Surgical resection of rectal cancer is the cornerstone of curative therapy. Typically during rectal cancer surgery the tumour and a contiguous segment of normal bowel are removed and the bowel tract reestablished. For surgically treated patients two unfortunate outcomes are permanent colostomy and local tumor recurrence. For various reasons the operating surgeon may deem it necessary to remove the rectum and anus rendering the patient dependent on a permanent colostomy. Local tumor recurrence is defined as tumour that recurs in the pelvis near the previous operative site [1–3]. It is especially feared since this outcome is usually inoperable and patients, as a result, can suffer a slow, painful death. This study is testing if these two important outcomes, rates of permanent colostomy and local tumour recurrence, can be improved at the hospital level using the surgeon-directed Quality Initiative in Rectal Cancer (QIRC) strategy.
Studies on rectal cancer surgery outcomes usually show that rates of permanent colostomy, local tumour recurrence, and even patient survival vary markedly at the surgeon, hospital, or region level. For example, among the regions of Ontario, Canada researchers showed that rates of permanent colostomy following rectal surgery varied from 31% to 41% . A study examining local tumour recurrence among surgeons operating in five hospitals in Edmonton, Alberta showed that rates varied from 10% to 45% based on surgeon specialty training and procedure volume . These Canadian results are similar to studies in Europe [6–8]. The presence of important outcome variations suggests variation in the quality of delivered surgery.
Total mesorectal excision is a refinement of traditional surgical techniques that stresses sharp dissection of the mesorectum – the lymph node-bearing portion of the rectum – with careful autonomic nerve preservation [9–11]. There are a growing number of single institution series describing improvements in outcomes with the introduction of total mesorectal excision- in particular local tumour recurrence rates as low as 5% and permanent colostomy rates of 10–15% [12–18]. Population based studies from Europe also detail positive changes when surgeons in large areas adopt total mesorectal excision principles [19, 20].
Knowledge translation research has identified interventions that may encourage physician behaviour change such as continuing medical education (e.g., workshops), use of opinion leaders, and audit and feedback [21–28]. It is also suggested that multiple interventions are more effective than single interventions. As well, behaviour change may be enhanced by using quality improvement principles such as a participatory and supportive approach that focuses on the system, not individuals; breaking processes down into definable steps that are more readily targeted for improvement; and, decreasing variation in process steps resulting in improved overall quality [29–32]. The QIRC strategy integrated such knowledge translation interventions and quality improvement concepts in an attempt to ensure that hospitals (i.e., the surgeons in the respective hospital) delivered optimal total mesorectal excision-type surgery to patients.
The purpose of this paper is to describe the methodology of our cluster randomized controlled QIRC trial, which is testing if the surgeon-directed QIRC strategy can improve patient outcomes at the hospital (i.e., cluster) level. We used a cluster design to minimize the chances of contamination at the patient and surgeon level. We surmised that patient-level randomization would not work since surgeons exposed to new information or techniques, such as those promoted through the QIRC strategy, would accept or reject such information or techniques for all of their subsequent patients. We surmised that surgeon-level randomization would not work since surgeons in a given hospital often share operative experiences through discussion or direct observation, and thus new information or techniques would likely be shared among surgeons in a given hospital. Since surgeons in Ontario rarely perform rectal cancer surgery at more than one hospital, we were confident that hospital-level cluster randomization would minimize the chances of contamination between the two arms of the QIRC trial.