Skip to main content
main-content
Top

03-11-2011 | Surgery | Article

Two methods best predict in-hospital mortality following abdominal surgery

Abstract

Free abstract

MedWire News: Study findings suggest that in-hospital patient mortality following gastric, hepatic, or pancreatic resection is best predicted using the All Patient Refined Diagnosis-Related Groups (APR-DRGs) and Disease Staging algorithms.

These two methods were found to be superior to the Charlson/Deyo and Elixhauser methods, which are also commonly used algorithms.

"The selection of a particular method for comorbidity risk adjustment after gastric, hepatic, and pancreatic resections has a significant impact on analysis of in-hospital mortality. This is explained by the reality that different methods classify the same patients into different risk categories," say Elijah Dixon (University of Calgary, Canada) and colleagues.

The team compared the performance of the four comorbidity algorithms for predicting in-hospital mortality following 46,395 gastric, 18,234 hepatic, and 15,443 pancreatic resections performed in adults aged 58-63 years on average during 2002-2007.

Overall, in-hospital mortality was 5.9% in gastric and pancreatic resections, and 4.4% in hepatic resections. Median length of stay ranged from 5 days for hepatic resections to 10 days for pancreatic resections.

Predicted in-hospital mortality rates following adjustment agreed for 43.8% to 74.6% of patients. The APR-DRGs and Disease Staging indices outperformed the Charlson/Deyo and Elixhauser, with the APR-DRGs showing more accuracy than Disease Staging, and Elixhauser showing more accuracy than Charlson/Deyo.

Compared with the Charlson/Deyo algorithm, the Elixhauser index was significantly more accurate for predicting in-hospital mortality among patients undergoing gastric resections (0.847 vs 0.792), hepatic resections (0.810 vs 0.757), and pancreatic resections (0.811 vs 0.741). These findings were not affected by diagnosis rank, multiple surgeries, or exclusion of transplant patients.

The researchers say that the greater accuracy of the APR-DRGs and Disease Staging algorithms (ranging from 0.836 to 0.945 in all subgroups) may be attributed to inclusion of some of the complications of care, in contrast to the other methods that included comorbidities only.

Further analysis showed that renal disease, congestive heart failure, coagulopathy, and fluid and electrolyte disorders were significantly associated with at least a two-fold increased risk for in-hospital mortality.

Indeed, fluid and electrolyte disorders were the most prevalent diseases, occurring in 12-20% of patients.

Cerebrovascular disease was another important predictor for increased mortality risk, length of stay, and hospital charges, and was associated with a 1.69 and 2.21-fold increased risk for death in gastric and pancreatic resections, respectively. Patients with cerebrovascular disease undergoing hepatic resections experienced a 10-fold increased risk for death.

Writing in the Archives of Surgery, Dixon and team conclude that provider-specific outcomes may be evaluated differently depending on the selected method and the importance of a particular comorbidity can also be misinterpreted because of inadequate risk adjustment.

By Ingrid Grasmo

Related topics