RESUMO
PURPOSE: The aim of this study is to define the significance of hyponatremia as a marker of anastomotic leakage after colorectal surgery. METHODS: All anastomoses in colorectal surgery performed at a single institution between July 2007 and July 2012 (n = 1,106) were retrospectively identified. Serum sodium levels and leukocyte values measured when an anastomotic leak was diagnosed by CT scan and/or surgical reintervention (n = 81) were compared to the values preferably on postoperative day 5 in the absence of an anastomotic leak (n = 1,025). RESULTS: The leak rate in anastomoses of the rectum was 9.0 %, while the leak rate of the other anastomoses was 5.4 %. Mean serum sodium level was 138.8 mmol/l in the group with an anastomotic leak and 140.5 mmol/l in the group without. Hyponatremia (<136 mmol/l) was present in 23 % of patients in the group with an anastomotic leak and in 15 % in the group without (p < 0.001). In multivariate analysis, leukocytes and serum sodium level remained as significant markers of an anastomotic leak. As a marker of an anastomotic leak, hyponatremia had a specificity of 93 % and a sensitivity of 23 %, while the presence of either leukocytosis or hyponatremia had a sensitivity of 68 %, a specificity of 75 %, a positive predictive value of 18 %, and a negative predictive value of 97 %. CONCLUSIONS: Hyponatremia could be a specific and relevant marker of anastomotic leakage after colorectal surgery. If hyponatremia and leukocytosis are present after colorectal surgery, anastomotic leakage should be suspected and a CT scan with rectal contrast dye is recommended.
