PURPOSE: Different predicting mortality models (PMM) are available: APACHE I-III, MODS and SAPS. With the exception of SAPS, emergency (EMS) or elective surgery (ELS) is not incorporated in these models. We compared mortality of patients admitted to the surgical ICU after EMS versus ELS.
METHODS: 2601 post-operative SICU patients (ELS=1167 / EMS=1434) were analyzed for 30-day hospital mortality that had PMM. Fisher’s test compared mortality. The discrimination of the four score systems were assessed by the areas under the receiver operating characteristic curves (AUC). P>.001
RESULTS: All patient mortality =11.6%; EMS =16.3% and ELS = 5.7%. (P= 0.0001). Gender had no significant impact on outcome. Discrimination was satisfactory for all scoring systems when used to assess the patients undergoing ELS. Statistically significant difference between the APACHE III and MODS (0.73 vs. 0.62 [P<0.05] occurred, APACHE III having a better predictive value for patients undergoing ELS. Discrimination was good for all scoring systems when used to assess the patients undergoing EMS, the AUC values ranging from 0.69 to 0.80. There was a statistically significant difference between the APACHE I-III and SAPS when they were compared to MODS (0.80, 0.79, 0.77 vs. 0.69 respectively (P<0.01)); with MODS having a worse predictive value for patients undergoing EMS. All Models were satisfactory in predicting the mortality of patients who underwent ELS, with APACHE III being a superior predictive model to MODS. MODS yielded the worst predictive value when compared to the three other models for EMS patients.
CONCLUSIONS: Emergency surgery is an independent factor of outcome in surgical critically ill patients. Mortality of patients undergoing emergency surgery was approximately three times higher than for patients who underwent elective surgery. APACHE III is a superior predictive model to MODS for elective surgical critically ill patients.
CLINICAL IMPLICATIONS: Patients undergoing EMS carry a higher mortality than ELS patients. The mortality of EMS patients is better predicted by the APACHE I-III, and SAPS as opposed to MODS.
DISCLOSURE: The following authors have nothing to disclose: Adam Khader, Gary Deutsch, Denis Knobel, Garry Ritter, Corrado Marini, Rafael Barrera
No Product/Research Disclosure Information