For the increasing numbers of patients undergoing fusion procedures for the degenerative lumbar spine, infection and re-operation can negatively impact outcomes. Numerous observational and retrospective reviews have shown advantages to para-median versus midline approaches; however, recent systematic reviews have shown a need for a well-powered, prospective randomized control trials comparing both exposures. As a step towards a long-term goal of an RCT to address this issue, the purpose of this pilot study is to gather initial data to examine whether operative approach impacts the short-term infection rate, re-operation rate, length of stay, and overall costs to the system. Patients deemed appropriate surgical candidates with single or two-level degenerative spondylolisthesis will be approached for participation, and randomized into either the midline or paramedian group. Initial follow-ups will be at 2 and 6 weeks, and 3 months. Infection rates, inpatient and outpatient adverse events, re-operation rates, radiation exposure and costs will be determined. Cost effectiveness analysis will be estimated comparing each procedure using a bottom-up estimation. Post-operative wound infection can have a significant effect on patient short and long term outcomes. If a significant difference in infection rate is demonstrated, as well as lower re-operation rates, shorter stays, and decreased overall costs, adoption of paramedian approaches to single or two-level fusions of the lumbar spine might be suggested, providing fuel for a full-scale RCT.

Paramedian lumbar interbody fusion has been hypothesized to be advantageous over midline approaches for several reasons, the most important of which is lower post-operative infection rates. To achieve our long-term goal of conducting a RCT to examine this issue in-depth, a pilot study is initially required to examine if this surgical approach may indeed lower infection rates and associated health care cost savings.

It is hypothesized that: (1) there will be a statistically- and clinically-significant difference in infection rates with each approach to the lumbar spine, with those for the paramedian approach being lower; and (2) as a result of lower infection and complication rates, and reduced length of stay, the costs associated with the paramedian approach to the spine will be substantially less.

All consecutive patients deemed appropriate surgical candidates with single or two-level degenerative spondylolisthesis will be approached for participation and inclusion/exclusion criteria will be evaluated. (Note: Target enrollments will not be set a priori, but determined through the course of this proposal.) Once patients are confirmed to be eligible for the study, baseline patient reported outcome data will be collected (ODI, SF-12, EQ-5D, PHQ-9 and generic demographic information). Recruits will be randomized into either the midline or paramedian group using a computer-generated randomization process. After surgery, the surgeon will complete a post-operative summary, discharge form and adverse events form. Follow-up appointments at 6 weeks and 3 months will include repeating the patient reported health outcome questionnaires (ODI, SF-12, EQ-5D, PHQ-D) as well as a cost diary and pain questionnaire. With the assistance of our institution's Research Methods Unit, data analysis will be conducted. Continuous variables (length of stay, SF-12, EQ-5D) will be compared using students T-test while adjusting for potential confounders (diabetes, primary diagnosis, etc) using multi regression models. Categorical data will be analyzed using Chi-squared analysis while controlling for confounders using Cox-proportional hazards ratios.

One of the primary outcomes will be infection rates. To avoid bias in this area, infection will be tracked using the Center for Disease Control (CDC) criteria for diagnosing both deep and superficial infections. Deep infections will be diagnosed with deep wound cultures. The CDC criteria for diagnosing superficial infection is erythema, drainage, dehiscence and "surgeon opinion", allowing for significant subjectively in superficial infections. As such, it is possible that surgeons may preferentially diagnose more midline or Wiltse incisions with superficial infection depending on whether they may have an inherent favoured approach. While it is impossible to truly blind surgeons, allied health care personnel who would have less bias in favour of one approach will assess all wounds. This is in-line with current practice, as wounds are generally assessed by nurses in the clinic, and staples or sutures are removed at the two-week follow-up visit. Surgeons are only asked to assess wounds that are questionable. This would decrease some of the potential bias as part of this un-blinded study. Surgeon opinion will have no impact on deep infection, because this will generally require irrigation and debridement in the operating theater which will be captured.

The benefits, in the event of a positive study outcome, will lead to substantial improvements in patient care with reduced infection rates. If the evidence is compelling from a pilot study, the goal would be to move to a prospective, multi-centered randomized control trial to further examine this research question and potential adoption of more midline sparing surgery for fusion procedures of the lumbar spine. In the event that significant cost savings are demonstrated, quality assurance officers and policy makers would be compelled to adjust remuneration codes and possibly limit the compensation in cases where a midline approach was chosen. The cost savings for the system would be substantial even if only a significant difference in infection rate was demonstrated. There are risks are for lumbar spine surgery in general. At present, these risks are assumed to be equivalent with the two surgical approaches to be examined in this study and patients joining in the study can be assured of high-quality care for their condition regardless of the treatment arm to which they are allocated. However, if this proposed study can show that infection rates can be lowered with surgical approach, then the benefits definitely outweigh the risks.

Inclusion Criteria:

• 1) surgical candidates with single or two level degenerative spondylolisthesis with the following clinical findings:

  1. a clinical history of back, buttock and leg pain with walking or standing that is improved when lying, sitting or bending forward
  2. a clinical history of leg symptoms that are greater than or equal to back symptoms with walking or standing
  3. greater than 6 months of symptoms with failed conservative care

• 2) sufficiently fluent in English to provide informed consent and complete questionnaires with or without the need of an interpreter.

Exclusion Criteria:

• 1) clinical history of osteoporotic fracture or chronic oral steroid use;
• 2) previous posterior lumbar spinal surgery (excluding prior microdiscectomy); and
• 3) patients who have evidence of neurological disorders that affect physical function (e.g. peripheral neuropathy), neuromuscular disorders (e.g. multiple sclerosis, Parkinson's etc.) or systemic illness (e.g. inflammatory arthritis) that affects physical function