COSTEFFECTIVENESS AND DURABILITY OF THE SHINGLES VACCINE

1
COST-EFFECTIVENESS AND DURABILITY OF THE SHINGLES
VACCINE

• Marc Brisson, PhD
• Canadian Research Chair in Mathematical Modeling
  and Health Economics of Infectious Disease,
  Université Laval

14th Annual Meeting of the IHMF Dubrovnik,
Croatia 10-11th October 2007
2
Background

• The Shingles Prevention Study (SPS) has shown
  that a live-attenuated varicella-zoster virus
  vaccine is effective against
• herpes zoster (HZ) ? VE 51
• post-herpetic neuralgia (PHN) ? VE 67
• burden of illness due to HZ ? VE 61
• Policymakers will be asked to make
  recommendations and decisions regarding the
  introduction of the HZ vaccine
• Criteria considered in such decisions include
  effectiveness and cost-effectiveness

3
Objective

• Estimate the duration of protection of the
  shingles vaccine
• Examine the potential cost-effectiveness of
  vaccination against HZ and PHN
• i.e. Gain insight into the potential benefits and
  costs of the shingles vaccine for different ages
  at vaccination

4
What is Cost-Effectiveness Analyses?

• Cost-Effectiveness Ratio
• (Cost of Vaccination Cost offsets by preventing
  HZ)
• (Gains in Health through HZ vaccination)

5
What is Cost-Effectiveness?

• Cost-Effectiveness Ratio
• (Costs)
• (Benefits)
• Costs
• e.g. Price of Vaccine, Physician visits,
  Hospitalization, Drugs, Treatment
• Benefits
• e.g. Cases prevented, Life-years gained,
  Quality-adjusted life-years (QALYs)

6
What is a QALY?

• Most common measure of benefit in
  cost-effectiveness analysis
• Developed to capture, in a single measure, both
  gains from reduced morbidity and reduced
  mortality
• QALY-weight ranges from 0 (statedeath) to 1
  (perfect health)
• The QALYs-gained is the difference, over time,
  between the QALY-weights with and without
  vaccination

7
Example QALY lost to PHN
8
QALY-gains through vaccination
9
Economic Analysis

• Compare Vaccination vs. No vaccination
• Perspective Health care provider
• Outcome Measure Cost per QALY gained
• Discounting 5 for costs and benefits
• Time Horizon Lifetime of vaccinee

10
Cohort Model

• Decision analytic cohort model
• Age-specific
• 3 HZ pain states no pain, mild/moderate severe
• Predicts HZ PHN incidence, healthcare resource
  use, QALYs and costs in a cohort of vaccinated
  and non-vaccinated individuals

11
Parameters References

• Epidemiological
• HZ Incidence Brisson, Epidemiol Infect 2001
• Hospitalization Canadian Institute for
  Health Information (CIHI)
• Length of Stay CIHI
• PHN incidence SPS Edmunds, Vaccine 2001
• Duration/severity Helgason, BMJ 2000
• Case-Fatality Stats Canada Edmunds, Vaccine
  2003
• QALY estimates Master SPS Bala, J Clin
  Epidemiol 1998 Oster, J of Pain 2005
• Unit Costs (2005 CAN)
• HZ GP consultation CIHI
• HZ Hospitalization Nowgesic, Can Commun Dis
  Rep 1999
• Prescription Costs IMS Aoki, Can J Clin
  Pharm 2003 Master
• PHN Costs White, ISID International Congress
  2006

12
HZ Vaccine Efficacy

• Vaccine efficacy is comprised of two
  components/parameters
• Take of individuals protected following
  immunization and the degree to which they are
  protected
• Waning rate rate of loss of protection
• To estimate vaccine efficacy parameter values
• fit the age-specific annual incidence of HZ
  predicted by the model with that observed in the
  vaccinated arm of the Shingles Prevention Study,
  using maximum likelihood

13
HZ Vaccine EfficacyFit to Vaccine Efficacy
Observed in the SPS
14
Vaccine Parameters

• Initial HZ Vaccine Efficacy AS (75 to 26)
• 50 yrs old ? 75
• 65 yrs old ? 63
• 80 yrs old ? 26
• Initial PHN Vaccine Efficacy 67 (95CI
  48,80)
• Vaccine duration Life (95CI 12yrs, Life)

HZ Vaccine Efficacy (Take) is Age-Specific and
decreases with age.
15
Sensitivity Analysis

• Probabilistic sensitivity analysis
• All parameters were assigned a probability
  distribution
• 20,000 possible combinations of these parameter
  values were drawn
• 20,000 simulations were conducted with 20,000
  different cost per QALY gained estimates
• Results are presented with the Base Case, 50 and
  90 Credibility Intervals (CrI)
• show the 5, 25, 75 and 95th percentile taken from
  the distribution of results from 20,000
  simulations

16
Number Needed to Vaccinate (NNV) (Base
Age65yrs, VEHZ63, VEPHN67, VDlife)
17
NNV Comparisons

• NNV to prevent 1 case of herpes zoster per year
• Skootsky NNV 175 (based on SPS data)
• Brisson NNV 200
• NNV to prevent 1 case of herpes zoster during
  lifetime
• Brisson NNV 13-64 depending on age
• CDC NNV 17
• NNV to prevent 1 case of PHN during lifetime
• Brisson NNV 58-78 depending on age
• CDC NNV 31

Skootsky. Proceedings of UCLA Healthcare 2006
10 1-3 Advisory Committee on Immunization
Practices. Considerations for shingles vaccine
recommendations in the U.S. record of the
proceedings. Atlanta Centers for Disease Control
and Prevention
18
Cost-Effectiveness of the Shingles Vaccine

• No widely accepted threshold for
  cost-effectiveness
• A commonly-cited rule of thumb is that
  interventions are very cost-effective if they
  have cost-effectiveness ratios less than the per
  capita GDP
• We use 40,000 per QALY gained as our threshold
  (Canadian per capita GDP)

19
Sensitivity to Vaccine Characteristics (Base
Age65yrs, VEHZ63, VEPHN67, VDlife,
Cost150)

Base 33,000 per QALY gained
20
Sensitivity to Vaccine Characteristics (Base
Age65yrs, VEHZ63, VEPHN67, VDlife,
Cost150)

Base 33,000 per QALY gained
21
Sensitivity to Vaccine Characteristics (Base
Age65yrs, VEHZ63, VEPHN67, VDlife,
Cost150)

Base 33,000 per QALY gained
22
Sensitivity to Epidemiological Parameters
(Base Age65yrs, VEHZ63, VEPHN67, VDlife,
Cost150)

Base 33,000 per QALY gained
23
Sensitivity to Epidemiological Parameters
(Base Age65yrs, VEHZ63, VEPHN67, VDlife,
Cost150)

Base 33,000 per QALY gained
24
Sensitivity to Economic Parameters (Base
Age65yrs, VEHZ63, VEPHN67, VDlife,
Cost150)

Base 33,000 per QALY gained
25
Sensitivity to Economic Parameters (Base
Age65yrs, VEHZ63, VEPHN67, VDlife,
Cost150)

Base 33,000 per QALY gained
26
Probabilistic Sensitivity Analysis(Cost per
course150)
The figure can be loosely interpreted as showing
the probability that HZ vaccination is
cost-effective for alternative threshold values
of cost per QALY-gained.
27
Probabilistic Sensitivity Analysis(Cost per
course100)
The figure can be loosely interpreted as showing
the probability that HZ vaccination is
cost-effective for alternative threshold values
of cost per QALY-gained.
28
Cost-Effectiveness Comparisons
. 90CrI5th and 95th percentile of probability
sensitivity analysis varying all parameters
including VE REF Edmunds Vaccine 2001
Hornberger Ann Intern Med 2006 Rothberg Clin
Infect Dis 2007
29
Hornberger Robertus

• Higher cost-effectiveness (CE) ratios
• Why?
• Very low QALY-weight lost used for HZ (0.05 vs.
  0.09-0.18 (SPS) or 0.07-0.28 (Brisson Bala
  Oster Coplan)).
• Lower proportion of HZ resulting in PHN (5 vs.
  12-34).

REF Brisson Can J Infect Dis Med Microbiol 2006
Bala J Clin Epidemiol 1998 Oster Journal of Pain
2005 Coplan J Pain 2004
30
Rothberg, Virapongse Smith

• Higher cost-effectiveness (CE) ratios
• especially for the youngest (age 60) and oldest
  (age 80) recipients examined
• Why?
• Shorter vaccine duration (10y vs. No waning) ?
  impacts youngest age group
• Erroneously assume no reduction in PHN beyond
  that afforded by reducing herpes zoster ? impacts
  older age groups where PHN incidence is high

REF Dworkin Am Geriatr Soc 2007
31
PHN Vaccine Efficacy beyond Shingles
efficacy(SPS data)
a. Analyses by Rothberg, unadjusted for age b.
Rothberg reduced the number of PHN cases in the
Placebo arm in order for the of PHN per case of
herpes zoster in the placebo group in year 1 to
be equal to the analogous observed in years 2
to 4
32
PHN Vaccine Efficacy beyond Shingles
efficacy(SPS data)
a. Analyses by Rothberg, unadjusted for age b.
We reduced the number of PHN cases in the Placebo
arm in order for the of PHN per case of herpes
zoster in the placebo group in year 1 to be equal
to the analogous observed in years 2 to 4 (to
be consistent with Rothberg and colleagues).
33
Discussion

• Limitations
• Did not include the potential impact of varicella
  vaccination on HZ incidence
• Did not include HZ complications other than PHN
• What happens if waning occurs? Should we give a
  booster?
• If waning vaccine efficacy occurs, adding a
  booster is unlikely to be cost-effective unless
  the cost of the vaccine is significantly less
  than 150.
• Assuming waning protection is 8 per year and a
  booster is given 10 years after the initial dose,
  the cost-effectiveness of vaccinating 65 years
  olds against HZ is estimated to be 127,000,
  79,000 and 32,000 per QALY-gained for a vaccine
  that would cost 150, 100 and 50 per dose,
  respectively.

34
Summary

• The average durability of the shingles vaccine is
  predicted to be greater than 12 years
• Vaccination of the elderly is likely to be
  cost-effective if given to adults between 65 and
  75 years of age
• Cost-effectiveness is most sensitive to
• Age at vaccination
• duration of vaccine protection
• cost of vaccination
• QALY-lost to PHN

35
Number Needed to Vaccinate (NNV)

• NNV Number of people that are needed to be
  vaccinated to prevent an HZ-related event during
  their lifetime lifetime
• NNV N P
• N Number of people vaccinated
• P Predicted Lifetime Prevented Events (i.e. HZ,
  PHN, etc)
• Illustrates the benefit of herpes zoster
  vaccination by combining
• Vaccine efficacy
• Background incidence of disease

36
Sensitivity to Vaccine Durability (Base
Age65yrs, VEHZ63, VEPHN67, VDlife,
Cost150)