Applicability of cost-effectiveness analysis to management of chronic hepatitis B
Version of Record online: 15 AUG 2007
Journal of Gastroenterology and Hepatology
Volume 22, Issue 9, pages 1357–1359, September 2007
How to Cite
Dan, Y. Y. and Lim, S. G. (2007), Applicability of cost-effectiveness analysis to management of chronic hepatitis B. Journal of Gastroenterology and Hepatology, 22: 1357–1359. doi: 10.1111/j.1440-1746.2007.05098.x
- Issue online: 15 AUG 2007
- Version of Record online: 15 AUG 2007
- Accepted for publication 9 June 2007.
Chronic hepatitis B (CHB) is a major health problem in Asia where up to 250 million people carry the hepatitis B virus. Of these, up to 20% of them may go on to develop complications of decompensated cirrhosis and hepatocellular carcinoma (HCC).1,2 The underlying basis of therapy is to suppress viral replication, and hence to halt or at least delay the onset of cirrhosis and reduce the incidence of development of HCC.3
To this end, four nucleoside/nucleotide analogs (lamivudine, adefovir, entecavir, telbivudine) and interferon/pegylated interferon regimes have been licensed by the Food and Drug Administration in the United States for the treatment of CHB.3 However, these drugs are not cheap and in many Asian countries, reimbursement is either nonexistent or limited. Coupled with the complicated uncertainty of natural history, durability of response, potential development of drug resistance, and need for salvage therapy, the impact of the various therapies on cost and benefit is an area of great interest, not only to health-care policy-makers but also to doctors and patients.
Advances in medical care have introduced a wide array of treatment options with varying efficacy and cost. How much is society, patient, or health payers willing to pay for the added benefit of a superior treatment option? Do treatment options that are more expensive in the short term lead to greater savings in the long term if one looks beyond the initial outlay? Cost-effectiveness analyses address these issues, and therefore play an important role in attempting to apply the best available evidence to estimate the cost–benefit of one treatment strategy over another.4
There have been several articles and abstracts over the recent years attempting to shed more light on this issue.5–7 In this issue of the Journal, Sullivan et al.8 reported their cost–benefit analysis derived by modeling a hypothetical 32-year-old HBeAg-positive patient with CHB (ALT > 2X ULN) in Taiwan. Treatment comparison was made between 48-week therapy of pegylated interferon and 48 weeks of lamivudine. HBeAg seroconversion data were obtained directly from the trial reported by Lau et al.9 Projection was then made on the long-term outcome, assuming that all patients stopped treatment after 48 weeks and those with no response to treatment progress similarly to patients who have not received any treatment. Outcome measures were based upon incremental cost-effectiveness ratios (ICERs), that is, the additional cost needed to save one additional quality-adjusted life-year (QALY) compared with the next cheaper option. Based on the Taiwan health-care cost-payer perspective, the authors found that pegylated interferon was actually cost-effective with a ICER of $NTD 313 819–485 262 ($US 9465–14 651) per QALY gained, which was lower than the threshold of $NTD 580 000 ($US 17 512) per QALY gained when compared with lamivudine treatment for 48 weeks.
The study contends that, from the health payer's perspective in the Taiwan setting, pegylated interferon is cost-effective compared with lamivudine. How applicable is this conclusion to our practice of treating Asian patients with CHB? In other countries where there may be no health subsidy policies to artificially limit the duration of treatment, particularly of antivirals, is this conclusion still valid or relevant?
Cost-effectiveness analysis in CHB has an inherent problem in that this is a disease in which complications and endpoints occur after a prolonged period of time, complicated by different patterns of remission and reactivation in a complex natural history characterized by clinical phenotypes that are likely to have different risk profiles.3 Without the benefit of real-life longitudinal data on the natural history of disease progression, projection of outcomes over cohort life expectancy usually involve broad unproven assumptions. For example, HBeAg seroconversion has been used as a surrogate marker of successful treatment outcome, but the risks of progressing on to liver cirrhosis and HCC despite HBeAg seroconversion, either spontaneously or secondary to treatment, has not been well established.10 Any errors, albeit small, can be magnified over a full life cycle simulation and must be interpreted with caution.
By using actual data from a reported 48-week trial, Sullivan et al. minimize any bias from assumption of treatment outcome, but this makes applicability to real-life scenarios more difficult. In real life, patients who do not seroconvert, are likely to continue on lamivudine. Patients who become resistant to lamivudine would then need salvage therapy. There is also good evidence that long-term therapy with nucleoside analogs in HBeAg-positive patients are associated with an accrual in HBeAg seroconversion rates, such that by year 3 of continuous nucleosides, the seroconversion rate approaches or betters that of 48 weeks of pegylated interferon.11 Similarly, patients who do not respond to pegylated interferon will likely also need nucleoside/nucleotide analogs if they continue to have active liver disease. Finally, Sullivan et al.'s study reports a cost analysis from the health-care payer's perspective. Cost-effectiveness analysis from a societal perspective may better reflect the net absolute cost of a treatment option.12 Additional hidden costs, such as transport for close follow up with pegylated interferon, economic productivity loss from pegylated interferon treatment may also weight downwards the actual cost-benefit of the treatment. The impact on the final cost and benefit may be different after factoring in all these components. Thus, comparison between 48 weeks of pegylated interferon and 48 weeks of lamivudine is unrealistic in real life, both in terms of the objectives of viral control and in terms of clinical practice that exists today. Sullivan et al.'s study conclusion must therefore be put in appropriate context before being applied to other countries and other populations.
Sullivan et al.'s study also highlights the evolving concepts and shifting strategies in treatment of CHB. Traditional approaches have debated the relative merits of nucleosides versus immunomodulators (which the article by Sullivan et al. also attempts to address), short-term versus long-term therapy, and monotherapy versus combination therapy. The REVEAL study in Taiwan clearly indicated that HBV-DNA is the most important prognostic marker for complications of liver disease such as cirrhosis and HCC13,14 and the inference is that long-term viral suppression may be the key to treating CHB. Pegylated interferon attempts to achieve control of HBV by inducing HBeAg seroconversion, hence allowing ‘natural immunity’ to ensure that serum HBV-DNA levels are low or undetectable. It is, however, a rather ineffective antiviral agent in terms of viral suppression.9 The long-term durability of low or undetectable HBV-DNA after ‘successful’ HBeAg seroconversion with pegylated interferon is unknown. Even with nucleoside analogs, the prospect of long-term antiviral therapy is being considered more seriously as clinicians grapple with reactivation of HBV after stopping defined short treatment regimens. The advent of new antivirals with lower resistance rates, such as entecavir and adefovir,15,16 has now make the prospect of life-long treatment possible for Asian patients.
Despite the inherent caveats, cost-effectiveness analysis allows us to use the best evidence available for understanding the cost dynamics in comparing different treatment strategies. A strong word of caution, though, is that the conclusions must be taken in reference to the unique conditions of that country. The ‘willingness to pay’ threshold, cost of medications, insurance coverage, and treatment options such as liver transplant, follow up, transport and all other costs accrued specific to the country can be very different between countries, and this must be taken into account. Real-life treatment strategies in each country also differ significantly from the study setup upon which Sullivan et al. based their analyses, limiting the applicability of the conclusion to other countries.
Chronic hepatitis B constitutes a significant health problem to most Asian countries and contributes significantly to overall health-care costs. Understanding the impact of cost and benefit of the various treatment options is important in helping physicians and patients make the most informed decision on their treatment choices. Unfortunately, the long-term data for both antivirals and interferon do not stretch beyond 5 years,17 and are based on small numbers of patients. Projection of outcomes over life expectancy needs to be interpreted with caution as small errors may be magnified and analysis must be individualized to each country. Duration of treatment for CHB is still evolving and is likely to be life-long for many patients. It is difficult to imagine that short finite duration of treatment as one-size-fit-all strategy will have a significant impact on treatment outcomes. Cost-effective therapy remains to be defined with better clinical studies that examine long-term outcomes prospectively, with long-term therapy.
- 17Efficacy and safety of 5 years lamivudine treatment of Chinese patients with chronic hepatitis B [Abstract]. J. Gastroenterol. Hepatol. 2001; 16 (Suppl. 1): A60., , et al.