M. Cheung1, R. LeDuc1, A. Cobb1, L. Gil1, H. Mundt1, R. P. Gonzalez1, M. J. Anstadt1 1Loyola University Chicago Stritch School Of Medicine,Department Of Surgery,Maywood, IL, USA
Introduction: Partial colectomy is used to treat a variety of conditions. Studies have shown a greater risk of mortality in partial colectomy performed in emergent versus elective cases. While one large database study identified a variety of factors that were predictive of inpatient mortality in colon and rectal surgery, elective and non-elective (urgent and emergent) were not analyzed as separate groups. In addition, a smaller scale study identified preoperative hypotension and ASA score as risk factors that influence outcomes of emergent colectomy. Given the relatively small sample size of that study, the goal of our study was to use a large database to identify what specific factors contribute to the difference in outcomes in emergent/urgent versus non-emergent colectomy. Additionally, we aimed to identify the relative importance of the risk factors to guide the discussion for family decision making when possible and limit the incidence of unfavorable outcomes.
Methods: Using the HCUP-SID from California , Florida , Iowa, New York, and Washington between 2006-2013, patients who underwent partial colectomy were identified by ICD-9 code. These patients were divided into an elective and non-elective group, with the non-elective group being defined by admission type (elective, urgent, emergent) and meeting the criteria of less than 24 hours to operation following hospitalization. Three independent decision trees were carried out for the elective, non-elective, and overall partial colectomy groups, respectively, with comorbid conditions and patient characteristics as the independent variables and mortality as the primary outcome. The importance of each of the variables was then determined and weighted using a variable importance function.
Results: We identified 181,130 patients who underwent partial colectomy. Of these patients, 31,978 (17.65%) were classified as non-elective cases. The two groups exhibited no clinical difference in age, gender, or race. The non-elective group demonstrated a statistically significant greater inpatient mortality rate than the elective cohort (7.28% vs 4.96%; p<0.001). Following decision tree analysis, insurance status (relative importance, non-elective: 70.07; elective: 51.90), age (non-elective: 100; elective: 43.77), and presence of congestive heart failure (non-elective: 34.41; elective: 37.52) were the most important factors for both elective and non-elective groups. Obesity and tobacco use were much more predictive of inpatient mortality in non-elective cases compared to elective operations.
Conclusion: The results of this study provide insight into what factors are predictive of inpatient mortality following partial colectomy in an elective and non-elective setting. These factors can be used as prognostic tools for predicting which patients will have better outcomes and in some cases may be useful in guiding the family discussion and decision making process in cases of emergent or urgent partial colectomy.