Assessment of the health impacts of H1N1 vaccination

From Opasnet
Revision as of 12:18, 26 August 2014 by Mori (talk | contribs)
(diff) ← Older revision | Latest revision (diff) | Newer revision → (diff)
Jump to: navigation, search


This assessment was completed in 2011, it is no longer actively updated.

Main message:
Question:

What was the overall health impact of the H1N1 (swine flu) vaccination in Finland in 2009-2010? Given current knowledge, which was the better decision between vaccinating as happened versus vaccinating no-one versus not vaccinating the population aged 5-19?

Answer:

Given current (2011) knowledge, the decision to vaccinate the whole population was the best decision even when narcolepsy is included in the assessment. Results of the Value of information analysis suggest that further knowledge about the uncertain variables considered very likely would not have changed the decision.


This assessment is about the total health effects the 2009 swine flu pandemic. It utilizes data from the Infectious disease registry (TTR) maintained by THL, the National narcolepsy task force report from 31.1.2011 and WHO. The current model is a simplification with no time dimension. The assessment has been evaluated. Evaluation results are displayed on the discussion page

Scope

  • What was the overall health impact of the H1N1 vaccination in Finland in 2009-2010?
  • Given current knowledge, which was the better decision between vaccinating as happened versus vaccinating no-one versus not vaccinating the population aged 5-19?
  • Monetary impact is not considered.

Participants

Result

Show results


Results

Distributions of the results for the different scenarios/decisions.
Comparison between scenarios and outcomes; onarc is DALYs from narcolepsy, osf is DALYs from having swine flu and osfm is DALYs from swine flu related mortality.
  • From initial results it would appear like swine flu is more significant than narcolepsy in terms of DALYs.
    • Vaccinating as planned would result in approximately 1850 DALYs due to swine flu and narcolepsy combined.
    • Vaccinating no-one would result in approximately 4400 DALYs due to swine flu.
    • Vaccinating everyone but people aged 5-19 would result in about a total of 2000 DALYs.
  • Probability of swine flu variable is revealed by both the sensitivity- and Value of information-analyses to have the most impact on the outcome.
    • The VOI analysis also reveals that further knowledge about the uncertain variables in the model is only worth up to ~80 DALYs, when considering the decision by age group, and less than 1 DALY when considering the decision as on/off as defined in the decision variable above. Which is only a small fraction of the total DALYs.
  • Suggested statement: Pandemrix should not be used any more anywhere because its narcolepsy risk is too high.
    • Resolution: Not accepted. Pandemrix is still an effective and safe vaccine. However, due to precautionary reasons, other alternatives should be used when available, because the occurrence of narcolepsy is not understood. R↻

Conclusions

Given current knowledge, the decision to vaccinate the whole population was the best decision. Results of the Value of information analysis suggest that further knowledge about the uncertain variables considered very likely would not have changed the decision. The total impact of the swine flu pandemic and related narcolepsy cases in Finland in terms of DALYs is slightly smaller than that of radon (~6700 DALYs yearly) and slightly larger than that of moisture damage (~650 DALYs yearly) for instance. It should be noted that only three variables had their uncertainty taken into account, although they should represent the major uncertainties present. Also, herd immunity is assumed not to affect the probability of a non immune subject to catch swine flu, this results in an underestimation of the number of swine flu cases in scenarios where the vaccination coverage is less than what was observed.

Rationale

Causal diagram.
Decisions
  • Vaccination decision
    • Vaccinate everyone (observed vaccination coverage)
    • Vaccinate no-one (0 vaccination coverage)
Variables
  • H1N1 vaccination coverage in Finland
  • ERF of H1N1 vaccination on Narcolepsy
    • Assumed lognormally distributed
  • A(H1N1)v immunity in the Finnish population
    • P(immune) = 1 - P(not vaccinated) * P(no base immunity)
  • Population of Finland
  • Disability weights
    • DALY weight of narcolepsy equals roughly that of epilepsy (0.065)
    • DALY weight of swine flu for the duration of the disease ~ 0.5 (~ 0.007 DALYs per case when adjusted for duration)
  • Life expectancy by age groups in Finland[1]
  • Probability of catching swine flu given subject is not immune
    • Estimated from data available (population, total immunity, number of cases) by fitting the number of cases to a poisson distribution and calculating probability from the mean estimate by dividing by the non-immune population
  • Fraction of all cases represented by lab confirmed cases (which we have data on)
    • Estimated as beta-distributed with mean of 0.2 and some sd
  • Probability of death due to swine flu given a subject has swine flu and belongs to a risk group
    • Estimated from data available
    • Assumed all deaths will be lab confirmed cases
    • Assumed that all deaths belonged to a risk group (had some base condition)
  • Fraction of population belonging to a risk group
    • Arbitrary values; trying to account for kids of age <1 and old folks with heart conditions etc.
  • Duration of swine flu
    • Assumed to be flat 5 days (mildly incapacitated for this duration)
  • Narcolepsy in Finland
  • AH1N1 cases in Finland
Indicators
  • DALYs from narcolepsy caused by vaccination
  • DALYs from having swine flu
  • DALYs from deaths caused by swine flu

R code

  • Basic model
    • Uncertainties of ERF of vaccine on narcolepsy, fraction of all cases represented by lab confirmed cases and probability of catching swine flu are implemented.

Disability weight for narcolepsy:

How many iterations do you want to run? (Full model is 1000 but it is slow):

+ Show code

See also

Scientific publications in Pubmed: [2] [3] [4] [5] [6] [7] [8] [9]

Keywords

Swine flu, influenza, vaccination, narcolepsy, risk perception, risk comparison

References

Related files

<mfanonymousfilelist></mfanonymousfilelist>

VipuvoimaaEU rgb1.jpg Eakr logo fi.png