The objective of this study was to determine if there is a significant difference in surgical site infection (SSI) when comparing the Wiltse and midline approaches for posterior instrumented interbody fusions of the lumbar spine and, secondarily, to evaluate if the reoperation rates and specific causes for reoperation were similar for both approaches.
A total of 358 patients who underwent 1- or 2-level posterior instrumented interbody fusions for degenerative lumbar spinal pathology through either a midline or Wiltse approach were prospectively followed between March 2005 and January 2011 at a single tertiary care facility. A retrospective analysis was performed primarily to evaluate the incidence of SSI and the incidence and causes for reoperation. Secondary outcome measures included intraoperative complications, blood loss, and length of stay. A matched analysis was performed using the Fisher's exact test and a logistic regression model. The matched analysis controlled for age, sex, comorbidities, number of index levels addressed surgically, number of levels fused, and the use of bone grafting.
All patients returned for follow-up at 1 year, and adverse events were followed for 2 years. The rate of SSI was greater in the midline group (8 of 103 patients; 7.8%) versus the Wiltse group (1 of 103 patients; 1.0%) (p = 0.018). Fewer additional surgical procedures were performed in the Wiltse group (p = 0.025; OR 0.47; 95% CI 0.23–0.95). Proximal adjacent segment failure requiring reoperation occurred more frequently in the midline group (15 of 103 patients; 14.6%) versus the Wiltse group (6 of 103 patients; 5.8%) (p = 0.048). Blood loss was significantly lower in the Wiltse group (436 ml) versus the midline group (703 ml); however, there was no significant difference between the 2 groups in intraoperative complications or length of stay.
The patients who underwent the Wiltse approach had a decreased risk of wound breakdown and infection, less blood loss, and fewer reoperations than the midline patients. The risk of adjacent segment failure in short posterior constructs is lower with a Wiltse approach.