Challenges in the investigation of adverse events following immunization. At National Seminar on New Vaccines for All: Why, Which, When. Organised by JSS SAMA NMJI, FMES

Jacob Puliyel

Thank you for inviting me for this national seminar on newer vaccines.
Slide 2
Chairperson Professor Imrana Quadir and distinguished delegates
Challenges in the investigation of adverse events following immunization.

I would like to do a quick question with you.
I tell you that I am interested in spreading a lie – some falsehood
I will divide this audience into two groups.
Those to my right I tell you I am willing to pay you 1 rupee every time you told someone this lie
To those on the left I tell you I am willing to pay 100 Rupees every time you lied for me.
Who do you think would do a better job for me?
No it is not the people who are paid 100 Rupees.
This is counterintuitive
Those who get 100 rupees say
I know it was a lie but the money was good so what the heck.
Those who get Rs 1 tell the lie all the same but feel so bad for selling themselves for 1 rupee that they convince themselves that they thought it was the truth. Thus because they have forced themselves to believe the lie they make more convincing liars.
This was first noticed in the Vietnam war.
During the war the Vietnamese took American prisoners and the American held Vietnamese prisoners. Prison condition for the American prisoners were far more basic than those provided to the Vietnamese.
Yet the Vietnamese Prisoners never denounced their country but evey few days American could be seen on Vietnamese Television denouncing America.
The Americans thought the Vietnamese has some very potent brainwashing technique till after the war was over and the prisoners were repatriated.
So what was the brainwashing technique. The Vietnamese would give the Americans an essay to write.What is wrong with the American capitalistic system.
The best essay would get the first prize. His essay would be read out. He would be awarded with some trinket – a lapel pin or an extra bowl of gruel.
The prize was so insignificant. The writer felt so bad about denouncing his country for so little that he would say he wrote the essay because he believed what he wrote was true. He would come on TV and say so.
This is cognitive dissonance. You know it is a lie but you convince yourself that it is true
I want you to hold that in mind for the present.
Slide 3
Adverse events after immunization. The investigations are the same as for a drug reaction or even for effects of environmental agents. Say smoking and cancer, or a child who dies soon after being injected penicillin.
Investigations are at two levels
Can smoking cause cancer?
Does it do so in society?
You have to do a cohort study or a case control study and see those exposed and those not exposed. If there is statistically significant increase among smokers (p value less than 0.05) then you say that smoking probably causes cancer.
When an individual who has smoked for 1 year has cancer can you say whether smoking caused cancer in him. Or a child has an adverse reaction – you have to look at a probability scale
Slide 4.
To be certain the drug produced the reaction and it was not a coincidental event one has to withdraw the drug and when he is well rechallenge him with the same drug . If the same reaction recurs we can be Certain. However if death is the adverse event one cannot revive him and rechallenge him to see if will die a second time. Very ofter this is unethical to rechallenge and all you can say is that the reaction is Probably due to the drug.
For this classification there must be a temporal relationship with use of the drug and there must not be any other plausible explanation for the adverse event.
If a temporal association exists but there is an alternate explanation the it is classifies as Possibly related to the drug and possibly to the alternate explanation
If there is no temporal relation and there are clearly other explanation it is unlikely or unrelated.
If no data is available to make tis evaluation it is unclassifiable.
Slide 5
I said there was no difference whether the reaction is after a drug or after a vaccine.
Vaccines are given often to the entire population of health children where as other drugs are given only in a small group of sick individuals. There is thus a very low tolerance for serious adverse events after vaccine.
Slide 6
There were 3 deaths with Pentavalent vaccine.
The WHO experts went to investigate.
They found clear temporal association and there was no alternate explanation present. Using the standard Brighton classification this was probably death related to vaccination.
So what did the WHO group do
Slide 7
They wrote in the report that they are deleting Probable and Possible from the Brighton classification and they reported the deaths were unlikely to be related to the vaccine.
Only the Summary was published online that the WHO experts declared the deaths were unlikely to be related to vaccine. The full report was presented to the Delhi high court in a vaccine case and that is how we got to know the methodology they used to arrive at the report.
Were the experts lying?
Cognitive dissonance.
The BMJ published the story of this innovative methodology to the embarrassment of the WHO
Slide 8
What did they do in response
The CIOMS/WHO set up a group of 40 of who 19 were representatives of vaccine manufacturers with conflicts of interest.
They developed a new Algorithm for AEFI.
They decreed that in future an adverse event following immunization will be considered an AEFI ONY IF there existed prior epidemiological evidence that the reaction can be caused by the vaccine
Slide 9
Ordinarily vaccines are tested in a small randomized controlled Phase 3 trials
If there are no problems in this small group the vaccine is licensed and given to the population at large in Phase 4 trial. Rare reaction that happen say 1 in 10,000 are picked up in Phase 4 trial
If such reactions happen repeatedly it forms a signal and then epidemiological studies and case controlled studies are done
Under the new scheme such reactions noticed for the first time in Phase 4 trials are all deleted as Not an AEFI
Slide 1
This is a schematic representation of the CIOMS WHO classification
On the right I have highlighted wha it would have been in the old Brighton scheme
The first question is Did the reaction occur after vaccine.
If there is no temporal association it was unlikely or unrelated.
In the new scheme it is not an AEFI
See the subtle change. It is actually an AEFI an adverse event after immunization but the AEFI is not causatively related. This is different from calling it Not an AEFI
The next question is whether the case has been investigated sufficiently.
In the old scheme if data is unavailable it was unclassifiable. Here it is Not an AEFI
If there is temporal relation and sufficient details available the next question is whether there is an alternate explanation for the even. If there is an alternate explanation it could still possibly be causatively related. Here it is classified as inconsistent with causal association.
If there is no alternate explanation this would have been classified as Probably related but in the new classification the next question is whether there is a known causal association in epidemiological studies.
What is Probably related to vaccine is classified as Not a case of [AEFI]
The next question is whether it fulfils CIOMS case definition for the known reactions before it is classified as consistent with causal association
You can see that what the CIOMS WHO did was exactly what the Sri Lankan team did. They have deleted Possible and probable from the classification.
Slide 11
In case you think I have made a parody of the classification this is the original algorithm
At the second level I have highlighted it Is there a known causal association with vaccine. If this answer is yes and only if the answer is yes can it move to the left and and bae classified as consistent causal association.

If the answer is No you can go 3 steps down and again you are asked Is it classifiable meaning whether it fulfils CIOMS case definition/
If the answer was no at level two obviously there will not a case definition developed by CIOMS and you will be stopped going further down. So everything to the Right is a waste of time and coloured ink.
Slide 12
Thus the CIOMS WHO has made Phase 4 trials redundant.. Any new reaction will simply be deleted as Not a case of AEFI and all memory of the signal will be erased
Remember Rotavirus intussusceptions were noticed in Phase 4 trials where 10 in 100,000 children developed the reaction.
In the new scheme of ting it would be ignored
Slide 13
After the Sri Lanka deaths there were 12 deaths in Vietnam with the same vaccine as used in Sri Lanka.
This was 2013. By then the CIOMs classification had been put in place and the Sri Lanka deaths were erased.
He WHO report says No fatal reaction has ever been associated with this vaccine.
Cognitive dissonance
Slide 14
The entire exercise is one of AEFI denial.
The reason all this exercise in changing the AEFI classification was undertaken was to deny deaths following Pentavalent and hexavalent vaccine. The CIOMS have decreed that no cognisance can be taken of AEFI signals unless there have been prior epidemiological studies
As it happens there is clear epidemiological evidence of death s with thee vaccines and I will explain them in the next few slides. Even with this epidemiological evidence AEFI denial continues in the case of Pentavalent vaccine
Cognitive dissonance on top of cognitive dissonance
Slide 15
The Periodic safety update reports are sent to the European Medical agency. For some reason these safety reports are considered confidential – suggesting to me that they are perhaps not safe.
An Italian court has made public the PSUR reports 15 and 16 and this is available on the net.
Slide 16
The PSUR 15
60 million doses had been used. The manufacturers estimate 9.4% or 5 million doses were used in kids above 1 year.
The SIDS rate in children under 1 is considered 0.4/1000 and it is less 0.06 in kids above 1
Slide 17From this they calculate the expected deaths for the doses used for each day after vaccination
You can see that day of vaccination 54 deaths are expected and double that in the first 2 days and so on.
You will see actual deaths are all less than expected.
But if you ignore the expected deaths you will see there were 42 deaths in the first 4 days and only 8 deaths in the next 4 days. This five fold clustering of deaths soon after vaccination suggests to me that there may be a relation between death and vaccination but I wont quibble about that
Skip Slide 18
Slide 19 See the table for PSUR 16
The same pattern is seen except that in the 2nd year there were more observed deaths on day 1 and day 2 than was expectedThere were 5 observed deaths instead of 3.96 expected on the day after vaccination
Slide 20
The PSUR 16 says that 20% of all doses were used in the 2nd year.
In all the PSUR till the 16th report it was estimated that 9.4% of all doses were used over 1 year.
Suddenly this rate was doubled to 20%
Slide 21
You can see why in this table. The expected deaths were nearly half the observed deaths so they doubled expected deaths. Even then observed were more than expected
Slide 22
The figure 20% is tenable only if every child who gets even 1 dose gets all 4 doses and 1 dose is after 1 year.
Italy the second largest user of the vaccine uses only 3 doses and all under 1 year.
So it is impossible to arrive at the 20% figure. Were they being dishonest in their expectation figure.
Only cognitive dissonance
Slide 23
Here is another epidemiological study that links sudden deaths to vaccine
The TOKEN study was done with funding by the German Government
This is Table 41 Model 1 shows the risk of sudden death in the period 0-3 days after Pentavalent vaccine was 8 times higher than expected. P value 0.006
Silde 24 delete
Slide 25
So we have epidemiological studies that associate sudden deaths with these vaccines.
Even if we use the new CIOMS classification these deaths must be considered as AEFI causatively related to vaccination. But AEFI denial continues. Cognitive dissonance
Slide 26
We spoke of 3 deaths in Sri Lanka and 12 in Vietnam
There have been 237 deaths within 72 hours of vaccination in India reported to the Government of India till August 2016
When compared to deaths with DPT deaths have doubled. We are only comparing DPT deaths where Pentavalent and DPT are being used. If deaths following DPT are the natural rate of deaths then the increase in deaths following Pentavalent vaccine must be attributed to Pentavalent vaccine
The Govt of India has the data but they continue vaccinating
Paediatricians all over the country are using the vaccine not for megere profits.
They believe they are helping the kids
Cognitive dissonance.
Slide 27
AEFI is no more challenging than the investigation of any other drug reaction
The challenge is to obfuscate
To make a case of AEFI into Not a case of AEFI
The CIOMS WHO classification is a perversion of science and militates against common sense
The Rota virus intussusceptions would never have been picked up with this classification
We need ideas and we need action
We have to work jointly to remove this perverse classification if we are to avoid unnecessary deaths