Remote Symptom Monitoring With Ecological Momentary Computerized Adaptive Testing: Pilot Cohort Study of a Platform...

Area of Science:

• Digital health
• Health informatics
• Psychometrics

Background:

• Remote patient-reported outcome measure (PROM) data capture offers valuable research and clinical insights.
• Frequent data collection, like ecological momentary assessment, is limited by patient burden from lengthy questionnaires.
• Computerized adaptive testing (CAT) algorithms can reduce this burden by selecting only relevant items.

Purpose of the Study:

• To determine the feasibility of ecological momentary computerized adaptive testing (EMCAT) for remote PROM administration.
• To assess EMCAT's effectiveness in reducing patient burden and facilitating high-frequency data capture via smartphone.

Main Methods:

• Developed an EMCAT web app using Concerto, an open-source CAT platform.
• Utilized item response theory models for CAT algorithms with a hand surgery PROM (patient evaluation measure).
• Piloted the app with 40 patients over 12 weeks, assessing symptoms thrice weekly, daily, or thrice daily.

Main Results:

• EMCAT significantly reduced PROM length (median 2 vs. 11 items) and completion time (median 8.8s vs. 1m 14s).
• Scores from EMCAT closely matched full-length PROM scores (mean error <0.01 on logit scale).
• Achieved a median response rate of 93% in the daily assessment group with high perceived usability (median 4.0/5.0).

Conclusions:

• EMCAT effectively reduces PROM assessment burden, enabling acceptable high-frequency remote data capture.
• Potential applications include studying symptom severity trajectories in research (e.g., post-surgery recovery).
• EMCAT can enhance clinical practice through remote patient monitoring and early intervention for concerning symptom trajectories.