Linking Patient-Reported and Clinician-Assessed Wound Status via Chatbot-Based Digital Surveillance for Wound...

Area of Science:

• Digital Health
• Wound Care Technology
• Patient-Reported Outcomes

Background:

• Digital wound monitoring via smartphones is feasible, but patient perception of acute wound changes post-discharge is understudied.
• Understanding patient self-assessment accuracy is vital for reliable digital health systems in wound care.

Purpose of the Study:

• Evaluate diagnostic performance of patient-reported infection symptoms for predicting callbacks in a chatbot wound monitoring platform.
• Identify wound features associated with patient-reported infection and compare acute vs. chronic wound populations.

Main Methods:

• Retrospective observational study using a chatbot-based monitoring system (Line app) at a tertiary medical center.
• Adults with acute or chronic wounds submitted daily symptom reports and photos; patient self-report of infection was the predictor.
• Logistic regression and GEE models analyzed data, accounting for wound type, age, and sex.

Main Results:

• 159 patients (88 acute, 71 chronic wounds) submitted 4764 photos.
• Diagnostic accuracy for acute wounds (AUC 0.702) was lower than for chronic wounds (AUC 0.907).
• Redness associated with infection in acute wounds; redness and skin darkening in chronic wounds.

Conclusions:

• Patient perception and reporting of infection symptoms differ significantly between acute and chronic wound populations.
• Lower diagnostic accuracy in acute wounds highlights challenges in patient self-assessment due to limited experience.
• Digital wound monitoring systems require tailoring to wound chronicity and patient experience, potentially using AI.