Authors (including presenting author) :
Cheung YS (1), Wong R (2), Lok HT (2), Lai PBS (2)(3)
Affiliation :
(1) Department of Surgery, North District Hospital, (2) Department of Surgery, Prince of Wales Hospital, (3) Department of Surgery, The Chinese University of Hong Kong
Introduction :
Operative risk assessment for emergency surgery could help clinicians in patient care and decision making. By classifying the estimated risk into different risk groups may help communication with patients and inform the operating team about the chances of post-operative intensive care unit support.
Objectives :
The aim of this study was to evaluate the feasibility of classifying patients into different risk group that may provide an objective range of 30-day mortality that can be easily understood by patients and relatives. The probability of post-operative intensive care unit admission in each group was also compared.
Methodology :
A retrospective study using patient registry from the Surgical Outcome Monitoring and Improvement Program (SOMIP) of Hospital Authority was performed. 30-day mortality risk of patients undergoing emergency operation captured by the SOMIP from July 2018 to June 2020 were estimated by the SOMIP risk model. Patients were then stratified into 4 subgroups according to their risk of 30-day mortality, namely Low risk (lowest 50 percentile), Medium risk (50-75 percentile) High risk (75-90 percentile) & Ultra-high risk (90 percentile - highest). The accuracy of the risk model was evaluated by area under the receiver operating characteristics curve. The mortality risk and probability of post-operative intensive care of each group were assessed by Kruskal Wallis Test & Chi-square Test where appropriate.
Result & Outcome :
8228 patients were included in this study. The rates of 30-day mortality and post-operative intensive care were 4% and 22.1% respectively. The median estimated risks of 30-day mortality of Low, Medium, High & Ultra-high risk group were 0.3%, 1.8%, 6.6% and 24% respectively (p-value < 0.001). The probability of post-operative intensive care admission of Low, Medium, High & Ultra-high risk groups was 5.7%, 24.5%, 43.8% and 65.1% respectively (p-value < 0.001). The area under curve of receiver operating characteristics curve in predicting 30-day mortality and ICU admission by the SOMIP risk model was 0.895 and 0.820 respectively.
In conclusion, the SOMIP risk model can be applied in patient undergoing emergency general surgery to provide an objective risk prediction. The estimated risk can then be classified into different group of varying 30-day mortality rate and post-operative intensive care unit admission rate. This can help the operating team in patient counselling and peri-operative care arrangement.
