How a Clinician-Designed Digital Tool Improves Adherence to Measurement-Based Care and Clinical Outcomes in Psychiatry
Y. Park, A. Vian-Lains, R. Frontini, C. Costa, S. Baptista, L. Nedel, F. Silva, A. Shahdid, N. KennedyIntroduction
Measurement-based care (MBC) uses patient-reported outcomes to guide treatment. Despite its robust evidence in improving engagement and outcomes it remains under-implemented due to time and workflow barriers. Tailored digital tools may enhance feasibility by automation of assessments and integration into workflows.
Objectives
To describe the design, implementation and clinical outcomes of Psymap, a clinician-developed software supporting MBC implementation.
Methods
Psymap is a web-based application that evolved from a Python script originally developed and refined over two years by one of the team’s psychiatrists. It streamlines MBC by integrating sociodemographic data, patient-reported outcome measures, and clinician-administered assessments into a single digital platform with automated scoring algorithms and interpretive text to support clinical evaluation. Psymap was introduced over 15 months in a private outpatient service. Patients completed validated digitized assessments. The tool produced patient-friendly summaries (Figure 1), automated documentation, and stored data securely. Baseline-to-endpoint (12 months or last observation) changes in PHQ9 and GAD7 were analysed. Patients who completed Psymap assessments at ≥90% of their visits were classified as High-Frequency Assessment (HFA) and the remainder as Low-Frequency Assessment (LFA). Kaplan–Meier and one-way ANOVA analyses compared rates of response (≥50% PHQ9 reduction) and remission (PHQ9<5), time to remission and mean score change across groups.
Results
728 patients completed 1,798 assessments; 349 with ≥2 assessments were analysed. Psymap use at first visit rose from 37.7% to 96%. Higher use correlated with greater reductions in PHQ9 and GAD7 scores (r=–0.16, r=–0.18; p<.001). The HFA group (n=153) showed higher response (48% vs 33%) and remission (33% vs 19%) rates (χ
2
=8.9, χ
2
=8.1; p<.01). Kaplan-Meier analysis showed response and remission curves differed significantly (log-rank=5.9 and 6.0; p<.05); mean time to response was 80 vs 140 days. (Figure 2) Mean time to remission was 210 days for the HFA group, while the LFA group did not achieve 50% response. One-way ANOVA showed greater PHQ9 (–5.67 vs –4.08; F=4.55, p<.05) and GAD7 (–6.01 vs –3.89; F=9.16, p<.01) reductions in HFA vs LFA. (Figure 3) No adverse events were reported. Feedback highlighted improved clinician workflow and patient engagement.
Image 1: Long description.