Publishing study protocols is increasingly being done, also in surgery [18, 19]. The aim is to improve transparency in clinical trials and thereby also contribute to alleviating the problem of publication bias. Furthermore, providing detailed descriptions of the methodology of a clinical trial to the scientific society prior to its conduction will help to increase reliability and validity of the results and findings as the adherence to the guidelines of good clinical practice can be adequately analysed. Finally, many details of a study design cannot be published in a results paper but these details may be of profound interest to researchers wanting to perform similar studies.
The primary objective of the LapConPouch-Study is to compare a totally laparoscopic with the standard conventional approach in regard to intraoperative blood loss and need for perioperative blood transfusions. We hypothesise that intraoperative blood loss and the need for preoperative blood transfusions are significantly higher in the conventional group. There is controversy in the literature on the actual benefit of minimal invasive techniques for extensive colorectal procedures. Up until now, only one prospective randomised clinical trial that compares conventional restorative proctocolectomy to a minimal invasive approach, in this case to a laparoscopically assisted technique has been published in the literature . This study could not find any advantage for the minimally invasive procedure in regard to postoperative quality of life. One may argue that the choice of estimated blood loss is not very objective or quantitative. Obviously, blood loss is a surrogate end point in this context. It would be preferable to use major complications as primary end point but when calculating the necessary sample size (at least 300 patients per group) and considering the paucity of this operation overall such a study would not be realistic and could not be done. In a recent published meta-analysis  on "short-term outcomes of laparoscopic versus open approaches to ileal pouch surgery" blood loss was estimated to be 84 ml less in the minimal invasive group supporting the results from our own series and the use of blood loss as a surrogate end point. In order to minimize bias the blood loss has to be independently estimated by the anesthesiologist, the surgeon and the scrub nurse. Moreover, the blood collected in the suction is measured (minus the rinsing fluid, if any has been used). We feel that these methods adequately minimize bias in regard to blood loss.
The number of blood units given perioperatively as the other primary endpoint is obviously dependant on the policy of blood transfusion and therefore also susceptible to bias. The policy of our department has become very restricted and standardized over the last 5 years. Patients are only transfused if the hemoglobin falls below 6 mg% (in known cardiac disease under 9 mg%) or if they become unstable. These rules are strictly adhered to, somewhat minimizing subjectiveness in this regard.
Another critical aspect of our study design is that the surgeons, in knowledge of the endpoint, may thus be biased in the method of conduct of surgery (conscious or unconscious). In principle, this assumption is correct but the strict requirement of blinding the surgeon in regard to the endpoint would virtually render most surgical trials for less common indications impossible. In the case of laparoscopic restorative proctocolectomy, the number of surgeons performing this procedure is limited and obviously these surgeons know the potential advantages and disadvantages of the different surgical techniques and cannot be blinded to the procedure they are performing.
The laparoscopically assisted technique was more costly than the open procedure. However, laparoscopic purists could argue that the laparoscopically assisted technique with a Pfannenstiel incision used in this trial reduced the potential benefit of a truly minimally invasive procedure thereby resulting in the differences failing statistical significance. A totally laparoscopic technique has not yet been investigated in a randomized trial thus justifying the execution of the LapConPouch study. Currently the conventional approach to restorative proctocolectomy is the gold standard; however, specialized centers are increasingly performing this operation laparoscopically. Patients are often told that the minimally invasive approach carries several clear advantages but these have not yet been adequately demonstrated in randomized controlled trials. In the contrary, at best equivalence has been shown in regard to complications and quality of life. After the learning curve for minimal invasive restorative proctocolectomy has been overcome in the specialized centers, a randomized study is now feasible and necessary in order to define the future role of the new approach.
One of the major problems encountered in the trial so far was the refusal of more than half of the potentially eligible patients to take part in the study. The majority of these patients had come with the explicit desire to have a laparoscopic procedure. These patients were often extremely well informed and had chosen our department due to our large published expertise in the field of minimally invasive restorative proctocolectomy. Even, after intensive informed consent on the current available data and the goal of this study, most of these patients were not prepared to change their decision for a laparoscopic procedure. In this context, a multicenter study would obviously be preferable in order to better achieve the calculated patient numbers, but this is not realistic because only very few centers worldwide perform laparoscopic restorative proctocolectomy with the totally laparoscopic technique as described above.
In statistical regard, a two tailed approach would obviously be the better choice but would also substantially increase the needed sample size resulting in the already discussed problems of inadequate patient recruitment. In the clinical setting one-sided statistical analysis are generally being more accepted albeit their known drawbacks. We argue that in this specific situation a one-sided approach is acceptable. One of the main prerequisites in this setting is the prior specification of the one-sided analysis which is abided to in this protocol. As to the expected association of blood loss and complications, the data of our study will hopefully help to clarify this issue.
The design of this single center RCT will allow the generation of results with adequate statistical power and internal validity. The results could potentially have a relevant impact on the future use of minimal invasive techniques for performing this and also other complex colorectal operation. If the results can show a significant reduction of blood loss and a lower rate of blood transfusions, patients will have a potential benefit in terms of less risk of overall complications and less infectious risk as these are known to be correlated to larger blood loss and transfusions . The conduction of this study will contribute to implementing the practices of "Evidence-based Medicine" (EBM) in surgery and hopefully help to clarify the role of laparoscopic restorative proctocolectomy.