John E. Scarborough

BACKGROUND: Although strong evidence exists for combined mechanical and oral antibiotic bowel preparation before elective colorectal resection, the utility of preoperative bowel preparation for patients undergoing sigmoid resection after endoscopic decompression of sigmoid volvulus has not been previously examined. The goal of this study was to evaluate the association between bowel preparation and postoperative outcomes for patients undergoing semielective, same-admission sigmoid resection for acute volvulus. STUDY DESIGN: Patients from the 2012 to 2019 Colectomy-Targeted American College of Surgeons NSQIP dataset who underwent sigmoid resection with primary anastomosis after admission for sigmoid volvulus were included. Multivariable logistic regression was used to compare the risk-adjusted 30-day postoperative outcomes of patients who received combined preoperative bowel preparation with those of patients who received either partial (mechanical or oral antibiotic alone) or incomplete bowel preparation. Effort was made to exclude patients whose urgency of clinical condition at hospital admission precluded an attempt at preoperative decompression and subsequent bowel preparation. RESULTS: Included were 2,429 patients, 322 (13.3%) of whom underwent complete bowel preparation and 2,107 (86.7%) of whom underwent partial or incomplete bowel preparation. Complete bowel preparation was protective against several postoperative complications (including anastomotic leak), mortality, and prolonged postoperative hospitalization. CONCLUSIONS: This study demonstrates a significant benefit for complete bowel preparation before semielective, same-admission sigmoid resection in patients with acute sigmoid volvulus. However, only a small percentage of patients in this national sample underwent complete preoperative bowel preparation. Broader adoption of bowel preparation may reduce overall rates of complication in patients who require sigmoid colectomy due to volvulus.

The comparative effectiveness of nonoperative management (NOM) vs immediate splenectomy (IS) for hemodynamically stable adult patients with grade IV or V blunt splenic injury (BSI) has not been clearly established in the literature. We performed a retrospective analysis of adult patients, from the 2013 to 2014 American College of Surgeons Trauma Quality Improvement Program (TQIP) Participant Use Data Files, who sustained grade IV or V BSI. Outcomes after IS vs attempted NOM were compared using propensity score analysis in order to adjust for patient- and injury-related variables. Nonoperative management was attempted in 1,489 (52.2%) of 2,746 patients who sustained grade IV or V BSI. Propensity matching techniques resulted in a cohort of 758 IS and NOM patients who were well matched for all known patient- and injury-related variables. In-hospital mortality was not different between the IS and NOM patients (11.5% vs 10.0%, p = 0.33), although IS patients had a higher incidence of infectious complications (21.4% vs 16.9%, p = 0.02). The rate of NOM failure in our sample was 20.1%. Independent predictors of failed NOM included the presence of a bleeding disorder, early blood transfusion requirement, and grade V injury. Splenic artery embolization was associated with a decreased risk of NOM failure. Patients who had failed NOM had a lower in-hospital mortality rate than IS patients (6.4% vs 16.4%, p = 0.004), but required longer hospitalization. Nonoperative management is as effective as IS for hemodynamically stable adult patients with grade IV or V BSI. The delay in operative intervention that results from failed attempts at NOM does not adversely affect the outcomes of patients who ultimately require splenectomy.

Because preoperative risk factor modification is generally not possible in the emergency setting, complication prevention represents an important focus for quality improvement in emergency general surgery (EGS). The objective of our study was to determine the overall impact that specific postoperative complications have in this patient population. Our study sample consisted of patients from the 2012-2013 ACS-NSQIP database who underwent an EGS procedure. We used population attributable fractions (PAFs) to estimate the overall impact that each of 8 specific complications had on 30-day physiologic and resource use outcomes in our study population. The PAF represents the percentage reduction in a given outcome that would be anticipated if a complication were able to be completely prevented in our study population. Both unadjusted and risk-adjusted PAFs were calculated. There were 79,183 patients included for analysis. The most common complications in these patients were bleeding (6.2%), incisional surgical site infection (SSI) (3.4%), pneumonia (2.7%), and organ/space SSI (2.6%). Bleeding was the complication with the greatest overall impact on mortality and end-organ dysfunction, demonstrating an adjusted PAF of 10.7% (95% CI 8.2%,13.1%, p < 0.001) and 15.9% (95% CI 13.9%, 16.7%, p < 0.001) for these respective outcomes. The only other complication with a sizeable impact on these outcomes was pneumonia (adjusted PAF of 7.9% for mortality and 13.2% for pneumonia). In contrast, complications such as urinary tract infection, venous thromboembolism, myocardial infarction, and incisional SSI had negligible impacts on these outcomes. Our study provides a framework for the development of high-value quality initiatives in EGS.