EP347 - Too Many Hartmann’s: A Decade of NSQIP Data on Perforated Diverticulitis

Purpose/Background: Perforated acute diverticulitis is a common surgical emergency historically managed with Hartmann's procedure (HP). Although effective for source control, Hartmann's carries substantial morbidity and frequently results in a permanent stoma. In contrast, primary anastomosis (PA), with or without a diverting loop ileostomy (PADLI or PAWOD), has emerged as a safe alternative. Over the past decade, randomized trials and meta-analyses have shown comparable or superior outcomes with PA compared to HP. In response, the American Society of Colon and Rectal Surgeons (ASCRS) updated its clinical practice guidelines in 2020, to endorse PA - with selective diversion - as the preferred option in most patients. Despite this guidance, national practice patterns remain incompletely defined.

Methods/Interventions: We conducted a retrospective cohort study using the ACS-NSQIP Procedure-Targeted Colectomy database (2014-2023) to evaluate operative trends in perforated diverticulitis. Adult patients undergoing emergent colectomy for acute perforated diverticulitis were identified and stratified into three groups: HP, PADLI, and PAWOD. Demographics, operative details, and 30-day postoperative outcomes were abstracted. Primary outcomes were 30-day mortality, overall morbidity, and anastomotic leak. Temporal trends were assessed over the study period and before versus after the 2020 ASCRS guideline publication. Multivariable logistic regressions were adjusted for age, sex, comorbidities, and disease severity.

Results/Outcomes: Of 400,018 patients, 9,089 were included. Median age was 64 years, 51.7% were female. Most underwent HP (6,848; 75.3%), while 832 (9.2%) had PADLI and 1,409 (15.5%) had PAWOD. Mortality rates for HP, PADLI, and PAWOD were 8.1%, 3.6%, and 8.5%, respectively (p < 0.001). Anastomotic leak rates for PADLI and PAWOD were 3.5% and 6.0% (p < 0.001). On multivariable analysis, compared with HP, PA with or without diversion was not associated with higher mortality (OR 1.12, 95% CI 0.89-1.39; p=0.331) or morbidity (OR 0.90, 95% CI 0.81-1.01; p=0.072).

From 2014-2023, PA use increased modestly (20.7% to 28.2%), while HP declined (79.3% to 71.8%) (Figure 1). Specifically, the rate of primary anastomosis increased from 22.8% pre-ASCRS guidelines, to 27.8% post-ASCRS guidelines. On multivariable analysis, the increase in primary anastomosis was not accompanied by an increase in 30-day mortality (OR: 1.27; 95% CI: 0.86-1.88; p = 0.230), 30-day morbidity (OR post- vs pre-ASCRS guidelines: 1.09; 95% CI: 0.90-1.32; p = 0.396), or leak rate (OR: 0.97; 95% CI: 0.64-1.47; p = 0.885).

Conclusion/Discussion: Primary anastomosis with or without diversion is a safe but underutilized alternative to Hartmann's procedure. Despite high-level evidence and supportive ASCRS guidelines, adoption of PA among NSQIP-participating hospitals has been limited, particularly for PADLI. Importantly, increased use of PA has not been associated with higher mortality, morbidity, or leak rates, reinforcing its safety and supporting wider implementation in the emergency management of perforated diverticulitis.

Practice patterns in emergency surgery for acute diverticulitis: A decade of evolution following randomized trials and ASCRS guidelines.