Defense Date


Graduation Date

Summer 1-1-2017


One-year Embargo

Submission Type


Degree Name



Pharmacy Administration


School of Pharmacy

Committee Chair

Jordan Covey

Committee Member

Khalid Kamal

Committee Member

Branden Nemacek


Objective: To analyze concordance with evidence-based clinical care guidelines in real world clinical practice in patients with cirrhosis and ascites, hepatic encephalopathy (HE), or spontaneous bacterial peritonitis (SBP).

Methods: A retrospective cohort analysis of the UPMC EMR database (2009-2014) with access to full outpatient and limited inpatient data was conducted to identify patients with cirrhosis and ascites, HE or SBP. Data regarding patient demographics, clinical characteristics, laboratory values and medication utilization were extracted. Analyses included examination of patient demographic and clinical characteristics, change in disease severity (via MELDNa scoring) from cirrhosis to complication development and outpatient/inpatient healthcare utilization patterns. Additionally, concordance with investigator-designed quality care indicators adapted from AASLD guidelines and other sources were assessed to understand real world clinical care. Patient- and physician- factors predicting concordance with pharmacotherapy recommendations were assessed via the use of logistic regression models.

Results: The inclusion/exclusion criteria yielded 4,116 patients with liver cirrhosis and 986, 665 and 148 patients with ascites, HE, and SBP respectively. Concordance with quality indicators ranged from 49.83% (recommended medication for HE) to 99.32% (MELD at SBP index). Body mass index and physician type were the only predictors that predicted concordance within the regression models for the selected indicators (prescription for recommended ascites and HE medications). A significant increase in MELDNa was observed from cirrhosis to complication index. No differences in healthcare utilization patterns were observed across complications.

Conclusions: Several opportunities for improvement in quality of care were noted. However, factors assessed in this study revealed limited information regarding opportunities to improve concordance to clinical guidance.