A reliable diagnostic method of surgical site infection after posterior lumbar surgery based on serial C-reactive protein

Abstract

Study design: A observational study. Purpose: To evaluate the detection of surgical site infection (SSI) after posterior lumbar surgery by the laboratory parameters. Materials and methods: The study included 189 patients undergoing posterior lumbar fusion or decompression between January 2018 and December 2019. Patients with chronic renal disease, rheumatoid arthritis, hepatic disease were excluded. Patients with tumor, trauma, cervical and thoracic lesions were also excluded. Leukocyte counts, segmented neutrophils, erythrocyte sedimentation rate (ESR), and C-reactive protein (CRP) were evaluated preoperatively and on postoperative days 1 (POD1), 3 (POD3), 7 (POD7), 14 (POD14), and 30 (POD30). Surgical site infection was defined as (1) purulent discharge at incision site, (2) organisms isolated aseptically from the incision site, and (3) evidence of infection based on MRI. Increasing CRP (ICRP) and decreasing CRP (DCRP) groups were classified according to the value of CRP on POD7. The change rate in CRP was calculated using the formula: (Late value−Previous value/Previous value)×100. Receiver operating characteristic curves were used to validate laboratory data involving SSI. Results: The mean age was 64.5±12.5 years. Male to female ratio was 91–98. The incidence of SSI was 8.8%. Leukocyte count, segmented neutrophil, and ESR were not specific. ICRP group were 43 patients. SSI was found in 14% of ICRP group and 18.6% was associated with secondary infection. Even 7.7% of the infection in DCRP group was identified as SSI. Based on the change rate of CRP, −50% of decreasing rate of CRP in ICRP group between POD7 and POD14 showed the validity that sensitivity was 90.9% and specificity was 68.1%. In DCRP group, −43.3% of decreasing rate of CRP in DCRP group between POD7 and POD14 showed the validity that sensitivity and specificity were 83.3% and 82.4%, respectively. Only 43% of all patients showed normalization of CRP can rule out the SSI. Conclusion: The serial change rate of CRP is a reliable method for detecting SSI more than the criteria of normal value. Nevertheless, close monitoring of patient’s signs and symptoms is important to detect early SSI.