Vital-sign Data Fusion Models for Post-operative Patients

Patients who undergo upper gastrointestinal surgery have a high incidence of post-operative complications, with up to 20% requiring readmission to the Intensive Care Unit (ICU) several days after surgery. Deterioration in these patients vital signs may be evident prior to adverse events. Failure to identify such deteriorations in a timely manner has led to the design of a clinical trial at the Oxford Cancer Hospital, in which ambulatory post-operative cancer patients are monitored using wearable sensors. A dataset comprising vital-sign data, acquired periodically from 128 post-operative patients by ward staff, was used to explore the statistical properties of the 5 main vital signs (breathing rate, heart rate, blood oxygen saturation, systolic blood pressure and temperature). Univariate and multivariate analyses of the trajectory of patients’ vital signs changes between their admission to the Upper Gastrointestinal Ward, mid-way through their stay, and near their time of discharge were used. A model of normality based on pre-discharge data from “normal” patients who had a “normal recovery” was constructed using a probabilistic method, and then tested with “abnormal” data from patients who deteriorate sufficiently after surgery to be re-admitted to the ICU. The vital-sign distributions from “normal” patients between admission to the Upper GI ward and subsequent discharge were found to be different, but no significant changes in the vitalsign distributions from halfway through their stay to the time of discharge were observed. The model of normality was able to identify patient deterioration when tested with “abnormal” test data. These preliminary results suggest that patients’ criticality can be assessed by looking at the distributions of their vital signs after their admission to the post-operative ward, following major surgery.