Prognostic factors in the management of the open abdomen
Le résumé fourni par la source
In the management of critically ill surgical patients with an open abdomen, predicting their probability of developing a certain complication or outcome might aid in their safe management until definitive fascial closure can be achieved. The objectives of this research were to perform a systematic review identifying prognostic factors regarding mortality, definitive fascial closure, and complications, to then develop and validate multivariable prediction models. A retrospective cohort study was conducted in hospitalised patients across three tertiary hospitals (Cairns, Townsville and Royal Brisbane and Women’s) from January 2000 to December 2017. Overall, 57 independent potentially predictive variables were used to develop the models. Outcomes included peri-operative, in-hospital, ninety-day and five-year mortality, definitive fascial closure, entero-atmospheric fistula, intra-abdominal abscess, ventral hernia, anastomotic leak, and bowel resection after index operation. Chi-squared test was used for descriptive statistics. Univariate and multivariable logistic regression analyses were performed for the developmental data set. Receiver operator characteristics analyses with area under the curve (AUC) and 95% confidence intervals (CIs) were included for both the developmental and validation data sets. Thirty-one studies were included in the final synthesis of the systematic review. Five-hundred and forty-eight patients were managed with an open abdomen. Median age of patients was 58 (interquartile range: 32 to 83) years. Male to female ratio was 3:2. Prediction models were externally validated for peri-operative mortality (AUC: 0.77; 95% CI: 0.66-0.88), in-hospital mortality (AUC: 0.79; 95% CI: 0.70-0.88), ninety-day mortality (AUC: 0.74; 95% CI: 0.64-0.84), ventral hernia (AUC: 0.75; 95% CI: 0.65-0.85)) and entero-atmospheric fistula (AUC: 0.79; 95% CI: 0.67-0.92). External validation of these models in a setting outside Australia could strengthen their usefulness.
Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.