Objectives
This study aimed to determine whether administration of lamivudine to pregnant women with chronic hepatitis B in the third trimester is a cost-effective strategy in preventing perinatal transmission.
Study Design
We developed a decision analysis model to compare the cost-effectiveness of 2 management strategies for chronic hepatitis B in pregnancy: (1) expectant management or (2) lamivudine administration in the third trimester. We assumed that lamivudine reduced perinatal transmission by 62%.
Results
Our Markov model demonstrated that lamivudine administration is the dominant strategy. For every 1000 infected pregnant women treated with lamivudine, $337,000 is saved and 314 quality-adjusted life-years are gained. For every 1000 pregnancies with maternal hepatitis B, lamivudine prevents 21 cases of hepatocellular carcinoma and 5 liver transplants in the offspring. The model remained robust in sensitivity analysis.
Conclusion
Antenatal lamivudine administration to pregnant patients with hepatitis B is cost-effective, and frequently cost-saving, under a wide range of circumstances.
Hepatitis B infection poses a significant global health problem with about 350 million chronically infected individuals. Responsible for approximately 30% of cirrhosis and over 50% of hepatocellular carcinoma worldwide, carriers of hepatitis B are at risk for complications of progressive liver disease. Although the prevalence of hepatitis B is relatively low in the United States, infection still results in significant morbidity and mortality. The Centers for Disease Control estimates that 1.2 million Americans are chronic carriers of hepatitis B, and an additional 5000 to 8000 individuals become chronically infected each year.
Perinatal transmission is the most common mode of hepatitis B transmission worldwide. Framing the importance of perinatal transmission in the overall health burden of hepatitis B, the risk of progression to chronic hepatitis B is inversely related to the age of acquired infection. In the absence of intervention, 90% of infants born to hepatitis B e antigen positive mothers become chronic carriers, with a subsequent 25-30% lifetime risk of serious liver disease. Current guidelines recommend that infants of women who are hepatitis B surface antigen positive receive both the hepatitis B immunoglobulin and first dose of the hepatitis B vaccine series within 12 hours of birth. Although the combination of passive and active immunization reduces the risk of perinatal hepatitis B infection by 85-95%, neonatal immunoprophylaxis does not prevent all vertical transmission and does not interrupt an infection that occurs in the antenatal period before birth.
Currently, antepartum antiviral prophylaxis to decrease hepatitis B perinatal transmission rates is not the standard of care. Previous small trials have evaluated the risk of perinatal transmission after maternal antenatal treatment with lamivudine and/or hepatitis B immunoglobulin. A recent metaanalysis by Shi et al demonstrated that lamivudine treatment in the third trimester reduced the risk of perinatal transmission in a Chinese population. The mechanism of action of lamivudine involves incorporation of the synthetic nucleoside analog into hepatitis B viral DNA by viral polymerase resulting in DNA chain termination. Usual dosing for chronic hepatitis B is 100 mg daily and common adverse effects include mild diarrhea, nausea, and headaches. Lamivudine appears to cross the term placenta by simple diffusion and interrupts intrauterine hepatitis B infection by decreasing maternal viral load.
For every 1000 pregnant women in the US, 1-2 have hepatitis B. Despite passive and active immunization of the neonate, 1 in 20 infants acquires hepatitis B through perinatal transmission. With a relative risk reduction of 0.38 provided by lamivudine treatment, approximately 50 women with hepatitis B would need to be treated with lamivudine to prevent 1 case of chronic hepatitis B in a child.
Although prevention of hepatitis B infection is undoubtedly desirable, it is uncertain if the resources needed to achieve this through routine antenatal administration of lamivudine are cost-effective. Therefore, using a decision analysis model, we chose to estimate whether third-trimester administration of lamivudine to pregnant patients with chronic hepatitis B is a cost-effective strategy in preventing perinatal transmission.
Materials and Methods
Using a decision analysis model, we compared the cost-effectiveness of 2 management strategies during the final trimester of pregnancy for women with hepatitis B. The first strategy involved expectant management without antepartum maternal prophylaxis. In the second strategy, women with chronic hepatitis B were treated with lamivudine in the third trimester. In both strategies, neonates received the currently recommended hepatitis B immunoglobulin and the hepatitis B vaccine series. The analysis was performed from the perspective of the health care system.
A Markov model was created to estimate lifetime costs in 2011 US dollars and quality-adjusted life years (QALYs) for offspring with vertically transmitted hepatitis B. We included a variety of shifting health states and assigned uses, costs, and probabilities to these specific health states. Both costs and QALYs were discounted at a 3% annual rate. A strategy was considered cost-effective at a ratio less than $50,000/QALY. The decision tree was created and analysis was performed using TreeAge Pro 2009 (TreeAge Software, Williamstown, MA).
The baseline probabilities and outcomes for each strategy were obtained based on a comprehensive English literature review and bibliographic survey. We used the following search terms in PubMed (from 1947-2011): hepatitis B, pregnancy, perinatal/in utero transmission, lamivudine treatment, chronic hepatitis B, cirrhosis, liver transplantation , and combinations of these terms. Point estimates were determined preferentially from systematic reviews and metaanalyses, and prospective and retrospective cohort studies. Review studies and other decision models were used when no other sources for the necessary data were available. In addition, we relied on data from large organizations that follow various long-term outcomes incorporated in the model (Organ Procurement Transplant Network, United Network for Organ Sharing).
In a recent metaanalysis, Shi et al demonstrated that the efficacy of lamivudine in reducing perinatal transmission was 62% as indicated by newborn hepatitis B surface antigen testing. Thus, our estimate of the relative risk of perinatal transmission was 0.38 in patients treated with antenatal lamivudine. In addition, because our model assumed that neonates born to mothers with hepatitis B received both the hepatitis B vaccine series and the hepatitis B immune globulin, we approximated the perinatal transmission rate to be 5% (a range of 3–10% was used in the sensitivity analysis).
We chose to offer lamivudine to all hepatitis B women regardless of e antigen positivity, appreciating that if the analysis for the whole group proved to be cost-effective, offering it to only those at highest risk (hepatitis B e antigen positive) would certainly be even more cost-effective. We also estimated that 90% of infants infected would potentially have long-term consequences of chronic hepatitis B. It is well-established in the literature that perinatal transmission leads to much higher rates of chronic hepatitis B than infections during adulthood.
The probability estimates and references used in support of our model are reported in Table 1 . Our model focused on long-term consequences of chronic infection such as cirrhosis, hepatocellular carcinoma, and need for liver transplantation. The rate of progression from 1 heath state to the next was determined by annual transition probabilities derived from the published literature.
| Variable | Base case | Range a | Reference |
|---|---|---|---|
| Risk of perinatal transmission in setting of neonatal immunoprophylaxis b | 0.05 | 0.03–0.10 | |
| Relative risk of perinatal transmission of hepatitis B with maternal lamivudine | 0.38 | 0.20–0.80 | |
| Risk of chronic hepatitis B after perinatal transmission | 0.90 | 0.85–0.95 | |
| Annual probability of disease progression | |||
| Chronic hepatitis B to compensated cirrhosis | 0.05 | 0.02–0.10 | |
| Compensated cirrhosis to decompensated cirrhosis | 0.04 | 0.02–0.10 | |
| Compensated cirrhosis to hepatocellular carcinoma | 0.03 | 0.01–0.10 | |
| Compensated cirrhosis to death | 0.02 | 0.01–0.10 | |
| Decompensated cirrhosis to compensated cirrhosis | 0.08 | 0–0.16 | |
| Decompensated cirrhosis to hepatocellular carcinoma | 0.07 | 0.01–0.10 | |
| Decompensated cirrhosis to liver transplant | 0.03 | 0.01–0.10 | |
| Decompensated cirrhosis to death | 0.13 | 0.05–0.25 | |
| Hepatocellular carcinoma to liver transplant | 0.10 | 0–0.40 | |
| Hepatocellular carcinoma to death | 0.43 | 0.20–0.60 | |
| Liver transplant to unstable disease (hepatitis B recurrence) c | 0.05 | 0–0.13 | |
| Liver transplant to death (1st year after liver transplant) | 0.14 | 0.10–0.20 | |
| Stable disease (no hepatitis B recurrence) after successful liver transplant to death | 0.05 | 0.02–0.12 | |
| Unstable disease (hepatitis B recurrence) after successful liver transplant to death | 0.18 | 0.05–0.20 |
a The specified ranges were used in the univariate sensitivity analyses and the Monte Carlo simulations;
b This risk assumes that the neonate exposed to maternal hepatitis B received the hepatitis B vaccine series and hepatitis B immunoglobulin;
c Adjusted to account for decreasing mortality over time from transplant.
All uses in our model were assigned a value from 0 to 1. Zero defined no quality of life (death) and 1 defined a perfect health state. Based on previously published cost-effectiveness analyses and systematic reviews, we derived uses for the various health states associated with hepatitis B.
Table 2 lists these health states and their uses. Chronic hepatitis B was assigned a use of 0.99. Although these patients are relatively healthy, they still require physician visits, surveillance of liver function tests, and imaging studies, and, furthermore, likely exhibit a slightly lesser quality of life because of the chronic nature of their disease. We assigned a use of 0.8 to compensated cirrhosis and a use of 0.7 to hepatocellular carcinoma. Decompensated cirrhosis was given a use of 0.6 to reflect the high level of morbidity and mortality associated with this disease state. Typical complications of decompensated cirrhosis include ascites, variceal hemorrhage, encephalopathy, and spontaneous bacterial peritonitis. These complications require immediate medical attention and acute interventions, which are reflected in the relatively low use score.
| Variable | Base case | Range | Reference |
|---|---|---|---|
| Healthy state | 1.0 | N/A | |
| Death | 0 | N/A | |
| Chronic hepatitis | 0.99 | 0.90–0.99 | |
| Compensated cirrhosis | 0.8 | 0.70–0.90 | |
| Decompensated cirrhosis | 0.6 | 0.50–0.70 | |
| Hepatocellular carcinoma | 0.7 | 0.50–0.80 | |
| Liver transplant | 0.86 | 0.70–0.90 | |
| Stable disease (no HBV recurrence) after liver transplant | 0.86 | 0.70–0.90 | |
| Unstable disease (HBV recurrence) after liver transplant a | 0.6 | 0.50–0.70 |
a Use of unstable disease following liver transplant was assumed to be equivalent to use of decompensated cirrhosis.
We designated a use of 0.86 for patients with liver transplants because they have received definitive therapy and have presumably overcome their end-stage liver disease. Nonetheless, their condition requires multiple medications (including antirejection medications), laboratory testing, and physician visits. After liver transplant, various outcomes are possible; patients may either continue to be stable or deteriorate and become unstable. Unstable disease reflects a worsening disease state, most likely because of damage from recurrent hepatitis B. Although the risk of hepatitis B recurrence is declining with the use of hepatitis B immunoglobulin and antiviral medications, this is still an important contributor to morbidity and mortality after transplantation. If patients remained stable after liver transplantation, the use remained constant at 0.86. However, once their clinical condition deteriorated, we assumed that the use of their unstable disease was equivalent to the use of decompensated cirrhosis as worsening disease after transplant is usually associated with a more rapidly progressive course of hepatitis B.
Costs were derived from the published literature and were adjusted to reflect 2011 US dollar amounts ( Table 3 ). The cost of lamivudine was estimated from the average wholesale price as listed by Cardinal Health in 2011 and the Red Book in 2010. Medication costs assumed that patients received 100 mg of lamivudine daily from 28 weeks until delivery (averaged to 3 months). In our model, we assumed that the cost of unstable disease after liver transplantation is similar to the cost of decompensated cirrhosis as both disease states similarly represent an acute progression of worsening liver disease and its complications. Both of these disease states require a multitude of tests, procedures, and therapies including hospital admission, imaging (ultrasound, computed tomographic scan, magnetic resonance imaging/MRCP), serial laboratory tests, endoscopy, liver biopsy, and medications, all of which contribute to the high costs associated with these health states.
| Variable a | Average, $ | Range, $ | Reference |
|---|---|---|---|
| Cost of death | 0 | N/A | |
| Cost of healthy life | 0 | N/A | |
| Cost of lamivudine (100 mg daily for 3 mo) | 1300 | 300–1800 | |
| Cost of diagnosis of chronic hepatitis B | 250 | 100–1000 | |
| Cost of liver transplant | 160,000 | 80,000–200,000 | |
| Annual costs | |||
| Stable chronic hepatitis B | 150 | 100–500 | |
| Compensated cirrhosis | 300 | 120–1200 | |
| Decompensated cirrhosis | 30,000 | 10,000–50,000 | |
| Hepatocellular carcinoma | 45,000 | 10,000–90,000 | |
| Liver transplant after initial first year costs/stable disease | 25,000 | 10,000–50,000 | |
| Unstable disease (HBV recurrence) after liver transplant | 30,000 | 10,000–50,000 |
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
