The EnROL trial (Enhanced Recovery Open versus Laparoscopic) is a phase III, multicentre, randomised controlled trial, with blinding of patients and outcome observers during the first week post-operatively or until the date of postoperative discharge if earlier. Patients will be randomised between open and laparoscopic surgery, with both groups treated within an ERP, see Figure 1. The trial is being run at 12 centres in the UK which encompass a mix of rural and urban populations, treated within district general and specialist hospitals. In contrast to most previous laparoscopic colorectal trials the surgeons will have extensive experience of both laparoscopic and open surgery, having completed at least 100 laparoscopic colorectal resections and 50 open total mesorectal excisions prior to participating in the study. This should minimise the number of operations that will need to be converted to the open approach. Additionally surgeons will be provided with video recordings of standardized laparoscopic procedures prepared for the UK laparoscopic colorectal preceptorship programme.

The primary endpoint is patient reported physical fatigue measured by the physical fatigue domain of the multidimensional fatigue inventory 20 (MFI-20) 20 at 4 weeks after surgery. The secondary endpoints include postoperative hospital stay; 30 day and in-hospital complications; 30 day re-admission and re-operation rates; the other four domains of the MFI-20; a generic and cosmetic health related quality of life assessments 212223; observer measures of performance 1024 and measures of resource use.

The study population is patients of age 18 and above who require elective resection of colonic or rectal cancer and are suitable for laparoscopic surgery. Exclusions to trial entry are acute intestinal obstruction, previous complex laparotomies, emergency admission and pregnancy. Patients are also ineligible when, in the opinion of the treating surgeon, conversion to open surgery is likely (such as in rectal cancer when preoperative imaging suggests tumour is present at the resection margin – ‘threatened margin’). Preoperative identification of the inability to insert an epidural catheter is also an exclusion criterion, as it would prevent patients who require it receiving postoperative epidural analgesia.

The EnROL trial has been approved by the National Research Ethics Service Committee South Central - Oxford B (REC reference: 07/H0605/150). The trial is co-ordinated by the Oncology Clinical Trials Office (OCTO) at the University of Oxford, with statistical support from the Centre for Statistics in Medicine, which together form Oxford Clinical Trials Research Unit (OCTRU). OCTRU is a UKCRN Registered Clinical Trials Unit. The trial is sponsored by the University of Oxford and North West London Hospitals NHS Trust. The Data Safety and Monitoring Committee and Trial Steering Committee, which includes a patient representative, will meet regularly in order to oversee the trial appropriately. Funding has come principally from the Bobby Moore Fund, Cancer Research UK (CR-UK) (Ref number: CRUK/07/019). Ethicon Endo-surgery have provided additional funding to facilitate provision of wound dressings and collection of pathology material.

Following consent, eligible patients will be randomised in a 1:1 ratio to open or laparoscopic resection using a central computer system at the Oncology Clinical Trials Office. Simple randomisation will be used for the first 50 patients then minimization with a random element (0.8) 25. The stratification factors used in the minimization are hospital, cancer site (colon/rectum) and age (<66 years, 66–75 years, >75 years). To facilitate blinding of outcome observers the randomisation allocation will be sent directly to the patient’s surgeon via email and will not be made available to the site staff member completing the randomisation process.

Surgery is to be carried out by one of the centre’s trial-accredited surgeons no later than 6 weeks after randomisation.

Surgery should be performed in a standard fashion by the trial surgeon(s) at each site, the only difference being the method of access. Rectal tumours within 10 cm of the anal verge should be treated by total mesorectal excision accompanied, whenever possible, by preservation of the hypogastric nerves. Tumours are defined as being rectal when at or within 15 cm of the anal verge on rigid sigmoidoscopy, performed with the patient awake on their left side. Conversion is defined as the inability to complete the dissection fully laparoscopically, including the vascular division, and it usually but not always, requires the use of a larger incision than that needed to remove the specimen.

Patients and outcome observers are to be blinded to the randomisation allocation until 7 days after surgery or the day of discharge if earlier. To facilitate this large Allevyn™ adhesive dressings will be provided for all trial patients. In addition, centres should ensure that all patient records which detail the randomisation allocation are stored in a sealed envelope within the patient’s notes until after the patient is unblinded.

Blinding is an important design feature of randomised controlled trials, reducing the risk of several forms of bias including reporting bias and observer bias. A successfully implemented blinding protocol improves the validity of trial results, however the success of blinding is infrequently tested or reported in randomised controlled trials. To investigate the role and success of blinding in surgical randomised controlled trials, and its impact on patient experience, the trial protocol was amended to include collection of quantitative data about the success of blinding using the Bang Blinding Index 26. This statistical measure calculates the proportion of un-blinded patients (and research staff) in the trial by asking them to state to which arm of the study they believe they (their patient) have been allocated.

The standardized perioperative care protocol (Enhanced Recovery Programme) is outlined in Table 1. It is possible for individual centres to amend this programme provided the treatment is standardized irrespective of randomisation, and the Trial Office is notified prior to the centre opening to recruitment. Compliance with the components of the Enhanced Recovery Programme will be recorded for all patients.

Postoperative hospital stay will be reported, counting the day of operation as day zero.

30 day morbidity will be assessed at 30 days post-operatively by the centre’s Research Nurse using a standardised definition of complications which have been modified from Lang et al. (2001) who reported a doubling of cost and hospital stay associated with postoperative complications 27. The standard definitions of complications are detailed in Table 2. 30 day re-admission and re-operation rates, along with 30 day and in-hospital mortality will also be recorded.

Major morbidity is defined as any of the following occuring within the hospital admission or 30 days of surgery: haemorrhage (requiring transfusion), any re-operation or readmission, anastomotic leakage, wound dehiscence, sepsis requiring at least high dependency support, HDU stay of > 5 days, unplanned admission to Intensive or Coronary Care Unit and death.

Health related quality of life will be assessed using the SF-36 questionnaire 21. Quality of life will also be analysed with the EQ-5D. Cosmetic outcomes will be performed using a body image assessment 2328. Time points for the completion of these questionnaires are shown in Table 3.

^* To be completed not more than 4 weeks prior to surgery.

An observer assessment of physical function will be measured using standardized, objective performance indicators (SPIs) 24 which comprise a test of lower limb strength, balance and endurance. SPIs will be measured prior to surgery and then at 4 weeks, 3 and 6 months postoperatively. The tests differs from those described by Guralnik et al., in that patients will be measured for walking speed along a 10 metre, instead of 8 foot course, with a 61.5 cm (2 foot) clear zone at each end as originally described.

Health Economics questionnaires and EQ-5D will be completed by all patients providing information about the use of healthcare post-operatively for 6 months. Data collected will include GP surgery visits, GP home visits, in- patient stays, out-patients visits, District Nurse visits, physiotherapy and stoma nurse usage. Data will also be collected from the trial records to calculate the relative cost of the operative and peri- operative periods. Costs to be calculated include in-patient hotel costs, in-patient treatments costs (including re-admission and re-operation within 30 days and convalescent care), operating theatre costs and drug costs 29.

The other four domains general fatigue, reduced activity, reduced motivation and mental fatigue will be analysed.

Patients will be seen at 2 and 4 weeks and then at 3 and 6 months post operatively. Follow up data will also be collected at 12 months after surgery. Research staff at the centres will collect the data and submit it to the Trial Office via the EnROL electronic data capture website.

Resection specimens will be photographed to judge the quality of surgery of the mesorectum 30, anal sphincters, 3132 and the mesocolon 3334. Dissection will be carried out according to the trial protocol using cross sectional slicing, careful inspection for distance of extramural spread, lymph nodes, extramural vascular invasion and peritoneal involvement. TNM version 5 and modified Dukes grading including C2 and stage D will be used and regression grading and data collected by protocol. The distance to the high tie from the tumour will be recorded. Quality will be assessed locally and centrally with the addition of measurements of the specimens’ physical characteristics performed centrally. The quality of laparoscopic and open surgery will be compared to previous studies to facilitate assessment of the results of the trial.

Sample size: The original sample size calculation required 266 patients (133 per treatment arm) to allow 90% power to detect a difference in the primary outcome of fatigue (as measured by physical fatigue domain of the MFI-20) of 0.45 standard deviations at p = 0.05 (two-sided). This includes a loss to follow up of 15% and a conversion rate of 8% from laparoscopic to open surgery. Following slower than expected recruitment the power was lowered to 80% in order to complete the study in a timely manner. With all other variables remaining the same the subsequent sample size of 202 (101 per treatment arm) is required. Additionally the study is powered to detect a 3 day difference in hospital stay at 80% power and p = 0.05 (two-sided).

All analyses will be on an intention-to-treat basis.

The primary outcome will be physical fatigue as measured by the physical fatigue domain of the MFI-20 at four weeks post operatively. If the primary outcome is normally distributed it will be compared using a t-test, ANCOVA adjusting for the minimization factors (primary analysis) and ANCOVA with further adjustment for prognostic factors.

Physical fatigue will also be compared at three and six months, the remaining quality of life variables at 4 weeks, 3 and 6 months, and the cosmetic outcome at 6 and 12 months, using the same analyses as for the primary outcome. Physical fatigue will be adjusted for the level of physical fatigue prior to surgery. The other measures of fatigue will also be reported.

For all continuous outcome variables, if there is severe departure from Normality the first approach will be transformation. If the data cannot be transformed to Normality, a Mann–Whitney test will be used. In the latter case no adjustment will be made for the minimization or prognostic factors.

Hospital stay will be compared using Kaplan Meier plots and log rank tests. The complications, re-operation and re-admission data will be binary and will hence be compared between the randomisation groups using a chi square test (or Fisher exact test if the data is sparse).

Quality of Life assessments at 4 weeks as measured by the SF-36 will be analysed as suggested in the SF-36 manual, comparison between treatments will use t-test and ANCOVA to adjust for prognostic factors. Mixed-effects models of repeated measures will be used to evaluate longitudinal comparisons (4 weeks, 3 months and 6 months).

The other four domains of the MFI will be analysed in the same way as the physical fatigue domain specified in the primary analysis.

The comparison between treatment groups of standardised performance indicators will use t-tests and ANCOVA adjusting for prognostic factors and baseline.

The tumour stage, quality of surgery grade, the physical measurements and the frequency of Dukes’ C2 cases will be compared between groups.

Analysis of the health economic data will be undertaken separately by those involved in the health economic analysis. Costs will be estimated from the perspective of the UK NHS and health benefits expressed in terms of quality-adjusted life-years (QALYs). The cases where laparoscopic surgery may be considered cost effective will be pre-defined and the mean differential costs and QALYs will be calculated in order to assess whether any of these conditions are satisfied. The relative costs of the operative and peri-operative period will be derived for all trial patients including the cost of consumables that are likely to differ between the two surgical procedures.

QALYs will be calculated for each patient in the trial, on the basis of their responses to the EQ-5D pre-operatively, at one month, three months and six months. Given that the time horizon of the analysis will be less than one year, total costs and QALYs remain undiscounted. To account for the skewed nature of the data, 95% CIs for the differential costs and QALYs will be estimated using bootstrapping methods. Missing data will be imputed using a multivariate multiple imputation procedure (Solas 3.0).

Statistical analysis will be undertaken using Stata (StataCorp LP).