B. Jiang2, V. P. Ho2, J. Ginsberg1, S. J. Fu1, Y. Perry1, L. Argote-Greene1, P. A. Linden1, C. W. Towe1 1University Hospitals Case Medical Center,Thoracic And Esophageal Surgery,Cleveland, OH, USA 2University Hospitals Case Medical Center,Surgery,Celeveland, OH, USA
Introduction:
A common post-operative complication after esophagectomy is anastomotic leak. Radiographic imaging of anastomotic leaks (by esophagram) is frequently inaccurate and is often performed after 5-7 days of observation. Prior study has supported the use of perianastomotic drain amylase (DA) on postoperative day 4 to identify patients at low-risk and high-risk for anastomotic leak. . The aim of this study was to determine if decision analysis supports the use of a DA-based accelerated care pathway to decrease hospital length of stay and cost.
Methods:
We designed a decision tree model to compare costs and lengths of stay of DA leak detection versus the standard of care (esophagram) using data extracted from cohort of post-esophagectomy patients from an academic medical center that has been routinely measuring DA. We assessed the model outcome using historical cost and length of stay from retrospective review of consecutive patients undergoing elective esophagectomy. We performed a Monte Carlo simulation to assess the effects of base-case variables on model outcomes. We also performed one-way sensitivity analyses to identified thresholds where a decision alternative dominated the model (both less expensive and shorter stay than the alternative).
Results:
Using DA cutoff value of 31U/L on post-operative day 4, 38% of patients were assigned to an accelerated recovery pathway, of which 10% were found to have a leak. Patients with DA over 125U/L were defined as ‘high risk’ for leak (20% of cohort), of which 50% were diagnosed with a leak. Decision analysis demonstrated that a DA-based accelerated recovery pathway was associated with an improvement in overall LOS of 0.96 days and a cost saving of $2,773.96. Monte Carlo simulation confirmed this finding, with a median saving of 0.78 days and $2078.46.
Conclusion:
Current methodologies to detect anastomotic leaks after esophagectomy radiographically are associated with prolonged hospitalization, but drain amylase can identify patients at low risk and high risk of anastomotic leak. Decision analysis supports the use of post-operative day 4 perianastomotic DA to predict anastomotic leak to reduce hospital length of stay and cost.
