Health economic evaluation of vaccine: the example of varicella … · 2019. 2. 26. · Impact on herpes zoster incidence-500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000

Health economic evaluation of vaccine: the example of varicella-zoster virus

Benoît DERVAUX,CNRS, Catholic University of Lille

Advances in infectious diseases modellingAnnecy, Les Pensières, December 10-12, 2007

Resource allocation in health care

� The Fourth hurdle to market:� Safety� Efficacy� Quality� Efficiency

� Economic evaluation = the «value for money» of new interventions

Types of economic evaluation

� How health consequences are valued ?

� Cost-effectiveness analysis: natural units� Cost-utility analysis: QALYs� Cost-benefit analysis: in monetary terms

� Ordered evaluations: CBA > CUA > CEA

Are vaccines different ? (1)

� Vaccination has direct and indirect effects. Such externalities may be positive or negative (herdimmunity)

� Epidemiological impacts of vaccination usually occur in the long term (sensitivity of the results to discounting rules)

IMPORTANCE OF MODELING

� Many vaccines prevent short-lived illness in veryyoung children (QoL measurement, valuation ofindirect costs)

� Some infections are eradicable

Are vaccines different ? (2)

� Key elements:� Herd immunity, externalities� Discounting, choice of end-points and time

windows: prevention vs cure

� Risk of underestimation of the value of vaccines

Discounting

� Should the health consequences be discounted?

� Do time preferences depend on decision time horizon? Should we use exponential or (quasi)-hyperbolic discounting models?

� What is the real objective of preventive programs? Should we consider risk reduction or long-term consequences on morbi/mortality as endpoints of vaccination programs?

� Are time preferences identical for all commodities? Should costs and health consequences be discounted at the same rate?

About varicella

� 90% of varicella cases occur in childhood, i.e. by the age of 15. Varicella is commonly regarded as a mild disease. Serious consequences are rare

� The incidence of severe complications increases with age

� Reactivation of dormant VZV results in herpes zoster (shingles)

� Immune response to VZV can be boosted via two mecanisms :� Sub-clinical reactivation of the virus (“endogeneous

boosting”)� Exposure to infectious individuals (“exogeneous boosting”)

Vaccination strategies (1)

� Mass childhood vaccination ± catch-up � May lead to an increase in adult cases

� Vaccination occurs in the youngest ages� As the risk of infection decreases, the average age at

which exposure to varicella occurs increases� Waning vaccine-induced immunity may lead to a pool of

susceptible older individuals

� Significant number of breakthrough cases may occur

� The lost of exogeneous boosting may lead to an increase on the incidence of zoster

RI 0.75(0.632-0.89)

-100

200

300

400

500

600

700

800

900

1,000

16-1

9

20-2

4

25-2

9

30-3

4

35-3

9

40-4

4

45-4

9

50-5

4

55-5

9

60-6

4

65+

Age group

Zos

ter

inci

denc

e ra

tepe

r 10

0,00

0 pe

rson

s-ye

ras

No children

Children < 16 years

Age specific incidence of zoster for adults with and without children in their household

Brisson et al. Vaccine 2002

See also case-control study by Thomas (2002)

Vaccination strategies (2)

� Targeted vaccinations: susceptible adoles-cents and adults, health care workers, immunocompromised individuals

� More recently, combined vaccinations� Varicella vaccination (in childhood) and zoster

vaccination (at older age)

What do we learn from US ?

� Routine childhood varicella vaccination program has been introduced in US since 1995 (single injection)

� The recommendations for vaccination were strengthened in 1999 (varicella vaccination required before entry of a child into day care or school)

� In June 2006, the vaccination schedule has changed (2 doses)

0

100

200

300

400

500

600

1 7 1 7 1 7 1 7 1 7 1 7 1 7 1 7 10

10

20

30

40

50

60

70

80

90

100

CasesVaccine coverage

1995 1997 19981996 1999 2000 2001 2002 2003

Jumaan- CDC 2005

Monthly Varicella Cases and Vaccine CoverageAntelope Valley, CA, 1995 – 2003

Reduction in varicella incidence

89785110-14

76

77

60

71

88

73

Antelope Valley

8686Total

547420+

788515-19

87805-9

87911-4

8787< 1

West Philadelphia

Travis

CountyAge

Seward JAMA 2002

Annual Hospitalization rates per 100,000 population Varicella Active Surveillance Project, 1995-2002

0

0,5

1

1,5

2

2,5

3

3,5

4

1995 1996 1997 1998 1999 2000 2001 2002

Hospitaliz

ation R

ate

/100,0

00

Jumaan- CDC 2005

Number of Varicella Deaths, 1990-2001

120

81

100 100

124115

8199

81

48 44

26

60

20

40

60

80

100

120

140

1990 1992 1994 1996 1998 2000 2002

Year

Num

ber

of d

eath

s

Average 107 deaths/year

Jumaan- CDC 2005

Gershon & Arvin, ACIP 2006

Adjusted annual rates of breakthrough varicella for children vaccinated between 12 mo-12 yrs:

Antelope Valley, 1995-2004

D.Guris ACIP 2006

Distribution of varicella cases by age:Antelope valley, CA, 1995 and 2004

Annual incidence of herpes zoster by age

CV 23%

CV 48%

CV 89%

Yih WK BMC Public Health 2005

Incidence : 2,77/ 1000 in 1999, 5,25/ 1000 in 2003 (+90%)

Epidemiological model (1)

Not vaccinatedProtected by

maternal antibodies

M(a,t) S(a,t) E(a,t) R(a,t)I(a,t)δδδδ σσσσ ννννλλλλ (a,t)

V(a,t) Sv(a,t) Ev(a,t) Rv(a,t)Iv(a,t)

t)

ττττ σσσσ ννννbλλλλ (a,t)

kλλλλ (a,t)

f(a,t) g(a,t)

Not vaccinated Not vaccinated Not vaccinated Not vaccinated

SusceptiblesLatent

infectedInfectious Immune

VaccinatedProtected by

the vaccine

Vaccinated Vaccinated Vaccinated VaccinatedPartly

SusceptiblesLatent


Not vaccinatedProtected by

maternal antibodies

M(a,t) S(a,t) E(a,t) R(a,t)I(a,t)δδδδ σσσσ ννννλλλλ (a,t)

V(a,t) Sv(a,t) Ev(a,t) Rv(a,t)Iv(a,t)

t)

ττττ σσσσ ννννbλλλλ (a,t)

kλλλλ (a,t)

f(a,t) g(a,t)

Not vaccinated Not vaccinated Not vaccinated Not vaccinated

SusceptiblesLatent


VaccinatedProtected by

the vaccine

Vaccinated Vaccinated Vaccinated VaccinatedPartly

SusceptiblesLatent


Realistic Age Structured Dynamic Model

Epidemiological model (2)

Susceptible Latent Infectious Immune

V Protected

V Suscept V Latent V Infectious V Immune

Sus BoostingZosterZos Immune

λ(a,t) σ α

Τ

W

bλ σ α

zλ(a,t) δ

α ρ(a)

kλ(a,t)

Published economic evaluations

Germany

Spain

France & Germany

Italy

Taiwan

Germany

Canada

Canada

Australia

Country

StaticNoNoNoHsu et al. (2003)

DynamicYesYesNoHammerschmidt et al (2007)

DynamicYesYesNoLenne et al. (2006)

DynamicYesYesNoCoudeville et al. (2005)

DynamicYesYesNoCoudeville et al. (2004)

DynamicYesYesNoBanz et al. (2003)

StaticNoYesNoGetsios et al (2002)

DynamicYesYesYesBrisson et al (2002)

StaticNoNoNoScuffham et al. (2000)

modelHerd immunity

WainingHerpes zosterSince 2000

Models predictions (1)

� Incidence of varicella would rapidly decline following the implemantation of vaccination

� After this “honeymoon period”, a rise in incidence would occur

� The post-vaccination equilibrium is always lower than the pre-vaccination level

� The age at infection is predicted to increase but the large reduction in incidence in children is likely to outweight any increase in incidence of chickenpox in adults

Impact on incidence:Base case scenario, VC = 90%

-

2,000

4,000

6,000

8,000

10,000

12,000

14,000

0 20 40 60 80

Year after start of vaccination

Var

icel

la c

ases

(per

1,0

00,0

00 p

erso

n-ye

ars)

Infant

Catch-up

Adolescent

Brisson et al. J Med Virol 2003

Impact on average age at infection:Infant vaccination, VC = 90%

-

2,000

4,000

6,000

8,000

10,000

12,000

14,000

-5 0 5 10 15 20 25 30 35 40 45 50 55 60 65 70

Year after start of vaccination

Cas

es p

er 1

,000

,000

per

son-

year

s

44+

19 to 44

5 to 18

0 to 4


Impact on hospitalizations:Base case scenario, VC = 90%

-

50

100

150

200

250

300

350

0 20 40 60 80Year after start of vaccination

Var

iclla

inpa

tient

day

s (p

er 1

,000

,000

per

son-

year

s)

Infant

Catch-up

Adolescent


Impact on herpes zoster incidence

-

500

1,000

1,500

2,000

2,500

3,000

3,500

4,000

4,500

5,000

0 20 40 60

Years after start of vacciantion

Her

pes

zost

er c

ases

(per

1,0

00,0

00 p

erso

ns-y

ears

)

Infant, immunity 20 years

95% IC, immunity 7-41 years

Adolescent, immunity 20 years



Impact of varicella vaccination on zoster incidence ?

No Yes

Varicella vaccination is less cost-effective from the health care payers perspective than from the societal perspective (time / leisure costs)

Routine vaccination of preteens is the preferred strategy under the health care payers perspective

Increases in zoster after routine vaccination of infants would render immunization highly cost-ineffective


5.094.444.905.24Society

0.160.590.730.600.61HC

11818846896185085086644503ICER($ / LYG)

Infant (zoster)

Infant (BV)

PreteenCatch-upInfantPersp



0.621.69CanadaGetsios et al. (2002)

0.342.06TaiwanHsu et al. (2003)

Italy

Germany

Canada

Australia

Country

1.203.50Coudeville at al. (2004)

1.754.12Banz et al. (2003)

0.615.24Brisson & Edmunds (2002)*

0.672.79Scuffham et al. (2000)

Health care payers

SocietalInfant vaccination(B/C ratio)

Without taking account zoster

Sensitivity analysis

� Epidemiological factors:� Vaccine efficacy� Vaccination coverage

� Economic factors� Vaccine price� Value of time/leisure lost (indirect costs)� Discount rate for health consequences and time

horizon

� Predictive value of anamnestic screening (for targeted vaccinations)

Indirect costs (1)� Should indirect costs be included in cost-

effectiveness analysis ?� If yes, how to measure productivity lost ?

� Human-capital method: � Reference to welfare economics (wage rate = opportunity cost

of time)� Easy to implement� Potential productivity costs

� Friction-costs method:� More realistic estimates of productivity costs� More data need to be collected

� The two methods lead to similar results when the period of absenteism is short (as in chickenpox case)

Indirect costs (2)

� Indirect costs = driver for varicella vaccine demand?

� Few WTP studies in the field� Revealed preferences versus real demand (eg

Australia)� Impacts on vaccination coverage (if CV is not high

enough => increase in average age at infection)

Comparison of studies

� Transferability versus generalisability

� Differences between studies:� Hypothesis on key parameters (vaccine efficacy /

waining / vaccination coverage)� Epidemiological factors� Organization of health care systems (impact on

costs)

FORCE OF INFECTION ACCORDING AGE

0

0.1

0.2

0.3

0.4

<1 year 1-2 years 3-4 years 5-9 years 10-14years

15-19years

>= 20years

For

ce o

f inf

ectio

n

GERMANY FRANCE

Treatment costs per varicella case according to age (€ 2002)

30.141.0

26.0 24.7

13

26.3

11.2 11.88

4.2

40.8

20.3533.8

0

0

1.9

0.2

€ 0

€ 20

€ 40

€ 60

€ 80

€ 100

€ 120

0-17 years >= 18 years 1-12 years > 12 years

Additionalexaminations

Hospitalisations

Pharmaceuticalprescriptions

Medicalconsultations

FRANCE GERMANY

Varicella treatment costs

100%100%70%63%% Supported by HS

59.5-81.149.3-65.760.0-174.143.4-52.5Range

€ 70.3€ 57.5€ 110.0€ 47.5

Varicella treatment costs

Indirect costs per varicella case according to age (€ 2002)

97,0

933,4

795,0

105,01,5

90,2€ 0

€ 100

€ 200

€ 300

€ 400

€ 500

€ 600

€ 700

€ 800

€ 900

€ 1000

0-17 years >= 18 years 1-12 years > 12 years

indi

rect

cos

ts

Societalperspective

Health systemperspective

FRANCE GERMANY

0%70%9,7%1,5%% Supported by HS

630-96084-126744-114283.8-110.7Range

€ 795,0€ 105,0€ 933,4€ 97,0Indirect costs

Coudeville et al. (2004)

Average annual costs related to varicella (millions of 2002 €) according to coverage rate

0

35,327

16,8

42,2

4,712,1

24,4

50,7

5,413,4

28,3

0

10

20

30

40

50

60

70

80

90

100

No va

ccina

tion

High c

overag

e rate

(90%)

Mediu

m cov

erage

(70%)

Low c

overag

e (45%

)

Sick leaves costs

Treatment costs

Vaccination costs

0

23,8

18

11

25

2,7

8,2

15,4

4,9 1,44,1

5,1

0

5

10

15

2 0

2 5

3 0

3 5

No vac

cinatio

n

High c

overag

e rate

(90%)

Medium

cover

age(70

%)

Low co

verage

(45%

)

Germany –Payer’s perspective:

France-Payer’s perspective:

Average annual costs related to varicella (millions of 2002 €) according to coverage rate

0

35,327

16,8

42,2

4,712,1

24,4

100,9

15,2

41,3

70,4

0

20

40

60

80

100

120

140

160

No va

ccina

tion

High c

overag

e rate

(90%)

Mediu

m cov

erage

(70%)

Low co

verag

e (45

%)

Sick leaves costs

Treatment costs

Vaccination costs

0

36,727,8

16,9

39,1

4,212,6

24,1

109,8

19,3

56,7

84,6

0

2 0

4 0

6 0

8 0

10 0

12 0

14 0

16 0

No vac

cinatio

n

High c

overag

e rate (

90%)

Medium

covera

ge (70

%)

Low co

verage

(45%)

Germany –Societal perspective:

France –Societal perspective:

Epidemiological model

Susceptible Latent Infectious Immune

V Protected

V Suscept V Latent V Infectious V Immune

Sus BoostingZosterZos Immune

λ(a,t) σ α

Τ

W

bλ σ α

zλ(a,t) δ

α ρ(a)

kλ(a,t)

VZ

Protected

VZ

Protected

δv

ezv

ezv

Varicella: 1 Year old

Zoster: 60 years old

Melegaro A, Jan Van Hoek A, Gay N, Edmunds J (2007)

0

200

400

600

800

1000

1200

0 20 40 60 80 100

Years after start of vaccination

Var

icel

la in

cide

nce

per

100,

000

Brisson mixing

Polymod mixing

-0.002

-0.0015

-0.001

-0.0005

0

0.0005

0.001

0.0015

0.002

0.0025

0 5 10 15 20 25 30 35 40 45 50

Years after start of vaccination

Zos

ter i

ncre

ase

in in

cide

nce

Brisson mixingPolymod mixing - infantPolymod mixing - infant + HZ

Melegaro A, Jan Van Hoek A, Gay N, Edmunds J (2007)

Combined strategies (2)Mixing patterns

Zoster vaccination

Preliminary results

Documents

Health economic evaluation of vaccine: the example of varicella … · 2019. 2. 26. · Impact on herpes zoster incidence-500 1,000 1,500 2,000 2,500 3,000 3,500 4,000 4,500 5,000