DIVERTI

Incomplete

Bridoux, V. "Hartmann’s Procedure or Primary Anastamosis for Generalized Peritonitis due to Perforated Diverticulitis: A Prospective Multicenter Randomized Trial". JACS (Journal of the American College of Surgeons). 2017. 225(6):798-805.
PubMed • Full text • PDF

[[Category:]]

Clinical Question

In patients with undergoing emergency sigmoidectomy with purulent/feculent peritonitis, does primary anastamosis with diverting ileostomy decrease mortality compared to Hartmann’s Procedure (discontinuity with end colostomy)?

Bottom Line

For operative approach to emergent purulent diverticulitis (Hinchey III or IV), primary anastomosis with ostomy likely leads to an increased likelihood of ostomy reversal with no difference in outcomes compared to Hartmann’s procedure.

Major Points

Diverticulitis is a common disease that can lead to life threatening sepsis and require surgery as often as 25% for source control . Historically, when gross contamination from perforation of the colon is present, the surgical approach had been diversion, most recently using a Hartmann’s procedure, where the sigmoid colon is resected, with the proximal end brought out in an end colostomy, and the distal end closed to create a rectal stump, thereby leaving the patient in anatomic discontinuity. The patient is then brought back at a later date for colostomy reversal and reanastamosis.

Prior data, hindered by suboptimal recruitment, emerged that supported primary anastamosis with diversion as a safe alternative to the Hartmann’s procedure, and appeared to demonstrate an improvement in mortality and hospital length of stay .

The DIVERTI trial is a landmark paper because it represents the best to date prospective, multicenter, RCT in emergent diverticulitis. The DIVERTI trial ultimately proved underpowered and failed to show the specific improvement in mortality, it did demonstrate a 30% absolute risk reduction in definitive stoma with primary anastamosis.

Guidelines

• No major guidelines have incorporated this work to date

Design

• N=102 (52 Hartmann Pouch, 50 primary anastamosis)
• Setting: 7 sites tertiary centers/affiliated centers in France
• Mean follow-up: 18 months
• Enrollment: June 2008 - May 2012
• Analysis: Intention-to-treat, T-test, Mann-Whitey, Fisher’s exacted, chi-squared as dictated by data
  • Stoma reversal analysis excluded patients without stoma
• Primary outcome: Rate of mortality after primary anastomosis , ostomy reversal

Population

Inclusion Criteria

• adults > 18 years old
• Hinchey stage III (purulent) or IV (feculant) peritonitis
• clinical assessment, CT, laparotomy confirming diagnosis of diverticulitis

Exclusion Criteria

• unable to provide consent
• physical states precluding operation (septic shock, multi-visceral organ failure)

Baseline Characteristics

“Data shown for all patients (n=102), mean and range displayed”

• Age 61 years (25-93)
• Male 50%
• BMI 26.8 (19.3-44.6)
• ASA grade > 1: 86.3%
• Mannheim Peritonitis Index (MPI) Score 26 (16-43)
• Hinchey stage IV 19.8%
• First surgeon resident 46.5%
• Night operation 40.2%

• • MPI > 26 is considered to have a high mortality rate

Interventions

• Primary anastomosis
  • via midline laparotomy
  • anastamosis mechanical or manual
  • end-to-side or end-to-end.
  • diverting ileostomy or colostomy were allowed
  • Stoma reversal 3 months after index operation, checked with barium enema
  • 15/45 patients in the primary anastamosis group did not have an ostomy placed, which constituted a protocol violation (23 colostomy, 7 ileostomy)

• Hartmann Pouch
  • sigmoid resection with rectal closure and end colostomy
  • rectal stump washed
  • reversal at 6 months (laparotomy or laparoscopy)

Outcomes

Comparisons are Hartmann’s procedure vs. Primary Anastamosis.

Primary Outcomes

Mortality

7.7% vs. 4.0%, P=0.68

Definitive stoma

35.4% (17/48) vs. 4% (2/48), P<0.001

Secondary Outcomes

• Presented outcomes at primary operation

Length of stay in ICU/hospital Stoma reversal rates

Operating time (minutes)

235 vs. 198, P=.26

ICU length of stay (days)

8.5 vs. 9.5 (P=0.20)

Hospital length of stay (days)

16 vs. 15.5, P=0.30

Overall morbidity

39.2% vs. 44%, P=.42

Serious complication (Clavien-Dindo III-IV)

17.3% vs. 16%, P=.85

• Secondary outcomes were reevaluated at secondary outcome. Operating time did ultimately differ in the second operation, significantly favoring reduced operative times during the primary anastomosis. No other differences in terms of complication rate, morbidity, or hospital stay were noted.

Subgroup Analysis

• No significant subgroup analysis was performed. The authors report morbidity and a non-operation related mortality in those without a diverting ostomy.

Adverse Events

• 1 patient underwent total proctocolectomy for malignant obstructing colon cancer
• 2 patients assigned to primary anastamosis had a Hartmann’s due to hemodynamic instability (which precludes an anastamosis)
• 3 patients assigned to primary anastamosis had a Hartmann’s due to technical factors

Criticisms

• Difficulty recruiting patients due to emergent conditions may have biased the cohort
• Hinchey III undergoing evaluation for peritoneal lavage
• No data on excluded patients
• High rate of protocol violation with primary anastamosis without ostomy

Funding

• French Ministry of Health