Predictive value of upper gastrointestinal studies versus clinical signs for gastrointestinal leaks after laparoscopic gastric bypass.

Author: Madan AK, Stoecklein HH, Ternovits CA, Tichansky DS, Phillips JC

Author affiliation: Department of Surgery, University of Tennessee Health Science Center, 956 Court Avenue, Room G210, Memphis, TN 38163, USA. amadan@utmem.edu

Publication date & source: 2007.02, Surg Endosc., 21(2):194-6. Epub 2006 Nov 21.

BACKGROUND: The utility of routine upper gastrointestinal (UGI) studies after laparoscopic Roux-en-Y gastric bypass (LRYGB) is a matter of great debate. Because the morbidity and mortality rates associated with an unrecognized postoperative leak are high after LRYGB, diagnosis of a postoperative leak earlier would be of benefit. Clinical signs, however, may predict the diagnosis of a postoperative leak more often. This study explored the hypothesis that UGI studies are more predictive than clinical signs for the early diagnosis of a postoperative leak after LRYGB. METHODS: All patients who underwent LRYGB at the authors' institution were included in this study. Charts were reviewed to examine immediate clinical signs (heart rate, temperature, and white blood cell count within the first 24 h), UGI studies, and clinical course. Sensitivity, specificity, positive predictive value, negative predictive value, and efficiency of clinical signs and UGI studies were calculated. RESULTS: This study included 245 patients with a 3% rate of leak. The positive and negative predictive value of UGI studies were 67% and 99%, respectively. Only an elevated white blood count had a better predictive value (100% for negative predictive value). The efficiency of UGI studies (98%) was better than that of heart rate (83%), white blood count (8%), or temperature (95%). CONCLUSIONS: According to our data, UGI studies are the most predictive of an early leak diagnosis. Clinical signs alone may not be as useful in predicting leaks early after laparoscopic gastric bypasses. Routine early postoperative UGI studies are a reasonable approach to predicting leaks after LRYGB.