobile app-based study. The number of participants in the study declined to 45 of 399 (11.3%) at week 12.

Conclusions: Despite the declining number of participants over time, the combination of social media and smartphone app with sensor integration was a feasible and cost-effective approach for the collection of patient-reported data in rheumatoid arthritis. Integral sensors within smartphones can be harnessed to provide novel end points, and the novel wrist range of movement test warrants further clinical validation.

Keywords: mobile applications; mobile phone; patient-reported outcome measures; rheumatoid arthritis; smartphone.

Figure 1

Wrist range of movement (ROM) exercise. (A) Instructions for the wrist ROM exercise provided to participants via the app. (B, C) Wrist ROM preliminary validation based on experimental test data. (B) Relative change in phone orientation (blue arrow) at full wrist extension with respect to initial orientation (black arrow). (C) Relative change in phone orientation at full wrist flexion. In both examples, the video frame from the task performance at the moment of measurement is displayed as a reference for visual inspection. Green dots correspond to phone orientation at previous measurements.

Figure 2

Patient Rheumatoid Arthritis Data From the Real World (PARADE) app study recruitment. *Defined as those completing all demographic questions.

Figure 3

Geographic distribution of participants within the United States.

Figure 4

Joint pain map. Percentage of patients reporting any pain in each of 55 joints at week 1 (n=336).

Figure 5

Association between patient wrist range of movement and reported level of wrist pain from the joint pain map assessment at week 1. Boxes represent the upper and lower quartiles; the line inside each box represents the median; the whiskers extending vertically from the boxes represent the range.

Figure 6

Proportion of patients retained within the study over 12 weeks.