Surgical site infections among patients receiving care per Surgical Care Improvement Project

Background: Current guidelines for prevention of surgical site infections (SSI) recommend appropriate selection, timing and duration of perioperative antibiotic(s). However, implementation of these measures has yielded variable results.

Objective: 1. To compare SSI rate in patients receiving care per peri-operative antimicrobial prophylaxis guidelines, with SSI rate in patients who did not.  2. To compare clinical and surgical characteristics of patients with and without SSI.

Methods: We used a retrospective observational cohort with nested case control study design. The setting was Parkland Hospital and Health System, Dallas, Texas, a large public academic health system. The study period was 1/1/2008-6/30/2009. Patients undergoing hip or knee replacement, hysterectomy, selected colon and vascular surgeries were included. Data sources were the SSI surveillance database and the surgical care improvement project (SCIP) database from the division of Clinical Quality Management. Definitions per the Centers for Disease Control- National Health Safety Network (NHSN) were used. Perfect antibiotic care was defined as correct choice, timing and duration of preventive antibiotic per SCIP guidelines. The rates of SSI among patients who received perfect care and those who did not were compared. Patients who developed SSI (cases) were compared to patients who did not develop SSI (controls) within the SCIP database who underwent surgery of the same ICD-9 code. Demographic, clinical, and microbiological data were collected upon review of medical records. Statistical analyses were performed with Chi-Square, Fisher’s Exact or t-test using SPSS^® v.17 for Windows. The critical value of α was 0.05.   

Results: During the 18-month study period, 815 patients undergoing surgery had available data on antibiotic use. Of these 815 patients, 37 (4.5%) developed SSI. The n (%) of SSI per procedure type was:  hip replacement 4 (2.9%); knee replacement 3 (1.3%); hysterectomy 10 (3.7%); colon 18 (13.4%); and vascular 2 (5.7%). Of the 815 patients, 613 (75%) received perfect antibiotic care. Thirty-two of these patients (5.2%) developed SSI vs. 5/202 patients (2.5%) who did not receive perfect care (p-value = NS). The NHSN risk index (NHSN-RI) was available for 734 cases. When the SSI rates were stratified per NHSN-RI, the rates among patients receiving perfect antibiotic care vs. those who did not were: 2.4% vs. 1.2% for NHSN-RI of 0 (p=1); 5.6% vs. 3.3% for NHSN-RI of 1 (p=0.746); 15.9% vs 4.2% for NHSN-RI of 2 (p=.277); NHSN-RI of 3 had no SSI.  28 patients with SSI were compared to 28 uninfected matched controls. None of the patient characteristics were significantly different except mean body mass index (31.9 among cases vs. 26 in controls; p=0.04).

Conclusions: Compliance with antibiotic selection and timing of administration did not impact the overall rates of SSI. Obesity was associated with a higher rate of SSI.