OBJECTIVES : The approval of sofosbuvir in December 2013 and subsequent introduction of other direct-acting antivirals changed the paradigm for treating Hepatitis C (HCV), but posed a challenge to payers seeking to anticipate characteristics of patients receiving treatment and their risk of outcomes. This study explored whether analyses of real-world data from patients treated on SOC prior to sofosbuvir approval could have helped payers anticipate baseline characteristics and outcome rates among sofosbuvir-treated populations.

METHODS : We identified adult patients initiating sofosbuvir (in 2014) or SOC (in 2012-2013) within Optum’s de-identified Integrated Claims-Clinical dataset. Patients entered the cohort upon treatment initiation and were required to have 180 days enrollment (baseline). In addition, we identified a broader population of patients with an HCV diagnosis in 2013 with 180 days baseline. Subgroups of interest included patients meeting prior authorization criteria for sofosbuvir (fibrosis stage and no recent alcohol abuse), and patients with compensated and decompensated cirrhosis. We compared demographics and clinical characteristics measured at baseline and rates of outcomes decompensation and hepatocellular carcinoma (HCC) during a 1-year follow-up.

RESULTS : During 2014, 2,895 patients initiated sofosbuvir and 2,226 initiated SoC during 2012-2013. Patients treated with sofosbuvir were slightly older (median age 59 vs 55) than those treated on historic SoC. Sofosbuvir-treated patients had a higher prevalence of compensated and decompensated cirrhosis (37.3% and 19.7% vs 28.9% and 11.1%), based on diagnoses and procedures during baseline. Rates of decompensation and HCC during follow-up were similar after subgrouping on cirrhosis status.

CONCLUSIONS : Drug innovations can significantly shift the characteristics of patient populations receiving treatment, due to both clinical and access-related factors. In cases where these factors can be measured, it may be possible to restrict the population on a prior SoC to more closely resemble that on the new SoC, thus leveraging historical RWE to represent a new SoC in forward-looking decisions.