• No results found

Cost-Effectiveness of Routine Childhood Vaccination for Hepatitis A in the United States

N/A
N/A
Protected

Academic year: 2020

Share "Cost-Effectiveness of Routine Childhood Vaccination for Hepatitis A in the United States"

Copied!
12
0
0

Loading.... (view fulltext now)

Full text

(1)

ARTICLE

Cost-Effectiveness of Routine Childhood Vaccination

for Hepatitis A in the United States

David B. Rein, PhDa, Katherine A. Hicks, MSa, Kathleen E. Wirth, BAa, Kaafee Billah, PhDb†, Lyn Finelli, DrPHb, Anthony E. Fiore, MD, MPHb, Thomas J. Hoerger, PhDa, Beth P. Bell, MD, MPHb, Gregory L. Armstrong, MDb

aRTI International, Atlanta, Georgia;bCenters for Disease Control and Prevention, National Center for Infectious Diseases, Division of Viral Hepatitis, Atlanta, Georgia

The author has indicated he has no financial relationships relevant to this article to disclose.

ABSTRACT

OBJECTIVES.Economic analysis is an important component in formulating national policy. We evaluated the economic impact of hepatitis A vaccination of all US children ages 12 to 23 months as compared with no vaccination and with current implementation of the preexisting (issued in 1999), regional policy.

METHODS.We developed a Markov model of hepatitis A that followed a single cohort from birth in 2005 through death or age 95 years. From the societal perspective, the model compared the outcomes that resulted from routine vaccination at age 1 year to 2 scenarios: no hepatitis A vaccination and hepatitis A vaccination at levels observed in 2003 under the preexisting policy. We evaluated the economic impact of vaccination nationwide, in areas where vaccination was already recommended, and in areas where no previous recommendation existed.

RESULTS.Without childhood vaccination, the⬃4 million children in the 2005 birth cohort would be expected over their lifetimes to have 199 000 hepatitis A virus infections, including 74 000 cases of acute hepatitis A and 82 deaths, resulting in $134 million in hepatitis A–related medical costs and productivity losses. Com-pared with no vaccination, routine vaccination at age 1 year would prevent 172 000 infections, at a cost of $28 000 per quality-adjusted life year saved. Compared with maintaining the levels of hepatitis A vaccination under the pre-existing regional policy, routine vaccination at age 1 year would prevent an additional 112 000 infections, at a cost of $45 000 per quality-adjusted life year saved.

CONCLUSIONS.The cost-effectiveness of nationwide hepatitis A vaccination compared with no vaccination, and the incremental cost-effectiveness of this recommenda-tion compared with preexisting recommendarecommenda-tions, is similar to that of other accepted public health interventions. In October 2005, the Advisory Committee on Immunization Practices recommended extending hepatitis A immunization to all US children ages 12 to 23 months.

www.pediatrics.org/cgi/doi/10.1542/ peds.2006-1573

doi:10.1542/peds.2006-1573

Deceased. Key Words

cost-effectiveness, hepatitis A, vaccination, public policy, economics, ACIP

Abbreviations FLF—fulminant liver failure HAV— hepatitis A virus ACIP—Advisory Committee on Immunization Practices QALY— quality-adjusted life year NNDSS—National Notifiable Disease Surveillance System

CDC—Centers for Disease Control and Prevention

(2)

H

EPATITIS A IScharacterized by abrupt onset of symp-toms such as fever, malaise, anorexia, nausea, ab-dominal discomfort, dark urine, and jaundice. The

syn-drome generally lasts 10 to 33 days1,2followed by a full

recovery, often after a prolonged convalescence. Severe cases can lead to fulminant liver failure (FLF) and

death.3Infection with hepatitis A virus (HAV),

particu-larly among children, can also result in less severe dis-ease, including a relatively mild, anicteric (without jaun-dice) illness or asymptomatic infection.

In the United States, an average of 26 000 hepatitis A cases were reported annually during the 1980s and 1990s, representing an estimated 270 000 infections per year after accounting for unreported cases, anicteric

ill-ness, and asymptomatic infections.4In 1995, highly

ef-fective hepatitis A vaccines were licensed in the United

States for use in adults and in childrenⱖ2 years old. In

2005, the Food and Drug Administration approved low-ering the minimum age for immunization to 12

months.5,6

Soon after hepatitis A vaccines became available, the Advisory Committee on Immunization Practices (ACIP) adopted a strategy of incremental implementation of recommendations for childhood hepatitis A vaccination.

The ACIP first issued recommendations in 1996,7

di-rected at children living in areas with the highest disease rates and periodic outbreaks. In 1999, the next incre-mental step was taken when recommendations were

expanded8 to include routine childhood immunization

in 11 states where hepatitis A rates were historically at least twice the national average. At that time, the ACIP also recommended that providers consider immunizing children in 6 additional states where hepatitis A rates were historically between 1 and 2 times the national average. No recommendation was made for the remain-ing 33 states where rates were historically below the national average. Subsequently, national hepatitis A in-cidence declined to historically low levels, with the larg-est declines observed in the areas in which routine

vac-cination of children was recommended.9,10Despite these

successes, hepatitis A vaccine coverage among children living in states covered by the recommendations has lagged behind the coverage of childhood vaccines for which there is a single nationwide recommendation, and little childhood vaccination has occurred in areas where there were no recommendations for statewide

vaccina-tion of children.11

In October 2005, the ACIP voted unanimously to take the final step in the incremental strategy by recommend-ing routine hepatitis A vaccination for all children aged

12 to 23 months nationwide.12The ACIP was concerned

about the long-term sustainability of the previous policy and the fact that the overwhelming majority of remain-ing US cases of hepatitis A were occurrremain-ing in parts of the country where no hepatitis A vaccination of children was recommended. In this article we describe the

cost-effectiveness of hepatitis A immunization in the United States in terms of the cost per year of life saved and the

cost per quality-adjusted life year (QALY) saved.13 This

analysis, provided to the ACIP as the updated policy was being considered, evaluates the direct (ie, excluding herd immunity) economic impacts of the vaccine, and finds that the economics of nationwide immunization against hepatitis A are favorable when compared both with no immunization and with immunization at current levels.

METHODS

Study Design

We developed a Markov model (a technical report out-lining the details of this model can be accessed on the internet at www.rti.org/pubs/hepa-model_report_rein. pdf) of hepatitis A that simulated clinical and economic outcomes among a single US birth cohort from birth in 2005 through age 95 years. We compared outcomes resulting from routine nationwide vaccination at age 1 to either no hepatitis A vaccination or hepatitis A vacci-nation at levels observed in 2003. The analysis was run nationwide and separately by region. Region 1 was de-fined as the 11 states where hepatitis A vaccination was recommended under the 1999 ACIP policy. Region 2 was defined as the 6 additional states in which the 1999 ACIP policy advised that vaccination should be consid-ered. Finally, region 3 was defined as the remaining 33 states for which no recommendation was made by the 1999 ACIP policy.

All of the cohort members were born susceptible to hepatitis A (Fig 1). In each subsequent time period with-out vaccination, cohort members could be uninfected and susceptible to future infection, actively infected with hepatitis A, recovered and thereby immune to infection, or dead from hepatitis A or other causes. The model incorporated the full range of potential hepatitis A symp-tomatic states from asympsymp-tomatic infection to FLF. With vaccination, patients would immediately develop immu-nity that would wane slowly over time.

Data and Assumptions

The probabilities of the model were derived from pub-lished studies, notifiable disease data, previously unpub-lished vaccine efficacy data, proprietary adult vaccine sales data, expert opinion, and assumptions (Table 1). Costs, life years, and QALYs were considered from the societal perspective and discounted on a present-value basis using a 3% annual rate. Costs were measured in 2005 dollars.

Incidence of Infection

US hepatitis A incidence trends have been characterized

by ⬃5- to 15-year cycles and by an overall decline. To

(3)

logarithm of reported hepatitis A rates against year, from 1966 to 2001, with incidence after 1994 adjusted to

compensate for the effects of immunization.9This

regres-sion showed an annual rate of decline of 1.4% after 1990. The model assumed that without immunization this rate of decline would have continued indefinitely from 1990. Historically, incidence also varied regionally

and by age.14 Regional age-specific incidence was

esti-mated by multiplying the estiesti-mated average incidence for each region by the ratio of age-specific incidence to overall incidence during 1990 –1995, the years

immedi-ately preceding the licensing of hepatitis A vaccine.14

Across all of the regions, we assumed 3.28 unreported

cases for each reported case of hepatitis A.4 All of the

reported cases were assumed to be icteric. The number of additional anicteric cases (asymptomatic or mildly symptomatic without jaundice) was estimated by

apply-ing an age-specific ratio, also determined by regression.4

Disease Severity, Service Use, and Costs of Care

The probability of receiving health care given icteric

infection was taken from published studies.15The

prob-ability of hospitalization given reported icteric infection was set equal to the average probability of hospitaliza-tion from 1990 through 1995 as reported by the Nahospitaliza-tional

Notifiable Disease Surveillance System (NNDSS)14 and

the National Hospital Discharge Survey.16Icteric patients

treated as outpatients were assumed to have more se-vere illness and to consume more health resources if their cases were reported to the health department; all hepatitis A hospitalizations were assumed to be reported. Half of anicteric infections were assumed to be asymp-tomatic and the other half to have mild, nonspecific symptoms of short duration requiring a single outpatient visit 50% of the time.

The number of deaths resulting from hepatitis A was estimated from NNDSS data. We assumed that death could occur only from hepatitis A after FLF and that all patients with FLF would be treated as inpatients. Values for the probability of death given FLF and no transplant and the probability of transplant given FLF were taken

from published studies.17The probability of death after

transplant was based on data from the United Network

for Organ Sharing.18The age-specific probability of FLF

was then calculated based on the known probabilities of inpatient admissions, transplants, and death from FLF with or without transplant and the total number of hepatitis A–related deaths identified in the NNDSS data. Medical costs of outpatient care, inpatient care, FLF, and liver transplants were taken from published stud-FIGURE 1

(4)

TABLE 1 Summary of Input Parameters for the Model

Variable Base-Case Value Source

Population and incidence No. of children in cohort

Region 1 942 913 National Vital Statistics41(2003)

Region 2 571 155 National Vital Statistics41(2003)

Region 3 2 507 658 National Vital Statistics41(2003)

Overall annual hepatitis A incidence per 100 000 in the absence of vaccination, 2005a

Region 1 22.6 Estimate based on Samandari et al9(2004)

Region 2 14.1 Estimate based on Samandari et al9(2004)

Region 3 6.7 Estimate based on Samandari et al9(2004)

Annual decline in incidence without vaccination, % 1.4 Samandari et al9(2004)

Vaccination

Probability of receiving first dose if recommended .93 Calculated from CDC26(2004)

Probability of receiving second dose if recommended .87 Calculated from CDC26(2004)

Probability of receiving first dose under previous recommendations, by regionb

Recommended .5 Fiore et al11(2005)

Considered .25 Fiore et al11(2005)

No recommendation .01 Fiore et al11(2005)

Antibody titer after vaccination with hepatitis A vaccine, mIU/mL

After first dose 141 Van Herck et al30(2000)

After second dose 1120 Van Herck et al30(2000)

Annual decline in antibody titer after vaccination, %

Years 0–4 20 Van Herck et al30(2000)

Years 5–95 5 Unpublished data

Antibody titer threshold for immunity, mIU/mL 20 Delem et al42(1993)

Outcomes of infection

Probability of nonspecific symptoms given nonicteric infection .5 Assumption Probability of icteric infection according to agec

0–4 y .076 Armstrong et al4(2002)

5–14 y .512 Armstrong et al4(2002)

15–95 y .832 Armstrong et al4(2002)

Probability of liver transplant given icteric infection

Dependent on age .00002–.00241 Schiodt et al17(2003)

Probability of death given icteric infection

0–4 y .00030 CDC14(2001)

5–14 y .00004 CDC14(2001)

15–39 y .00054 CDC14(2001)

40–59 y .00436 CDC14(2001)

60–95 y .01276 CDC14(2001)

Probability that an icteric case is reportedc .23 Armstrong et al4(2002)

Probability of hospitalization given that an icteric infection is reported according to age

0–4 y .05 CDC14(2001)

5–14 y .17 CDC14(2001)

15–39 y .23 CDC14(2001)

40–59 y .19 CDC14(2001)

60–95 y .2 CDC14(2001)

Costs of medical intervention, $

Average cost of childhood vaccine per dose 17.96 CDC27(2006)

Average cost of adult vaccine per dose 55.49 CDC27(2006)

Vaccine administration cost per dose according to age

1 yd 6 CDC14(2001), assumption

Symptomatic anicteric infection requiring outpatient visit 84.16 American College of Physicians21

Unreported outpatient icteric 274 Berge et al2(2000), assumption

Reported outpatient icteric 821 Berge et al2(2000)

Inpatient icteric 8111 Dalton et al19(1996) and Berge et al2(2000)

Fulminant disease without transplant 24 138 Berge et al2(2000)

Liver transplant (first year) 285 900 Hauboldt et al20(1999)

Liver transplant (subsequent years) 25 598 Berge et al2(2000)

Public health cost per reported case 667 CDC Division of Viral Hepatitis

Quality-of-life values

Quality-adjusted value of life during icteric hepatitis A 0.41 Jacobs et al24(2002)

Quality-adjusted value of life during anicteric hepatitis A with symptoms 0.83 Mittmann et al25(2001)

Duration of outpatient icteric hepatitis A, d 34.4 Berge et al2(2000)

Duration of inpatient icteric hepatitis A, d 67.8 Berge et al2(2000)

Duration of anicteric hepatitis A with symptoms, d 3 Assumption

Range and distributional assumptions can be found in the technical report (www.rti.org/pubs/hepa-model_report_rein.pdf). aRelative risk of infection varies widely by age and region.

bScenario 2 compares current immunization levels with increased levels under routine recommendations.

cBecause the values of several other variables were a function of the probability of icteric infection and the probability that an icteric case is reported, these values were held constant in the simulations.

(5)

ies.2,19,20The costs of unreported outpatient icteric cases were assumed to be one third those of reported cases. We assumed that patients with symptomatic anicteric infections who used medical services used only 1 outpa-tient visit and that the cost of that visit was the same as that for a visit for fever and malaise associated with

influenza.21 Productivity losses, based on the expected

duration of the episode of illness and the expected wage of the patient or caretaker, were included in the model to account for the work loss incurred by caregivers of cohort members infected during childhood and by

co-hort members themselves when infected as adults.2,22

Only work losses incurred by caregivers of cohort mem-bers who were infected while children were included in the numerator used to calculate the cost per QALY saved ratio, whereas all of the productivity losses were in-cluded in the numerator used to calculate the cost-per-life-year-saved ratio. This is because patients were asked to consider work losses that they would experience as adults in the survey used to determine QALY values.

Public Health Costs

As a reportable infectious disease, hepatitis A cases cre-ate public health costs associcre-ated with surveillance, con-tact tracing, and outbreak response that would be largely eliminated through routine childhood vaccination. We estimated the cost and probability of hepatitis A–related physician and patient interviews, contact tracing, im-mune globulin distribution, and public notification from

expert opinion and published reports.23

Health Utilities

Relative quality-of-life values were taken directly from the tables found in a published time tradeoff estimation

of utility losses associated with hepatitis A.24,25Full-year

QALY values were then adjusted for partial-year disease duration, with different durations for different

symp-tomatic states.2These adjusted QALY weights were

mul-tiplied by age-specific background QALYs to account for the background prevalence of chronic disease and dis-ability in the population.

For vaccine studies, cost-effectiveness ratios based on traditional time tradeoff utility estimates are directly comparable to the majority of other vaccine cost-effec-tiveness studies. However, more recently, some econo-mists have been concerned about the accuracy of

tradi-tional QALY estimation methods.26 To account for

uncertainty in the precise value of QALYs associated with hepatitis A, we used a low and high method to calculate QALYs in our sensitivity analysis. Our low estimate simply divided the QALY losses for each hepa-titis A health state in half compared with the baseline. Our high estimate assumed no loss of background QALYs with age, a method commonly used in earlier vaccination cost-effectiveness studies.

Childhood Vaccine Coverage and Costs

Vaccine coverage for 1 and 2 doses was set equal to the

average coverage in 2003 forHaemophilus influenzaetype

b vaccine and pneumococcal conjugate vaccine, both of which are routinely administered between ages 1 and 2

years.27For the scenario comparing routine vaccination

to vaccination at current levels, current coverage levels

were taken from a 2003 survey.11 Vaccine costs were

based on the reported public and private prices paid for childhood vaccine by the Vaccines for Children

pro-gram,28 assuming that 59% of the vaccine would be

purchased at the lower public contract price (Centers for Disease Control and Prevention [CDC], unpublished data, 1995–2003). Excise taxes were excluded. For each

vaccine dose, patients had a 1 in 200 chance29of a mild

adverse reaction that imposed only slight productivity costs and a 1 in 1000 000 chance of a severe adverse reaction requiring some medical attention.

Vaccine Efficacy and Duration of Vaccine-Acquired Immunity

Not all patients develop immunity from vaccination and in those that do, vaccine-acquired immunity wanes over time. In the model, the establishment of initial immunity and the years of duration of immunity after vaccination was estimated from the mean and distribution of geo-metric mean titers after 1 and 2 doses of vaccine

ob-served in published studies30,31and in primary data.

Us-ing these data, our model predicted that 91% of those

vaccinated would develop immunity forⱖ1 year after 1

dose of vaccine, and 100% of those vaccinated would

develop immunity for ⱖ1 year after 2 doses.

Subse-quently, the model used an estimated 20% annual de-cline in antibody concentrations for the first 5 years after

vaccination,30a 5% annual decline in the years

thereaf-ter (P. Van Damme, MD, PhD, written communication,

2005), and a threshold for immunity of 20 mIU/mL31to

calculate the duration of immunity among those who exhibited an initial response. After vaccination with 1 or 2 doses, the vaccinated patient cohort was distributed according to the estimated number of years of vaccine-acquired immunity remaining. For patients who received 1 dose of vaccine that developed an initial immune response, the model estimated that 37% maintained immunity for 1 to 10 years, 20% for 11 to 20 years, 30% for 21 to 40 years,

and 13% for ⱖ41 years. The model estimated that all of

the patients who received 2 doses of vaccine developed an

immunity forⱖ20 years: 10% for between 21 and 40 years

and 90% forⱖ41 years. The model assumed that patients

who lost vaccine-acquired immunity had the same risk of infection as those who were susceptible and never vacci-nated.

Adult Vaccination

Based on vaccine sales data,9unvaccinated adults in the

model aged⬎18 years were vaccinated at a rate of 1.3%

(6)

in region 3. Most of this adult vaccination would be unnecessary if routine childhood immunization were implemented.

Sensitivity Analyses

In univariate sensitivity analyses, we tested the sensitiv-ity of our model to changes in the discount rate, the baseline and annual rates of decline of hepatitis A inci-dence, the long-term decline in antibody to HAV after immunization, the rate of adult vaccination, the value of QALY decrements associated with illness, public health costs, and symptomatic adverse events. In a multivariate probabilistic sensitivity analysis, we tested the sensitivity of the model to the combined uncertainty of all of the model parameters. This was done by randomly and in-dependently altering all of the model parameters within their margin of error in each of 10 000 iterations of the model and then evaluating the range of cost-effective-ness ratios that were generated.

RESULTS

Disease Burden and Costs of Hepatitis A

Our model estimated that without childhood vaccina-tion, 198 751 members (4.9%) of the 2005 birth cohort would be infected with HAV during their lifetimes. Of these infections, 124 939 would be asymptomatic or in-volve mild nonspecific symptoms and 73 812 would re-sult in acute hepatitis A with jaundice. Of the jaundiced cases, 222 would progress to FLF, resulting in 33 liver transplants and 82 deaths (Table 2). These cases would produce $133.5 million in total economic costs,

includ-ing $86.8 million in health care and adult vaccination costs, $31.6 million in productivity losses among parents of infected children, and $15.3 million in productivity losses among adults with hepatitis A.

Compared with no childhood vaccination, routine vaccination at age 1 year would result in 172 334 fewer infections and 32 fewer deaths nationwide. This reduc-tion would produce a gain of 247 discounted life years or 2154 discounted QALYs (Table 3). Routine vaccination at age 1 year would cost $49.3 million more than no vaccination, resulting in cost-effectiveness ratios of $284 per infection averted, $199 000 per life year gained, and $28 000 per QALY saved. Compared with coverage

lev-els in 2003,10 routine nationwide vaccination at age 1

year would result in 112 411 fewer infections and cost an additional $55.8 million, with incremental cost-effec-tiveness ratios of $496 per infection averted, $338 000 per life year gained, and $45 000 per QALY saved. From the perspective of the health care system (ie, excluding productivity losses), nationwide routine vaccination costs $40 000 per QALY saved compared with no vacci-nation and $57 000 per QALY saved when compared with coverage levels in 2003.

Cost-effectiveness varied by region. Compared with no vaccination, routine vaccination at age 1 year was

TABLE 2 Expected Hepatitis A Disease Outcomes and Costs with No Immunization

Variable ACIP Region Regions

1 and 2a

Region 3b

All Regions

Outcomes

No. of disease events

Symptomatic anicteric 50 239 12 230 62 469 Icteric (acute hepatitis A) 55 318 18 493 73 812

FLF 160 62 222

Deaths 60 22 82

LYs lost from hepatitis Ac 268 101 369

QALYs lost from hepatitis Ac 1959 619 2577

No. of medical services

Outpatient visits 143 189 45 587 188 775

Hospitalizations 2608 910 3518

Liver transplants 23 9 33

Costs

Total costs for cost per LY saved, $ 76 500 000 35 900 000 112 400 000 Total costs for cost per QALY saved, $d 90 200 000 43 300 000 133 500 000

aRegions 1 and 2 include 11 states for which the ACIP recommended hepatitis A vaccination and 6 states for which the ACIP called for consideration of vaccination.

bRegion 3 includes 33 states and the District of Columbia, for which the ACIP made no recom-mendation regarding hepatitis A immunization.

cDiscounted to 2005 value.

dExcludes productivity losses other than those associated with caregivers of infected children.

TABLE 3 Impact of Childhood Immunization at 1 Year of Age Compared With No Vaccination in 2 US Regions and Compared Nationally With No Vaccination

Outcomes ACIP Region

Regions 1 and 2a

Region 3b

All Regions

Reduction in disease events,n

Infections 135 486 36 848 172 334

Deaths 22 10 32

LYs savedc 178 70 247

QALYs savedd 1640 514 2154

Reduction in medical services,n

Outpatient visits 116 522 36 847 153 369

Hospitalizations 1996 703 2699

Liver transplants 14 6 20

Change in costs, $

Total healthcare costs 12 800 000 73 500 000 86 300 000 Vaccination 46 300 000 84 200 000 130 500,000 Hepatitis A medical costs (29 300 000) (9400,000) (38 700 000) Public health costs (4200,000) (1300,000) (5500,000) Total productivity costs (28 400 000) (8600,000) (37 000 000) Total costs for cost per LY savedc (16 000 000) 65 300 000 49 300 000

Total costs for cost per QALY savedd (7300,000) 68 200 000 60 900 000

Cost per LY savedc Cost-saving 933 000 199 000

Cost per QALY savedd Cost-saving 133 000 28 000

aRegions 1 and 2 include 11 states for which the ACIP recommended hepatitis A vaccination and 6 states for which the ACIP called for consideration of vaccination.

bRegion 3 includes 33 states and the District of Columbia, for which the ACIP made no recom-mendation regarding hepatitis A immunization.

cCost per LY saved is equal to discounted total costs for cost per LY saved divided by discounted LYs saved.

(7)

cost saving in regions 1 and 2 and cost $933 000 per life year saved and $133 000 per QALY saved in region 3. Compared with immunization at 2003 vaccine coverage levels, routine vaccination was cost saving in regions 1 and 2 and cost $927 000 per life year saved and $132 000 per QALY saved in region 3. From the perspec-tive of the health care system (ie, excluding productivity losses), routine vaccination costs $8000 per QALY saved in regions 1 and 2 and $143 000 per QALY saved in region 3 when compared with no vaccination and $9000 per QALY saved in regions 1 and 2 and $143 000 per QALY saved in region 3 when compared with 2003 vaccine coverage levels.

Sensitivity Analyses

The cost-effectiveness of vaccination was fairly sensitive to the combined cost of a vaccine dose and administra-tion, the baseline incidence value, the discount rate, and the use of a more conservative QALY decrement associ-ated with illness (Fig 2) If the total vaccination cost per dose, including acquisition and administration costs, de-creased to $17 (roughly the public sector cost of vacci-nation at age 1 year), the incremental cost per QALY of

routine vaccination at age 1 would fall to approximately $7000 compared with no vaccination and approximately $18 000 compared with vaccination at 2003 vaccine coverage levels. Decreasing the baseline incidence in each region by 25% resulted in a cost-effectiveness ratio of $73 000 per QALY. Likewise, increasing the discount rate decreased the cost-effectiveness of vaccination. Re-ducing the QALY decrements associated with symptom-atic hepatitis A illness by half increased the cost-effec-tiveness of nationwide vaccination to $71 000 per QALY compared with no vaccination.

Our results were moderately sensitive to the rate of decline in vaccine-induced antibody to HAV. Assuming a vaccine that was half as effective as what has been observed in clinical trials thus far would yield a cost-effectiveness ratio of $62 000 per QALY. Relative to these other variables, the model was fairly insensitive to the inclusion of public health costs, the annual rate of decline in HAV incidence, and the inclusion of adult vaccination costs.

In 10 000 simulations of the probabilistic sensitivity analysis for immunization at age 1 year compared with no vaccination, the cost per QALY ratio fell below

FIGURE 2

(8)

$50 000 in 82% of the simulations, between $50 000 and $75 000 in 15% of the simulations, between

$75 000 and $100 000 in⬍3% of the simulations, and

above $100 000 in⬍1% of the simulations. Vaccination

was essentially cost-neutral or cost-saving in 8% of the simulations (Fig 3).

DISCUSSION

Hepatitis A causes relatively few deaths but often results in prolonged, temporarily disabling illness and is, thus, responsible for substantial morbidity. In the model pre-sented here, without immunization, the 2005 US birth

cohort could expect ⬃200 000 HAV infections and

$133.5 million of related economic losses. In this sce-nario, the cost-effectiveness ratio of introducing routine childhood hepatitis A immunization ($28 000 per QALY saved) is comparable to that of other public health in-terventions, such as expanded HIV screening among

high-risk patients ($36 000 per QALY)32and the general

population ($42 000 per QALY, excluding secondary

benefits)33 and diabetes screening among patients with

hypertension ($34 000).34 This favorable

cost-effective-ness ratio is similar to findings from a previous analysis of the economics of nationwide hepatitis A vaccination

of children.35 However, that analysis depended on the

inclusion of benefits from averted secondary infections whereas this one does not.

Compared with no vaccination, the cost-effectiveness of nationwide routine hepatitis A vaccination at age 1 year is similar to that of other recently recommended

vaccine policies.36,37Even compared with the preexisting

regional policy, which focused vaccination efforts on areas with the highest disease burden, the cost-effective-ness of nationwide routine vaccination is more favorable than that of nationwide routine adolescent vaccination

for meningitis ($138 000 per QALY),38also recently

rec-ommend by the ACIP. Nationwide hepatitis A vaccina-tion would be less cost-effective than vaccinavaccina-tion against

varicella39 or hepatitis B,40 which are both cost-saving.

Cost-effectiveness varied substantially by region, and the cost-effectiveness of vaccinating only in region 3 was less favorable ($132 000 per QALY) but comparable to

rou-tine adolescent vaccination for meningitis.37

Both the cost-effectiveness ratio comparing nation-wide vaccination to no vaccination and the incremental cost-effectiveness ratio of expanding immunization rec-ommendations from the 1999 regional strategy support a policy of expanded vaccination. Although some econ-omists might consider the incremental analysis the best reflection of the costs and benefits of the new policy, many policy-makers doubt its relevance in public health practice. From its inception, the regional strategy was considered to be an interim step and not a permanent recommendation. Concerns about its sustainability

com-FIGURE 3

(9)

plicates the definition of a status quo against which to compare the new policy. For this analysis, we assumed that without a change in policy, immunization would continue at 2003 levels. At these levels, half of children in region 1, 75% of children in region 2, and 99% of children in region 3 would remain unprotected from hepatitis A. If immunization rates were to fall, hepatitis A incidence would likely increase. HAV infection has not been eradicated and, except for the presence of immu-nization, the epidemiological conditions that allow the spread of HAV in the United States have not changed.

This study has several limitations. First, the modeled hepatitis A incidence represents an averaged value ob-served over a 5-year period, whereas the actual inci-dence of hepatitis A is cyclical and varies greatly within

each region.14 This variation is important, because the

actual incidence will influence the degree of economic benefit derived from vaccination in any given year and locality. Second, in practice, it may take several years after issuance of any new vaccination recommendation for coverage to reach that of other childhood vaccines. We did not consider this phase-in period for nationwide hepatitis A vaccination in our analysis. During this pe-riod, the costs and the benefits will be lower than those used in this analysis, because fewer doses of vaccine will be administered. The effect of these differences on the interim cost-effectiveness of the policy is unknown. Third, except for the productivity costs associated with caretakers of infected children, our model excludes any costs incurred outside of the birth cohort, nor does it include the benefits of vaccination associated with herd immunity. Previous research suggests substantial eco-nomic benefits from a reduction in secondary infections among household and social contacts of immunized

chil-dren.35 The cost-effectiveness of immunization against

hepatitis A becomes substantially more cost-effective over the first 10 years of implementation when these

out-of-cohort effects are included in our model.40

Finally, some authors have argued that the standard method of assessing QALY decrements may overvalue

the benefit of preventing short-term health states.26

However, even halving the decrement, as was done in the sensitivity analysis, results in a nationwide cost-effectiveness ratio that is comparable to that of other recently recommended vaccinations.

From an epidemiological perspective, the interim re-gional ACIP recommendations were highly successful,

reducing hepatitis A rates to all-time lows.10 However,

even at current low rates, there remains a substantial burden of disease from hepatitis A, with the majority in states not covered by the regional recommendations. These and other factors, including the ability now to readily incorporate the vaccine into the routine early childhood vaccination schedule, prompted the ACIP to bring recommendations for using this vaccine into line with those for vaccines for other common infections.

The results of the analysis presented here, indicating that the economics of an expanded vaccination policy are reasonable, formed an essential component of the ACIP conclusion that the time was right for finalizing national universal hepatitis A vaccination of children in the United States.

REFERENCES

1. O’Connor JB, Imperiale TF, Singer ME. Cost-effectiveness analysis of hepatitis A vaccination strategies for adults. Hepa-tology.1999;30:1077–1081

2. Berge JJ, Drennan DP, Jacobs RJ, et al. The cost of hepatitis A infections in American adolescents and adults in 1997. Hepa-tology.2000;31:469 – 473

3. Craig AS, Schaffner W. Prevention of hepatitis A with the hepatitis A vaccine.N Engl J Med.2004;350:476 – 481 4. Armstrong GL, Bell BP. Hepatitis A virus infections in the

United States: model-based estimates and implications for childhood immunization.Pediatrics.2002;109:839 – 845 5. Centers for Disease Control and Prevention. Notice to readers:

FDA approval of VAQTA (hepatitis A vaccine, inactivated) for children agedⱖ1 year.MMWR Morb Mortal Wkly Rep.2005;54: 1026

6. Food and Drug Administration. Havrix product approval infor-mation. Available at: www.fda.gov/cber/approvltr/havgsk101705L. htm. Accessed November 3, 2005

7. Advisory Committee on Immunization Practices. Prevention of hepatitis A through active or passive immunization: recom-mendations of the Advisory Committee on Immunization Prac-tices (ACIP). [published correction appears in MMWR Morb Mortal Wkly Rep.1997;46:588].MMWR Recomm Rep. 1996;45: 1–30

8. Advisory Committee on Immunization Practices. Prevention of hepatitis A through active or passive immunization: recom-mendations of the Advisory Committee on Immunization Prac-tices (ACIP).MMWR Recomm Rep.1999;48:1–37

9. Samandari T, Armstrong GL, Bell BP. Quantifying the impact of hepatitis A immunization in the United States, 1995–2001.

Vaccine.2004;22:4342– 4350

10. Wasley A, Samandari T, Bell BP. Incidence of hepatitis A in the United States in the era of vaccination.JAMA.2005;294:194 –201 11. Fiore A, Bell B, Barker L, Darling N, Amon J. Hepatitis A vaccination coverage among children aged 24 –35 months: United States, 2003. MMWR Morb Mortal Wkly Rep.2005;54: 141–144

12. Reuters Health Information. CDC committee recommends rou-tine hepatitis A vaccination for children. Available at: www. medscape.com/viewarticle/515745. Accessed November 3, 2005

13. Gold MR, Siegel JE, Russell LB, Weinstein MC, eds. Cost-Effectiveness in Health and Medicine. New York, NY: Oxford Uni-versity Press; 1996

14. Centers for Disease Control and Prevention. Summary of no-tifiable diseases.MMWR Morb Mortal Wkly Rep.2001;48:89 –98 15. Bownds L, Lindekugel R, Stepak P. Economic impact of a hepatitis A epidemic in a mid-sized urban community: the case of Spokane, Washington. J Community Health. 2003;28: 233–246

16. National Center for Health Statistics. National Hospital Dis-charge Survey 1990 –95. Public use data file and documenta-tion. Hyattsville, MD: National Center for Health Statistics. Available at ftp://ftp.cdc.gov/pub/Health_Statistics/NCHS/ Datasets/NHDS. Accessed November 8, 2006

(10)

18. United Network for Organ Sharing. Liver Kaplan-Meier patient survival rates for transplants performed 1990 –2003. Primary diagnosis hepatitis A. Available at: www.unos.org/data. Ac-cessed November 10, 2004

19. Dalton CB, Haddix A, Hoffman RE, Mast EE. The cost of a food-borne outbreak of hepatitis A in Denver, Colo.Arch Intern Med.1996;156:1013–1016

20. Hauboldt RH. CostImplications of Human Organ and Tissue Trans-plantations, an Update: 1999. Seattle, WA: Milliman, Robertson; 1999

21. American College of Physicians. Preventive strategies for influenza: is influenza prophylaxis cost-effective? Available at: www.acponline.org/flu/cost_effectiveness_evidence.htm. Ac-cessed December 1, 2004

22. Current Population Survey BoLS. Table 2. Median usual weekly earnings of full-time wage and salary workers by age, race, Hispanic or Latino ethnicity, and gender, fourth quarter 2003 averages, not seasonally adjusted. Available at: www.bls. gov/schedule/archives/wkyeng_nr.htm#2003. Accessed No-vember 8, 2004

23. Meltzer MI, Shapiro CN, Mast EE, Arcari C. The economics of vaccinating restaurant workers against hepatitis A. Vaccine.

2001;19:2138 –2145

24. Jacobs RJ, Moleski RJ, Meyerhoff AS. Valuation of symptom-atic hepatitis A in adults: estimates based on time trade-off and willingness-to-pay measurement.Pharmacoeconomics.2002;20: 739 –747

25. Mittmann N, Chan D, Trakas K, Risebrough N. Health utility attributes for chronic conditions.Dis Manage Health Outcomes.

2001;9:11–21

26. Centers for Disease Control and Prevention.2003 National Im-munization Survey. Hyattsville, MD: National Center for Health Statistics; 2004

27. Centers for Disease Control and Prevention. Vaccines for Chil-dren Program: vaccine price list. Available at: www.cdc.gov/ nip/vfc/cdc_vac_price_list.htm. Accessed November 10, 2004 28. Innis BL, Snitbhan R, Kunasol P, et al. Protection against

hepatitis A by an inactivated vaccine.JAMA.1994;271:1328 – 1334

29. Van Damme P, Banatvala J, Fay O, et al. Hepatitis A booster vaccination: is there a need?Lancet.2003;362:1065–1071 30. Van Herck K, Beutels P, Van Damme P, et al. Mathematical

models for assessment of long-term persistence of antibodies after vaccination with two inactivated hepatitis A vaccines.

J Med Virol.2000;60:1–7

31. Paltiel AD, Weinstein MC, Kimmel AD, et al. Expanded screen-ing for HIV in the United States: an analysis of cost-effectiveness.N Engl J Med.2005;352:586 –595

32. Sanders GD, Bayoumi AM, Sundaram V, et al. Cost-effectiveness of screening for HIV in the era of highly active antiretroviral therapy.N Engl J Med.2005;352:570 –585 33. Hoerger TJ, Harris R, Hicks KA, Donahue K, Sorensen S,

En-gelgau M. Screening for type 2 diabetes mellitus: a cost-effectiveness analysis.Ann Intern Med.2004;140:689 – 699 34. Jacobs RJ, Greenberg P, Koff RS, Saab S, Meyerhoff AS.

Re-gional variation in the cost-effectiveness of childhood hepatitis A immunization.Pediatr Infect Dis J.2003;22:904 –914 35. Shepard CW, Ortega-Sanchez IR, Scott RD, Rosenstein NE.

Cost-effectiveness of conjugate meningococcal vaccination strategies in the United States.Pediatrics.2005;115:1220 –1232 36. Lee GM, Lebaron C, Murphy TV, Lett S, Schauer S, Lieu TA. Pertussis in adolescents and adults: should we vaccinate? Pedi-atrics.2005;115:1675–1684

37. Lieu TA, Cochi SL, Black SB, et al. Cost-effectiveness of a routine varicella vaccination program for United States chil-dren.JAMA.1994;271:375–381

38. Margolis HS, Coleman PJ, Brown RE, Mast EE, Sheingold SH, Arevalo JA. Prevention of hepatitis-B virus transmission by immunization: an economic analysis of current recommenda-tions.JAMA.1995;274:1201–1208

39. Armstrong GL, Billah K, Rein DB, Hicks KA, Wirth KE, Bell BP. The economics of routine childhood hepatitis A immunization in the United States: the impact of herd immunity.Pediatrics.

2007;119(1). Available at: www.pediatrics.org/cgi/content/ full/119/1/e22

40. Lee GM, Salomon JA, LeBaron CW, Lieu TA. Health-state valuations for pertussis: methods for valuing short-term health states.Health Qual Life Outcomes.2005;3:17

41. National Vital Statistics Reports. Table 10. Number of births, birth rates, fertility rates, total fertility rates, and birth rates for teenagers 15–19 years by age of mother: United States, each State and territory, 2002.Nat Vital Stat Rep.2003;52:10 42. Delem A, Safary A, De Namur F, Hauser P, D’Hondt E.

Char-acterization of the immune response of volunteers vaccinated with a killed vaccine against hepatitis A. Vaccine. 1993;11: 479 – 484

(11)

DOI: 10.1542/peds.2006-1573

2007;119;e12

Pediatrics

Anthony E. Fiore, Thomas J. Hoerger, Beth P. Bell and Gregory L. Armstrong

David B. Rein, Katherine A. Hicks, Kathleen E. Wirth, Kaafee Billah, Lyn Finelli,

United States

Cost-Effectiveness of Routine Childhood Vaccination for Hepatitis A in the

Services

Updated Information &

http://pediatrics.aappublications.org/content/119/1/e12

including high resolution figures, can be found at:

References

http://pediatrics.aappublications.org/content/119/1/e12#BIBL

This article cites 31 articles, 3 of which you can access for free at:

Subspecialty Collections

b

http://www.aappublications.org/cgi/collection/infectious_diseases_su Infectious Disease

following collection(s):

This article, along with others on similar topics, appears in the

Permissions & Licensing

http://www.aappublications.org/site/misc/Permissions.xhtml

in its entirety can be found online at:

Information about reproducing this article in parts (figures, tables) or

Reprints

http://www.aappublications.org/site/misc/reprints.xhtml

(12)

DOI: 10.1542/peds.2006-1573

2007;119;e12

Pediatrics

Anthony E. Fiore, Thomas J. Hoerger, Beth P. Bell and Gregory L. Armstrong

David B. Rein, Katherine A. Hicks, Kathleen E. Wirth, Kaafee Billah, Lyn Finelli,

United States

Cost-Effectiveness of Routine Childhood Vaccination for Hepatitis A in the

http://pediatrics.aappublications.org/content/119/1/e12

located on the World Wide Web at:

The online version of this article, along with updated information and services, is

by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397.

Figure

FIGURE 1
TABLE 1Summary of Input Parameters for the Model
TABLE 2Expected Hepatitis A Disease Outcomes and Costs with NoImmunization
FIGURE 2
+2

References

Related documents

The studies—which originated as World Bank regulation “social safeguards”—are meant to identify major concerns of informal settlement residents regarding their living conditions

Since pick list validation will be done on the DCF platform, rather than in the published XML schema (see 3.4.2- Validation of pick lists in DCF), pick list validation will remain

Bike sharing sustav ima za cilj potaknuti veći broj ljudi u korištenju električnih bicikala kao transporta, te učestalije korištenje postojećih korisnika kako bi

Since years 1980 the Transportation Engineering Department (PTR) has dedicated important efforts to the development of research on concrete pavements with special concerns

This CDI Practice Paper, written by Melanie Punton, Isabel Vogel and Rob Lloyd, reflects on the first year of a three-year realist impact evaluation, examining the Building Capacity

Training increases wage and salary earnings in the formal sector but not in the informal sector for both women and men, indicating that the wage and salary earnings gains

This compounds is also adopted for wine by the Codex Alimentarius in the General Standard for Food Additives

It gives us great pleasure to introduce this brochure detailing over 300 Adult Education classes, Programmes and Services provided by Kildare and Wicklow ETB within the three