APACHE II score: a useful tool for risk assessment and an aid to decision-making in emergency operation for bleeding gastric ulcer.

BACKGROUND: Operating for bleeding gastric ulcer remains controversial. Gastric resection bears a higher surgical risk while limited operation may result in more postoperative hemorrhage. There has been little discussion of effective risk assessment of patients. The aim of this study is to define surgical risk by using the APACHE II scoring system, and to determine optimal management. STUDY DESIGN: Records from October 1990 to December 1996 were retrospectively reviewed for patients (n=101) with bleeding gastric ulcer who had undergone emergency operation after failed endoscopic therapy. Mortality rates were examined according to different APACHE II scores, and the surgical risk was defined. From January 1997 to December 1997, 35 consecutive patients were enrolled for prospective study. Partial gastric resection (PGR) was performed for patients with huge ulcers (>2 cm) and for low-risk patients with ulcers at the antrum or angularis, while limited operation (oversewing or excision of bleeding ulcer) was reserved for others. The results were compared with the retrospective study. RESULTS: In the retrospective study, the mortality rates for the group with a score < 15 and > or = 15 were 5% (3 of 63) and 58% (22 of 38), respectively (p < 0.05). In the group with a score < 15, PGR was performed on 27 patients, and one died. For those patients with a score > or = 15, PGR carried a lower mortality than limited operation, although this was not statistically significant (47% vs 65%). Limited operation resulted in an overall rate of 22% postoperative hemorrhage and 12% reoperation rate, in which all patients with a score > or = 15 died. In the prospective study, the mortality rates in those scoring <15 and > or = 15 were 6% and 50%, respectively. This is not significantly different than the retrospective study. However, the rate of postoperative hemorrhage was diminished (5%). CONCLUSIONS: APACHE II score is a useful tool for assessing risk in patients with bleeding gastric ulcer. The mortality is minimal in those with a score <15, and PGR can be performed with low risk. Although high-risk patients have dreadful outcomes, limited operation cannot improve them if postoperative hemorrhage occurs. Decision making in emergency operation for such patients should be based on the ulcer conditions and the patient's hemodynamic status.