To avoid targeting sites with minimal procedure volumes, we stated that participating hospitals in Ontario, and the respective surgeon group operating in them, must have an annual procedure volume for major rectal cancer resections of 15 or greater. We identified 33 such hospitals using administrative data for the period April 1, 2000 to March 31, 2001. Hospitals were then excluded if they had participated in the QIRC pilot study 33 (3 sites), if the majority of rectal surgery was performed using laparoscopic techniques (2 sites), or if surgeons at the site were involved in the QIRC trial as rectal surgery experts (4 sites). These two latter criteria were needed since a key intervention of the QIRC strategy was operative demonstrations delivered by rectal surgery experts using open versus laparoscopic techniques (See Interventions below). Nine hospitals were excluded leaving 24 potential sites. Hospitals were then eligible if 60% or more of surgeons at the respective site consented to participate and if the respective research ethics board approved the study. Sample size calculations indicated the need for 16 sites. For the first 18 hospitals approached, nearly every surgeon consented to participate and 16 hospital ethics boards approved the study protocol. (Figure 1)

Patients were eligible if they underwent major rectal cancer surgery (i.e., partial or complete segmental resection of rectum with or without an anastamosis) for a diagnosis of primary rectal cancer. Thus patients who received local excisions were not eligible. We defined a rectal versus colon cancer as a tumour located within 15 cm of the anal verge by rigid sigmoidoscopy, or, a tumour at or below the level of the sacral promontory as seen at the time of surgery. These anatomic landmarks ensured the inclusion of patients that would potentially benefit from total mesorectal excision -type surgery. Tumour stage was not used as an exclusion criterion since issues of permanent colostomy and local recurrence are relevant to all patients undergoing surgery.

These were used to introduce to participating surgeons all aspects of the study and new information. For the latter, topics included rectal cancer surgery outcomes in Ontario and elsewhere, total mesorectal excision, and, knowledge translation and quality improvement. The intent of the workshop was to promote thought in three areas: 1) There is a need to improve population level rectal cancer surgery outcomes in most jurisdictions. 2) The QIRC strategy may help surgeons improve these outcomes. 3) The pursuit of surgical excellence is an active and continuous process.

At each workshop, participating surgeons selected an opinion leader for their hospital using a validated approach 34. In summary, the selection was based on the opinion leader having three attributes including a high level of clinical expertise, a willingness to share knowledge, and being educationally influential. The opinion leader acted as a local resource person on issues pertinent to the study. For example, the opinion leader encouraged their colleagues to participate in operative demonstrations.

Participating surgeons invited a study team surgeon to assist them with a case of rectal cancer surgery. The main intent was to demonstrate total mesorectal excision techniques. Surgeons were encouraged to participate in a minimum of two operative demonstrations. The participating surgeon retained full control over all peri-operative and operative decision-making.

This questionnaire was completed after each rectal cancer surgery. The questions were designed to prompt surgeons to re-examine key total mesorectal excision operative steps [see Additional file 1].

Data (e.g., rates of permanent colostomy) were provided to individual surgeons and in a group forum for overall hospital results. The intent of feedback was to encourage individual surgeons to constantly self-examine surgical decision-making.

We wish to test the efficacy of the QIRC strategy to improve hospital rates of permanent colostomy and local tumour recurrence among patients surgically treated for rectal cancer. We hypothesize that hospitals exposed to the QIRC strategy (intervention) will have better outcomes – lower rates of permanent colostomy and local tumour recurrence – compared to hospitals in the normal practice environment (control).

We wish to test if the QIRC strategy can lead to improved quality of life, and improved bladder, bowel, and sexual function among patients surgically treated for rectal cancer. We hypothesize that patients treated in hospitals exposed to the QIRC strategy (intervention) will have higher scores on quality of life measures, and bowel, bladder and sexual function questionnaires compared to patients treated in control arm hospitals.

Since the hospital is our unit of analysis it is imperative that consecutive patients be included to prevent surgeon selection bias (e.g., excluding patients with difficult tumours perceived as high-risk for negative outcomes). Three methods of patient accrual were used to ensure consecutive patient accrual for primary outcomes – regular (weekly) phone calls to surgeons' offices, review of operating room booking logs, and hospital health records queries. Patient consent was not obtained for primary outcomes since data were collected via retrospective chart review and involved no patient contact. Primary outcomes included:

It is usually known at the time of rectal cancer surgery if a person has received a permanent colostomy. However, in cases where a patient receives a temporary stoma 18 months will be allowed for temporary stoma closure, after which time the patient will be designated as having a permanent colostomy. An anastomosis of the rectum must also be functioning for at least three months to be considered a non-permanent colostomy. This will prevent a positive outcome being marked for a patient who subsequently receives a colostomy for problems such as an anastomotic leak or severe incontinence.

Local tumor recurrence is usually defined as tumour that recurs in the pelvis near the previous operative site 123. Most, though not all, local recurrences manifest within two years of surgery. In this study local recurrence in the pelvis will be defined in three ways:

1. Definite recurrence – positive histology.

2. Probable recurrence – a postoperative mass in the area of previous pelvic surgery with any of the following patient signs (hydronephrosis, invasion of pelvic structures, or bleeding rectal mass) or symptoms (deteriorating sexual function, deteriorating bladder function, deteriorating bowel function, or persistent and worsening lower back, perineal or sciatic pain).

3. Possible recurrence – based strictly on a patient developing any of the symptoms mentioned above (deteriorating sexual function, deteriorating bladder function, deteriorating bowel function, or persistent and worsening lower back, perineal or sciatic pain).

Surgeons determined which of their respective patients may be candidates to participate in secondary outcomes data collection. Surgeons introduced this part of the study to appropriate patients and if the individual was agreeable, the study team approached the respective patient to obtain written informed consent. Patients were not approached if they were too ill, frail or psychologically vulnerable, as determined by the participating surgeon. Patients required sufficient command of the English language and capacity to give informed consent. Secondary outcomes included:

To assess overall impact of disease and treatment on patient quality of life we used the following: the European Organization for Research and Treatment of Cancer (EORTC) QLQ-C30 35, an instrument to assess quality of life in cancer patients; the supplemental EORTC QLQ-CR38 36, a colorectal cancer-specific quality of life questionnaire module; and, the 36-Item Short Form Health Survey, Version 2 (SF-36) 37.

At the time of trial initiation we felt there was no validated instrument(s) that adequately addressed these three important domains. To this end a brief set of questions were developed [see Additional file 2].

At the completion of accrual surgeons were mailed a survey. Intervention arm surgeons were asked to comment on their involvement in the trial, to evaluate the individual QIRC strategy interventions and the overall strategy, and to provide insights on issues that may be relevant to the uptake of a new innovation such as the QIRC strategy. A similar survey was sent to control arm surgeons, however, they were not asked to evaluate the individual QIRC strategy interventions. Participation in the survey was voluntary. Surgeons who agreed to participate were given the option to identify themselves or remain anonymous.

Qualitative telephone interviews were conducted with opinion leaders in the intervention arm hospitals. At least one other surgeon from each of the intervention arm hospitals was also interviewed. Participation in the interviews was voluntary. The interviews will assist in assessing the role of the local opinion leader and the overall QIRC strategy