RESUMO
BACKGROUND AND AIM: Medically refractory upper gastrointestinal hemorrhage (UGIH) is a complex clinical problem. Selection of patients suitable for surgery is difficult and often involves subjective clinical judgment. The simplified acute physiology score (SAPS) II is a validated predictor of mortality in the intensive care setting. Our aim was to assess the SAPS II score in patients with medically refractory UGIH who were referred for a surgical opinion. Patients were subsequently classed as 'accepted' or 'declined' for surgery and SAPS II scores were compared between these two groups. METHODS: From July 1996 to July 1999, patients referred for surgical intervention with UGIH were included (varices excluded). The SAPS II was calculated at the time of surgical referral. This was converted into a mortality probability using multiple regression analysis. Clinical outcome was defined as either survival to discharge or death while an inpatient. RESULTS: Ninety-nine patients were referred for surgical review. Sixty-five patients were accepted for surgery and 34 were declined. The mean SAPS II score for those who were accepted was 31.7 (mortality probability 0.16, actual mortality 15.4%, 10/65) and 30.2 (mortality probability 0.15, actual mortality 29.4%, 10/34) for those who were declined. CONCLUSION: The SAPS II scores were no different between the two groups. Observed mortality was consistent with mortality predicted in the operated group but twice that predicted in those where surgical intervention was declined. This suggests that clinical selection criteria for patients undergoing surgery for UGIH are inconsistent.