Fig. 1: DETECT study results23.
a–d, The DETECT study evaluated the added utility of wearable data to symptom questionnaires. a, Screenshot of symptom questionnaire on the DETECT study platform. b, The algorithm based on self-reported symptoms alone had a ROC-AUC value of 0.72. c,d, Passive wearable data from any wrist sensor connected to the Apple HealthKit or Google Fit (c) was then added to the subjective symptom questionnaire, and demonstrated a ROC-AUC value of 0.80 (d). CI, confidence interval; SE, sensitivity; SP, specificity; PPV, positive predictive value; NPV, negative predictive value.
