Resection for perforated diverticulitis After termination of the LOLA arm

Transcription

1 Laparoscopic peritoneal lavage or resection for generalised peritonitis for perforated diverticulitis: a nationwide multicenter randomised trial [NTR 2037] Resection for perforated diverticulitis After termination of the LOLA arm Version 8. First version published on April 12th, 2009 Version date April , protocol according to Good Clinical Practice Adjusted for the termination of laparoscopic lavage treatment. Project leaders Trial coordinators Prof. dr. W.A. Bemelman Prof. dr. J.F. Lange Drs. G.D.Musters Drs. S. Vennix LADIES-trial, version 8, dd

2 Contents Investigator approval...3 Summary Background Patients and Methods Study design: Study population Inclusion criteria: Exclusion criteria: Procedure Intervention(s): Study outline Standardisation of indication for reintervention Stoma reversal Data collection and follow-up Patient accrual Patient withdrawal Data sampling Outcome parameters Primary endpoint Secondary endpoints Economic evaluation Statistical analysis Scientific justification for group size calculation Group size calculation Cost-effectiveness Accounting for missing data Data and record keeping Deviations from statistical plan Patient Safety Serious Adverse Events Reporting Serious Adverse Events Predicted Serious Adverse Events Adverse events Data safety monitoring committee Termination of the study Ethical and legal requirements Insurance Quality control Publication policy References...23 Verwijderd: 4 Verwijderd: 5 Verwijderd: 7 Verwijderd: 9 Verwijderd: 9 Verwijderd: 10 Verwijderd: 10 Verwijderd: 10 Verwijderd: 10 Verwijderd: 10 Verwijderd: 10 Verwijderd: 11 Verwijderd: 12 Verwijderd: 12 Verwijderd: 12 Verwijderd: 13 Verwijderd: 13 Verwijderd: 14 Verwijderd: 14 Verwijderd: 14 Verwijderd: 14 Verwijderd: 15 Verwijderd: 15 Verwijderd: 17 Verwijderd: 17 Verwijderd: 18 Verwijderd: 19 Verwijderd: 19 Verwijderd: 19 Verwijderd: 19 Verwijderd: 20 Verwijderd: 20 Verwijderd: 21 Verwijderd: 21 Verwijderd: 22 Verwijderd: 22 Verwijderd: 22 Verwijderd: 22 Verwijderd: 23 Verwijderd: 24 Met opmaak: Engels (Groot-Brittannië) LADIES-trial, version 8, dd

3 Investigator approval I certify that I have read and understood the protocol of: Laparoscopic peritoneal lavage or resection for generalised peritonitis for perforated diverticulitis: a nationwide multicenter randomised trial. I agree to conduct the study in accordance with this document. Version 8. First version published on april 25th, 2013 Version date March 15 th 2010, protocol according to Good Clinical Practice LADIES-trial, version 8, dd

4 Summary Perforated diverticulitis generally requires emergency surgery. Regardless of selected strategy, sigmoid resections for acute perforated diverticulitis are associated with substantial morbidity (up to 50%) and mortality (15 to 25%). Just recently, excellent results were reported with laparoscopic lavage and drainage only, in patients with purulent peritonitis. Mortality and morbidity figures were less than 5%, and a colostomy was avoided in the majority of these patients. Potentially, this alternative brings a large gain in health and reduction of costs. Nevertheless, since sigmoidectomy is still considered the standard of care for perforated diverticulitis by most surgeons, implementation might be variable. Some surgeons will embrace laparoscopic lavage because of its technical simplicity, other might be reluctant fearing failure of this recent strategy. Only a head to head comparison of the several treatment options for perforated diverticulitis with purulent peritonitis will provide sufficient evidence for implementation. If it is decided to perform a sigmoidectomy for perforated diverticulitis, the optimal strategy is still a matter of debate. The available literature suggests equality of sigmoid resection with or without anastomosis regarding postoperative mortality and morbidity. If a sigmoidectomy with anastomosis is done, a protective loop ileostomy probably diminishes the number of anastomotic leakages and its complications. The available literature suggests that the likelihood of stoma closure is higher after resection, anastomosis and ileostomy (90%) in comparison to Hartmann s with end colostomy (60%), but evidence is lacking. The first objective of this integrated trial (LOLA) was to determine whether laparoscopic lavage led to better clinical outcomes compared to sigmoidectomy in patients with perforated diverticulitis with purulent peritonitis in terms of mortality and morbidity. However the data safety board commission has advised to stop inclusion of the LOLA arm after 96 patients. This advice was followed by the investigational group and therefore the resections for purulent and faecal peritonitis remained. The current main objective of the DIVA arm is continued. The current primairy objective is to determine whether sigmoidectomy with anastomosis and ileostomy or sigmoidectomy with end colostomy is the superior approach in patients with perforated diverticulitis with either LADIES-trial, version 8, dd

5 purulent or faecal peritonitis in terms of stoma free survival. The study is designed as a multicenter and randomised trial. Patients diagnosed as having perforated diverticulitis with free air on plain abdominal X ray or CT scan fulfilling the in and exclusion criteria are randomised during laparoscopy via a central computer. In case of purulent diverticulitis patients are randomised between two arms: (a) sigmoidectomy with colostomy or (b) sigmoidectomy with anastomosis with defunctioning loop ileostomy in ratio of 1:1. In case of faecal peritonitis or an overt perforation of the sigmoid, the patient will be randomised 1:1 to sigmoidectomy with colostomy or sigmoidectomy with anastomosis with or without defunctioning loop ileostomy. The first primary outcome parameter consists of stoma free survival one year after initial surgery. Secondary endpoints are number of days alive and outside the hospital, health related quality of life, health care utilisation and associated costs. A sample size of 2x118 patients are needed to significantly demonstrate a difference of 30% in stoma free survival between both treatment arms (log rank test two sided alpha of 5% and power of 90%) in favour of the patients with a protective loop ileostomy after primary anastomosis. More than 35 hospitals have agreed to participate in this study with an estimated total inclusion of 100 patients per year must be realistic over a 4 years period. Patients will be followed for one year. The study will be executed in concordance with the protocol, the Good Clinical Practice guidelines and regulatory requirements. LADIES-trial, version 8, dd

6 1. Background Diverticular disease is an important condition in terms of healthcare utilisation and it is one of the five most costly gastrointestinal disorders in westernised countries.[1] Despite this high prevalence, treatment of all different stages of diverticular disease is still hardly evidence based, hence containing a lot of controversies. High quality studies are lacking which makes clinical decision making hardly evidence based. Four trials with different research questions all involving important issues concerning the treatment at different manifestations of diverticulitis have evolved in 2008 in the Netherlands (flow chart figure 1). This has led to a joint Dutch initiative the Dutch Diverticular Disease Collaborative Study Group" or the "3D Collaborative Study Group. Two of these trials with different, but correlated research questions aim to include patients with perforated diverticulitis. For this reason, the project leaders of the LOLA trial (purulent perforated diverticulitis: LaparOscopic LAvage or resection) and the DIVA trial (faecal perforated DIVerticulitis: Hartmann procedure or primary Anastomosis with ileostomy) have decided to cooperate and to combine the two studies in one proposal. A joint effort will improve hospital participation, patient accrual, data quality and study efficiency in this nationwide multicenter study. In the present proposal the study objective and research questions of both the LOLA trial and the DIVA trial are incorporated. However, as previously stated, the LOLA arm objective was preliminary terminated, after an interim analysis of the Data Safety Board Commission of 96 patients, with the continuation of the DIVA arm. Perforated diverticulitis is a perforation of a diverticulum of the large bowel, mostly the sigmoid, resulting in either a purulent or faecal peritonitis (Hinchey stadia 3 and 4). Both conditions require emergency surgery. [2,3] Regardless of selected strategy emergency operations for acute perforated diverticulitis are associated with substantial morbidity (up to 50%) and mortality (15 to 25%).[3 8] Primary sigmoidectomy with or without an anastomosis has become the standard practice for patients with generalised peritonitis complicating diverticulitis.[6 10] For many surgeons Hartmann s procedure still remains the favoured option, meaning that patients will have end colostomy. Reversal of stoma after Hartmann s however, is only done in 50 to 60% of the patients due to high mortality and morbidity rates [12,13,19], thereby LADIES-trial, version 8, dd

7 compromising quality of life and increasing costs. [14] Recently laparoscopic lavage emerged as an effective alternative in patients with perforated diverticulitis with purulent peritonitis. [18] Laparoscopic lavage for perforated diverticulitis with purulent peritonitis has first been described by O'Sullivan in 1996.[22] Just recently, Myers et al [18] reported excellent results in a series of 92 patients. Mortality and morbidity figures are less than 5% and a (permanent) colostomy was avoided in the majority of these patients.[18, 22 28] So there is a potential large gain in health and costs to be expected applying laparoscopic lavage for perforated purulent diverticulitis. Nevertheless, since sigmoidectomy is still considered the standard of care for perforated diverticulitis by most surgeons, implementation might be variable. Some surgeons will embrace laparoscopic lavage because of its technical simplicity, other might be reluctant fearing failure of this recent strategy. Only a head to head comparison of the several treatment options for perforated diverticulitis with purulent peritonitis will provide sufficient evidence to guide surgeons what to do. In case of faecal peritonitis there is no evidence for alternatives other than sigmoidectomy with an ostomy. If it is decided to perform a sigmoidectomy for perforated diverticulitis, the optimal strategy is still a matter of debate. The available literature suggest equality of sigmoid resection with or without anastomosis regarding postoperative mortality and morbidity.[5,7,9,29,30] If anastomosed, a defunctioning loopileostomy might diminish the number of anastomotic leakages and its complications.[15] The available literature suggests that the likelihood of stoma closure is higher after resection, anastomosis and ileostomy (85%) in comparison to Hartmann s (60%) (DIVA) [14,16], but robust evidence is lacking. The first objective (LOLA) was to determine whether laparoscopic lavage was the superior approach compared to sigmoidectomy with end colostomy or with anastomosis and ileostomy in patients with perforated diverticulitis with purulent peritonitis in terms of mortality, morbidity, quality of life, health care utilisation and associated costs. Due to the preliminary termination of the LOLA arm, the main objective of the DIVA remains. Therefore the primairy objective (DIVA) is to determine whether sigmoidectomy with anastomosis and ileostomy or sigmoidectomy with end colostomy is the superior approach in patients with perforated diverticulitis with purulent and LADIES-trial, version 8, dd

8 faecal peritonitis in terms of stoma free survival, quality of life and cost effectiveness. 2. Patients and Methods This multicenter randomised trial will be executed in concordance with the protocol, the Good Clinical Practice guidelines and relevant legal requirements. After the termination of the LOLA arm in this study, we will continue to compare the two surgical strategies for patients with faecal or purulent perforated diverticulitis. All patients will be followed up to one year and in the period we will measure clinical relevant outcomes (combined endpoint of mortality and major morbidity), health related quality of life, number of days alive and spent outside the hospital and costs. A specific outcome when comparing surgical strategies is stoma free survival. Furthermore we will determine which treatment strategy is the most costeffective. 2.1 Study design The design of the study is randomised and multicenter. Patients presenting with signs of generalised peritonitis will have a CT scan. If there is free intra abdominal air with suspicion of perforated diverticulitis, the patient is potentially eligible for this study. If the in and exclusion criteria are fulfilled, the patient will have a diagnostic laparoscopy to confirm the diagnosis In case of purulent peritonitis the patients enters the purulent arm of the study, in case of faecal peritonitis, the patient enters the faecal arm of the study. Randomisation is performed during laparoscopy or laporotomie, whereby the patient are randomised via the trial website according to the attached flowchart (figure 2). In case of purulent diverticulitis sigmoidectomy with colostomy is compared with sigmoidectomy with anastomosis with defunctioning loop ileostomy. The best evidence indicates that the two resectional strategies are equal in terms of morbidity and mortality in case of generalised peritonitis [8]. For this reason a three way 1:1 randomisation is proposed. In case of faecal peritonitis or an overt perforation of the sigmoid, the patient will be randomised between a sigmoidectomy with colostomy or a sigmoid resection with primary LADIES-trial, version 8, dd

9 anastomose with or without a deviating ileostomy. 2.2 Study population Inclusion criteria patients suspected of diverticulitis age in between 18 and 85 years informed consent faecal peritonitis or an overt perforation of the sigmoid, the patient will be randomised to undergo either sigmoidectomy with colostomy or sigmoidectomy with anastomosis with or without defunctioning loop ileostomy. In this way all patients with perforated diverticulitis fulfilling the in/exclusion criteria can be included in this study. with free air on plain abdominal X ray or CT scan OR with peritonitis and diffuse gas or fluid on CT scan Exclusion criteria dementia prior sigmoidectomy steroid treatment > 20 mg daily prior pelvic irradiation preoperative shock: requirement of inotropics due to circulatory insufficiency 2.3 Procedure Intervention(s) In case of purulent diverticulitis sigmoidectomy with colostomy is compared with sigmoidectomy with anastomosis with or without defunctioning loop ileostomy. In case of Study outline Patients fulfilling the inclusion and exclusion criteria will have a diagnostic laparoscopy. After the pneumoperitoneum is installed, two additional 5 mm trocars are inserted in the left and right lower abdomen. A careful inspection of the abdominal cavity particularly the stomach, duodenum and sigmoid, is done to localize the site of perforation. In case of peritonitis due to a perforated diverticulum it must be attempted gently to locate the site of perforation. Careful removal of adherent omentum or bowel is tried. If clearly adherent, it should be left in place. If no obvious perforation is apparent and in the absence of faecal content, the patient is included in the purulent arm. In case of an overt perforation or intra abdominal contamination with faeces, the patient enters the faecal arm of the study. Abdominal content is aspirated for culture LADIES-trial, version 8, dd

10 and randomisation is done via the website. When there is doubt on the peritonitis being purulent or faecal, the abdominal content should be aspirated and the colour or the content should be examined. Only brown coloured fluid can be marked as faecal content. When aspiration alone is not sufficient for determining the faecal content, the aspirated fluid should be sprayed on a gauze to see if there are any brown bits on the gauze. If present, there is faecal peritonitis. Sigmoidectomy and anastomosis and loop ileostomy: Sigmoidectomy is done according to the guidelines of the American Society of Colon and Rectal Surgeons. [31,32] The distal transsection margin has to be on the proximal rectum, the proximal margin is determined by the absence of wall thickening due to diverticulitis. The type of anastomosis is done according to the preference of the operating surgeon. A loop ileostomy can be fashioned in order to ensure faecal deviation, this is upon the discretion of the surgeon. Sigmoidectomy with colostomy (Hartmann s procedure): Hartmann's procedure is stated as a two stage procedure with the intention to close the colostomy in a second stage. During the primary surgery, only the perforated diseased part must be resected. There is no need of having the distal transsection line on the proximal rectum. Leaving a Douglas drain is at the discretion of the operating surgeon. Intravenous or oral antibiotics are administered for seven days in both groups. Postoperatively, oral diet and mobilisation are advanced as soon as possible. Within four to six weeks after surgery a sigmoidoscopy is performed to exclude malignancy as the underlying cause of the perforation Standardisation of indication for reintervention Relaparotomy is indicated in patients with clinical deterioration or lack of clinical improvement with a likely intraabdominal cause within hours after the index procedure. Other (intercurrent) infectious foci (e.g. pneumonia) must be ruled out using laboratory tests, imaging modalities, or both. The decision to perform a relaparotomy must be made by the multidisciplinary medical team. To guide the decision for reoperation, the Sequential Organ Failure Assessment (SOFA) score can be used as help for decision making.[33] Prespecified surgical emergencies which might require reintervention, are abdominal compartment syndrome, intra abdominal LADIES-trial, version 8, dd

11 bleeding with persistent decrease in haemoglobin despite replacement, hemodynamic instability, burst abdomen, perforation of visceral organ, anastomotic leakage, intra abdominal abscess that cannot be drained percutaneously, and ischemia/necrosis of a visceral organ Stoma reversal After sigmoidoscopy is performed and the surgeon finds the patient eligible for stoma reversal, the surgeon can offer the patient reversal of the stoma. When the patient accepts, he or she can be planned for surgery. When a patient does not show to the outpatient department, when the patient is not eligible for stoma reversal, when the patient refuses stoma reversal or when the anesthesiologist refuses stoma reversal, the patient can not be planned for surgery. The reason for not reverting the stoma needs to be filled in in the Case Record Form Data collection and follow up Data collection includes patient characteristics, POSSUM score, preoperative APACHE score, surgical parameters, Hinchey score, Mannheim peritonitis index, morbidity, length of stay, quality of life at 2, 4, 13 and 26 weeks (SF 36, GIQLI, EQ 5D), consumption of hospital resources and health care at 4, 13, 26, 39 and 52 weeks, early and late reinterventions and stoma rate at end of follow up. Pre, per and postoperative data have to be filled in online via the trial website. The data will be filled accompanied by the patients randomisation code. No names or contact information appears on the Case Record Forms. The trial coordinators do not have access to the patients information. The participating hospitals keep the list for decoding the randomisation codes in the investigators site file, located in the participating hospital Patient accrual As the ratio between purulent (Hinchey III) and faecal (Hinchey IV) peritonitis due to perforated diverticulitis is estimated 2:1, it is expected that 212 patients have to fulfil the in/exclusion criteria and complete the follow up. In this way, 106 patients will be randomised for faecal peritonitis and will enter the faecal arm. The patients having had resectional therapy in purulent arm together with the patients from the faecal arm with an estimated loss of patients of 10% (due to mortality and lost in follow up) will be enough to power for the outcome "stoma LADIES-trial, version 8, dd

12 free survival at one year" (see flow chart). This gives a total number of 233 patients. Up to now more than 35 hospitals agreed to participate in this study. Extrapolation of the current included patients in the LADIES trial will lead to a completion of the inclusion at the end of 2016 (DIVA arm). Due to the termination of the laparoscopic lavage, evaluation of the LOLA arm can be done after the one year follow up, which is to be expected in Patient withdrawal Any patient maintains the right to revoke their consent at any point and withdraw from the trial. In that event the investigator would attempt to ascertain the reason for withdrawal and record this on the patient termination form. As much follow up data as possible will be collected to that point and all data recorded will be included in the final analysis provided the surgery was attempted. If the patient withdraws prior to attempted surgery then this patient will be excluded from the analysis. Additional patients may be enrolled if patients withdraw from the study prior to closure of recruitment, to ensure the sample size is maintained Data sampling Data is sampled by the clinical investigator in the online Case Record Form at four times. This data is recorded in a source document at the centre and will be copied to the online CRF. The first time is after informed consent by the patient (preoperative data), the second time is immediately after surgery (per operative data), the third time is at discharge (postoperative data), the fourth time is after the patient s visit of the outpatient department. All other data will be retrieved from questionnaires, that are sent to the patient by the trial coordinator at 2, 4, 13, 26, 39 and 52 weeks and will be recorded in the CRF directly by the trial coordinator or sent to the data management by fax or mail and thereby recorded automatically. When the patient does not show up at the outpatient department, he or she will be contacted by telephone by the trial coordinator to check the patient s compliance and verify whether the patient has received medical care from a different center. All patients who show bad compliance will be contacted by the trial coordinator by telephone. All patients who indicate on the questionnaire Health Care Use that they have received medical care from a medical center different from the primary hospital, LADIES-trial, version 8, dd

13 will be contacted by telephone by the trial coordinator. 2.4 Outcome parameters Primary endpoint Stoma free survival one year after initial surgery, is the only relevant and primary outcome for the comparison of the resectional strategies Secondary endpoints Secondary endpoints are operating time, hospital stay, number of days alive and outside the hospital, incisional hernia, reinterventions within twelve months, health related quality of life, health care utilisation, associated costs and major morbidity includes any of the following events or conditions: reintervention, wound dehiscence, incisional hernia, abscess needing percutaneous drainage, urosepsis, myocardial infarction, renal failure and respiratory insufficiency Economic evaluation Background and rationale: The key question in the economic evaluation is to assess whether to what extent a more effective surgical strategy also reduces health care utilisation and associated costs. A more effective strategy in terms of health status is expected to incur less costs due to health care utilisation (direct medical costs) and less productivity loss due to absence from paid work (indirect costs). Comparisons of the different surgical strategies in the economic evaluation will be analogous to the analyses of the clinical endpoints. The economic evaluation will be performed from a societal perspective, with the costs per unit improvement on the primary clinical endpoints, defined as poor outcome and stoma free survival for the DIVA arm as the primary outcome measure. The appropriate type of economic evaluation is be conditional on the results (Gold, 1996). We hypothesise that a more effective intervention will be associated with less health care utilisation as well absence from paid work (productivity costs). Therefore, the primary analysis will be a costeffectiveness analysis that evaluates costs associated with an improved surgical outcome in terms of stoma free survival. In addition, a secondary analysis, will evaluate cost differences in relation to differences in quality adjusted life years (QALY s). This cost utility analysis, resulting in an incremental costeffectiveness ratio expressed in costs per QALY, will be included to allow comparison with other health related interventions or programs. With a study LADIES-trial, version 8, dd

14 horizon of twelve months, no discounting will be applied. We will differentiate between direct medical, direct nonmedical and indirect costs. Direct medical costs are associated with health care utilisation related to surgical interventions, percutaneous drainages, diagnostic procedures, medications, materials (blood products, gauze packing), readmissions and reinterventions, visits to outpatient, primary care and paramedical health care providers, and home care. Direct non medical costs are generated by travel to and from health care providers. Indirect costs are associated with lost productivity due to absence from paid work. 2.5 Statistical analysis Scientific justification for group size calculation A. Systematic review: Mortality of sigmoidectomy for perforated diverticulitis. Primary Resection With Anastomosis vs. Hartmann s Procedure in Nonelective Surgery for Acute Colonic Diverticulitis: A Systematic Review. Vasilis A. Constantinides et al. Diseases of the Colon and Rectum 2006 Jul;49(7): [8] Search terms: comparative studies and diverticular disease / diverticulitis, comparative studies and peritonitis, diverticular disease / diverticulitis and primary and resection / anastomosis, onestage and diverticular disease / diverticulitis and operation / resection, diverticular disease / diverticulitis and Hartmann s and procedure / operation. Population: Fifteen studies matched the selection criteria and were suitable for meta analysis. These included a total of 963 adult patients, of which 547 underwent primary resection and anastomosis (PRA 57%) and 416 Hartmann s procedure (HP 43%). The study design was retrospective in 13 studies and prospective, nonrandomized in 2. Outcome: The primary end point was postoperative mortality. Secondary endpoints included surgical and medical morbidity, operative time, and length of postoperative hospitalisation. Methodological filters: Comparative filters Databases used and number of manuscripts retrieved: A MEDLINE, Ovid, Embase, and Cochrane database search was performed on all studies published between 1984 and Twenty four studies published between 1984 and 2004 matched the search criteria, comparing PRA vs. HP and reporting the incidence of mortality, postoperative complications, operative time, or length of postoperative hospital LADIES-trial, version 8, dd

15 stay. Nine studies were excluded because they contained a patient group undergoing PRA vs. HP that was indistinguishably mixed for cancer as well as diverticulitis. Selection procedure, validity assessment: Two reviewers (VC and PT) independently extracted the data Results: Overall mortality was significantly reduced with primary resection and anastomosis (4.9 vs. 15.1%; odds ratio = 0.41). Subgroup analysis of trials matched for emergency operations showed significantly decreased mortality with primary resection and anastomosis (7.4 vs percent; odds ratio = 0.44). No significant difference in mortality was observed in trials matched for severity of peritonitis Hinchey > 2 (14.1 vs percent; odds ratio = 0.85). Overall morbidity was 29% for PRA and 33% for HP. Conclusions: Patients selected for primary resection and anastomosis have a lower mortality than those treated by Hartmann s procedure in the emergency setting and comparable mortality (14%) under conditions of generalised peritonitis (Hinchey > 2). The retrospective nature of the included studies allows for a considerable degree of selection bias that limits robust and clinically sound conclusions. B. Probability of mortality and morbidity analysis of sigmoidectomy for perforated diverticulitis: Operative strategies for diverticular peritonitis: a decision analysis between primary resection and anastomosis versus Hartmann's procedures. Constantinides VA et al. Ann. Surg. 2007, 245: [9] Summary of results: A total of 135 primary resection and anastomosis (PRA), 126 primary anastomoses with defunctioning stoma (PADS), and 6619 Hartmann s procedures (HP) were considered. The probability of morbidity and mortality was 55% and 30% for PRA, 40% and 25% for PADS, and 35% and 20% for HP, respectively. C. Systematic review (unpublished): Laparoscopic lavage for generalised peritonitis due to perforated diverticulitis. [18, 22 28] Search terms: diverticular disease / diverticulitis and peritonitis and laparoscopy, diverticular disease / diverticulitis and peritonitis and laparoscopy and lavage Outcome: The primary end point was postoperative mortality. Secondary end points included surgical and medical morbidity. Databases used and number of manuscripts retrieved: A MEDLINE, Ovid, Embase, and LADIES-trial, version 8, dd

16 Cochrane database search was performed on all studies published between 1984 and Population: Eight studies matched the selection criteria and were suitable for analysis. The study design was retrospective in six studies and prospective in two. These studies included in total 207 patients. Results: A weighted mean mortality and morbidity rate was calculated amounting resp. 1.3% and 6.7%. Conclusions: Mortality and morbidity rates are low for laparoscopic lavage for generalised purulent peritonitis due to perforated diverticulitis. The retrospective design of most studies inevitably causes bias. A prospective study comparing lavage with the standard of care e.g. sigmoidectomy is warranted Group size calculation One year stoma free survival (the former DIVA arm): using log rank statistics with a power of 90% and a type I error of 5%, 212 evaluable patients are needed to significantly demonstrate a difference in stoma free survival between both treatment arms. When corrected for loss to follow up a total of 233 may be warranted. Wich means 116 patients per study arm (feacal or purulent peritonitis)the power analysis is based on the existing literature, in which the suspected postoperative mortality after Hartmann s procedure and sigmoid resection with primary anastomosis are equally high (+ 15%).[7] About 60% of the patients that had undergone Hartmann s will have their stoma reversed.[12,13]. Corrected for expected mortality before reversal results in + 50%. Patients with a protective loop ileostomy after primary anastomosis will have their enterostomy reversed in over 85%.[13] Corrected for expected mortality before reversal results in 72% of the initial patients Cost effectiveness Measurements: health care utilisation, travel and lost productivity due to absence from work or decreased performance at work will be collected with a clinical report form, complemented with a patient questionnaire based on the Trimbos/iMTA Labour and Health questionnaire (Hakkaart van Rooijen, 2000) administered at each follow up visit. Data on general well being (health related quality of life) will be also collected at each visit with the EQ 5D (Brooks, 2004). Health state utilities to estimate QALY s will be derived from an EQ 5D measurement at baseline, as well as the follow up assessments. Utility values for EQ 5D scores will be based on UK LADIES-trial, version 8, dd

17 estimates (Dolan,1997). Utility scores will be uniformly interpolated, assuming constant health state between subsequent assessments. Unit costs: unit costs will be estimated according to the Dutch guideline on (unit) costing in health care (Oostenbrink, Bouwmans, Koopmanschap, & Rutten, 2004). For surgical and diagnostic procedures, we will estimate unit costs for one academic and one non academic hospital. GP visits, medical specialist visits, paramedical care, home care, and travelling will be valued based on the guideline prices. Medication costs will be valued by their market prices (van Loenen, 2007). The friction cost method will be used to estimated the duration of lost productivity, age adjusted average daily wages will be used to value this duration. Statistical analysis: as most volumes of resource utilisation follow a skewed distribution, differences between the two study groups will be statistically evaluated with bias corrected accelerated bootstrap analyses (Barber, 2000). Mean total costs for the two groups will be reported, as mean costs per patient are relevant from a societal perspective. Economic evaluation: cost effectiveness will be evaluated by calculating incremental the cost effectiveness ratios (costs /effects), with proportion of poor outcome as the effect parameter. In addition, in a cost utility analysis, costs per QALY will be estimated based on costs / QALY s. A cost acceptability curve will be generated to indicate the probability that the more effective treatment is also costeffective for a range of willingness to pay values. Robustness of the results for uncertainty in the assumptions will be evaluated in sensitivity analyses, including Dutch health states (Lamers, 2005) instead of the UK based model in the main analyses; a linear interpolation between EQ 5D measurements instead of uniform interpolation; and varying unit costs for pertinent volumes of health care utilisation (e.g. therapy costs, productivity costs). In model based analyses using data from literature about these middle to longterm effects and costs associated with PTSD patients we will extrapolate the results to estimate the effectiveness and cost effectiveness for both treatment options on long term (3 to 5 years) Accounting for missing data All clinical data will be entered by the clinical investigator in an online Case Record Form. Data from questionnaires sent to the patients will be entered by the trial coordinator in the online Case Record From. Missing data will be revealed by LADIES-trial, version 8, dd

18 data validation of the CRF and will be delivered for resolution with Data Clarification Forms (DCFs). The trial coordinator will receive this form and hereafter contact the clinical investigator to complete the online CRF by means of the DCF. Completed DCFs will be returned by fax or mail to data management for updating the database. Missing data on the sent questionnaires will be solved by the trial coordinator contacting the patient by telephone and hereafter completing the CRF Data and record keeping The clinical investigator is asked to keep a list with full names and addresses of all patients participating in the trial, giving reference to the patient records. All investigation related essential documents will be retained by the clinical investigator for at least five years. The structure of the database will be based on the online Case Record Form. Discrepancies revealed by data validation will be delivered for resolution with a Data Clarification Form (DCF). Completed DCFs will be returned by fax or mail to data management for updating the database. The clean database will be locked after all detected discrepancies have been solved and database has been updated accordingly. Only authorised and documented updates are possible after the database lock Deviations from statistical plan When certain data is missing and can not be completed by the procedure described above, the initial statistical plan must be deviated to be able to perform statistical analysis. Deviations from the statistical plan will be recorded in the notes to file section of the Trial Master File. 3. Patient Safety 3.1 Serious Adverse Events In accordance to the guidelines for Good Clinical Practice, all Serious Adverse Events (SAEs) need to be reported to the Medical Ethics Committee in the Academic Medical Center. SAEs are specified as any medical event that - leads to death - is life threatening - makes admission to a hospital or an extended hospital stay inevitable. - leads to serious or persistent disability (for work). Any other important medical event that is not mentioned above but can threaten the patient or requires intervention to prevent any of the above, is considered a SAE. LADIES-trial, version 8, dd

19 3.2.1 Reporting Serious Adverse Events The principal investigator is required to report to the Medical Ethics Committee, to all participating centres and to competent authorities all findings that can - threat the safety of the patients, - threat the continuation of the study - likely change the initial positive advice from the Medical Ethics Committee on continuation of the study. The local investigator is required to notify the principal investigator in every case of a Serious Adverse Event, within 24 hours after becoming aware of the event. These SAEs include the previously defined mortality and major morbidity: mortality reintervention wound dehiscence incisional hernia abscess needing percutaneous drainage urosepsis myocardial infarction renal failure respiratory insufficiency This can be done at the online CRF, by filling in the SAE section. The principal investigator gets notified and will inform the medical ethics committee from the Academic Medical Center within 24 hours Predicted Serious Adverse Events Not every SAE needs to be reported. Predicted SAEs need to be filled in in the regular Case Record Form, at the time of discharge of the patient, or 14 days after initial surgery, whichever comes first. These SAEs will not need to be directly reported. Based on the Case Record Forms, the principal investigator will create linelisting and report predicted SAE s to the Medical Ethics Committee every six months. Predicted SAEs include: wound infection non septical urinary tract infection pneumonia without respiratory insufficiency cardiac disease abscess without drainage ileus without reintervention thrombosis delirium decubitus urine retention cellulitis as a consequence of infusion electrolyte disorder with suppletion overhydration with diuretic use The consequences of all SAEs will need to be reported in the Case Record Form. This consequences can be reported as: additional therapy, reintervention, ICUadmission and result: LADIES-trial, version 8, dd

20 - patient died from SAE - patient is left with permanent damage from SAE - patient is cured from SAE 3.3 Adverse events Adverse events will only be recorded at the visit of the outpatient department. Adverse events that occurs during the clinical episodes, will not be recorded, because this group of events is not relevant for patient safety nor for final analysis of trial data. The group of investigated patients is highly ill. Therefore recording of all signs and symptoms will put a high amount of administrative labour to the clinical investigators, while the patients safety will not benefit from these recordings. At the visit of the outpatient department postoperatively, adverse events can be recorded by the clinical investigator in the online CRF. At this time remaining signs and symptoms deem relevant. 3.4 Data safety monitoring committee A data safety monitoring committee (DSMC) will guard the safety of the included patients, give advice on continuation of the study upon superiority of one of the types of treatment, and will guard the methodological quality of the study. The DSMC will consist of clinical, epidemiological and statistical experts, that are independent and have no personal interest in the results of the trial. Reports on meetings from the committee are confidential to prevent influence on the course of the trial. De DSMC of the Ladies trial will consist of Prof.dr. D. Legemate (surgeon and clinical epidemiologist) Prof.dr. J.F. Bartelsman (gastroenterologist) Dr. D.W. Meijer (contract research organisator) Dr. J.B. Reitsma (clinical epidemiologist) The clinical epidemiologist of the Ladies trial will create a report on the study after every 25 included patients, and present this tot the DSMC in an initial face to face meeting. Following reports will be written, unless the committee decides to organise a face to face meeting. The report consist on the number of included patients, in relation to elapsed time and specified per participating hospital. Further on the baseline characteristics from the included patients, and an overview from the major adverse events, classified as: - mortality within four weeks after initial surgery LADIES-trial, version 8, dd

21 - major morbidity as defined in the protocol. The DSMC will write a short report in which the considerations are summarised. This report will be sent to the project leaders. 3.5 Termination of the study The entire clinical investigation will be terminated if a serious treatment related adverse event occurs, making it impossible to recruit new patients or to continue the treatment of patients already recruited for medical or ethical reasons. The clinical investigation in an individual patient will be terminated in case of 1. insufficient compliance 2. a patients request 3. investigators request because of the onset of life threatening adverse event. 4. changes in health status incompatible with continued participation in the clinical investigation, as judged by the clinical investigator. 3.6 Ethical and legal requirements The protocol will be submitted to the appropiately structured Ethics Committe. Approval for the clinical investigation plan at each center has to be obtained prior to start of enrollment. 3.7 Insurance All patients participating in this study need to be insured for death or injury caused by the study that has occured within four years after participation. Patients included in one of the participating centers are to be ensured by the Academic Medical Center, unless the participating center is attached to the ensurance company CentraMed B.V. or is a university medical center. The insurance certificate is provided by the Academic Medical Center Medical Research bureau or by the participating center (for CentraMed members and university medical centers), and is issued to the Medical Ethics Committee of the Academic Medical Center by the investigator. The Medical Ethics Committee can only agree on participation of a center upon receival of this certificate and a positive advice of the board of directors from the participating center. 4. Quality control Monitoring will ensure that the investigation is conducted in accordance with this protocol, including all amendments, Good Clinical Practice and relevant regulatory requirements. LADIES-trial, version 8, dd

22 Monitoring will involve frequent visits to participating centers to verify good management of patients and the clinical investigation treatment, to observe procedures, to supervise the investigation for quality control purposes, to check presence of required documents, informed consent and for purpose of source data verification. There will also be frequent telephone contact and written communication between monitor and clinical investigator. The participating centers will permit trial related monitoring, audits, IRB/EC review, and regulatory inspections, providing direct access to source data and documents. 5. Publication policy The project leaders are committed to presenting the results of this trial in an appropiate scientific peer reviewed journal. No single center data will be published prior to the publication of the clinical trial data in its entirety. LADIES-trial, version 8, dd

23 6. References 1. Sandler RS, Everhart JE, Donowitz M, Adams E, Cronin K, Goodman C, Gemmen E, Shah S, Avdic A, Rubin R. The burden of selected digestive diseases in the United States. Gastroenterology 2002;122: Kang JY, Hoare J, Tinto A, Subramanian S, Ellis C, Majeed A, Melville D, Maxwell JD. Diverticular disease of the colon on the rise: a study of hospital admissions in England between 1989/1990 and 1999/2000. Aliment Pharmacol Ther 2003;17: Morris CR, Harvey IM, Stebbings WSL, Hart AR. Incidence of perforated diverticulitis and risk factors for death in a UK population. Br J Surg. 2008;95: Vermeulen J, Gosselink MP, Hop WCJ, Lange JF, Coene PPLO, Harst E van der, Weidema WF, Mannaerst GHH. Prognostische factoren voor ziekenhuissterfte na een spoedoperatie van acuut geperforeerde diverticulitis. Ned Tijdschr Geneeskd 2009 [in press] 5. Constantinides VA, Tekkis PP, Senapati A. Prospective multicentre evaluation of adverse outcomes following treatment for complicated diverticular disease. BJS 2006;93: Abbas S. Resection and primary anastomosis in acute complicated diverticulitis, a systematic review of the literature. Int J Colorectal Dis Apr;22(4): : Chirurgie Sep;123(4): Salem L, Flum DR. Primary anastomoss or Hartmann s procedure for patients with diverticular peritonitis? A systematic review. Dis Colon Rectum 2004;47: Constantinides VA, Tekkis PP, Athanasiou T, Aziz O, Purkayastha S, Remzi FH, Fazio VW, Aydin N, Darzi A, Senapati A. Primary resection with anastomosis vs. Hartmann's procedure in nonelective surgery for acute colonic diverticulitis: a systematic review. Dis Colon Rectum Jul;49(7): Constantinides VA, Heriot A, Remzi F, Darzi A, Senapati A, Fazio VW, Tekkis PP. Operative strategies for diverticular peritonitis: a decision analysis between primary resection and anastomosis versus Hartmann's procedures. Ann Surg. 2007;245: Gooszen AW, Tollenaar RA, Geelkerken RH, Smeets HJ, Bemelman WA, Van Schaardenburgh P, Gooszen HG. Prospective study of primary anastomosis following sigmoid resection for suspected acute complicated diverticular disease. Br J Surg May;88(5): Gooszen AW, Geelkerken RH, Tollenaar RA, Timmermans DR, Kievit J, Gooszen HG. Operatieve strategie bij de acute of electieve sigmoidresectie in Nederland; enquete op basis van een marketing model. NTVG 1994;138: Maggard MA, Zingmond D, O Connell JB, Ko CY (2004). LADIES-trial, version 8, dd

24 What proportion of patients with an ostomy (for diverticulitis) get reversed? Am Surg;70: Vermeulen J, Coene PP, Van Hout NM, van der Harst E, Gosselink MP, Mannaerts GH, Weidema WF, Lange JF. Restoration of Bowel Continuity after Surgery for Acute Perforated Diverticulitis Should Hartmann's procedure be considered a one stage procedure? Colorectal Dis Aug 21 [Epub ahead of print] 14. Banerjee S. Leather AJM, Rennie JA, Samano N, Gonzalez JG, Papagrigoriadis S. Feasibility and morbidity of reversal of Hartmann s. Colorectal Dis 2005;7: Gooszen AW, Geelkerken RH, Hermans J, Lagaay MB, Gooszen HG. Quality of life with a temporary stoma: ileostomy vs. colostomy. Dis Colon Rectum. 2000;43: Matthiessen P, Hallböök O, Rutegård J, Simert G, Sjödahl R. Defunctioning stoma reduces symptomatic anastomotic leakage after low anterior resection of the rectum for cancer: a randomized multicenter trial. Ann Surg. 2007;246: Bell C, Asolati M, Hamilton E, Fleming J, Nwariaku F, Sarosi G, Anthony T. A comparison of complications associated with colostomy reversal versus ileostomy reversal. Am J Surg. 2005;190: Myers E, Hurley M, O'Sullivan GC, Kavanagh D, Wilson I, Winter DC. Laparoscopic peritoneal lavage for generalised peritonitis due to perforated diverticulitis. Br J Surg Jan;95(1): Dutch hospital statistics at Janes S, Meagher A, Frizelle F. Management of diverticulitis. BMJ 2006;332: Makela J, Kiviniemi H, Laitinen S. Prevalence of perforated sigmoid diverticulitis is increasing. Dis Colon Rectum 2002;45: O'Sullivan GC, Murphy D, O'Brien MG, Ireland A. Laparoscopic management of generalised peritonitis due to perforated colonic diverticula. Am J Surg Apr;171(4): Da Rold AR, Guerriero S, Fiamingo P, Pariset S, Veroux M, Pilon F, Tosato S, Ruffolo C, Tedeschi U. Laparoscopic colorrhaphy, irrigation and drainage in the treatment of complicated acute diverticulitis: initial experience. Chir Ital Jan Feb;56(1): Franklin ME Jr, Portillo G, Treviño JM, Gonzalez JJ, Glass JL. Long term experience with the laparoscopic approach to perforated diverticulitis plus generalised peritonitis. World J Surg Jul;32(7): Faranda C, Barrat C, Catheline JM, Champault GG. Two stage laparoscopic management of generalised peritonitis due to perforated sigmoid diverticula: eighteen cases. Surg Laparosc Endosc Percutan Tech Jun;10(3):135 8; discussion Bretagnol F, Pautrat K, Mor C, Benchellal Z, Huten N, de Calan L. Emergency laparoscopic management of perforated sigmoid diverticulitis: a LADIES-trial, version 8, dd