Graduation Date

Summer 8-14-2020

Document Type

Dissertation

Degree Name

Doctor of Philosophy (PhD)

Programs

Epidemiology

First Advisor

Paraskevi A. Farazi, PhD

Second Advisor

Tricia D. LeVan, PhD

Abstract

The incidence and mortality of liver cancers are increasing in the United States. In addition to the classic risk factors of hepatic viral infection and high alcohol consumption, non-alcoholic Fatty Liver Disease (NAFLD), has emerged as a risk factor for hepatocellular carcinoma (HCC). The most effective curative treatment for HCC is liver resection (hepatectomy), however, hepatectomy is a major operation with potential risks of mortality, hospital readmission and other adverse outcomes such as postoperative complications and longer hospital stay. The reported 90-day mortality and 30-day readmission rates following a hepatectomy vary widely across studies, most of which have limitations due to small sample size, localized study sites and lack of use of national representative samples. Furthermore, very few studies have focused on the impact of NAFLD on postoperative outcomes in HCC patients undergoing hepatectomies. The long-term goal of this proposal was to improve 90-day mortality, unplanned 30-day readmission and other adverse outcomes among HCC patients undergoing hepatectomies. The specific aims of this work were to: (1) develop and test the discriminatory ability of a risk score model to group patients into low, medium, high and excessive risks groups for 90-day mortality following hepatectomy based on the weighting of identified demographic and hospital related risk factors; (2) investigate risk factors associated with 30-day unplanned readmission and address its impact on conditional 90-day mortality; (3) describe 10-year (2005-2014) national trends and patterns of NAFLD prevalence among all hospitalized HCC patients and compare the impact of NAFLD on adverse outcomes (increased occurrence of postoperative complications, in-hospital mortality and longer hospital stay) in HCC patients who underwent hepatectomy. To address the first aim, a retrospective, cross-sectional study was conducted using one of the largest clinical cancer database in the world – the National Cancer Database (NCDB) to investigate the rate of 90-day mortality and associated risk factors and to build a risk score model based on the weighting of identified risk factors to predict 90-day mortality. For the second aim, a cross-sectional study design was used to identify risk factors associated with 30-day unplanned readmission and assess how a history of 30-day unplanned readmission would affect patients’ conditional 90-day mortality using NCDB. For aim 3, another national representative database - Nationwide Inpatient Sample (NIS) was used for a retrospective study to describe the national trend and pattern of NAFLD prevalence among hospitalized HCC patients from 2005 to 2014 and to address the impact of NAFLD on adverse outcomes including postoperative complications, in-hospital mortality and hospital length of stay (LOS) in HCC patients undergoing hepatectomies. Our results indicated that an integer-valued risk score can be developed based on 10 identified risk factors to predict 90-day mortality. The discriminatory ability of this risk score, measured by AUC of 0.69, was comparable with other risk models of prior studies. Our study also showed that patients with a history of 30-day unplanned readmission carried higher risk of conditional 90-day mortality. Interestingly, NAFLD did not seem to increase the risk of postoperative adverse outcomes. Future studies are needed to further improve the risk model by linking NCDB to Midwest CoC accredited hospital databases to add detailed information on comorbidity and complications, by addressing the roles of family care and medication compliance between discharge and readmission, and by repeating study 3 in multi-institutional setting with more accurate and complete diagnosis of NAFLD cases.

Included in

Epidemiology Commons

Share

COinS