Atrial Fibrillation Is Highly Prevalent Yet Undertreated in Patients With Biopsy-proven Nonalcoholic Steatohepatitis

Maureen Whitsett; Jane Wilcox; Amy Yang; Lihui Zhao; Mary Rinella; Lisa B. VanWagner


Liver International. 2019;39(5):933-940. 

In This Article


This is a retrospective cohort study of all adult (age ≥18 years) patients with biopsy-proven NASH who were evaluated at a large tertiary care centre between February 2002 and December 2015. Clinical and demographic information was extracted from patient charts through the Northwestern Medicine Enterprise Database Warehouse, a repository of clinical and research data for the Northwestern Medicine health system. The diagnosis of NASH/NAFLD was determined using an International Classification of Diseases, the 9th Revision, Clinical Modification (ICD-9-CM) algorithm adopted from Reddy et al[18] All patients with ICD-9-CM billing code diagnosis of NASH/NAFLD were then additionally searched for the presence of liver biopsy pathology reports to identify the biopsy-proven NASH cohort. Pathology reports from liver biopsies performed at our institution were independently reviewed (MW; LV), and those with a diagnosis of biopsy-proven NASH, defined as the presence of steatohepatitis on pathology read, or biopsy-proven cirrhosis thought secondary to NASH (eg, cirrhosis on biopsy plus an ICD9 code for NAFLD/NASH), were included in the biopsy-proven NASH study cohort. Prevalent comorbidities, including AF (ICD9-CM 427.31x), were identified from the electronic medical record. Medication and documented reasons for not prescribing indicated therapy were manually reviewed. Guideline-recommended therapy for AF was defined as a patient having a Class I indication for warfarin or a direct thrombin or factor Xa inhibitor according to Clinical Practice Guidelines for patients with atrial fibrillation.[19] Laboratory data were used to calculate CHA2DS2VASc score, a risk stratifying tool which incorporates stroke risk factors and modifiers to determine if anticoagulation is indicated for the reduction of stroke risk.[20] To determine hospitalization data, intensive care unit (ICU) admission, and rate of re-hospitalization and discharge diagnoses were searched. To determine AF prevalence among a comparator cohort of patients without NASH seeking care at a tertiary care centre, we identified all adult patients age <65 years who were evaluated at our institution between February 2002 and December 2015 without a diagnosis of NAFLD/NASH, stratified by AF status. AF prevalence in the general population for adults age <65 years was determined from previously published literature.

The primary outcomes were prevalence of AF and the prevalence of patients on guideline-recommended therapy for AF.[21] Secondary outcomes included total hospitalizations, total number of ICU stays and length of hospitalization. The study was approved by the Northwestern Institutional Review Board. A waiver of informed consent was granted due to the retrospective study design and minimal risk to participants.

Statistical Methods

To summarize the clinical characteristics of patients with NASH, descriptive statistics (count and percentage for categorical variables and mean (standard deviation) for continuous variables) were presented for each of the characteristics. To compare the difference between patients with NASH with and without AF, we performed chi-square or Fisher's exact test for categorical variables and two-sample t tests for continuous variables. All analyses were performed using R 3.5.0. A P value of <0.05 was considered statistically significant.