1389. Risk Identification for 30-Day Readmission Following Percutaneous Transhepatic Biliary Drainage Catheter Placement
Authors * Denotes Presenting Author
  1. Alice Kim; Virginia Commonwealth University
  2. Kunal Patel; Virginia Commonwealth University
  3. John Nestler *; Virginia Commonwealth University
  4. Malcolm Sydnor; Virginia Commonwealth University
Literature has shown readmission following biliary drainage catheter (PTBD) placement to be relatively high, which we hypothesize is due to inadequate and inconsistent post-procedural care. Our study’s purpose was to retrospectively analyze all-cause-30-day readmission rates for patients undergoing PTBD placement and implement new post-procedure care guidelines as compared to the current protocol in the hopes of readmission rate reduction. Furthermore, specific variables were analyzed in order to try and identify predictive factors that could help determine pre-procedural risk.

Materials and Methods:
A total of 228 patients who underwent a total of 243 PTBD procedures were retrospectively identified from our electronic health records between 01/2013 to 05/2019. Numerous variables were analyzed, which included the primary diagnosis, benign vs. malignant disease, ASA classification score, and lab values prior to the procedure (i.e. creatinine, liver transaminases (AST/ALT), alkaline phosphatase, total/direct bilirubin, white cell count, hemoglobin, platelets and INR). Numeric lab values were reviewed for correlation to 30-day admission risk using Mann-Whitney U tests.

A total of 107 (44%) females and 136 (56%) males were included. Patients were 63.8% (n=155) Caucasian, 30% (n=73) African American, with 6.1% (n=15) of another race. Access was primarily from the right side (n=152, 62.6%), followed by left (n=82, 33.7%) and bilateral access (n=9, 3.7%). Reasons for biliary drains included 116 (47.7%) benign causes and 127 (52.3%) malignant causes. Drain sizes were primarily 8.5 French (n=131, 53.9%) and 10.2 French (n=99, 40.7%). In total, 72 (31.6%) of patients were readmitted. Median creatinine, AST, ALT, total bilirubin, direct bilirubin, and INR were 0.83 (Q1-Q3: 0.635-1.12), 105.5 (Q1-Q3: 65.75-172.5), 107 (Q1-Q3: 53.75-179.25), 6.25 (Q1-Q3: 2.30-11.60), 4.8 (Q1-Q3: 1.78-9.23), and 1.2 (Q1-Q3: 1.1-1.4) respectively. Readmission was significantly correlated to total bilirubin and trended towards significance for direct bilirubin values, p=0.044 and 0.073 respectively. The medians in readmitted and non-readmitted patients were 4.25 and 6.90 for total bilirubin vs. 3.30 and 5.85 for direct bilirubin. All other laboratory values showed no significant correlation with readmission. On logistic regression, factors such as age, gender, indication (benign vs. malignant), race, insurance status, drain size, or ASA risk did not predict for readmission.

Total bilirubin showed a statistically significant correlation to 30-day readmission with a trend towards significance with direct bilirubin. However, the direction of correlation was opposite than expected with lower medians for total and direct bilirubin levels in the readmission cohort. Better identification of risk allows for improvements in overall quality and safety of patient care, but it remains difficult to predict those patients at higher risk of readmission based on lab values alone in PTBD catheter placement.