PREDICTORS OF SEVERE MORBIDITY AND MORTALITY FOLLOWING ESOPHAGECTOMY: A MULTIVARIATE VASQIP ANALYSIS
Eitan Neidich*, Mary Whooley, Lygia Stewart
San Francisco VA Medical Center/ University of CA, San Francisco, San Francisco, CA
Introduction: Patient and operative risk factors associated with severe complications and mortality following esophagectomy were studied to improve patient risk stratification and identify quality improvement targets.
Methods: We analyzed esophagectomy cases in the VA Surgical Quality Improvement Program (VASQIP) database (1998-2018). Multivariate analysis, using logistic regression, was used to identify risk factors associated with mortality and severe morbidity [Clavien–Dindo grade 4 (CD4)]. The VA population was selected because of known high patient co-morbidity which can influence morbidity and mortality.
Results: 3872 patients underwent esophagectomy (1998-2018). Mean age 63 years (25-93 years), male (99%), 72% white; ASA class: 1-2 (8%), 3 (75%), 4-5 (17%). Notable co-morbidities included: smoking (41%), diabetes (22%), COPD (20%), dyspnea (16%), alcohol use (11%), and partial/complete dependency (5%)(Table 1). Overall mortality was 5.6%.
On Multivariate analysis, factors associated with mortality included: 1) patient comorbidity: metastatic cancer (AOR: 2.58,P<0.01), dyspnea (AOR: 1.6,P<0.01), EtOH use (AOR: 1.6,P=0.025), advancing age (AOR: 1.05,P<0.01); and 2) Operative factors: blood transfusion >4 units (AOR: 2.5,P=0.023)(Table 2).
On Multivariate analysis, factors associated with CD4 morbidity included: 1) Patient comorbidity: pre-op partial/complete dependency (AOR: 1.6,P<0.01), bleeding disorder(AOR: 1.58,P<0.01), dyspnea (AOR: 1.4,P<0.01), diabetes (AOR: 1.3,P<0.01), COPD (AOR: 1.3,P<0.01), smoking(AOR: 1.2,P<0.01), advancing age(AOR: 1.02,P<0.01); and 2) Operative factors: blood transfusion >4 units (AOR: 1.9, P<0.01) and increased operative time (AOR: 1.1,P<0.01) (Table 2). Operative approach was not associated with changes in morbidity or mortality.
On Multivariate analysis, factors associated with decreased mortality included: increased pre-op albumin (AOR: 0.6,P<0.01) and operations performed in second decade (7.29 vs 3.77%, 1998-2008 vs 2009-2018;AOR: 0.56,P<0.05). Similarly, factors associated with decreased CD4 morbidity included: pre-op radiation therapy (AOR: 0.75, P<0.01) and operations in second decade (2009-2018) (AOR: 0.63,P<0.01).
Conclusion: This study confirmed high pre-op co-morbidity among VA patients (e.g., 92% ASA >3, 41% smokers, 20% COPD). Yet there was a significant decrease in morbidity and mortality over the two decades, without significant change in pre-op co-morbidity. This is possibly due to the VASQIP program. The current study identified patient and operative factors which can be utilized to guide patient optimization and risk stratification. Important factors for patient optimization included nutrition (albumin), functional status, and neoadjuvant therapy. This is the largest study to date of patient and peri-operative factors associated with severe morbidity and mortality following esophagectomy.
Table 1: Factors associated with esophagectomy severe morbidity and mortality
Table 2: Univariate and Multivariate Analysis of patient and peri-operative factors associated with severe morbidity and mortality
Back to 2020 Abstracts