Laparoscopic Lavage in the Management of Hinchey III/IV Diverticulitis

Abstract

Diverticulitis manifestations may cover a spectrum of mild local inflammation to diffuse feculent peritonitis. Up to 35% of patients presenting with diverticulitis will have purulent (Hinchey grade III) or feculent (Hinchey grade IV) contamination of the abdomen, with a high-associated morbidity and mortality. Surgical management may involve segmental resection with or without restoration of bowel continuity. However, emergency resection for diverticulitis can be associated with high mortality rates, as well as low stoma reversal rates at 1 year. Therefore, laparoscopic peritoneal lavage has been proposed for use in selected patients with purulent peritonitis. The topic of laparoscopic peritoneal lavage for the treatment of perforated diverticulitis in the literature has been controversial. Our review of the recent data show that laparoscopic lavage may be safe and feasible in select patients with similar rates of mortality and major morbidity. There is, however, a concern regarding an associated higher rate of postoperative abscess and early reintervention risk.

Acquired colonic diverticulosis is a disease process characterized by outpouchings of the colon which have the potential for complications that may lead to serious morbidity and sometimes death. It is an increasingly common condition in the Western world. Given that it affects 5 to 10% of the population over the age of 45 years, and almost 80% of those over 85 years, complications of diverticulosis are becoming more prevalent as the population in these industrialized countries age. ^{1 2 3} In these nations, diets low in fiber and high in refined carbohydrates contribute to increased colonic transit time and intracolonic pressures. ^{4 5} These changes, in combination with risk factors, such as obesity and smoking, lead to the formation of diverticula of colonic mucosa and submucosa at areas of weakness where the bowel wall is penetrated by the vasa recta. ² Clinical diverticulitis represents micro- or macroperforation of these thin-walled diverticula. These patients have progressed through the sequence of obstruction of the diverticular neck, mucous secretion and bacterial overgrowth, increased intraluminal pressure, and ischemia, ultimately resulting in perforation. ^{6 7}

Diverticulitis manifestations cover a spectrum from mild local inflammation to diffuse feculent peritonitis. Complicated diverticulitis refers to disease involving phlegmon, abscess, fistula, stricture, or obstruction. Hinchey classification, originally described in 1978, suggested a disease severity staging which is still used today to classify acute diverticulitis ⁸ ( Fig. 1 ). Up to 35% of patients will have purulent (Hinchey grade III) or feculent (Hinchey grade IV) contamination of the abdomen, with a high-associated morbidity and mortality. ⁹

Fig. 1.

Hinchey classification of diverticulitis.

Current evidence-based practice parameters published in 2014 by the American Society of Colon and Rectal Surgeons (ASCRS) state that the standard of care for Hinchey grade III and IV diverticulitis is segmental resection with or without restoration of bowel continuity. ¹⁰ However, the mortality associated with emergency resection for diverticulitis can be as high as 35% with a colostomy reversal rate of only 63% at 1 year. ^{11 12} Therefore, laparoscopic peritoneal lavage without resection has been proposed for use in select patients with purulent peritonitis. We aim to review the history of surgical management of Hinchey grade III and IV diverticula, and the evidence basis for use of laparoscopic lavage.

History of Surgical Techniques for Management of Hinchey Grade III/IV Diverticulitis

The first description of surgical management of generalized peritonitis from diverticulitis was by Lockhart-Mummery in 1910. ^{13 14} His approach involved abdominal washout, drainage, and suture repair of the colonic perforation. The addition of selective use of a diverting proximal colostomy in the setting of obstruction was advocated by William Mayo in a series of case reports. ¹⁵ If medical treatment was not successful, he then recommended elective sigmoid resection to prevent complications of diverticulitis. ^{15 16} Proximal transverse colostomy with abdominal washout but without the resection of the diseased segment of bowel was a common method of treatment in the early 1900s because of the morbidity associated with sigmoid resection. ¹⁶ If proximal diversion resulted in successful healing of the inflamed bowel, patients would often undergo closure of the colostomy without resection. However, it was often observed that closure of the colostomy without resection did not resolve the inflammatory process in a significant proportion of patients and thus the traditional three stage approach was developed. ¹⁷ The three stage surgery was as follows: (1) proximal colostomy with drainage of abscess, (2) sigmoid resection with maintenance of the proximal colostomy to protect the anastomosis, and (3) closure of the colostomy. ^{17 18 19}

Although this approach was the standard of care until the late 1940s, it was associated with high morbidity and mortality and an extended duration of therapy. ^{13 20} In addition, improvements in antibiotic therapy in the first half of the 20th century promoted the idea of eliminating the source of sepsis at the first operation, a strategy that later came to be known as obtaining “source control.” ^{21 22} In the 1970s, experimentation with multiple other approaches to the three-stage procedure led to the popularization of a technique originally developed by the French surgeon Henri Hartmann in 1923. ²³ Hartmann's procedure, pioneered for use in rectosigmoid malignancy, consisted of sigmoid resection and a terminal colostomy, with oversewing of the distal rectal stump. ²⁴ This was later modified to a two-stage technique with colostomy closure and restoration of intestinal continuity. ^{25 26} In the 1980s and 1990s landmark, systematic reviews and randomized controlled trials were published in support of Hartmann's procedure for diverticulitis, and this approach became the standard used to the present day. ^{27 28 29} Successive randomized controlled trials also explored the role of resection with primary anastomosis and proximal diverting loop ileostomy, and found no significant difference between this and Hartmann's procedure in terms of morbidity and mortality. ³⁰ However, they did find a significantly higher probability of stoma reversal in the ileostomy versus colostomy group (90 vs. 57%, p  = 0.005). ³⁰ Current ASCRS practice parameters espouse the use of segmental colectomy in the first operation, with or without immediate restoration of bowel continuity depending on patient factors, intraoperative conditions, and surgeon preference. ¹⁰

The Emergence of Laparoscopic Lavage

Contemporary advances in laparoscopy coupled with enhanced patient recovery and reduced morbidity compared with an open surgical approach has led to a generalized adoption of more minimally invasive therapies in surgery. Laparoscopic lavage for diverticulitis with generalized peritonitis has been proposed as a valid alternative to immediate segmental colectomy for those patients in whom the colonic perforation has already sealed. ³¹ Theoretically, this approach would significantly reduce the inflammatory environment around the intestine and avoid the morbidity of bowel resection and open surgery in the emergency setting. The majority of patients later undergo colonic evaluation and interval colectomy after the acute phase of disease has passed. However this approach has proven controversial.

A method of intraoperative irrigation of the peritoneal cavity with saline solution was first described by Rehn in 1902 and was shown to reduce mortality by 38%. ³² Lavage significantly reduces the level of endotoxin in the peritoneal fluid of a septic patient and has a dialysis-like effect to improve physiologic parameters. ³³ The first laparoscopic lavage for purulent diverticulitis was described by O'Sullivan et al in 1996. ³⁴ His case series performed in Ireland, featuring eight patients with purulent diverticulitis diagnosed by laparoscopy. The peritoneum was “thoroughly and vigorously” irrigated with warm 0.9% saline until clear of purulent fluid. Pelvic drains were then placed to drain any pelvic abscesses. All patients made a complete recovery with no mortalities and a mean hospital stay of 10 days. ³⁴

There have since been several observational studies done on this subject. Myers conducted a prospective multi-institutional study in 2008 of 100 patients presenting with diverticulitis causing generalized peritonitis. ³⁵ Ninety-two of these patients were managed with laparoscopic lavage. Eight patients with Hinchey grade IV diverticulitis were converted to open Hartmann's procedure. Risk of morbidity and mortality was low at 4 and 3%, respectively, with only 2% of patients representing with diverticulitis during the median 36-month follow-up. ³⁵ The authors concluded that laparoscopic peritoneal lavage was feasible and safe in the short-term postoperative period.

One of the first large retrospective population studies to investigate outcomes after laparoscopic lavage for diverticulitis was published in 2012. ³⁶ This study analyzed an Irish national database spanning a 14-year interval. During this time period, there were 28,480 emergency admissions for acute diverticulitis of which 2,455 (8.6%) patients underwent surgery for diverticulitis. Of patients who had surgical intervention, a total of 427 patients underwent laparoscopic peritoneal lavage. They found that patients selected for laparoscopic lavage compared with laparotomy and surgical resection had lower mortality (4 vs. 10.4%, p  < 0.001), less complications (14.4 vs. 25%, p  = 0.001) and shorter length of stay (10 vs. 20 days, p  = 0.0001). ³⁶ Postoperative complications included urinary tract infections, myocardial infarctions, pneumonia, postoperative wound infections, respiratory failure, atelectasis, Clostridium difficile infection, and venous thromboembolism. Patient factors predictive of mortality included those older than 65 years of age (odds ratio [OR] = 4.1, p  = 0.0001), chronic kidney disease (OR = 8.0, p  = 0.0001), and rheumatological disease (OR = 7.3, p  = 0.05). ³⁶ This study was the largest series of lavage cases at the time but was limited by its retrospective nature and methodology and thus subject to selection bias. The patients in the lavage group were younger (60.7 vs. 64.8 years) and had lower Charlson's comorbidity scores. ³⁶ In addition, disease severity was not included in the analysis likely indicating patient with more severe disease (Hinchey grade IV) underwent surgical resection. Nonetheless, the authors were able to show in this large population study that laparoscopic lavage is gaining popularity with the proportion of patients undergoing lavage increasing from 8 to 17% in their study period. ³⁶ They do not conclude that laparoscopic lavage was superior to colonic resection in acute diverticulitis but in those carefully selected patients, this approach could be done with low morbidity and mortality compared with those who underwent resection.

The 2014 ASCRS practice parameters for the treatment of sigmoid diverticulitis reviewed data on laparoscopic lavage and concluded that “in patients with purulent or feculent peritonitis, operative therapy without resection is generally not an appropriate alternative to colectomy,” which was a strong recommendation based on low-quality evidence. ¹⁰ The authors reviewed the lavage literature including previous smaller retrospective reports on the subject through 2011, as well as the large retrospective population study from 2012, using data from the Irish national database as above. They concluded that the poor quality of the existing lavage literature and inherent selection biases in observational studies are major obstacles in advocating the widespread adoption and practice of laparoscopic peritoneal lavage and that the safety of lavage for patients with purulent or feculent peritonitis has not been proven or disproven by the published studies to date. ^{10 37 38}

Since the publication of the ASCRS practice parameters, there have been four randomized controlled trials (The Ladies Trial, ^{37 38} SCANDIV, ³⁹ DILALA, ^{40 41} and Lap-LAND ⁴² ) that have been registered of which three have published their results ( Table 1 ).

Table 1. Summary RCT trial data on laparoscopic lavage for Hinchey grade III diverticulitis.

Trial	Study design	Sample size	Interventions	Outcomes (lavage vs. control)	Conclusions
Ladies LOLA-arm 37 38	RCT superiority	45 lavage 42 control Trial terminated early	Laparoscopic lavage (6 L) vs Hartmann's procedure	1) No superiority of composite endpoint of major morbidity and mortality within 12 months (67 vs. 60%, p  = 0.58) 2) Shorter operative time (60 vs. 120 minutes, p  = 0.001) 3) Increased risk of reintervention (36 vs. 7%, p  = 0.004) 4) No significant difference in length of stay	Lavage is not superior
SCANDIV 39	RCT superiority	101 lavage 96 control	Laparoscopic lavage (4 L) vs resection (lap/open ± stoma)	1) No significant difference in severe postoperative complications, length of stay, or quality of life at 90 days 2) Shorter operative times (72 vs. 149 minutes, p  < 0.001) 3) Increased risk of reoperation (20 vs. 6%, p  = 0.01)	Lavage is not superior
DILALA 40 41	RCT	38 lavage 35 control	Laparoscopic lavage (3 L) vs open Hartmann's procedure	1) Mortality and severe adverse events did not differ between groups. 2) Shorter total length of hospital stay within 12 months in the laparoscopic group (14 vs. 18 days) 3) Fewer patients in the laparoscopic group than in the Hartmann group had at least 1 reoperation (including elective cases) within 12 months. (27.9 vs. 62.5%, with relative risk reduction of 59%)	Laparoscopic lavage may be an appropriate treatment of choice for acute perforated diverticulitis with purulent peritonitis
LapLAND 42	RCT	300 total	Laparoscopic lavage (4 L) vs resection (±anastomosis)	Pending	Pending

Abbreviation: RCT, randomized controlled trial.

Randomized Controlled Trial Data for Laparoscopic Lavage

The Ladies Trial

The Ladies trial was a two-armed multicenter randomized controlled trial which included 34 teaching hospitals and 8 academic hospitals in Belgium, Italy, and the Netherlands in an attempt to provide evidence for laparoscopic peritoneal lavage and on the optimal resection strategy for patients with purulent and feculent diverticulitis. ^{37 38} The LOLA-arm (LaparOscopic LAvage) of the trial's objectives was to determine whether laparoscopic lavage is a safe and effective treatment for patients with purulent peritonitis. The study included all patients with perforated diverticulitis. Patients then underwent laparoscopy to determine if they had purulent or feculent diverticulitis. Patients with purulent diverticulitis were treated with laparoscopic lavage with >6 L warmed saline and drainage, Hartmann's procedure or sigmoid resection with primary anastomosis in a 2:1:1 ratio (LOLA-arm). Those with feculent diverticulitis were randomized in a 1:1 ratio in Hartmann's procedure or resection with primary anastomosis (DIVA-arm: perforated DIVerticulitis: sigmoid resection with or without Anastomosis).

The LOLA-arm was the first of the laparoscopic lavage randomized controlled trials to report their findings that were published July 2015 in the Lancet . ³⁸ The trial was conducted with an intention to treat principle with a calculated sample size of 264 patients (132:66:66 in each group) to detect an assumed difference of 15% in the composite endpoint of major morbidity and mortality (as determined from prior literature) between the two groups (laparoscopic lavage versus resection). ³⁸ The published trial included 90 patients who underwent randomization but was terminated early by their data and safety monitoring board due to high–short-term morbidity and reintervention rate. Major morbidity events included surgical reintervention, abdominal wall dehiscence, abscesses needing percutaneous drainage, urosepsis, myocardial infarction, renal failure, and respiratory insufficiency. The trial found that the rate of short-term serious adverse events (defined as within 30 days of surgery or in hospital) was much higher in the laparoscopic lavage arm than the surgical resection arm (39 vs .19%, p  = 0.04). ³⁸ The reoperation rate in the laparoscopic peritoneal lavage group was 20 versus 7% in the resection arm ( p  = 0.12). ³⁸ Prior to termination of the trial, early analysis showed that major morbidity and mortality at 12 months occurred similarly between the groups (67 vs. 60%, p  = 0.58). ³⁸ Of note, the recurrent diverticulitis rate in the laparoscopic arm was 20 versus 2% in the resection arm ( p  = 0.03) at 12 months. ³⁸

Due to the early termination of the study, the study concluded that laparoscopic lavage is not superior to sigmoid resection for the treatment of purulent diverticulitis.

SCANDIV Trial

The SCANDIV (Scandinavian Diverticulitis) trial was a two-group multicenter randomized clinical superiority trial published in the Journal of the American Medical Association (JAMA) in 2015. ³⁹ It also could not support laparoscopic lavage for treatment of perforated diverticulitis. This study enrolled 199 adult participants between February 2010 and June 2014 from 21 surgical centers in Sweden and Norway who were suspected clinically of having perforated diverticulitis requiring urgent surgery. Patients had peritonitis on exam with abdominal computed tomography (CT) scans showing free air with colonic thickening and pericolic inflammation. Patients with bowel obstruction or who were pregnant were excluded. Patients were randomized by a center-stratified block randomization approach to either laparoscopic peritoneal lavage or primary colon resection and analyzed on an intention-to-treat basis. The resection group could be laparoscopic or open, with or without anastomosis, and diverting ileostomy. Irrespective of preoperative randomization, all patients with Hinchey grade IV diverticulitis underwent a Hartmann's procedure.

The primary outcome of this study was the risk of severe complications 90 days after surgery. Severe complications were defined as those greater than Clavien–Dindo IIIa including complications that lead to a reintervention requiring general anesthesia (IIIb), a life-threatening organ dysfunction (IV) or death (V). ³⁹ Secondary outcomes included other postoperative complications, reoperations, length of operating time, length of postoperative stay, and 90-day postoperative quality of life.

The authors found no significant difference between the laparoscopic lavage and resection groups (all Hinchey grades) in severe postoperative complications at 90 days (31 vs. 26%, p  = 0.53). ³⁹ Interestingly, the observed severe complications trended toward being lower in the resection group than in the laparoscopic lavage group. The rate of death although not statistically significant was 13.9% (14/101 patients) in the laparoscopic lavage group and 11.5% (11/96 patients) in the resection group. The rate of reintervention requiring general anesthesia was also higher in the laparoscopic group than the resection group (12.9 vs. 8.3%). ³⁹

In the group of patients without feculent peritonitis (Hinchey grades I–III), statistically significant differences were only seen in the secondary outcomes. Patients undergoing laparoscopic lavage had significantly shorter operative times (72 vs. 149 minutes, p  < 0.001), lower blood loss (3 vs. 175 cc, p  < 0.001) and lower rates of superficial wound infection (1 vs. 9%, p  = 0.008). ³⁹ However, these did not translate into shorter hospital length of stay (6.5 vs. 7.5 days, p  = 0.19) or need for blood transfusion. In fact, laparoscopic lavage patients demonstrated an increased risk of requiring reoperation (20.3 vs. 5.7%, p  = 0.01) primarily for secondary peritonitis (40%) but also for missed carcinoma (27%). ³⁹ They also demonstrated a trend toward increased intra-abdominal abscess (16 vs. 7%, p  = 0.07), though this did not reach statistical significance. Although more resection patients received a stoma (69 vs. 16%, p  < 0.001), there was no significant difference in quality of life scores at 90 days between the groups. ³⁹

The group concluded that the use of laparoscopic lavage versus primary resection did not statistically reduce severe postoperative complications and had significant differences seen in secondary outcomes including higher rates of reoperation seen similarly in the early terminated LOLA-arm of the Ladies trial. They stated that their findings do not support laparoscopic lavage for treatment of perforated diverticulitis.

This study's methodology differed with previous studies' in its preoperative randomization to groups, thereby eliminating the risk of selectively excluding patients whose intra-abdominal condition may be deemed “unsuitable” for the study by the operating physician. Its study design was also pragmatic in that surgeons who would normally have operated on the patients did in fact do so, with no requirement for a minimum number of cases performed prior to joining the trial, making its results more generalizable to the greater provider population. This study was limited by lack of adjustment for multiple comparisons, the inclusion of Hinchey grade I and II patients, and the presence of more experienced surgeons in the resection group, all of which could be sources of bias.

DILALA Trial

The treatment of acute DILALA (DIverticulitis LAparoscopic Lavage) versus resection trial ^{40 41 43 44} was a prospective, multicenter, randomized controlled trial with initial results first published in the Annals of Surgery in 2016. ^{40 41} This trial included three publications with short- and long-term results. ^{41 43 44} The initial publication included only short-term outcomes. ⁴¹ Their primary endpoint was the number of reoperations within 12 months including elective sigmoid colectomies after laparoscopic lavage and continuity restoration after Hartmann's procedure and was reported after all patients have reached full follow-up. ⁴³

Patients were recruited from February 2010 to 2014 from nine surgical departments in Sweden and Denmark. Inclusion criteria were patients with clinical and radiological signs of diverticulitis in whom surgery was possible in light of comorbidities. Surgical emergency was decided by the attending surgeon and eligible patients were taken to the operating room for a diagnostic laparoscopy. Patients with Hinchey grade III diverticulitis were then intraoperatively randomized to laparoscopic lavage versus Hartmann's procedure. All other grades of diverticulitis or those with alternative pathology to explain the patient's presentation were excluded. There were 139 patients taken for diagnostic laparoscopy of which 83 had Hinchey grade III diverticulitis and underwent randomization. After further exclusions, 39 patients underwent laparoscopic lavage and 36 patients had a Hartmann's procedure. Laparoscopic lavage was performed with 3 L or more of warmed saline in all four quadrants until clear fluid was returned. In the Hartmann's procedure group, an open sigmoid resection with colostomy was performed via a midline incision. In both groups a drain was left in place during surgery and removed one or more days postoperatively, and antibiotics were continued.

The initial results from this trial compared short-term results of laparoscopic lavage versus Hartmann's procedure. The results of secondary endpoints in the laparoscopic lavage versus Hartmann's group included reduced operative time (68 vs. 154 minutes, p  < 0.0001), shorter hospital stay (6 vs. 9 days, p  < 0.04), but longer number of days with a drain in situ (3 vs. 2 days, p  < 0.02). ^{40 41} Mortality and reoperation within 30 days was not significantly different between the two groups. There were fewer colonic perforations reported in the laparoscopic group (5 vs. 50%, p  < 0.0001) suggesting the possibility of selection bias, though the authors felt this could be explained by the need for greater handling of the tissue to perform a resection versus a simpler lavage procedure.

Long-term follow-up from the DILALA study looked at the primary outcome of all reoperation within 12 months including elective operations. Secondary outcomes were number of reoperations, hospital readmissions, total length of hospital stay during the 12-month follow-up, postoperative adverse events, incisional hernia, bowel obstruction, mortality, persisting stoma at 12 months, and quality of life.

The authors concluded that patients in the laparoscopic arm had fewer reoperations within the 12 months (27.9 vs. 62.5%) than the Hartmann group. ⁴³ Of the 25 patients in the Hartmann group who needed reoperations, 21 of the patients underwent elective stoma reversal. Total length of hospital stay within 12 months was shorter for the laparoscopic group (14 vs. 18 days) than the Hartmann group. ⁴³ Seven patients in the laparoscopic group were readmitted because of diverticulitis and 21 in the Hartmann group were readmitted because of stoma reversal. ⁴³ Mortality and severe adverse events did not differ between groups. Quality of life data showed no significant differences between groups at any time point in the follow-up.

The authors conclude that laparoscopic lavage that reduced the need for reoperations had a similar safety profile to the Hartmann procedure, and may be an appropriate treatment of choice for acute perforated diverticulitis with purulent peritonitis. ⁴³

Evidence from Systematic Review and Meta-analysis

Several reviews have been published on this subject ( Table 2 ). These analyses generally agree that the quality of evidence is low because of risks of bias, issues with randomization in the emergency setting, and imprecision and heterogeneity in the inclusion criteria and endpoints. In some instances, this makes the trials barely combinable for analysis. ^{45 46} This may partially explain the difference in conclusions drawn by the studies despite analysis of the same trials.

Table 2. Summary of systematic reviews and meta-analyses comparing laparoscopic lavage to segmental colectomy for Hinchey grade III diverticulitis.

Study	Cirocchi et al 45	Penna et al 49	Ceresoli et al 46	Angenete et al 50	Marshall et al 51	Shaikh et al 52
Design	Systematic review and meta-analysis	Systematic review and meta-analysis	Meta-analysis	Meta-analysis	Systematic review	Systematic review and meta-analysis
Trials included	3 RCTs 37 38 39 40 41	3 RCTs 37 38 39 40 41 4 comparative studies 53 54 55 56	3 RCTS 37 38 39 40 41	3 RCTs 37 38 39	3 RCTs 37 38 39 40 41	3 RCTs 38 39 41 43
Main results (lavage vs. controls)
Operative time	NA	−79 minutes (95% CI: −105 to −52, p  < 0.001)	NA	NA	NA	−72 minutes (95% CI: −88 to −55.9, p  < 0.0001)
Postoperative intra-abdominal abscess	RR = 2.54 (95% CI: 1.34–4.83)	OR = 2.85 (95% CI: 1.52–5.34, p  = 0.001)	OR 3.50 (95% CI: 1.79–6.86)	NA	NA	OR = 4.1, (95% CI: 1.89–8.99, p  = 0.0004)
Postoperative wound infection	RR = 0.10 (95% CI: 0.02–0.51)	OR = 0.15 (95%CI: 0.05–0.43, p  < 0.001)	OR 0.14 (95% CI: 0.04–0.45, p  = 0.0009)	NA	NA	NA
Length of stay	−2.03 days (95% CI: −2.59 to −1.47)	−8.3 days (95% CI: −12.5 to −4.1, p  < 0.001)	NA	NA	NA	−2 days (95% CI: −2.6 to −0.5, p  = 0.005)
Morbidity (90 days)	NSD	NSD	NSD	NSD	NSD	NSD
Mortality (30 days)	NSD	NSD	NSD	NSD	NSD	NSD
Surgical reintervention (30–90 days)	At 1 month: NSD	At 1 month: NSD At 1 year: OR = 3.32 (95% CI: 1.73–6.38, p  < 0.001)	Index admission: OR = 3.75 (95% CI: 1.45–9.69) At 1 year: OR = 0.32 (95% CI: 0.17–0.60)	At 3 months: NSD At 1 year: OR = 0.54 (95% CI: 0.38–0.76)	At 1 month: RR = 3.03 (95% CI: 1.16–7.89)	NSD
Stoma rate (1 year)	NSD	Stoma risk: 14% (lavage) vs. 90% (resection). Stoma reversal by 12 months: 48% (lavage) vs. 74% (resection)	NSD	NA	NSD	NA
Quality of life	NA	NSD	NA	NA	NA	NA
Conclusions	Lavage not fundamentally inferior to traditional surgical resection though rate of postoperative abscess and ongoing sepsis is concerning	Lavage associated with increased risk of persistent abdominal sepsis, and should be used only on case-by-case basis	Lavage comparable in terms of mortality but has higher abscess and early reoperation risk. More studies required to define safety and appropriateness	Lavage is a valid alternative to resection with lower rate of 1-year reoperation after laparoscopic lavage	Lavage comparable in terms of mortality but has higher 30-day re-intervention rates	Lavage without significant differences in terms of mortality or major morbidity. However, lavage was associated with higher rates of postoperative abscess formation requiring more percutaneous drainage interventions

Abbreviations: CI, confidence interval; NSD, No statistical difference; NA, Not assessed through meta-analysis; OR, odds ratio; RCT, randomized controlled trial; RR, relative risk.

Nevertheless, some overall conclusions can be drawn. There is no significant difference between lavage and control groups in postoperative morbidity and mortality. However, it is concerning that given the purpose of the surgery is to control sepsis, the rate of postoperative abscess is reproducibly higher in the lavage group. In addition, there does appear to be a significantly higher risk of unplanned reoperation in the lavage group within the immediate postoperative 30-day period. It is likely that these two issues are related. The fact that the reoperation risk inverts at 1 year, with the lavage group having significantly fewer reoperations, likely reflects the role of elective stoma reversals in the resection group. Arguably there is a clinical difference between the significance of nonelective versus elective procedures to a patient's course. In addition, heterogeneity in the reoperation data could be explained by differential classification of reoperation in the trials. In some cases, postoperative percutaneous abscess drainage was not included as a reoperation, and in others, elective stoma reversals were. The SCANDIV trial made a clear distinction between elective and nonelective reintervention, and posited a significantly higher reoperation risk in the lavage versus resection group at 3 months (20 vs. 6%, p  = 0.01). ³⁹

Although all the trials included data on quality of life, the methods of measurement were not comparable and could therefore not be examined via meta-analysis. Nevertheless the trials have individually indicated no significant difference between groups.

Economic Analysis

Cost analyses comparing laparoscopic lavage to resection in Hinchey grade III diverticulitis have been performed within the framework of both the Ladies and the DILALA trials. Both concluded a significant cost reduction supporting the use of laparoscopic lavage. The Ladies analysis utilized prospectively collected actual resource use per individual patient to calculate costs. ⁴⁷ It found a mean cost difference of €3,513 at 1 year postoperatively. ⁴⁷ The DILALA analysis derived clinical effectiveness and resource use data from the trial, and unit costs from Swedish sources, and analyzed them from the health care sector perspective. ⁴⁴ This study found a mean cost difference of €8,983 at 1 year postoperatively, and €19,794 over the patient's expected lifespan. ⁴⁴ Both studies' estimates behaved robustly during sensitivity analysis. The main contributors to differences in costs were noted to be postoperative reinterventions in the lavage group, and stoma reversals in the resection group.

Conclusions and Recommendations

The role of laparoscopic peritoneal lavage in the surgical management of diverticulitis presenting with generalized peritonitis is controversial. Limitations of lavage clearly include the risk of missing an incompletely sealed perforation (30%), the risk of missing contained fecal peritonitis (10%), and the risk of missing a cancer (10%). ⁴⁸ There is general consensus that Hinchey grade IV peritonitis with visible perforation and feculent contamination is the best treated with segmental resection. However, the data surrounding the use of lavage in Hinchey grade III purulent peritonitis is less straightforward and more heterogeneous, with conclusions coming out of large randomized controlled trials and meta-analyses seeming contradictory.

However after careful review, it is evident that when the results of these studies are appropriately qualified, some trends clearly appear. Overall, postoperative morbidity shows no significant difference between groups. Superficial wound infections are 10 times less likely to occur in the lavage group but deep space abscesses are approximately three folds more common in this group. This likely contributes to a three-fold increased risk of reintervention in some analyses within 30 days in the lavage group. Despite this, lavage patients do not appear to have increased mortality compared with the resection group. They also benefit from an overall reduction in length of stay, which likely contributes to a greater cost effectiveness of this technique. However, at 1 year, the reintervention risk in some analyses was shown to be two to three folds lower in the lavage group, likely reflecting the weight of elective stoma reversals in the resection group. These produced an overall result of no significant difference between groups in stoma presence at 1 year. Essential data on how these differences in early versus late reinterventions affect overall quality of life indicators are currently lacking.

Conclusion

Randomized controlled studies with more consensus endpoints, randomization schema less susceptible to bias, a stricter definition of “postoperative reintervention,” and use of consensus tools to measure quality of life would be very helpful in attempting to clarify the data. Nevertheless, it does appear that there may be a role for laparoscopic lavage in the management of Hinchey grade III diverticulitis and this should be discussed on an individualized basis. If offered as a treatment in these selected cases, patients should have informed consent regarding the data on laparoscopic lavage including the possible higher need for reintervention. However, the existing available evidence for this technique does not currently meet the criteria required to recommend routine lavage for standard practice.