BACKGROUND: Understanding risk factors of surgical site infections (SSIs) in gastrectomy is important to provide the best treatment for the patients with gastric cancer. METHODS: This is a retrospective observational study using the medical records of 790 patients with gastrectomy from 2005 through 2009. SSIs were classified into incisional SSIs (iSSIs) and organ/space SSIs (o/sSSIs). RESULTS: iSSIs and o/sSSIs were detected in 41 (5.2%) patients and 68 (8.6%) patients, respectively. Open surgery was the only independent risk factor (p = 0.028) for iSSIs, while open surgery (p = 0.004), concurrent splenectomy (p < 0.001), operative time ≥220 min (p = 0.009), preoperative body mass index ≥20.8 kg/m2 (p = 0.004) and male gender (p = 0.028) were the independent risk factors for o/sSSIs. We created a risk model for o/sSSIs using these independent risk factors. The C-index model discrimination was 0.84 (p < 0.001), and the calibration of the models demonstrated a linear correlation between the predicted and observed probability. CONCLUSION: We reported the risk factors of SSIs for gastrectomy. The risk model developed in this study for o/sSSIs pertaining to gastric cancer surgery would contribute to provide guidance for the development of best practices.
BACKGROUND: Understanding risk factors of surgical site infections (SSIs) in gastrectomy is important to provide the best treatment for the patients with gastric cancer. METHODS: This is a retrospective observational study using the medical records of 790 patients with gastrectomy from 2005 through 2009. SSIs were classified into incisional SSIs (iSSIs) and organ/space SSIs (o/sSSIs). RESULTS: iSSIs and o/sSSIs were detected in 41 (5.2%) patients and 68 (8.6%) patients, respectively. Open surgery was the only independent risk factor (p = 0.028) for iSSIs, while open surgery (p = 0.004), concurrent splenectomy (p < 0.001), operative time ≥220 min (p = 0.009), preoperative body mass index ≥20.8 kg/m2 (p = 0.004) and male gender (p = 0.028) were the independent risk factors for o/sSSIs. We created a risk model for o/sSSIs using these independent risk factors. The C-index model discrimination was 0.84 (p < 0.001), and the calibration of the models demonstrated a linear correlation between the predicted and observed probability. CONCLUSION: We reported the risk factors of SSIs for gastrectomy. The risk model developed in this study for o/sSSIs pertaining to gastric cancer surgery would contribute to provide guidance for the development of best practices.