The DOSA model uses EHR input data from a patient’s Duke Health history initiated at the time of first opioid administration following inpatient admission. Upon meeting inclusion requirements, the model runs hourly and estimates the probability that the patient will experience an opioid induced respiratory depression (OIRD) event requiring rescue naloxone administration and provides recommendations for level of monitoring to prevent such events.
It was developed in 2021 and 2022 by the Department of Anesthesiology, Duke Institute for Healthcare Innovation, and approved by the Algorithm-Based Clinical Decision Support (ABCDs) committee in 2023.
Model Facts
View the Model Fact Sheet.