Impact On Hospital Systems By Esophagectomy (Eg) and Pancreaticoduodenectomy (Pd): Objective Comparison of Surgical Complications Using the Accordion Severity Grading System
Donald E. Low*, Madhankumar Kuppusamy, Yasushi Hashimoto, L. William Traverso
General Surgery, Virginia Mason Clinic, Seattle, WA
Introduction. EG and PD are complex GI operations that can impact a hospital system if the complication rate is high. The recently reported Accordion Severity Grading System of Surgical Complications (ASGS) quantifies the amount of treatment (Tx) required to manage surgical complication by using readily available variables with minimal subjectivity (Ann Surg. 2009 Aug;250(2):177-86). For the first time the ASGS will allow grading and comparison of complications between individual case series to determine the impact of each operation on a hospital supportive systems. Methods.From prospective IRB-approved databases we retrospectively assigned ASGS categories 1-6 to all post-operative complications documented in consecutive cases through December-2008 (EG=463 and PD=507) by a single esophageal or pancreatic surgeon respectively, in a tertiary-referral, resident-training hospital. Grade 1 - mild, Grade 2 - moderate with pharmacologic Tx, Grade 3 - severe with invasive Tx not requiring general anesthesia (GA), Grade 4 - severe invasive Tx with GA, Grade 5 - severe with organ failure, Grade 6 - postoperative death.Results.(Table. 1&2)Standard outcome measures were similar for EG and PD. As ASGS grade increased the number of cases decreased and this decrease was similar (significant correlation) for EG vs. PD. As ASGS increased so did LOS and the increase was similar for EG vs. PD. Conclusions.EG and PD have a remarkably similar incidence of standard outcome measures. As expected the new ASGS correlates with LOS. For the first time severity grading of a complication, based on the amount of intervention required to treat, provides a clinically relevant tool for comparing complication rates between very different complex operations. Surprisingly their impact on the hospital system was the same.
Table.1
| Traditional outcome measures | EG (n=463) | PD (n = 507) |
| Perioperative Mortality | 2 (0.4%) | 5 (1%) |
| Overall morbidity | 211 (46%) | 222 (44%) |
| Median length of hospital stay (LOS) in days (range) | 10 (6 - 49) | 9 (6 - 79) |
| Patients with operation specific complications (bleeding, leak etc.) | 121 (26%) | 135 (27%) |
| Patients with complications not specific to operation (pneumonia, DVT etc.) | 90 (19%) | 88 (17%) |
Table.2 Comparison of Complication Grades Between Procedures
| Operation | Highest Accordion Grade for Any Type of Complication | ||||||
| 1 | 2 | 3 | 4 | 5 | 6 | ||
| % in Grade | EG (n=463) | 12% | 22% | 5% | 5% | 2% | 0.4% |
| PD (n=507) | 12% | 20% | 10% | 0.4% | 1% | 1% | |
| LOS(days) | EG (n=463) | 9 | 10 | 15 | 18 | 27 | 8 |
| PD (n=507) | 9 | 11 | 14 | 29 | 43 | 9 | |
Statistically significant correlation as % in Grade or LOS varied similarly for EG vs. PD, p<0.0001
Back to Program
