|
|
Back to 2017 Program and Abstracts
WHAT ARE RATES OF SIGMOID RESECTION WITH PRIMARY ANASTOMOSIS AND DIVERTING LOOP ILEOSTOMY VERSUS END COLOSTOMY IN THE REAL WORLD? RESULTS OF A COLORECTAL SURGERY COLLABORATIVE AUDIT
Christy E. Cauley*1, Ruchin Patel1, Peter Fagenholz1, Sadiqa Mahmood3, David L. Berger1, Deborah Schnipper4, Pardon Kenney2, David W. Rattner1, George Velmahos1, Liliana G. Bordeianou1
1Surgery, Massachusetts General Hospital, Boston, MA; 2Surgery, Brigham and Women's Surgical Specialties at Faulkner , Jamaica Plains, MA; 3Quality Safety and Value, Partners Health, Boston, MA; 4Surgery, Newton-Wellesley Hospital, Newton, MA
Purpose: Current recommendations suggest that immediate anastomosis with proximal diversion is safe in select patients who need emergent surgery for acute diverticulitis. Patients who undergo an anastomosis with proximal diversion are more likely to ultimately have restoration of intestinal continuity when compared to end colostomy. The aim of this study is to 1) evaluate the patterns of care of acute diverticulitis requiring emergent surgical intervention and 2) compare outcomes of patients who undergo primary anastomosis with proximal diversion versus end colostomy across a healthcare network.
Methods: Multi-institution cohort study of 6 years of prospectively collected NSQIP data merged and validated with electronic health record data on patients who underwent surgery for diverticulitis in a Healthcare Network-wide Colorectal Surgery Collaborative of 5 hospitals (community: 3, academic: 2). Patients treated with emergent or urgent surgery for the diagnosis of diverticulitis were identified. Those treated with a colostomy (C) were compared to those treated with a protected anastomosis (PA). In order to compare like C and PA patients we created a propensity score based on the variables found most disparate in univariable comparisons. These variables were: age, BMI, DM, Emergent Surgery, and ASA Class. We then used this score to create a 1:1 matched cohort (1 PA patient matched with 1 C patient)
Results:
A total of 1,423 patients underwent surgical management of diverticulitis in the cohort over the study period (2010-2016) with 317 patients presenting to the OR emergently. Only 38 (12%) had a protected anastomosis. Of these 38, 21 (67.8%) where performed in a teaching hospital. Rates of PA varied from 0-22.6 % in community hospitals. Prior to matching PA patients were older, had higher BMI, had less severe diabetes, and less sepsis, septic shock and SIRS. Patients receiving PA were also healthier with CCI 2.2 vs 3.7, p=0.029 and were less likely to need emergent (instead of urgent) procedures (PA: 21.1% emergent vs C: 59.1%) or be in extremis as witnessed by lower ASA rates (ASA -4 5.3% in PA versus 15.1% in C). After performing the 1:1 propensity score match this led to a cohort of 31 patients in each group (31/38=82% of PA cohort). In the matched cohort PA and C were quantitatively similar in mortality (3.2% vs 3.2%, p=.99) and LOS (9.8days vs 10.4 days, p=.741). There were qualitative differences in SSIs (22.6 vs 12.9).
Conclusion
Carefully selected healthy patients with hemodynamically stable sepsis appear to have equivalent postoperative outcomes after protected anastomosis or colostomy. Surgeon preference and hospital affiliation, rather than true outcomes seem to dictate performance of this more complex operation in patients who present with this urgent surgical problem. This suggests opportunity for quality improvement.
Propensity Matched Outcomes After Operation for Diverticulitis
| | Protected Anastomosis n=31 | Colostomy n=31 | p-value | | Surgical Site Infection | 22.6% | 12.9% | 0.318 | | Unplanned re-operation | 9.7% | 3.3% | 0.316 | | Anastomotic Leak | 47.4% | 25.0% | 0.191 | | Length of Stay | 9.8 +/- 7.2 | 10.4 +/- 5.7 | 0.741 | | Death | 3.8% | 3.8% | 0.99 |
Back to 2017 Program and Abstracts
|
