Introduction: The majority of pancreatic resections for malignancy are performed in older patients with major co-morbidties. The aim of this study was to develop a pre-operative nomogram based on the presence of co-morbdities to predict risk of peri-operative mortality. Materials and Methods: The National Inpatient Sample (NIS) database was queried (2000-2005) to identify adult patients that underwent resection for pancreatic malignancy. The pre-operative co-morbidties, identified as predictors were used and a nomogram was created with multivariate logistic regression using the Taylor expansion method in SAS callable SUDAAN 10.0.(Figure:1) Patient Sample A (2000- 2004) was utilized to develop the model and patient Sample B (2005) was utilized to validate this model. Results: A total of 4482 and 999 patients were included in Samples A and B. The overall peri-operative mortality rate for Samples A and B was 6.3% and 5.2% respectively. The mean total points calculated for Sample A by the nomogram was 131.7 (SE= 1.54, range: 7.7 to 339.8) that translates to a nomogram-predicted mortality rate of 4.9 %, which is similar to the actual mortality. The mean total points for Sample B was 128.1 (SE= 1.62, range: 7.7 to 367.9) which translates to a nomogram-predicted mortality rate of 4.6%. The similarity of mortality rates of Samples A and B as predicted by the nomogram and a concordance index of 0.76 shows good agreement between the data and the nomogram. Conclusion: This nomomgram consisting of pre-operative co-morbdities has been shown to accurately predict the risk of peri-operative mortality. The nomogram can be used for risk assessment and counseling in the pre-operative setting for patients with pancreatic cancer considering surgical intervention.


Pre-operative nomogram to predict risk of peri-operative mortality following pancreatic resections for malignancy