How do we know vaccines are safe ? Safety inferred because of an - - PowerPoint PPT Presentation

• We detect Adverse Events Following Immunisation
• Then perform cause specific classification to

differentiate an event which is a reaction from a co-incidental event.

How do we know vaccines are safe ?

Safety inferred because of an absence of adverse reactions

An AEFI that is caused or precipitated by a vaccine due to one or more

• f the inherent

properties of the vaccine product. An AEFI that is caused or precipitated by a vaccine that is due to one

• r more quality

defects of the vaccine product including its administration device as provided by the manufacturer. An AEFI that is caused by Inappropriate vaccine handling, prescribing or administration. An AEFI arising from anxiety about the immunization. An AEFI that is caused by something

• ther than the

vaccine product, immunization error or immunization anxiety

CIOMS/ WHO cause specific definition of AEFIs

How do we differentiate an adverse vaccine/ immunisation reaction from a co-incidental event ?

A/Professor Michael Gold Head Allergy and Immunology Women’s and Children’s Health Network and Discipline of Paediatrics University of Adelaide

Outline

• AEFI surveillance systems

– Components

• Causality assessment

– For a clusters of cases (Aggregate) – For individual cases (Case Based)

Why should we have a country specific or regional AEFI surveillance system ?

– Incomplete safety information because;

• Country specific schedules
• Genetic predisposition

– Eg Narcolepsy Adjuvant Pandemic Influenza Vaccine – HLA DQB1-0602

• Novel vaccine use in limited countries

Nohynek H1 AS03 adjuvanted AH1N1 vaccine associated with an abrupt increase in the incidence of childhood narcolepsy in Finland. PLoS One. 2012;7(3):e33536. doi: 10.1371/journal.pone.0033536. Epub 2012 Mar 28.

Why should we have a country specific or regional AEFI surveillance system ?

– Responding to local community concern

Why should we have a country specific or regional AEFI surveillance system ?

– Vaccine quality defects

• Vaccine manufacture - batch related defects

– Immunisation error

• Programme error

– Differentiate local co-incidental events from reactions

• Background rates of disease
• Eg Introduction of a new vaccine in a malaria endemic region vs

same vaccine introduced into non-malaria region

Components of an AEFI surveillance system

• Ascertainment of AEFI reports
• Ascertainment of AEFI reports
• Ascertainment of AEFI reports

– Passive, Active, Passive Stimulated

• Report management

– Coding, classification, data entry, storage

• Analysis

– Analysis, causality assessment, signal detection – Hypothesis generation

• Action

– Investigation, corrective action, communication (feed-back),

Current estimates are that there will be approximately 30 vaccine products "newly pre-qualified" and introduced into immunisation programmes in the middle/low income countries and will include Rotavirus, HPV, PCV, Influenza vaccine 1995-2008 – manufacturers doubled to 136

• Q1. How can industry compliance with

GMP be strengthened ?

• Resources

– AEFI reporting, evaluation and action (investigation, education, research, communication)

• Involves and co-ordination of all stakeholders

– NRA – Immunisation Programme (EPI) – Government – Industry – Experts - Academia

• Ownership
• Q2. How can vaccine pharmacovigilance be

strengthened ?

• Universally recommended, healthy individuals,

vulnerable populations..........

• Higher level of trust
• Cannot have

trust without good governance

• Q3. How can vaccine pharmacovigilance be

governed and include all stakeholders (industry) ?

South Australian officials 'hiding' flu vaccination statistics Tory Shepherd, Health reporter From: The Advertiser April 30, 2010 1:14PM

• Q4. How can Post Licensure Surveillance

systems be strengthened and complementary systems developed ?

Types of AEFI surveillance

Passive surveillance

• Reporting of AEFI

by HCW and the public

• AEFI needs to be

recognized

• Under-reporting is

common Stimulated passive

• Passive

surveillance +

• ngoing reminders

& education of HCW for AEFI cases

• May be focused on

selected AEFI Active surveillance

• Systematic search
• r scan for defined

AEFI in target populations (clinic, hospital, community)

Stimulated passive surveillance system

• Useful where AEFI

surveillance is new

• Involving doctors in clinical

settings can increase reporting of serious AEFI

• “Zero-reporting” may be a

useful

Periodic reminders, education & feedback to detect & report cases

Active surveillance

• Active search for safety signals:

cases, risks or syndromes (HHE, intussusceptions, etc)

• Time and resource consuming
• Higher sensitivity, but lower

coverage

• Prone to biases unless blinded
• bservers

Systematic search/scan for defined AEFI in target populations (health unit, clinic, hospital, community)

• Review hospital and clinic records for

selected AEFI

• Complete reporting form when a

patient meets the criteria

Based on search for AEFI

• Follow up patients after specific

vaccination

• Inquires about any AEFI (specific

targets) and complete reporting form

Based on search for vaccine received Collation and analysis of completed report forms

How to do active surveillance continued

To assess the effect of: (i) sending weekly SMS, or (ii) weekly supervisory visits on AEFI reporting rate during a meningitis immunization campaign conducted in Cameroon in 2012 using the meningitis A conjugate vaccine (MenAfriVac™).

A total of 348 (77.2%) of 451 health facility were included, and 116 assigned to each of three groups

The incidence rate of reported AEFI per 100 health facility per week was

– 20.0 (15.9-24.1) in the SMS group, – 40.2 (34.4-46.0) in supervision group – 13.6 (10.1-16.9) in the control group.

Vaccines safety; effect of supervision or SMS on reporting rates of adverse events following immunization (AEFI) with meningitis vaccine (MenAfriVac™): a randomized controlled trial.

Ateudjieu J1. Vaccine. 2014 Sep 29;32(43):5662-8. doi: 1016/j.vaccine. 2014.08.012. Epub 2014 Aug 23.

“For all vaccines, the attitude of the physician, nurse…is very influential in the decision to vaccinate a child… *

• Q5. How to improve health care and

vaccine provider knowledge, attitude and practice about safety and surveillance ?

* Swennen B et al. Analysis of factors influencing vaccine uptake: Vaccine 2002;20 S5-S7.

* Ansari M et al. Reducing resistance against polio drops. JRSH 2007;127:276-9 Hak E at al.. Vaccine 2005; 23: 3103-7.Pearce et al. BMJ 2008 336 (7647): 754.

• 29 Health Professionals
• 46% of ED consultants not aware of a system
• f reporting, all nurses aware
• Variation in what to report as an AEFI

– Nurse, GP, ED consultant – Eg ED only report life threatening, did not report febrile seizures

• GP and ED no training, all nurses trained

Vaccine Product Related Reaction

Limitations of passive surveillance

Vaccine Product Related Reaction

Limitations of passive surveillance

Febrile seizures per 1,000 vaccines Fluvax Jr Western Australia 9 Australia 5 Panvax Australia 0.08-0.17 USA CDC Vaccine Datalink 0.16

• Global implications of

safety concerns

– MMR and autism (Industrialised) – HPV vaccine and Fiji (ID-NID) – Pentavalent vaccine and death

• (NID-ID)
• Global Surveillance and

communication

• Q6. How can we co-ordinate and strengthen

global safety surveillance efforts ?

• Q7. How do we engage the community to

communicate and monitor vaccine safety ?

Solutions Pharmacovigilance in Africa

• Improving knowledge

– Community, health providers

• Improving co-ordination

– NRA/EPI/Government/NGO/Industry

• Novel methods of surveillance

– Passive vs Stimulated vs Active sentinel

• Mobile phone technology
• Establishing regional networks

– Regional causality assessment committee, NITAGS

• Sustaining surveillance capacity after clinical trials
• Opportunities with the polio end game

Causality and Causality assessment

Causality*

• Is the relationship

between two events (the cause and the effect), where the second event is a consequence of the first

Causality Assessment

• Determining if such a

relationship exists and if so to what extent *A direct cause is a factor in absence of which the effect would not occur (necessary cause). *Sometimes, there are multiple factors that can precipitate or function as co- factors for the effect (event) to occur.

Causality assessment for AEFI

• Why ?

To avoid automatically concluding that

“the event happened after vaccination, therefore it happened because of vaccination”…

• For a collection of AEFI reports

(aggregate)

• For an individual AEFI report (case based)

Co-incidental or causal (reaction) ?

Passive AEFI surveillance does not establish

causality

Event/Illness/Syndrome yes no Vaccination

no yes

• Unique lab result
• Unique clinical syndrome
• Epidemiological study
• NEED A RELATIVE RISK
• ABSOLUTE RISK

Adverse vaccine reactions

Unique Laboratory results

• Culture of live attenuated vaccine organism

Adverse vaccine reactions

Unique Laboratory results Culture of live attenuated vaccine organism

• Meningoencephalitis
• Mumps, Measles
• Disseminated infection
• BCG, Rotavirus, Varicella (CP and Zoster)

Adverse vaccine reactions

Unique clinical syndrome

Adverse vaccine reactions

Unique clinical syndrome

– Injection site reaction – Hyporesponsive Hypotonic Episode (HHE)

• Any vaccine
• Usually Pertussis containing vaccine
• < 12 months of age

– Viscertrophic disease

• Yellow Fever vaccine

An AEFI that is caused or precipitated by a vaccine due to one or more

• f the inherent

properties of the vaccine product. An AEFI that is caused or precipitated by a vaccine that is due to one

• r more quality

defects of the vaccine product including its administration device as provided by the manufacturer. An AEFI that is caused by Inappropriate vaccine handling, prescribing or administration.

.

CIOMS/ WHO cause specific definition of ISR

• F. Buser, M.D. Side Reaction to Measles Vaccination Suggesting the Arthus
• Phenomenon. N Engl J Med 1967; 277:250-251August 3, 1967

Froehlich H, Verma R. Arthus reaction to recombinant hepatitis B virus vaccine. Clin Infect Dis. 2001 Sep 15;33(6):906-8. Epub 2001 Aug 21.

Injection Site reaction Product related reaction

Arthus reaction

• Difficult to diagnose without a skin biopsy
• “Arthus like” reaction
• Unlikely with priming doses
• Onset < 4 days
• Symptoms inflammation
• ISR with skin necrosis

Injection Site reaction

Product related

Non-Inflammatory -Urticaria

Injection Site reaction Product related Non-Inflammatory - Angioedema

Extensive Limb Swelling

Per 100,000 doses

Gold MS, Noonan S, Osbourne M, Precepa S, Kempe A. Local reactions following the fourth- dose of acellular pertussis vaccine in South Australia. Med J Aust 2003; 179: 191-194.

Ultrasound examination

Control arm Vaccinated arm

Injection Site Reaction – Immunisation error (Administration)

2/12 old, Dose 1 of DtaP/HepB/Hib/IPV, ISR noted within hours, image taken at 24 hours, infant irritable ++, febrile – immunisation administered by a “student nurse”

Gold M et al. Med J Aust 1998; 168 (9): 471. Wawryk A, Mavromatis C, Gold M. Electronic monitoring of the vaccine cold chain in a metropolitan area. BMJ 1997; 315: 518.

Injection Site Reactions – Immunisation error (Handling)

Effect of adverse vaccine storage - < 20 C, vaccines rarely freeze Cold exposure uncouple antigen from adjuvant

Injection Site reaction

Product related or vaccine quality defect

Adverse vaccine reactions

All other reactions can only be demonstrated by epidemiological studies because condition occurs without vaccination that is a BACKGROUND RATE

– Need to calculate a relative risk or odds ratio

Co-incidental or causal (reaction) ?

Passive AEFI surveillance does not establish

causality

Event/Illness/Syndrome yes no Vaccination

no yes

• Unique lab result
• Unique clinical syndrome
• Epidemiological study
• NEED A RELATIVE RISK
• ABSOLUTE RISK

Best evidence is randomized trial comparing AEFI in vaccinated and non vaccinated groups.

But…never large enough to detect rare events and comparator group not usually unvaccinated

Unvaccinated group Vaccinated group

Example epidemiological investigation

Intussusception risk and disease prevention associated with rotavirus vaccines in Australia's National Immunization Program. Carlin JB1. Clin Infect

• Dis. 2013 Nov;57(10):1427-34. doi: 10.1093/cid/cit520. Epub 2013 Aug 26.

Farrington CP, Pugh S, Colville A, Flower A, Nash J, Morgan-Capner P, Rush M, Miller E 1995. A new method for active surveillance of adverse events from diphtheria/ tetanus/pertussis and measles/mumps/rubella vaccine. Lancet 345(8949:567-569

Gold M, Dugdale S, Woodman RJ, McCaul KA. Use of the Australian Childhood Immunisation Register for vaccine safety data linkage. Vaccine. 2010 Jun 11;28(26):4308-11. doi:

For example;

• Ascertainment

– School based programme – direct observation – 12 suspected cases

• Definition

– Brighton Collaboration CD – 7 met the BC for anaphylaxis

• Denominator

– 269 680 doses administered

2.6 per 100,000 doses (95% CI 1.0-5.3 per 100,000) 0.1 per 100,000 (95% CI 0.003-0.7) - Men C Conjugate

Individual Causality

Serious AEFI

Potential immunizatio n error

Clust er

Parent al concer n

Communit y concern

Unexpected relationship with vaccination

Unexpecte d frequency

AEFI Reporting*

Reporter should NOT assess causality * Events to be reported according to context - Routine surveillance, new vaccine, mass campaign etc

Seriou s AEFI

Clusters & events above expected rate/ severity

Evaluation

• f

suspected Signals

Other AEFI*

Case selection for formal causality assessment

Death, Hospitalization, Significant disability, Life threatening, or Congenital anomaly/ Birth defect As decided by reviewing team / committee  If immunization error is suspected  Significant events of unexplained cause within 30 days of vaccination  Events causing significant parental

• r community concern (e.g.

Hypotonic Hyporesponsive Episode (HHE), febrile seizures)

Who should assess AEFI causality

• A reviewing team/ committee that should

– Be Independent

• free of real or perceived government, industry conflicts of

interest – Have broad range of expertise

• infectious diseases, epidemiology, microbiology, pathology,

immunology, neurology, vaccine program expertise, other… – Have written terms of reference (ToR)

• Can make use of an existing drug

causality assessment team BUT modify and adapt to fit vaccines.

Causality Assessment Steps

Eligibility Checklist Algorithm Classification

Why 4 steps?

To categorize the association

• f the AEFI to

vaccine / vaccination Step 4: Classification To obtain a trend on causality with the checklist information Step 3: Algorithm To systematically review available information Step 2: Checklist To determine that the AEFI case satisfies minimum criteria for causality assessment Step 1: Eligibility

AEFI case

• Ensure AEFI investigation is completed and all details of the case

is available

• Retain case details in a retrievable database

Identify vaccine(s)

• Identify one vaccine (implicated) administered before this event

Valid Diagnosis

• Select the unfavorable or unintended sign, abnormal laboratory

finding, symptom or disease you want to check causality.

Case definition

• Brighton Collaboration definition, Standard literature definition,

National definition or other approved definition

Step 1: Eligibility

Brighton Collaboration

https://brightoncollaboration.org/ Develops AEFI case definitions Promotes global implementation

• f case

definitions

• Translation into

regional languages

Brighton Collaboration-Case Definition Format

Level 1 of diagnostic certainty Level 2 of diagnostic certainty Level 3 of diagnostic certainty

Creating a causality question*

Examples of causality questions

• “Has the vaccine A caused hepatomegaly?” (An example of an unfavorable or

unintended sign)

• “Has the vaccine B caused thrombocytopenia?” (An example of a laboratory

finding)

• “Has the patient complained that the vaccine C caused itching and redness?” (An

example of a symptom)

• “Has the vaccine D caused meningitis?” (An example of a disease).

Has the ____________ vaccine / vaccination caused ____________? * Should address the unfavorable or unintended sign or laboratory finding or symptom or disease.

Vaccine name

Valid Diagnosis

Step 2 Checklist

• I. Is there strong evidence for other causes?
• Relationship with vaccine ingredients
• Immunization error
• Relationship with vaccine administration
• II. Is there a known causal association with the Vaccine /

Vaccination II (Time). Was the event within the time window of increased risk?

• III. Is there a strong evidence against a causal association?
• IV. Other Qualifying Factors

Y N UK NA

    

AEFI causality of a single case - The Scientific Basis

I A. Inconsistent causal association to immunization II A. Consistent causal association to immunization

IV B. Indeterminat e IV D. Unclassifiabl e

IV C. Inconsistent causal association to immunization

• II. Is there a

known causal association with the vaccine/ vaccination

IV A. Consistent causal association to immunization Yes Yes Yes Yes No

III A. Inconsistent causal association to immunization

Yes

• IV. Review
• ther

qualifying factors

• III. Is there a

strong evidence against a causal association? II (Time). Was the event within the time window of increased risk?

• I. Is there

strong evidence for

• ther causes?

Is the event classifiable?

Step 3 Algorithm (summary)

• A. Consistent causal

association to immunization

• A1. Vaccine product-related

reaction (As per published literature)

• A2. Vaccine quality defect-

related reaction

• A3. Immunization error-

related reaction

• A4. Immunization anxiety-

related reaction

• B. Indeterminate
• B1. *Temporal relationship

is consistent but there is insufficient definitive evidence for vaccine causing event (may be new vaccine-linked event)

• C. Inconsistent causal

association to immunization

• C. Coincidental

Underlying or emerging condition(s), or condition(s) caused by exposure to something

• ther than vaccine

Unclassifiable

Specify the additional information required for classification Adequate informatio n available Adequate information not available

*B1 : Potential signal and maybe considered for investigation

• B2. Qualifying factors result

in conflicting trends of consistency and inconsistency with causal association to immunization

Step 4: Classification

• A. Consistent

causal association to immunization

• A1. Thrombocytopenia

after MMR vaccination.

• A2. Paralytic polio

caused by failure of manufacturer to inactivate IPV completely .

• A3. Transmission of

infection by contaminated multidose vial.

• A4. Vasovagal syncope

in an adolescent following vaccination.

• B. Indeterminate
• B1. Irritable Bowel

Syndrome after TT vaccine (Hypothetical and unproved so far)

B2. Thrombocytopenia after MMR vaccine in a dengue endemic area

• C. Inconsistent

causal association to immunization

• C. Coincidental

Child dies after DPT vaccine and autopsy shows congenital heart disease

• r

Fever occurs after vaccination (temporal association) and malarial parasite isolated from blood.

Classification - Examples

Key considerations

The "critical" part is the valid diagnosis. The first conclusion may not be final: as more information becomes available, the causality can change

KL 12 months of age

• Previously well
• No known allergies
• Immunised with MMR vaccine at 10:30AM
• 10 minutes later developed widespread

urticarial rash, persistent cough and wheeze.

• Treated ventolin
• Hospital rash noted, chest clear
• Diagnosis Anaphylaxis

https://brightoncollaboration.org/public.html

Are reported symptoms/signs consistent with anaphylaxis ?

MA Yes

http://www.who.int/vaccines- documents/ DocsPDF06/843.pdf

Anaphylaxis MMR

Has the MMR vaccine / vaccination caused Anaphylaxis? (The event for review in step 2)



No details available on the

• ther tests



YES Anaphylaxis is a recognized event



No

       

Within 4-6 hours

No other cause of anaphylaxis

      

No

 

Unknown MMR can cause anaphylaxis Inadequate information Inadequate information Inadequate information Yes

Notes for Step 3: Notes for step 3: II A: With available information, it seems likely that the vaccine caused the event. This is because MMR vaccine is a known to cause the event and the time window is suitable.





Summarize the classification logic: With available evidence, we could conclude that the classification is consistent because: With available information, it seems likely that the vaccine caused the event.

Multiple Vaccines, Multiple AEFI and Clusters

Each vaccine should be assessed separately

Each event should be listed separately and specific eligibility question (step 1) asked and independently evaluated.

Each patient in the cluster should be separately evaluated

Multiple Vaccines in the same patient Multiple AEFI in the same patient AEFI cluster

What can go wrong?

Causality assessment not done, not systematic, not done by trained personnel and/or not done timely Information in AEFI report is so limited that causality assessment cannot be done Lack of expertise and/or independence of the review committee responsible for formal causality assessment undermines credibility Non analysis of AEFI data may delay recognition of clusters and programme errors Lack of communication of findings, not addressing all target audiences,

• r lack of diplomacy and/or cultural sensitivity

Conclusion

• AEFI surveillance systems

– Components

• Causality assessment

– For a clusters of cases (Aggregate) – For individual cases (Case Based)

michael.gold@adelaide.edu.au

EB 3 year old

Ex preterm 34 weeks Downs syndrome with AVSD repaired Immunised Fluvax Jr 15/3/2010 Onset of symptoms 5 hours Irritable, sudden onset pallour, felt cold, cyanosed lips, difficulty breathing (fast breathing but not noisy, no cough) Ambulance called Adrenaline 90ug administered Transfer Hospital No - Rash, RDS, Wheeze, T-39.3c Diagnosis – Anaphylaxis to vaccine

OJ 8 weeks of age

– Routine Immunisations DTaP-IPV-HepB-HiB, RVV – Acute respiratory distress following oral RVV – At provider given RVV, gagged, started to cough, developed respiratory distress. – IMI adrenaline – Ambulance called - noted “increased work breathing” – At hospital - No RDS, wheeze, rash – CXR-normal – Diagnosis Anaphylaxis

LM 2 months of age

• Previously well
• Breast fed
• Immunised with DTwP-HepB-Hib, RRV,

PNC

• 12 hours noted to be unresponsive, pale,

circulatory compromise (poor peripheral perfusion), respiratory distress

• Demise
• Diagnosis Anaphylaxis