Late Breaking Abstract - Impact of increasing notification thresholds for remote respiratory monitoring in patients with chronic lung disease

Background: A motivation to expand remote patient monitoring (RPM) use is to allow for intervention earlier in the course of deterioration of chronic disease. A component of respiratory RPM is to algorithmically identify and selectively escalate patients with clinically significant changes in their respiratory status. Objective: The aim of this study is to determine the impact of increasing respiratory rate (RR) notification threshold values on the predictive accuracy of RPM to determine clinically significant respiratory changes. Methods: Chart and physiologic data from patients in a large pulmonary practice enrolled in respiratory RPM (Spire Health) was assessed for evidence of clinical deterioration following RR notifications. Simulations of multiple threshold settings were computed to evaluate impact on predictive value. Results: 40 patients were seen within 7 days of a ≥10% increase in daily RR above the patient9s baselines. Of these, 31 (77.5%) were found to have respiratory deterioration requiring medical intervention. When 15%, 20%, and 25% thresholds were simulated, clinical deterioration would have been found in 81.3%, 90.9%, and 100% of patients, respectively. However, these thresholds resulted in 18, 21, and 26 instances where physiologic deviations associated with clinical deterioration were not detected. Conclusion: Increasing the RPM system’s elevated daily RR notification threshold increased the proportion of true positives. However, this also increased the proportion of false negatives.