Abstract
Abstract
Background
Shift-work nurses experience substantial variability in work schedules, health behaviors, and recovery patterns, which traditional retrospective surveys fail to capture due to recall bias. Although mobile health (mHealth) tools offer ecological momentary data capture, existing off-the-shelf survey platforms lack shift-synchronized notification logic and impose excessive cognitive friction on fatigued clinicians, limiting their applicability in nursing research.
Objective
This study aimed to propose and demonstrate a transferable, dual-perspective evaluation framework for mHealth data collection tools in high-burden occupational settings, using the newly developed “Nurses’ Work-Life and Health” app as a tailored exemplar.
Methods
A 3-phase, user-centered iterative design was used: (1) needs assessment, (2) app development with alpha testing (n=5) and beta testing (n=16), and (3) a 14-day feasibility and process evaluation. The feasibility study evaluated 5 shift-work nurses who completed daily near–real-time journal entries over 14 consecutive days, capturing shift characteristics, sleep, nutrition/hydration, physical activity, and acute fatigue/stress. To complement end-user evaluation (n=5), an expert nurse researcher participated in a follow-up semistructured interview assessing methodological rigor. Evaluation was guided by the technology acceptance model and system usability scale. Objective system logs and subjective surveys were integrated to analyze adherence, completion, and user burden.
Results
Phase 1 needs assessment identified key functional requirements, including shift-synchronized notifications, low cognitive burden timeline entries for postshift fatigue, and automated time-stamping to verify contemporaneous logging. In phase 2, the app demonstrated high usability and acceptance, with total mean scores of 3.47 (SD 0.53) in alpha testing and 3.48 (SD 0.55) in beta testing (range 1‐4). In the phase 3 feasibility study, the app demonstrated 100% retention and 94.3% adherence (mean 13.2, SD 1.1 d). The data entry completion rate was 88.5%, with an average entry time of 3.4 minutes. Participants reported high overall satisfaction and low user burden, reflected by mean scores of 3.82 and 3.65, respectively. Time-stamped logs supported the feasibility of near–real-time data capture without disrupting clinical workflows. Qualitative feedback from the expert nurse researcher highlighted the app’s methodological strengths, including its potential to reduce recall bias and support for self-monitoring, while also identifying areas for refinement.
Conclusions
By integrating subjective end-user perceptions, objective usage logs, and qualitative scrutiny from an expert researcher, this study established a transferable evaluation framework for mHealth research instruments. The findings indicate that shift-tailored mobile platforms can achieve high temporal fidelity data capture in complex clinical settings, providing a replicable evaluation protocol for digital health investigators prior to full-scale deployment.