Professional Documents
Culture Documents
Cost Effectiveness of Maternal Treatment To.23
Cost Effectiveness of Maternal Treatment To.23
ment, in addition to standard neonatal immunoprophylaxis, regions of the world.1,2 Even in the United States,
for the prevention of perinatal hepatitis B virus transmission. which overall is a low-prevalence region (less than
METHODS: A decision-tree model was created to esti- 2%),2 certain immigrant populations have much
mate the cost-effectiveness of maternal administration of higher rates, reflective of their countries of origin.2,3
either lamivudine or HBIG in the third trimester to Perinatal transmission is an important cause of
prevent perinatal hepatitis B transmission compared with chronic hepatitis B, accounting for 35–50% of the
no maternal treatment. The model was first estimated for
disease in endemic countries.4 Whereas adults in-
each treatment using overall transmission rates, and then
fected with hepatitis B virus infections have less than
stratified by maternal hepatitis B virus DNA viral load.
a 5% chance of developing chronic infection, 90% of
RESULTS: The model estimated that for each 100 hepa- perinatally infected children become chronic carri-
titis B surface antigen positive pregnant women treated
ers.5–9 Chronically infected individuals are at high risk
with lamivudine, 9.7 cases of chronic hepatitis B virus
for cirrhosis and hepatocellular carcinoma. Approxi-
infections are prevented, with a cost-savings of $5,184
and 1.3 life-years gained per patient treated. For HBIG,
mately 25% of chronic hepatitis B carriers die prema-
9.5 cases of chronic hepatitis B virus infections are turely as a result of one or both of these complica-
prevented for each 100 pregnant women treated, with a tions,2 and this rate may be up to 40% for perinatally
cost-savings of $5,887 and 1.2 life-years gained per pa- infected individuals.10
tient treated. Under baseline assumptions, lamivudine Currently recommended active-passive immu-
remains cost-saving unless the reduction in perinatal noprophylaxis for all newborns of women positive for
transmission is less than 18.5%, and HBIG remains cost- hepatitis B surface antigen is 85–95% effective in pre-
saving unless the reduction in perinatal transmission is venting perinatal transmission.11 However, multiple
less than 9.6%. studies have shown that women with high levels of viral
CONCLUSION: In this decision analysis, administration replication remain at elevated risk. Perinatal transmis-
of lamivudine or HBIG to hepatitis B surface antigen sion rates despite appropriate neonatal prophylaxis in
positive pregnant women for the prevention of perinatal these highest-risk women have been reported to be
transmission of hepatitis B is cost-savings across a wide 13– 62.5%,12–19 with most estimates between 15% and
range of assumptions. 30%. The transition point in maternal viral load at which
(Obstet Gynecol 2011;118:655–62) transmission rates begin to rise dramatically is approxi-
DOI: 10.1097/AOG.0b013e31822ad2c2 mately 106–108 copies/mL.12,13,16
LEVEL OF EVIDENCE: III A significant proportion of perinatal transmission
in women with high viral loads is now believed to be
From the Departments of Obstetrics and Gynecology, Pediatrics, and Health Science intrauterine.13,14,17,20 –22 Therefore, studies have been
and Research, Medical University of South Carolina, Charleston, South Carolina.
conducted to evaluate the benefit of maternal admin-
Corresponding author: Elizabeth Ramsey Unal, MD, 96 Jonathan Lucas Street, istration of either lamivudine or hepatitis B immune
Suite 634 MSC 619, Charleston SC 29425; e-mail: unal@musc.edu.
globulin (HBIG) in the third trimester to decrease
Financial Disclosure
The authors did not report any potential conflicts of interest.
transmission rates. Most of these studies have been in
Asia, and the results have been promising. In 2010,
© 2011 by The American College of Obstetricians and Gynecologists. Published
by Lippincott Williams & Wilkins. Shi et al published two meta-analyses on the use of
ISSN: 0029-7844/11 third-trimester maternal lamivudine or HBIG in
No Yes
Fig. 1. Decision tree for maternal
treatment with lamivudine or hep-
No hepatitis B Hepatitis B No hepatitis B Hepatitis B atitis B immune globulin to pre-
transmission transmission transmission transmission
vent perinatal hepatitis B virus
transmission.
Resolved Chronic hepatitis B Resolved Chronic hepatitis B
Unal. Preventing Perinatal HBV
Transmission. Obstet Gynecol 2011.
women with high viral loads.18,19 These meta-analyses Medical University of South Carolina Office of Re-
found a 69% and 67% reduction in infant hepatitis B search Integrity states that a study of this kind is not
surface antigen positivity at age 9 –12 months with the considered to be human subjects research and thus
use of lamivudine and HBIG, respectively. Despite does not require University IRB approval.
this evidence, use of these agents to reduce perinatal Our baseline assumptions are outlined in Table 1.
hepatitis B virus transmission is not standard of care The overall perinatal hepatitis B transmission rate
in the United States. Our objective was to estimate the without maternal treatment (15.7%) was obtained by
cost-effectiveness of maternal treatment with lamivu- averaging the transmission rates in the control groups
dine or HBIG. from two recent meta-analyses, one on lamivudine
and the other on HBIG, both by Shi et al.18,19 Both
MATERIALS AND METHODS meta-analyses included randomized controlled trials
We created a decision-tree model to estimate the of maternal lamivudine or HBIG compared with no
cost-effectiveness of maternal administration of lami- maternal treatment, in English language or Chinese
vudine or HBIG in the third trimester to prevent peer-reviewed literature. Transmission rates without
perinatal hepatitis B transmission compared with no treatment stratified by maternal viral load were taken
maternal treatment (Fig. 1). The decision tree was from a study by Yuan et al.12 The odds ratios for
created and run using Microsoft Excel 2008 12.2.3. perinatal transmission with maternal treatment com-
This project was not submitted to our institutional pared with no treatment (0.31 and 0.33 for lamivu-
review board, as all data used were extracted from dine and HBIG, respectively) were taken from the
published studies, and no institutional or individual meta-analyses18,19 and applied to the averaged trans-
patient information is included in this analysis. The mission rate without maternal treatment to obtain
Lamivudine
Overall 15.7 4.9 0.31 0.06–0.94 (ie, 6–94% 18, 19
HBV DNA less than 108 3.3 1.0 0.31 reduction) 12, 18
HBV DNA 108–109 14 4.3 0.31 12, 18
HBV DNA 109–1010 42.9 13.3 0.31 12, 18
HBV DNA 1010–1011 53.3 16.5 0.31 12, 18
HBV DNA more than 1011 62.5 19.4 0.31 12, 18
HBIG
Overall 15.7 5.2 0.33 0.06–0.84 (ie, 16–94% 18, 19
HBV DNA less than 108 3.3 1.1 0.33 reduction) 12, 19
HBV DNA 108–109 14 4.6 0.33 12, 19
HBV DNA 109–1010 42.9 14.2 0.33 12, 19
HBV DNA 1010–1011 53.3 17.6 0.33 12, 19
HBV DNA more than 1011 62.5 20.6 0.33 12, 19
HBV, hepatitis B virus; HBIG, hepatitis B immune globulin.
Data are % unless otherwise specified.
* Defined as positive for hepatitis B surface antigen at 9 –12 months of age.
VOL. 118, NO. 3, SEPTEMBER 2011 Unal et al Preventing Perinatal HBV Transmission 657
ical severity of chronic hepatitis B infection, and infections for each 100 patients treated, well within the
lifelong cost of chronic hepatitis B. We also varied the limits of what is commonly considered to be cost-
percentage of women who would warrant treatment effective. For each of the other cases considered, use of
after viral load determination, to evaluate the cost of lamivudine or HBIG provided a cost savings, in addi-
checking viral loads in all patients to determine need tion to preventing chronic hepatitis B and its sequelae.
for treatment in some. Detailed results for the base-case estimates are outlined
Because HBIG and lamivudine are hypothesized in Table 4.
to have different mechanisms of action in reducing One-way and multi-way sensitivity analyses were
perinatal hepatitis B transmission, it has been sug- performed by altering estimates for the percentage of
gested that additive benefit may be expected if both women whose viral load would warrant treatment (ie,
HBIG and lamivudine are administered in combina- the cost of checking viral loads in all women to
tion.20 To our knowledge, this hypothesis has not determine need for maternal treatment in some), the
been evaluated in prospective studies. To evaluate degree of reduction in perinatal transmission with
whether this strategy would be cost-effective, we maternal treatment, the medical severity and life
estimated a model using both HBIG and lamivudine expectancy in chronic hepatitis B infection, and life-
in combination, assuming varying levels of additional long cost of chronic hepatitis B. The results are
benefit in reducing perinatal hepatitis B transmission. summarized in Table 5.
Adjustment of parameters related to percentage
RESULTS of chronic hepatitis B patients developing cirrhosis,
Using our base-case estimates, use of either lamivudine decompensated cirrhosis, or hepatocellular carcinoma
or HBIG was not only cost-effective, but also cost- had modest effects on lifetime cost of chronic hepatitis B
saving, for all situations except lamivudine at a maternal infection. Altering these parameters yielded a lowest
viral load of less than 108 copies/mL. For every 100 lifetime cost of $72,045 and highest lifetime cost of
pregnant women positive for hepatitis B surface antigen $74,600. If both life expectancy (ranging from 50 years
treated with lamivudine, 9.7 cases of chronic hepatitis B to 64.9 years) and distribution of chronic hepatitis B
virus infections are prevented, with a cost-savings of health states were altered, the lowest lifetime cost was
$5,184 and 1.3 life-years gained per patient treated. For $64,295 and highest was $74,600.
HBIG, 9.5 cases of chronic hepatitis B virus infections Altering the proportion of women who would
are prevented for each 100 pregnant women treated, have a viral load drawn, but not warrant maternal
with a cost savings of $5,887 and 1.2 life-years gained treatment, was also evaluated in the sensitivity analy-
per patient treated. The only instance in which no cost sis. The base-case estimates assumed that all women
savings was achieved (lamivudine at a maternal viral having a third-trimester viral load would be treated
load of less than 108 copies/mL) had a cost per patient of with lamivudine or HBIG. Because most experts
$390 to prevent two cases of chronic hepatitis B virus recommend reserving third-trimester maternal treat-
Lamivudine
Overall 9.7 $5,184 1.3
HBV DNA less than 108 2.0 ⫺$390* 0.3
HBV DNA 108–109 8.7 $4,419 1.1
HBV DNA 109–1010 26.6 $17,410 3.5
HBV DNA 1010–1011 33.1 $22,085 4.3
HBV DNA more than 1011 38.8 $26,220 5.0
HBIG
Overall 9.5 $5,887 1.2
HBV DNA less than 108 2.0 $475 0.3
HBV DNA 108–109 8.4 $5,145 1.1
HBV DNA 109–1010 25.9 $17,760 3.4
HBV DNA 1010–1011 32.1 $22,299 4.2
HBV DNA more than 1011 37.7 $26,314 4.9
HBV, hepatitis B virus; HBIG, hepatitis B immune globulin.
* Additional cost.
Lamivudine
Base estimates $72,384 9.7 $5,184
Altering chronic hepatitis B health states All lowest estimates $72,045 9.7 $5,150
across the ranges given in Table 3 All highest estimates $74,600 9.7 $5,400
Altering chronic hepatitis B health states Lowest cost equals life expectancy of $64,295 9.7 $4,395
across the ranges given in Table 3, as 50 y for chronic hepatitis B
well as varying life expectancy of cirrhosis, highest estimate for
patients with chronic hepatitis B proportion of chronic hepatitis B
cirrhosis from 50–64.9 y having compensated cirrhosis,
lowest estimate for proportion of
decompensated cirrhosis and
hepatocellular carcinoma
Highest cost equals life expectancy of $74,600 9.7 $5,400
64.9 y for chronic hepatitis B
cirrhosis, highest estimates for all
chronic hepatitis B health states
Altering degree of reduction in perinatal 6% reduction 0.8 ⫺$1,260*
transmission with maternal treatment 94% reduction 13.3 $7,741
across range given in Table 1
Altering the proportion of women 10% 9.7 $3,078
warranting treatment after third- 100% 9.7 $5,184
trimester viral load determination
Altering the proportion of women Lowest benefit equals 6% reduction 0.8 ⫺$3,366*
warranting treatment after third- and 10% warrant treatment
trimester viral load determination, as Highest benefit equals 94% reduction 13.3 $7,741
well as varying benefit from treatment and 100% warrant treatment
HBIG
Base estimates $72,384 9.5 $5,887
Altering chronic hepatitis B health states All lowest estimates $72,045 9.5 $5,855
across the ranges given in Table 3 All highest estimates $74,600 9.5 $6,097
Altering chronic hepatitis B health states Lowest cost equals life expectancy of $64,295 9.5 $5,122
across the ranges given in Table 3, as 50 years for chronic hepatitis B
well as varying life expectancy of cirrhosis, highest estimate for
patients with chronic hepatitis B proportion of chronic hepatitis B
cirrhosis from 50–64.9 years having compensated cirrhosis,
lowest estimate for proportion of
decompensated cirrhosis and
hepatocellular carcinoma
Highest cost equals life expectancy of $74,600 9.5 $6,097
64.9 y for chronic hepatitis B
cirrhosis, highest estimates for all
chronic hepatitis B health states
Altering benefit from treatment across the 16% reduction 2.3 $671
ranges given in Table 1 94% reduction 13.3 $8,649
Altering the proportion of women 10% 9.5 $3,781
warranting treatment after third- 100% 9.5 $5,887
trimester viral load determination
Altering the proportion of women Lowest benefit equals 16% reduction 2.3 ⫺$1,435*
warranting treatment after third- and 10% warrant treatment
trimester viral load determination, as Highest benefit equals 94% reduction 13.3 $8,649
well as varying benefit from treatment and 100% warrant treatment
* Additional cost.
VOL. 118, NO. 3, SEPTEMBER 2011 Unal et al Preventing Perinatal HBV Transmission 659
ment for women at highest risk of perinatal transmis- tal prophylaxis, beginning at maternal viral loads
sion from high viral activity, all women would need to around 106–108 copies/mL.12,13,16 The transmission
have a viral load drawn to determine need for treat- rates range from 5% to 13% between 106 and 107
ment. The proportion of women with viral loads high copies/mL, and rise to the 40 –50% range in women
enough to warrant treatment would vary by popula- with viral loads of 108 copies/mL or higher.12,13
tion. Even if only 10% of women qualified for third- Our model does have limitations. The model
trimester maternal treatment based on high viral assumes no significant maternal or fetal risk from
loads, one-way sensitivity analysis demonstrates that lamivudine or HBIG. Based on currently available
this strategy remains cost-saving for both lamivudine data, this is a reasonable assumption, but potential
and HBIG. concerns remain. Studies in both pregnant and non-
The parameter most sensitive to changes in base- pregnant patients show that short courses of lamivu-
line estimates was the degree of reduction in perinatal dine (12–16 weeks) are well tolerated and safe.18,37
transmission with treatment. One-way sensitivity Because a strategy of lamivudine to prevent perinatal
analysis for lamivudine showed that for a baseline transmission would not be initiated until the third
transmission rate without treatment of 15.7%,18 if the trimester, birth defects should not be a concern,
risk reduction with treatment is at least 18.5% (ie, although data regarding teratogenesis are also reas-
from 15.7% to 12.8%), this strategy remains cost- suring.38 One serious theoretical concern regarding
saving. For HBIG, using the baseline transmission use of lamivudine is development of resistance. Long-
rate without treatment of 15.7%, treatment remains term lamivudine use is associated with high rates of
cost-saving if the reduction in perinatal transmission is resistance: 12.8% after 1 year and 76% with 5 or more
at least 9.6% (ie, from 15.7% to 14.2%). When two- years of use.39,40 Resistance has not been reported to
way sensitivity analysis was performed, as long as the
be an issue after short-term use of lamivudine in
degree of risk reduction was at least 39.5% for lami-
pregnancy for perinatal transmission prevention, but
vudine and at least 30.6% for HBIG, treatment re-
it remains a theoretical concern.
mained cost-saving even if only 10% of women tested
Another potential concern is postpartum flare of
with a third-trimester viral load warrant treatment.
disease activity in hepatitis B virus–positive women
When combination therapy of HBIG and lami-
treated with lamivudine. In one Dutch study,41 against
vudine was evaluated using baseline estimates, this
a historical background flare rate of 10 –27% per year,
strategy remained cost-saving even if the degree of
36% of women not treated with lamivudine experi-
reduction was only 1% more than with either lamivu-
dine or HBIG alone. Assuming a baseline transmis- enced a flare (3-fold increase in alanine aminotrans-
sion rate without treatment of 15.7%, and a reduction ferase) in the 6 months after delivery. Among women
in transmission of 70% (compared with 69% for treated with lamivudine, this incidence was 62%. This
lamivudine), the cost savings are $2,278 per patient study was retrospective, and the women in the lami-
treated. If only 10% of patients in whom a viral load vudine group were at greater risk of postpartum flares
is drawn warrant treatment, using lamivudine and by virtue of higher disease activity. Nonetheless,
HBIG in combination remains cost-saving with $1,225 withdrawal flares are known to occur in 17–25% of
saved per patient treated. nonpregnant individuals after cessation of lamivu-
dine therapy,42,43 and this risk will likely need to be
DISCUSSION evaluated in a prospective fashion in women treated
Despite evidence of benefit, use of lamivudine or to prevent perinatal transmission. The women in the
HBIG to reduce perinatal transmission of hepatitis B study described above41 stopped lamivudine imme-
virus is not common practice in the United States. diately after delivery, and the now more common
Our analysis demonstrates that adoption of maternal regimen of continuing therapy until 4 weeks post-
third-trimester treatment with either agent in women partum may provide some protection from post-
with high viral loads would be not only cost-effective, treatment flare.
but also cost-saving. With regard to HBIG, its safety profile is also
Perinatal transmission of hepatitis B is uncom- good. There has been no reported increase in hepa-
mon in women with low viral loads, with some series titis B virus mutation among infected newborns of
even reporting no cases of transmission in women mothers treated with HBIG.19 However, because
with viral loads less than 105 copies/mL.12,16 However, HBIG is made from human plasma, there is a small
the literature is consistent in revealing a dramatic rise theoretical risk of infectious disease transmission, and
in perinatal transmission, despite appropriate neona- anaphylactic reactions are also theoretically possible.
VOL. 118, NO. 3, SEPTEMBER 2011 Unal et al Preventing Perinatal HBV Transmission 661
23. 2010 Red Book: Pharmacy’s Fundamental Reference. Mont- 35. Marx G, Martin SR, Chicoine JF, Alvarez F. Long-term
vale (NJ): Thomson Medical Economics; 2010. follow-up of chronic hepatitis B virus infection in children of
24. Xu JQ, Kochanek KD, Murphy SL, Tejada-Vera B. Deaths: different ethnic origins. J Infect Dis 2002;186:295–301.
final data for 2007. National Vital Statistics Reports, vol. 58, 36. Bortolotti F, Jara P, Crivellaro C, Hierro L, Cadrobbi P, Frauca E,
no. 19. Hyattsville (MD): National Center for Health Statistics; et al. Outcome of chronic hepatitis B in Caucasian children
2010. during a 20-year observation period. J Hepatol 1998;29:184 –90.
25. HCUP Florida State2006 and 2007 Discharge Data. Available 37. Dienstag JL, Perrillo RP, Schiff ER, Bartholomew M, Vicary C,
at: http://www.hcup-us.ahrq.gov/db/state/siddist/sid_multivar. Rubin M. A preliminary trial of lamivudine for chronic
jsp. Retrieved March 2011. hepatitis B infection. N Engl J Med 1995;333:1657– 61.
26. Lee TA, Veenstra DL, Iloeje UH, Sullivan SD. Cost of chronic 38. Antiretroviral Pregnancy Registry Steering Committee. The
hepatitis B infection in the United States. J Clin Gastroenterol Antiretroviral Pregnancy Registry Interim Report 1 January
2004;38(10 suppl 3):S144 –7. 1989 through 31 July 2010. Wilmington (NC): Registry Coor-
27. Chen CJ, Yang HI. Natural history of chronic hepatitis B dinating Center; 2010. Available at: http://www.apregistry.
REVEALed [published online ahead of print February 16, 2011]. com/forms/interim_report.pdf. Retrieved March 12, 2011.
J Gastroenterol Hepatol doi:10.1111/j.1440-1746.2011.06695.x. 39. Zhao S, Tang L, Fan X, Chen L, Zhou R, Dai X. Comparison
of the efficacy of lamivudine and telbivudine in the treatment
28. Fattovich G, Giustina G, Schalm SW, Hadziyannis S, Sanchez-
of chronic hepatitis B: a systematic review. Virol J. 2010;7:211.
Tapias J, Almasio P, et al. Occurrence of hepatocellular
carcinoma and decompensation in western European patients 40. Ghany MG, Doo EC. Antiviral resistance and hepatitis B
with cirrhosis type B. The EUROHEP Study Group on therapy. Hepatology 2009;49(5 suppl):S174 – 84.
Hepatitis B Virus and Cirrhosis. Hepatology 1995;21:77– 82. 41. ter Borg MJ, Leemans WF, de Man RA, Janssen HL. Exacer-
29. Fattovich G, Pantalena M, Zagni I, Realdi G, Schalm SW, bation of chronic hepatitis B infection after delivery. J Viral
Christensen E; European Concerted Action on Viral Hepatitis Hepat 2008;15:37– 41.
(EUROHEP). Effect of hepatitis B and C virus infection on the 42. Honkoop P, de Man RA, Niesters HG, Zondervan PE, Schalm
natural history of compensated cirrhosis: a cohort study of 297 SW. Acute exacerbation of chronic hepatitis B virus infection
patients. Am J Gastroenterol 2002;97:2886 –95. after withdrawal of lamivudine therapy. Hepatology 2000;32:
30. Fattovich G. Natural history of hepatitis B. J Hepatol 2003; 635–9.
39(suppl 1):S50 – 8. 43. Dienstag JL, Schiff ER, Wright TL, Perrillo RP, Hann HW,
31. Fattovich G, Stroffolini T, Zagni I, Donato F. Hepatocellular Goodman Z, et al. Lamivudine as initial treatment for chronic
carcinoma in cirrhosis: incidence and risk factors. Gastroenter- hepatitis B in the United States. N Engl J Med 1999;341:
ology 2004;127(5 suppl 1):S35–50. 1256 – 63.
32. Fattovich G, Olivari N, Pasino M, D’Onofrio M, Martone E, 44. Haber BA, Block JM, Jonas MM, Karpen SJ, London WT,
Donato F. Long-term outcome of chronic hepatitis B in McMahon BJ, et al. Recommendations for screening, monitor-
Caucasian patients: mortality after 25 years. Gut 2008;57: ing, and referral of pediatric chronic hepatitis B. Pediatrics
84 –90. 2009;124:e1007–13.
33. Ikeda K, Saitoh S, Suzuki Y, Kobayashi M, Tsubota A, Koida 45. United States Department of Labor, Bureau of Labor Statistics.
I, et al. Disease progression and hepatocellular carcinogenesis Medical Cost Specific Consumer Price Index. Available at:
in patients with chronic viral hepatitis: a prospective observa- http://data.bls.gov/cgi-bin/surveymost. Retrieved March 5,
tion of 2215 patients. J Hepatol 1998;28:930 – 8. 2011.
34. U.S. Department of Health and Human Services. 2009 Annual 46. Practice Management Information Corporation (PMIC). Med-
Report of the U.S. Organ Procurement and Transplantation ical Fees in the United States 2008. Los Angeles (CA): PMIC;
Network and the Scientific Registry of Transplant Recipients: 2008.
Transplant Data 1999 –2008. Rockville (MD): Health Resources 47. Liaw YF, Lin DY, Chen TJ, Chu CM. Natural course after the
and Services Administration, Healthcare Systems Bureau, Divi- development of cirrhosis in patients with chronic type B
sion of Transplantation; 2009. hepatitis: a prospective study. Liver 1989;9:235– 41.