HOME
Return to Table of Contents
COMORBIDITY SCALE IN ENDOSCOPIC RETROGRADE CHOLANGIOPANCREATOGRAPHY
A Rahman, W Depew, LC Hookey
Queen’s University
PURPOSE AND OBJECTIVES: We intend to determine probability of safety of endoscopic retrograde cholangiopancreatography (ERCP) on the basis of pre-procedure co-morbidity which may provide a means to identify patients who are safe for outpatient procedure, patients who could be transferred back to peripheral centers, and patients who would require inpatient admission.
METHODS: We are performing a retrospective chart review of patients who have undergone ERCP at Queen’s University for 5 calendar years starting from January 1, 2001. To quantify co-morbidity, we are using the Charlson Co-morbidity Index (CCI), a validated method of co-morbidity assessment. The primary endpoint is correlation between the CCI to complications of ERCP using univariate and multivariate analysis. Secondary endpoints include correlation of comorbidity to specific adverse sequelae such as acute pancreatitis, bleeding, perforation, cholangitis, hemodynamic instability and abdominal pain. We are also recording death in-hospital from any cause. We identified CCI score and complication rates from the first ERCP only, if multiple procedures were performed. To ensure a high capture rate, we have stringent protocol for inclusion, understanding this may artificially inflate our complication rate.
RESULTS: To date, we have reviewed 159 individual patient charts. 40 patients were excluded from review. Of the 119 remaining charts, 92 patients underwent ERCP without complication, 24 experienced complications as defined above, and 3 patients died in hospital. The primary indication for ERCP was likely choledolithiasis, cholangitis or gallstone pancreatitis in 73.11%. The remaining 26.89% underwent ERCP for likely cancer, idiopathic pancreatitis, or bile leak. In the non-complication group the mean CCI score was 2.99. Of the 24 complications, 17 were pancreatitis, 2 were pain, 2 were bleeding, 2 were hemodynamic instability, and 1 was perforation. Overall, the complication group had a lower mean CCI score of 2.71 (p=NS). Subgroup analysis revealed patients with serious non-pancreatitis complication (bleeding, instability, perforation) had a mean CCI of 6.4 which was significantly higher than the non complication group (p= 0.005743).
CONCLUSION: To date, we have demonstrated no significant difference between non-complication patients and complication patients undergoing ERCP. However, we have demonstrated a significant difference in those who experience serious, non-pancreatitis complications. We expect with completion of the chart review to construct a probability model of CCI index and specific complications from ERCP which we will then test prospectively.