PUBLIC HEALTH · ADOPTION

« How do you restore ROR vaccination coverage in lagging areas without triggering national polarization? »

For five years, a national health authority has observed a continuous decline in ROR vaccination coverage. The national average at 24 months fell from 91 % to 84 % ; in around twelve more vulnerable territories, it dropped below 76 %. Two national campaigns were launched over three years, widely recognized, without any measurable movement in intentions.

The question is not how to convince opponents. It is which intervention architecture produces which reaction in which population — and at what point an intervention intended to restore adoption turns diffuse hesitation into lasting conflict.

In an ordinary waiting room, a parent reads a health record booklet with a young child sitting on their lap.
DECISION
Structure a plan to restore ROR vaccination coverage across twelve territories where coverage is declining
POPULATION
2,4 million parents of children aged 12 to 24 months, 18 hesitation typologies
WHAT IS TESTED
3 structuring options, 34 communication formats, 8 possible messengers
HORIZON
24 months of coverage evolution, territory by territory

THE PROBLEM

A public intervention
does not produce the same effect
across the whole population.

The same campaign can reassure, inform, make an appointment easier and reactivate existing trust. In some groups, it can also reinforce a sense of pressure, shift the debate from health to institutions, and turn hesitation into a firmly held position.

The usual levers had been deployed without significant effect. Impact studies of the two previous campaigns showed high message recognition and almost no movement in intentions. Kits sent to treating physicians were rarely used in consultations. Territorial catch-up programs reached only a limited share of uncovered children.

The decision-relevant question was therefore not which message was most convincing, but which intervention produced which reaction in which population — including unintended reactions.

WHAT IS THE SAME FOR EVERYONE

  • a national health authority and twelve territories with declining coverage
  • national coverage at 24 months down from 91 % to 84 % on average
  • coverage below 76 % in the most vulnerable territories
  • three limited but visible measles outbreaks in 18 months
  • a €42 million budget over three years
  • a deliverable expected within five weeks, before presentation to the minister

WHAT DIFFERS FOR EACH GROUP

  • personal healthcare experience over the past five years
  • level of trust in health institutions
  • scientific cultural capital
  • family and social environment
  • media exposure
  • local political context

HESITATION IS NOT OPPOSITION

A territory with declining coverage
is not
a territory in opposition.

In the source case, only 8 % of hesitant parents are convinced opponents. Grouping all situations of declining coverage into a single category makes the decision impossible to assess : these are different positions that do not react to the same interventions.

  1. 01DELAYvaccination was not refused: it was postponed, then forgotten in the household calendar
  2. 02QUESTIONa specific question remains without a clear answer, often raised with a healthcare professional who did not have time to return to it
  3. 03HESITATIONintermediate trust: neither immediate support nor refusal, with the decision kept open
  4. 04DISTRUSTa poor healthcare experience resurfaces at the moment of the decision, without any direct connection to vaccination
  5. 05STRUCTURED OPPOSITIONan already formed, reasoned position connected to organized networks: around 8 % of hesitant parents in the source case
PRACTICAL DELAYESTABLISHED POSITIONDELAYQUESTIONHESITATIONDISTRUSTSTRUCTURED OPPOSITION

These positions lie between practical delay and an already established position. This is not a pathway : no one moves through these stages automatically.

What the simulation makes visible is that the same intervention does not encounter the same thing depending on where a person sits along this spectrum.

ADOPTION AND POLARIZATION

A campaign can increase
average adoption
and harden a minority.

WHAT A CAMPAIGN MEASURES

Message recognition. The two previous national campaigns achieved high levels.

WHAT IT DOES NOT MEASURE

Actual movement in intentions. It was almost nil, despite substantial budgets and careful execution.

WHAT REALLY DIFFERENTIATES

Lived healthcare experience, relationship with the healthcare professional, political context of the territory.

WHERE CONFLICT FORMS

Where a firm national message encounters structural distrust of central authorities.

WHAT A NATIONAL AVERAGE TELLS YOU

A level of coverage, and how it changes. It indicates an aggregate result, not how that result is distributed.

WHAT A TERRITORIAL DISTRIBUTION TELLS YOU

The same program can be positive in eight territories and negative in four. Average progress can coexist with localized deterioration.

Adoption and polarization are two distinct dimensions. In this case, the firm national campaign increases coverage in eight territories and lowers it in four : a positive average can mask a negative political dynamic.

WHAT IS TESTED

The same objective.
Several architectures
of intervention.

Three structuring options were on the steering committee's table : firm national campaign, targeted territorial plan, equipping treating physicians. Each was tested on the actual populations of the territories with declining coverage, then developed into variants by kit, messenger and media treatment.

  1. 01FIRM NATIONAL CAMPAIGNNational messaging centered on reminding people of the vaccination requirement, uniform across the whole territory.
  2. 02TARGETED TERRITORIAL PLANDifferentiated programs deployed across the twelve departments with declining coverage, with local adaptation.
  3. 03EQUIPPING TREATING PHYSICIANSIndirect approach relying on the treating physician as the main messenger, with dialogue kits and catch-up appointments.
  4. 04GENERIC OR DIFFERENTIATED DIALOGUE KITFive variants tested: generic, safety-oriented, empathy-oriented, healthcare-experience-oriented, intergenerational-transmission-oriented.
  5. 05NATIONAL MEDIA COVERAGE OF CATCH-UP SUCCESSNational media storytelling around successes achieved in the fastest-moving territories.
  6. 06ACTIVE MEDIA DISCRETIONLocal and professional communication, without national storytelling around territorial success.

4 800 synthetic parents distributed across the 18 typologies were exposed to the 34 communication formats tested, in randomized order, with an average of eight follow-up questions per person. Coverage trajectories were then projected over 24 months, territory by territory, producing 96 trajectories for the eight finalist scenarios retained by the committee.

Heterogeneity
is not
a matter
of being for
or against.

SIMULATED POPULATION

The reconstructed population covers 2,4 million parents of children aged 12 to 24 months, distributed proportionally across the twelve territories with declining coverage, calibrated on public data from INSEE, Santé publique France and annual vaccination-coverage surveys. No personally identifiable data entered the system.

It is structured into 18 hesitation typologies built by combining six dimensions : personal healthcare experience, institutional trust, scientific cultural capital, family and social environment, media exposure, and territorial political context.

  • ESTABLISHED TRUST

    vaccination schedule followed without discussion, stable relationship with the treating physician

  • PRACTICAL DELAY

    appointments postponed, schedule delayed, no disagreement in principle expressed

  • UNRESOLVED QUESTION

    a specific question left unanswered, reactivated at the moment of the decision

  • MODERATE HESITATION

    intermediate trust, decision kept open, high sensitivity to the messenger

  • POOR HEALTHCARE EXPERIENCE

    contradictory information, insufficient listening, poorly explained complication

  • LOW INSTITUTIONAL TRUST

    structural distrust of central authorities, often in a territory already polarized

  • HIGH TRUST IN THE PHYSICIAN

    distrust of national institutions coexisting with willingness to listen to the treating physician

  • STRUCTURED OPPOSITION

    around 8 % of hesitant parents: a reasoned position connected to organized networks

These configurations reveal part of the population's heterogeneity. The simulation operates on synthetic individuals, not a handful of persona types.

REACTIONS

It is not the message
that varied the most.

  1. 01

    WITHDRAWAL IS EXPERIENTIAL BEFORE IT IS IDEOLOGICAL

    Only 8 % of hesitant parents are convinced opponents. The remaining 74 % have experienced one or more poor healthcare interactions in the past five years: a missed appointment, insufficient listening, contradictory information, a minor complication poorly explained. These experiences, unrelated directly to vaccination, crystallize at the moment of the decision.

  2. 02

    THE SAME CAMPAIGN PRODUCES TWO OPPOSITE EFFECTS

    Projected territory by territory, the firm national campaign raises coverage by 2 to 4 points in eight territories, and lowers it by a further 3 to 7 points in four others. These four territories were already politically polarized on other issues: the reminder to comply becomes a trigger for opposition.

  3. 03

    THE MESSENGER IS THE ONLY CROSS-CUTTING PARAMETER

    Eight messengers were tested across the 18 typologies. Only one receives favorable attention from all of them: the treating physician. Every other messenger is rejected by at least three typologies. A strategy that systematically equips treating physicians achieves +23 points of coverage in 18 months in the targeted territories.

  4. 04

    NATIONAL VISIBILITY OF SUCCESS HARMS THE REMAINING TERRITORIES

    Giving successful catch-up efforts national media coverage harms performance elsewhere: success is read as proof that the authorities are instrumentalizing public health. Discreet, professional and local communication progresses faster in the territories that were initially most vulnerable.

Two parents can receive exactly the same information and draw opposite conclusions. For one, “ I need to make an appointment ”. For the other, “ they are telling me what to do again ”. The reaction cannot be read from the content alone : it depends on what that content encounters.

THE MESSENGER

Same information.
Different messenger.
Different reaction.

Eight possible messengers were tested across the 18 hesitation typologies. The content does not vary : the attention it receives does. This reading applies to this case and these populations ; it is not a general rule.

  • NATIONAL HEALTH AUTHORITYrejected by several typologies, particularly in territories already polarized on other issues
  • REGIONAL AUTHORITYvaried receptiveness by territory, without working across all typologies
  • TREATING PHYSICIANthe only messenger receiving favorable attention across all 18 typologies tested
  • OTHER MESSENGERS TESTEDpediatricians, engaged celebrities, scientific experts, other parents, family associations: each rejected by at least three typologies
In a general-practice office, a physician speaks with a parent holding a young child on their lap.
The decision is not made in front of a campaign. It is made within a relationship, with someone already known.

SOCIAL DIFFUSION

A campaign does not remain identical
after it has been received.

  1. 01INITIAL MESSAGEA statement, a letter, an ad, a consultation: identical content for all recipients.
  2. 02INDIVIDUAL RECEPTIONEach parent reads it through their own healthcare experience and level of trust.
  3. 03DISCUSSIONIt is then discussed within the family and social circle, where it encounters other interpretations.
  4. 04REINTERPRETATIONIn some territorial contexts, its meaning changes: a health reminder becomes an order to comply.
  5. 05CIRCULATIONIt is this discussed version, not the original message, that circulates and produces the observed reaction.

ADOPTION

coverage can increase at the aggregate level

POLARIZATION

while the intensity of rejection increases in specific territories

STRUCTURE OF OPINIONS

diffuse hesitation can turn into an established position

REVERSIBILITY

an established position unwinds much more slowly than it forms

Making this circulation visible does not indicate which health policy to adopt. It indicates where an intervention changes not only behavior, but the very structure of existing positions.

COMPARISON

Five architectures,
assessed across four dimensions.

ARCHITECTUREADOPTIONTRUSTREACTANCE RISKPOLARIZATION RISK
01Firm national campaignlowlowhighhigh
02Targeted territorial planmediummediummediummedium
03Equipping treating physicians, generic kitmediumhighlowlow
04Treating physicians with kits differentiated by typologyhighhighlowlow
05Treating physicians, differentiated kits and national media discretionhighhighlowlow

Comparative reading derived from simulated reactions. No architecture is cost-free. The national campaign is the most visible and the most exposed to reactance. The territorial plan is better adjusted to local causes, but slower and less legible nationally. Facilitating catch-up resolves practical delays and has little effect on already established opposition. The treating-physician approach requires co-design with the profession; without it, actual kit use remains below 30 %.

THE MOST ROBUST

Equipped treating physicians dialogue kits differentiated by typology paid catch-up consultation national media discretion

THE MOST FRAGILE

Firm national campaign neither territorial differentiation nor local relays

PROJECTION, THEN OBSERVATION

At eighteen months,
coverage restored
without national polarization.

The steering committee selected systematic support for treating physicians : dialogue kits differentiated by typology, a specifically paid catch-up consultation, automated reminders, and national media discretion around territorial progress. The €42 million budget was reallocated : a smaller share for general-public communication and a larger share for treating physicians and territorial infrastructure.

The kits were co-designed with representatives of general practice, then tested in three pilot territories for six months before broader rollout. This co-design, not initially planned, had been identified as critical in the simulations : an imposed kit achieves an actual usage rate below 30 %.

COVERAGE ACROSS THE TWELVE TERRITORIES AT 18 MONTHS
76 % → 87 %, above the 85 % projection
MOST POLARIZED TERRITORIES
+8 to +12 points, where the firm campaign was projected to cause deterioration
NATIONAL POLARIZATION
no movement detected by media-monitoring systems
PHYSICIAN ADOPTION OF THE KIT
78 % use it at least monthly, versus 65 % projected

TAKEAWAY

Part of the decline
was not ideological.

It was experiential and practical : postponed appointments, unanswered questions, poor healthcare experiences resurfacing at the moment of the decision. Among hesitant parents, only 8 % held an established oppositional position. The risk was therefore not failing to persuade enough : it was that firm national messaging would politicize a withdrawal that was not yet political.

An adoption policy is not tested on the persuasive power of its message. It is tested on the distribution of positions in the population concerned, on how much attention each messenger receives, and on what each version of the program changes for situations that do not stem from disagreement.

DECISION

What the decision
selected.

TO FACILITATE
Catch-up for practical delays: dedicated appointment, automated reminder, simplified access.
TO DIFFERENTIATE
The dialogue kit by typology, using a three-question questionnaire before the consultation.
TO PRIORITIZE
The existing relationship of trust with the treating physician, co-designed with the profession.
TO AVOID
Uniform national messaging, and national media storytelling around territorial successes.

POSSIBLE FUTURES

The same public-health objective.
Three ways to pursue it.

A

SCALE UP

A firm national campaign, uniform across the territory

  • maximum visibility and consistency of the national message
  • projected increase of 2 to 4 points in eight territories
  • additional decline of 3 to 7 points in four already polarized territories
  • risk of remobilizing organized opposition networks

B

TARGET

A differentiated territorial plan across the twelve departments with declining coverage

  • better fit with local causes of declining coverage
  • catch-up programs reaching only a limited share of uncovered children
  • lower national legibility, more complex management
  • real effect, but insufficient on its own for the most distrustful typologies

C

WORK THROUGH THE EXISTING RELATIONSHIP

Equipped treating physicians, differentiated kits, easier catch-up, media discretion

  • +23 points of coverage projected over 18 months in the targeted territories
  • a single messenger listened to by all 18 hesitation typologies
  • co-design with the profession is essential: an imposed kit remains below 30 % actual use
  • longer and more intensive deployment, with a territorial pilot phase

METHOD

Before recommending,
we tested reactions.

  1. DECISION
  2. POPULATION
  3. STRATEGIES
  4. DYNAMIC REACTIONS
  5. COMPARISON
  6. DECISION

This case adds a capability to the library : observing that an intervention changes not only behavior, but the structure of positions. Before: diffuse hesitation. After a poorly calibrated intervention: more firmly established positions. The synthetic population makes it possible to explore the sequence action → reaction → discussion → polarization or depolarization, not only action → adoption.

A real case.
An unnamed health authority.

This case is based on a simulation carried out for a national health authority. The institution is not named, no personally identifiable data entered the system, and the detailed results remain the property of the commissioning client. The comparisons between architectures published here are qualitative ; the quantitative values cited are those documented by the commissioning client.

The simulation concerns behavioral reactions to different public-intervention architectures. It is neither a clinical study, nor a medical recommendation, nor an individual prediction, nor an assessment of a vaccine's effectiveness or safety.

Your next decision

Which decision do you want to explore?

Describe your need. We can point you to the right level of support.

What if you tested
your next decision?

State your decision. See the future it produces.

Explore the product