An Observational Study of the Safety of Direct-acting Antivirals in Patients With Hepatitis C

The investigators will assess whether patients with the Hepatitis C virus (HCV) who are prescribed direct-acting antiviral (DAA) medications experience higher rates of adverse events than patients with HCV who are untreated. The investigators hypothesize that patients receiving DAAs do not experience higher rates of adverse events compared to patients who have not received DAAs. The study population is adults between the ages of 18 and 88 with any indication of a diagnosis of HCV. An intervention group (those receiving a DAA) and comparison group (those who are not treated) will be created using medication dispensing data. Eligibility for the study will be determined from January 1, 2011 through December 31, 2017. Covariates will be collected from January 1, 2011 through December 31, 2017. Individual study sites may have access to historical data prior to 2011 that can be used as covariates or to identify individuals with HCV. The primary outcomes of interest include acute myocardial infarction, neurological outcomes (e.g. acute stroke, intracranial bleed), acute kidney failure, acute on chronic liver failure, hepatic decompensation, multiple organ dysfunction syndrome, cancer, bradyarrhythmia, and death. The secondary outcomes include decompensated cirrhosis, hospitalization, emergency department visit, and arrhythmia. Outcomes will be assessed from January 1, 2011 through December 31, 2017. The investigators will use two different analytic approaches to answer the question of interest: a Poisson regression model and marginal structural modeling (MSM). The simpler Poisson model is an extension of tabular rate of event analysis. The more complicated MSM model incorporates modeling of the treatment decision to more flexibly control for confounding by indication. For each outcome, the investigators will only record the first date an outcome occurs. Each outcome will be modeled separately.