Skip to Main Content

The Science of Vaccine Acceptance

January 22, 2021
  • 00:00So I'll be talking about
  • 00:02the science of acceptance,
  • 00:04and just to Orient everyone that
  • 00:06we have a set of interventions.
  • 00:09But one way of thinking about vaccine
  • 00:12acceptance is that it's a Russian dog model,
  • 00:15and so some of us have been proposing this.
  • 00:18And then it's used by several entities
  • 00:21to think about this is that you have
  • 00:25thoughts and feelings that overlay
  • 00:27our attitudes and cognitive biases.
  • 00:30That sit on top of our trust social
  • 00:32Norm believes experiences and
  • 00:33underlying all of that is moral
  • 00:35values and ideology and worldviews.
  • 00:41And so therefore vaccine
  • 00:42hesitancy is complex and nuanced.
  • 00:44So before we go to it,
  • 00:46we just you know,
  • 00:48why do we worry about vaccine hesitancy?
  • 00:50Generally speaking,
  • 00:51so these are data from measles.
  • 00:55That looked at the post elimination
  • 00:58scenario in the US of what was happening
  • 01:01in terms of measles Epidemiology.
  • 01:04After you know the endemic transmission
  • 01:07was illuminated from this country
  • 01:09and we showed that a significant
  • 01:12proportion as much as 69% or
  • 01:15approximately 70% of the cases of
  • 01:17measles in this country in the post
  • 01:21elimination area era were coming from
  • 01:23those who were non medically exempt.
  • 01:27And this of those who are
  • 01:29eligible to get vaccinated,
  • 01:30even if you account for those who
  • 01:33were too young to be vaccinated
  • 01:35when they got the infection,
  • 01:37we were getting 43% of all those cases and
  • 01:40so forth so far as measles is concerned.
  • 01:43That is true for a lot of a few
  • 01:46other vaccine preventable diseases.
  • 01:48That vaccine hesitancy is the single
  • 01:50most preventable cause of of cases of a
  • 01:53lot of these vaccine preventable diseases.
  • 01:56Then we did.
  • 01:57Another analysis where we looked at.
  • 02:00We created this cumulative
  • 02:02epidemiological or epidemic curve
  • 02:04and we found that vaccine refusers
  • 02:07were disproportionately represented
  • 02:08in the early parts of epidemics.
  • 02:11In the post elimination era for Mesos,
  • 02:14and they were providing that
  • 02:16critical mass that tender to start
  • 02:19the fire in a lot of these places,
  • 02:22and therefore these clusters of vaccine
  • 02:26refusers were driving the overall.
  • 02:29Allergy So what are some of
  • 02:31the ways of doing it?
  • 02:33We have been working on an one of those
  • 02:36sort of things that some of us feel
  • 02:39strongly about that vaccine acceptance
  • 02:41science should be as rigorous as
  • 02:45vaccine development and evaluation science.
  • 02:47So as someone who has worked on trials
  • 02:50has have LED trials vaccine trials,
  • 02:52we try to bring the same level of rigor
  • 02:55to vaccine communication approaches,
  • 02:58and so we evaluate that.
  • 03:00From that perspective,
  • 03:01not every assessment is based
  • 03:03on randomized control design,
  • 03:04but a lot of these are,
  • 03:06so one way of looking at it is to.
  • 03:10See to talk to people's values,
  • 03:13and there's this theory that postulates
  • 03:15the so called moral foundations theory
  • 03:18that postulates that we have these moral
  • 03:22test just like our regular taste buds.
  • 03:25You know where the combination
  • 03:27of you know stimuli give us give
  • 03:30each food as distinct flavor.
  • 03:32We have these moral taste
  • 03:35bugs that are universal.
  • 03:36These are six moral taste buds.
  • 03:39There's our care, harm,
  • 03:41fairness, cheating.
  • 03:42For example, the Golden rule in terms of
  • 03:46fairness is universal in all cultures,
  • 03:50loyalty and betrayal.
  • 03:51Authority,
  • 03:52subversion, meaning,
  • 03:53sort of the positive valences authority.
  • 03:56Subversion is in negative valence purity
  • 03:59and degradation and so purity can be
  • 04:03very secular or religious as well,
  • 04:05and then Liberty oppression is another one,
  • 04:09and that's why we emphasize.
  • 04:13You know human rights almost universally,
  • 04:16and individual Liberty and so and so forth.
  • 04:19These values are emphasized
  • 04:21differentially in different populations,
  • 04:23and each stimulus triggers
  • 04:24them differentially. So it is.
  • 04:26Populations vary in terms
  • 04:28of their moral matrices,
  • 04:30and so we wanted to see our vaccines,
  • 04:33decisions, moral decisions.
  • 04:34So in this nature, human behavior paper.
  • 04:37We actually were the showed that
  • 04:40actually vaccine decisions can be moral.
  • 04:43Decisions are often.
  • 04:44Our moral decisions, and so this.
  • 04:47These data show the hesitancy high
  • 04:50versus low hesitancy above the above.
  • 04:53The the one Odds ratio line is
  • 04:56high hesitancy below that is
  • 04:59lower hesitancy and we found that
  • 05:02values of purity and Liberty.
  • 05:05Were associated with high level of hesitancy,
  • 05:09so people who generally emphasized
  • 05:13purity and generally emphasized
  • 05:15Liberty were more hesitant and
  • 05:18similarly deference to authority was.
  • 05:21More associated with vaccine acceptance.
  • 05:23So these things were helpful in
  • 05:26developing further interventions,
  • 05:27and I'll give you an example
  • 05:29of one of those interventions,
  • 05:31and the idea is not to change
  • 05:34people's values.
  • 05:34It's not.
  • 05:35I would postulate that it is not our
  • 05:38place to change people's values.
  • 05:40The key is to speak to people's values.
  • 05:43So we did this randomized control trial,
  • 05:46for example,
  • 05:47where we said that young adults
  • 05:49who missed their vaccines.
  • 05:51In teens and now you know that HPV vaccine.
  • 05:56IS has sub optimal uptake in this
  • 05:59country and actually many countries
  • 06:01and a lot of teens missed their
  • 06:03doses and then they get to be young
  • 06:06adults and they get to be of college
  • 06:08age and people have tried in that
  • 06:11group to talk about that a cancer,
  • 06:13salient message and so talking about
  • 06:15cancer and cervical cancer and
  • 06:17treat and messaging through that.
  • 06:19I think overall it is a very good.
  • 06:22It's a very good and reasonable strategy,
  • 06:24but in a lot of places when you
  • 06:27have tried that.
  • 06:28There's a huge chunk of people,
  • 06:31especially young adults,
  • 06:32who do not respond to that message,
  • 06:35so we did a trial where we showed
  • 06:38people images of genital warts so
  • 06:40it's a less severe outcomes and
  • 06:43had a veneer to go with it that
  • 06:46focused on violation of purity and
  • 06:48what we found that compared to the
  • 06:51control group we had 20 percentage
  • 06:53points absolute percentage points.
  • 06:55Higher impact of this message
  • 06:57in terms of peoples.
  • 06:59A willingness to accept the HPV vaccine.
  • 07:03And so then one policy tool
  • 07:05is mandates or requirements,
  • 07:07but the way we think about it is
  • 07:10whether they can be used as behavioral
  • 07:13interventions just to remind everyone
  • 07:15for childhood vaccination.
  • 07:17School immunization requirements
  • 07:18are based on state laws.
  • 07:21They have played a major role
  • 07:23in keeping rates of vaccine
  • 07:25preventable diseases low,
  • 07:27but they allow for three types of exemptions,
  • 07:30means medical,
  • 07:31religious and personal beliefs
  • 07:33or philosophical exemptions.
  • 07:34And there's a lot of heterogeneity
  • 07:37across different states,
  • 07:38etc and these.
  • 07:41Requirements are some places allow
  • 07:43unknown non medical exemptions,
  • 07:45others allow only reduce religious
  • 07:47exemptions and and a few others allow
  • 07:49religious and philosophical exemptions.
  • 07:51This is a slightly dated map but
  • 07:54I'm showing this because you know
  • 07:56some of these contracts were used in
  • 07:59our analysis and this one where we
  • 08:02showed that not only the different
  • 08:04kinds of exemptions we are,
  • 08:06you know 8 nine years ago showed that
  • 08:09there was a lot of heterogeneity.
  • 08:12In the kinds of exemptions that
  • 08:15each state was implementing,
  • 08:16and so these data were initially published
  • 08:19in the New England Journal of Medicine.
  • 08:22But it has been updated since.
  • 08:25So based on the new information,
  • 08:27but essentially there was substantial
  • 08:29activity based on what are the
  • 08:32bureaucratic requirements to obtain
  • 08:34the vaccine exemption in each of these
  • 08:37States and what we found was that
  • 08:39the rate of pertussis in this case,
  • 08:42or whooping cough.
  • 08:44Was associated even after adjusting
  • 08:46really aggressively accounting
  • 08:48for state level differences was
  • 08:50associated with how easy it was to
  • 08:53get an exemption an so if you have a
  • 08:56pathway where your default option if
  • 08:58you are moving people away from non
  • 09:01vaccination to vaccination it not only
  • 09:04has an impact on your immunization
  • 09:06rates but also on disease rates.
  • 09:10These rates are changing and we
  • 09:12found that in a subsequent analysis
  • 09:15that you have states that allow
  • 09:19philosophical exemptions had the
  • 09:21highest rates of non medical exemptions,
  • 09:24but states that had religious only
  • 09:27exemptions were catching up and the
  • 09:30other thing we found actually this
  • 09:33was earlier on in a 2000 union Journal
  • 09:36paper where we describe that in various
  • 09:39states there's substantial geographic.
  • 09:41Heterogeneity in in exemption rates,
  • 09:44by County,
  • 09:45by subregion,
  • 09:46and so it's important to keep.
  • 09:49Keep in mind that state level rates
  • 09:53themselves are not sufficient to account
  • 09:56for disease risk at the lower local level.
  • 10:00Because it's these clusters of refuzors
  • 10:03that are driving a lot of these outbreaks.
  • 10:06So we had a natural experiment in
  • 10:08California where three in policy
  • 10:10interventions were implemented.
  • 10:12The first one was an implementation
  • 10:14of a requirement to say that if you're
  • 10:17going to get a non medical exemption,
  • 10:19philosophical or religious exemption,
  • 10:21you have to have a requirement.
  • 10:23You have to have a physical health
  • 10:26care provider consultation and then
  • 10:28you have to go through a process
  • 10:30someone explaining to you what are
  • 10:32the consequences of your action.
  • 10:34Then the other one was this conditional.
  • 10:37Mentoring program where they crack
  • 10:39down on this provision better.
  • 10:40Even if you hadn't provided proof
  • 10:43of exemption or vaccination,
  • 10:44as schools would allow you temporarily
  • 10:47and or conditionally and then there
  • 10:49was this famous blog called SB 277 that
  • 10:52illuminated all non medical exemptions.
  • 10:54What we essentially found and without
  • 10:57going into too much detail that
  • 10:59there was a replacement effect.
  • 11:01So non medical exemptions did go
  • 11:03down but there was within 2 years
  • 11:06a near total replacement.
  • 11:08Of those who work,
  • 11:10seeking not just medical exemptions,
  • 11:12but those who are just kind of ignoring
  • 11:14the law and those who were home
  • 11:17schooling or finding other pathways
  • 11:20of stepping away from the law itself.
  • 11:22And so that was concerning.
  • 11:24And we also found that in California,
  • 11:27subsequent analysis that came
  • 11:29out in 2019 that this law
  • 11:31did have an impact on the clusters of,
  • 11:34even though there was a
  • 11:37replacement with the third.
  • 11:39Legal intervention of eliminating
  • 11:40all non medical exemptions.
  • 11:42It did reduce the clustering of,
  • 11:45or illuminated some of the
  • 11:47clusters of vaccine refusers.
  • 11:49So we now propose based on this
  • 11:52Goldilocks approach not too hard not
  • 11:55to called use of mandates and described
  • 11:58in Nature best practices around it.
  • 12:01Essentially set out the criteria based
  • 12:03on which states should consider or
  • 12:06countries should consider using mandates.
  • 12:08And when they use it,
  • 12:11it should be paired with soft penalties
  • 12:13rather than hard draconian measures.
  • 12:16And in the context of COVID-19
  • 12:18we laid out in back in July and
  • 12:21Union Journal the criteria for.
  • 12:26What do you think about like when entities
  • 12:29are imposing and considering mandates,
  • 12:31they should meet six criteria that
  • 12:34ensure that there is access there is.
  • 12:37You know some behavioral
  • 12:40principles that are followed.
  • 12:42So there are several key
  • 12:45challenges in a COVID-19 vaccine.
  • 12:48Acceptance and we did with you
  • 12:50with colleagues that have college.
  • 12:52We did an experiment where we found
  • 12:54that if the approval had happened
  • 12:57and this was done in August,
  • 12:59September before the election,
  • 13:00one week before the election,
  • 13:02the uptake and confidence in vaccines
  • 13:04would have been much lower and
  • 13:06compared to a December approval.
  • 13:08So this would help us going forward because
  • 13:11it ended up being a December authorization.
  • 13:14We then found that elite so-called
  • 13:16elite endorsers like Doctor Fouchy.
  • 13:18Will, if they are negatively
  • 13:20endorser vaccine,
  • 13:21then we can forget about it.
  • 13:23But if they positively endorsed the vaccines,
  • 13:25then it will move people both
  • 13:27Democrats as well as Republicans,
  • 13:29although Democrats would be moved
  • 13:31a little bit more in terms of
  • 13:33increasing their uptake and confidence.
  • 13:35And we also highlighted the
  • 13:37value of bipartisan endorsement.
  • 13:39So these recommendations.
  • 13:40The preprint was available earlier
  • 13:42on the paper came out more recently,
  • 13:44but this was we shared these finding on
  • 13:47an ongoing basis with various policy makers.
  • 13:50We also identified the using modeling
  • 13:53and using the data we collected
  • 13:57in May we we also identified the.
  • 14:01The characteristics of high vaccine
  • 14:04refusal refusal so that we can target
  • 14:07a lot of these interventions and one
  • 14:10of the other things that we're doing
  • 14:13with the colleagues at Hill College
  • 14:16and Baconi in Italy is looking at
  • 14:19behavioral policy level behavioral
  • 14:21decisions meaning when to give
  • 14:23vaccine to others or went to support
  • 14:26external programs and we're finding
  • 14:29that it is in our indicted self.
  • 14:32Interest and not just to support building
  • 14:34up of vaccine coverage in this country,
  • 14:37but to also support other countries.
  • 14:40And sometimes there will be.
  • 14:43Total level of coverage.
  • 14:45The marginal value of supporting
  • 14:47other programs would be so much
  • 14:49that that would become a pretty
  • 14:52high priority and there are few
  • 14:54sources that we have generated.
  • 14:56Resources we have generated for messaging,
  • 14:58so this one is a messaging guide
  • 15:01with UNICEF that we developed and
  • 15:03this is with WHO that we developed
  • 15:06an overall acceptance and uptake
  • 15:08of COVID-19 vaccine paradigm that
  • 15:11focuses on the whole 360 degree
  • 15:13behavioral insights based.
  • 15:15Framework we are also creating this
  • 15:17vaccine demand Observatory with
  • 15:19a few partners including UNICEF,
  • 15:21First Draft and an entity called PGP,
  • 15:23so I'll pause here and I think
  • 15:26my 15 minutes are up.