Predictors of anastomotic leak following esophagectomy with gastric reconstruction: a contemporary National Surgical Quality Improvement Project (NSQIP) database analysis
Benjamin A. Palleiko*, Kevin M. Dickson, Allison Crawford, Isabel Emmerick, Karl Uy, Mark W. Maxfield, Feiran Lou
UMass Chan Medical School, Worcester, MA
Objectives: Anastomotic leak (AL) is a serious complication after esophagectomy and is associated with worse outcomes. This study utilized the National Surgical Quality Improvement Project (NSQIP) esophagectomy specific database to identify preoperative risk factors associated with AL.
Design: This was a retrospective analysis of a large, national, surgical database. Patients in the 2016-2021 NSQIP database who underwent elective esophagectomy with gastric reconstruction for malignant indications were included. Multivariable logistic regression identified risk factors associated with AL.
Setting: A large, nationwide surgical database
Patients: There were 6813 patients in the 2016-2021 NSQIP esophagectomy database. There were 4281 patients were included in the analysis after excluding patients who underwent esophagectomy for non-malignant indications, non-elective surgery, and patients who did not undergo gastric reconstruction.
Interventions: This was a non-interventional study
Main outcome measures: The occurrence of anastomotic leak
Results: There were 597 patients who experienced AL (13.9%) and 3684 (86.1%) who did not. Multivariable logistic regression revealed increased risk of AL with American Indian, Alaskan Native, Native Hawaiian or Other Pacific Islander race (OR 4.08, CI 1.48-11.25, p=0.007), smoking within 1 year of surgery (OR 1.25, CI 1.02-1.52, p=0.035), and modified frailty index-5 (mFI-5) of 1 (OR 1.50, CI 1.23-1.84, p<0.0001) or 2 (OR1.60, CI 1.25-2.06, p=0.0002). For every 1,000/microliter increase in white blood cell (WBC) count, there was a 6% increased risk (OR 1.06, CI 1.02-1.10, p=0.001). Each 10,000/microliter increase in platelet count corresponded with a 2% reduced risk (OR 0.98, CI 0.96-0.99, p=0.002).
Conclusions: In a retrospective analysis of a large quality and outcomes-based database, we found that race, smoking status, frailty index, WBC and platelet counts all significantly correlated with risks for AL in patients who underwent elective esophagectomies for cancer. The results help to inform surgeons and patients of the true risks of AL and potentially improve outcomes by optimizing preoperative characteristics like frailty. Additionally, the disproportionate elevated AL risk in Native Americans warrants further investigation.
Multivariable Logistic Regression Predicting Anastomotic Leak
| Variable | Odds Ratio | 95% Confidence Interval | p-value |
| Race (versus white) | 0.030 | ||
| Black or African American | 0.77 | 0.42-1.42 | 0.40 |
| Asian | 1.45 | 0.91-2.31 | 0.12 |
| American Indian, Alaskan Native, Native Hawaiian or Other Pacific Islander | 4.08 | 1.48-11.25 | 0.007 |
| Unknown | 1.09 | 0.87-1.37 | 0.46 |
| Smoking | 1.24 | 1.02-1.52 | 0.035 |
| Frailty Index (versus 0) | <0.0001 | ||
| 1 | 1.50 | 1.23-1.84 | <0.0001 |
| 2 | 1.60 | 1.25-2.06 | 0.0002 |
| 3 or 4 | 0.92 | 0.43-1.96 | 0.84 |
| Preop WBC, per 1-unit increase | 1.06 | 1.02-1.10 | 0.001 |
| Preop platelets, per 10-unit increase | 0.98 | 0.96-0.99 | 0.002 |
Back to 2023 Posters