Short- and long-term outcomes for primary anastomosis versus Hartmann's procedure in Hinchey III and IV diverticulitis: a multivariate logistic regression analysis of risk factors.

PURPOSE: The management of perforated diverticulitis with generalized peritonitis is still controversial and no preferred standardized therapeutic approach has been determined. We compared surgical outcomes between Hartmann's procedure (HP) and primary anastomosis (PA) in patients with Hinchey III and IV perforated diverticulitis. METHODS: Multicenter retrospective analysis of 131 consecutive patients with Hinchey III and IV diverticulitis operated either with HP or PA from 2015 to 2018. Postoperative morbidity was compared after adjustment for known risk factors in a multivariate logistic regression. RESULTS: Sixty-six patients underwent HP, while PA was carried out in 65 patients, 35.8% of those were defunctioned. HP was more performed in older patients (74.6 vs. 61.2 years, p < .001), with Hinchey IV diverticulitis (37% vs. 7%, p < .001) and in patients with worse prognostic scores (P-POSSUM Physiology Score, p < .001, Charlson Comorbidity Index p < .001). Major morbidity and mortality were higher in HP compared to PA (30.3% vs. 9.2%, p = .002 and 10.6% vs. 0%, p = .007, respectively) with lower stoma reversal rate (43.9% vs. 86.9%, p < .001). In a multivariate logistic regression, PA was independently associated with lower postoperative morbidity and mortality (OR 0.24, 95% CI 0.06-0.96, p = .044). CONCLUSIONS: In comparison to PA, HP is associated with a higher morbidity, higher mortality, and a lower stoma reversal rate. Although a higher prevalence of risk factors in HP patients may explain these outcomes, a significant increase in morbidity and mortality persisted in a multivariate logistic regression analysis that was stratified for the identified risk factors.