To investigate the main risk factors of surgical site infection after coronary artery bypass grafting (CABG), and to construct a quantitative nomogram risk prediction model, and put forward targeted management strategies.

Total of 302 patients with CABG admitted to Yancheng Third People’s Hospital Affiliated to Nanjing Medical University from May 2015 to may 2019 were retrospectively summarized as the model group. The clinical data of infection and non-infection patients were compared, including gender, age, basic disease history, body mass index, American Society of Anesthesiologists (ASA) score, routine skin preparation, unreasonable use of perioperative antibiotics, operation time > 4 h, operating room visit and continuous use of the same operating room, and the main risk factors were screened by multivariate Logistic regression analysis, and the risk prediction model of nomogram was established. A total of 211 patients with CABG from June 2019 to December 2020 were enrolled as the validation group and accepted the infection management strategy. Finally, receiver operating curve (ROC) was used to evaluate the efficacy of nomogram model for infection in model group and validation group.

In model group, 11 patients with infection (3.64%, 11/302) were diagnosed. Univariate analysis showed that patients with infection were older, with higher proportions of basic disease history (hypertension and diabetes) and obesity, higher ASA score, higher rates of routine skin preparation, irrational use of antibiotics during perioperation, operation time > 4 h, operating room visitation and continuous use of the same operating room than non-infected patients, with significant differences (all P < 0.05). Logistic regression analysis showed that elder (OR = 1.58, 95%CI: 1.12-2.53, P = 0.011), basic disease history (OR = 2.63, 95%CI: 2.12-3.06, P = 0.001), unreasonable use of antibiotics during perioperation (OR = 2.01, 95%CI: 1.55-2.69, P = 0.002), operation time > 4 h (OR = 3.11, 95%CI: 2.68-3.59, P = 0.001) and operating room visitation (OR = 1.24, 95%CI: 1.01-1.85, P = 0.024) were the main risk factors of surgical site infection after CABG, all with significant differences. The nomogram model was established by R software according to Weight (β value) of the main risk factors (elder, basic diseases, unreasonable use of antibiotics, operation time > 4 h, operating room visitation). There were two patients with infection in validation group (0.95%, 2/211), which was significantly lower than that of model group (Fisher’s exact probability method, taking one side P = 0.047). ROC analysis showed that the accuracy of nomogram model in predicting infection in model group and validation group were 0.895 and 0.864, respectively. The Hosmer-Lemeshow test showed a good fit.

Surgical site infection after CABG is related to many clinical factors, such as elder, basic disease history, unreasonable use of antibiotics during perioperation, operation time > 4 h and operating room visitation. Medical staff should fully understand these risk factors and take strict infection management measures to reduce the occurrence of infection.