Purpose: Quantifying the economic burden of disease is important for public health policy. Our aim was to estimate long-term health care costs attributable to hepatitis C virus (HCV) infection using administrative database.
Method: A population-based linkage of serologic database including over 95% of all identified cases of HCV in British Columbia (BC) to the BC Linked Health Database identified 20,670 HCV seropositive cases. Cases were matched 1:4 to HCV seronegative controls (n=71,709) based on age, gender, socioeconomic status, place of residence, co-morbidity, and phase of illness. Subjects were followed from January 1, 1997 to December 31, 2004. The observation time was divided into three phases: the year preceding death (pre-death phase, n=418); the period following the diagnosis of decompensated liver disease (late phase, n=537); and the time between cohort entry and either of the preceding phases or end of follow-up (initial phase, n=19,041). Costs of care attributable to HCV (net costs, mean cost in cases minus mean cost in controls) for each phase of illness were estimated using Generalized Estimating Equations. Per-patient total costs and net costs were applied to Kaplan Meier survival estimates of all HCV cases to estimate 5-year costs. Costs were discounted at 5% annually. We captured costs for physicians’ and other fee-for-service practitioners’ services, drug costs reimbursed by the Provincial Plan, hospital costs, and same day surgery costs.
Result: Mean annual net costs were $1,852 (95% confidence intervals (CI), $1,728-$1,972, 2005 CAD) in the initial phase and $5,997 (95% CI, $4,756-$7,243) in the late phase. Mean costs in the pre-death phase were not substantially different between cases and controls. The undiscounted and discounted mean 5-year total costs for HCV cases were $24,243 and $18,995, respectively. The respective mean (95% CI) 5-year net costs were $7,653 ($7,586-$7,721) and $5,997 ($5,944-$6,050). Older age, female gender, co-morbidity, and HCV viremia were significantly associated with higher costs of HCV care in all phases of illness. HIV was associated with higher costs in the initial and pre-death phases.
Conclusion: The costs of HCV care in this population are substantial and vary by phase of illness. Predictors of costs were consistent with known factors affecting resource utilization in other chronic conditions. These results may prove useful for future economic models evaluating hepatitis C prevention and care programs.
Candidate for the Lee B. Lusted Student Prize Competition