Treatment with direct antiviral agents (DAA) is highly effective at curing Hepatitis C virus (HCV) infection in HCV mono-infected and HIV-HCV co-infected individuals. However, its long-term effects on hepatic and extra-hepatic morbidity and HCV reinfection are still unknown. Observational data collected in routine clinical practice provide a unique opportunity to monitor post-DAA outcomes in large and heterogenous populations. I will discuss some methodological challenges in estimating the long-term effect of DAA treatment, particularly when the outcome is a biomarker. I will illustrate how the target trial approach and causal inference methods can guide the choice of the study design and statistical analyses and lead to meaningful and interpretable findings.