A simple pre-operative prediction tool for surgical site infection in vascular surgery

Background:

Several scoring systems have been developed for risk assessment of surgical site infections (SSIs) but are based on a combination of pre- and peri-operative variables, and have limited ability to predict individual patient risks prior to elective surgery. The National Nosocomial Surveillance Systems (NNIS) risk index is the most commonly used risk stratification tool. However, NNIS does not perform equally well for all types of surgeries.

Objective:

1) to construct a model for calculation of SSI risk based on noninvasive pre-operative intrinsic patient factors, and 2) to compare the relative predictive ability of the above model with the NNIS risk index, among patients undergoing elective vascular surgery.

Methods:

Nested case-control study among patients who underwent elective vascular (abdominal aortic and peripheral arterial) surgery at Mayo Clinic, Rochester, in a 5 year period between January 1, 2003 and December 31, 2007. Patients with SSI were identified using prospectively collected surveillance data by Infection control practitioners; all infections requiring hospitalization for management were included as cases. From the cohort of patients who underwent surgery but did not develop SSI, controls were selected by matching (1-2/ case) on type of procedure and date of surgery. Clinical data were collected by chart review. Risk factors for SSI were assessed using univariate and multivariable logistic regression analyses; a prediction model was developed using multivariable logistic regression and bootstrap re-sampling. The c-statistic, equivalent to the area under the receiver operating characteristics (ROC) curve, was used to assess the model's prognostic ability.

Results:

A total of 87 cases met inclusion criteria, and were compared with 166 controls. Relative to controls, cases were more likely to report difficulty climbing stairs, use of assistive ambulation device, have COPD, critical ischemia, history of previous SSI, and previous peripheral revascularization (all P < 0.05). There was no significant difference between cases and controls in age, sex, diabetes, smoking, chronic kidney disease, liver disease, weight loss, immunosuppressive therapy, and presence of skin ulcers. From multivariable analysis, previous SSI, previous peripheral revascularization, critical ischemia, and COPD were significantly associated with SSI risk. A prediction model containing these variables was developed with a c-statistic of 0.73. In comparison, a prediction model using NNIS risk index reflected a c-statistic of 0.504 (Table 1).

Conclusions:

This prediction model based on non-invasive pre-operative patient factors could serve as a simple risk assessment aid to patients and physicians prior to elective vascular surgery. Data from this retrospective study indicate fairly good predictive ability, and an improvement over NNIS in this type of surgery. This tool needs prospective validation.