You don't need to be signed in to read BMJ Blogs, but you can register here to receive updates about other BMJ products and services via our site.

Jacob Puliyel on the pentavalent study in Kerala

21 Dec, 12 | by BMJ

jacob_puliyelThe 14 December 2012 marked the end of a year long pentavalent vaccine study in Kerala, India. Kerala is a small state in the south with some of the best healthcare indicators in the country. The pentavalent vaccine is to replace the diphtheria, pertussis, and tetanus (DPT) vaccine and combines H influenza B and Hepatitis B with the older trivalent vaccine.

The pentavalent vaccine study in neighboring Tamil Nadu has about a month more to run. The final results may take another three months to be collated and analysed, but data on the first six months in Kerala are now available under the local right to information laws. The results are of interest not only to India, but also to the Asian countries, where the introduction of the vaccine has been controversial.

The controversy surrounding the vaccine relates to the fact that when introduced in neighbouring countries Sri Lanka, Bhutan, and Pakistan there were numerous “isolated instances” of unexplained or “sudden death” following vaccination. The deaths in Bhutan were said to be unrelated to vaccine, but probably related to viral encephalitis; the evidence provided for this was that the vaccine had not increased the death rate from viral encephalitis, when deaths following vaccination were added to the encephalitis deaths. However, after the vaccine was suspended in Bhutan there were “no further deaths from the encephalitis” in children under one year of age.

The WHO report on adverse events following immunisation in Sri Lanka found no alternate cause for deaths that would fall in the category “probably related” to vaccine under the Brighton Classification. The category “probably related” was removed, and the report stated that the deaths were unrelated to the vaccine.

The deaths in Pakistan were declared as “sudden infant death.” While these deaths were indeed sudden, it certainly is not SIDS, which refers only to unexplained death.

In view of these reports the National Technical Advisory on Immunization (NTAGI) in India suggested that the pentavalent vaccine be introduced only in two states and the harms and benefits evaluated after a year before it is considered for further rollout in the country.

A statistician friend analysed the deaths in Kerala against background mortality in the State. In the first six months of the immunisation programme 40,000 children were vaccinated with the new vaccine. Five of them died soon afterwards (in the next 36 hours or so). Four of the five occurred after the first dose of the vaccine and one after the second dose. Infant mortality in the state is 13 per 1000, and neonatal mortality is 7 per 1000. Post neonatal infant mortality of 6 per 1000 during 337 days (365-28 days of neonatal period) is 0.0178 per day per 1000 children. Using the Poisson probability theory my friend tells me that an observation of four deaths among the vaccinated 40000 vaccinated for the first time is highly inconsistent with the background mortality.

If five children were to die for every 40,000 vaccinated it means we can anticipate 3125 deaths after vaccination in India’s birth cohort of 25 million a year. A meticulously done population based study of Hib meningitis in India has shown that the incidence of haemophilus influenzae type b meningitis is 7/100,000 children under 5. This suggests there are 1750 cases of haemophilus influenzae type b meningitis in the nationwide birth cohort of 25 million. Given the estimated mortality rate 10% there would be 175 deaths from haemophilus influenzae type b meningitis in this cohort.  The vaccine related deaths appear to be far higher than the deaths likely to be prevented from meningitis.

The government standard operating procedure for AEFI recommends that all serious adverse events including seizures and hypotonic pathology be recorded. It states that approximately 570 cases of seizure, 570 cases of HHE and 20 cases of encephalopathy per 1 million doses of the DPT. The fact that none of these have been recorded after 40,000 children were immunised is suggestive that the recordings of adverse events are inadequate.

Two AEFIs related in time, place and/or by vaccine is considered a cluster.  When similar reactions occur with the same vaccine, across different countries, and with vaccines from a variety of manufacturers, it is difficult to consider them mere coincidences.  It will be interesting to see how the final arguments pan out in this debate.

Jacob Puliyel is a pediatrician and head of department at St Stephen’s Hospital in Delhi.

By submitting your comment you agree to adhere to these terms and conditions
  • Wendy Lydall

    The vaccine industry will continue to say
    that the deaths are mere coincidence.

  • Jacob Puliyel Author of blog

    I have responded to these comments in the rapid response column of the electronic
    BMJ http://www.bmj.com/content/341/bmj.c4001?tab=responses
    Please permit me to cut and paste those responses here

    Addendum to the
    response of the Author – for the record

    12 February 2013

    In view of the interest and responses to the controversy raised by the deaths following immunization with Pentavalent vaccine in India, I would like to update the earlier response.

    Intention of NTAGI to Study AEFI

    Drs Phadke and Menon suggest that the NTAGI introduced Pentavalent vaccine in just two States, because “simultaneous nationwide rollout (would have been) logistically challenging.” In our response we referred to the minutes of the NTAGI meeting that stated its rationale quite explicitly. The minutes state
    clearly as follows: “As the vaccine has not been introduced in India, there is not enough data on vaccine safety therefore; vaccine should be initially used in the states with better AEFI management and surveillance systems to monitor the vaccine safety.”

    In its ‘Recommendations’ the NTAGI states: “Pentavalent vaccine to be introduced in Immunization programme in the states of Tamil Nadu and Kerala. Thereafter, data may be
    reviewed after 1 year of introduction before expanding the vaccine into other states.”

    It was not clear why the correspondents brought up the matter of the intentions of the NTAGI. This is easier to understand now that it is being stated that the deaths following vaccination are merely coincidental deaths. Like with hypothesis testing, if
    the intention of the NTAGI was to examine AEFI and subsequently it was found that 15 children died soon after immunization, it will be difficult to pass them off as mere coincidences.

    Proof of Causal Relationship

    The difficulty in establishing causality in these deaths also needs to be explained. In 1999 with rotavirus vaccine, when there was an increase in incidence of intussusceptions of 1 per 10,000 children immunized, the vaccine was pulled off the shelves voluntarily by the manufacturers. No one thought of
    asking for a re-challenge with vaccine, to prove that there was a causal relationship. Now, when 1 in 10,000 are dying following Pentavalent vaccine, we have people asking for proof of a causal relationship which literally will require the dead child to be resuscitated and then re-challenged with the vaccine
    again, to see if he will react in the same way.

    The Myth of Pentavalent Safety in Private Practice

    Another argument that is being brought up is that the vaccine has been used in private practice for many years and it has not caused AEFI. http://week.manoramaonline.com/cgi-bin/MMOnline.dll/portal/ep/theWeekCon

    The implication is that the vaccine itself is safe but the procedures employed in the public heath system are defective and they may be responsible for the many deaths.

    There are several problems with this logic.

    1. Firstly private practice doctors see a very limited number of cases. Only 1 in 10,000 children had a reaction. In all of Kerala, only 15 doctors who had prescribed the vaccine had children dying soon afterwards. So the vast majority of doctors in Kerala will say they have had no deaths from the vaccine. But that does not negate the fact that 15 children died.

    2. Very often the deaths from vaccine in private practice are not recorded as AEFI. I am aware of one such death in February 2012 in Kollengode, Kerala in a private hospital. When I
    asked for a list of all children who had died following immunization with
    Pentavalent vaccine in Kerala (under the Right to Information Act (RTI) 2005), this child’s name was not included in the list sent to me. As I knew all the details of the death, I wrote to the public information officer. He then wrote to me that they had information on the death but they did not include it, as it
    was in private practice!

    3. The convener of the Indian Academy of Pediatrics (IAP) Committee on Immunization, has also asserted that the
    vaccine has been used in the private market for about 10 years and so it was safe. http://www.thehindubusinessline.com/industry-and-economy/govt-needs-to-c

    He stated that the IAP had conducted a study last January on the vaccine, across 1,000 pediatricians. Their responses showed that 80 per cent of those reviewed found the Hib vaccine was safe and over 40 per cent had not observed “untoward” side-effects.

    This figure may have been provided to allay public anxiety but paradoxically, it is indeed alarming. It would appear that 20% pediatricians did not find Hib was safe. I hope this is not true.

    I am aware that all the AEFI investigations into the 4 children who died last month may not have been completed as yet, and it is too early to jump to conclusions. In the meantime it may be prudent to be cautious about recommending this vaccine.

    Jacob Puliyel, Pediatrician St Stephens
    Hospital, Delhi

    Author response to
    Peter J Flegg Re: “Antivaccine Lobby” replies to the BMJ

    7 February 2013

    We appreciate the comment by Dr Peter Flegg that deaths following immunization do not always imply a cause and effect relationship.

    I know more about the deaths in India following Pentavalent vaccine and will base my response on that experience.

    The autopsy reports suggest hypersensitivity reaction in many of the cases. There is no test that can be performed on the vials to look for the hypersensitivity reaction. We depend on
    the Brighton Classification of AEFI to confirm causation.

    If a person gets a rash after taking some drug (medication), the way to be certain of a cause and effect relationship – is to give the same person the medicine again when he is well, and see if an identical rash develops. This is called a re-challenge to prove causation.

    However if the reaction results in death, the person cannot be re-challenged again and the best we can say is that there was no other cause for the death and so it was probably related to drug administered. If 2 such deaths occur, it is called a ‘cluster’
    by the WHO and it makes cause and effect more certain. In Kerala we have 15 deaths already. No alternate cause for the deaths could be identified in spite of a diligent investigation.

    Lung edema is often seen on autopsy in hypersensitivity reactions.

    Peter Flegg writes that ‘data emerging from some countries can be quite unreliable, and in the absence of good clinically verifiable information can be subject to wild speculation’.
    This disbelieving of data from developing countries somehow implies that a death in such countries counts for less than one in the West. Personally I think, many will find this rather offensive.

    The older version of the Rotavirus vaccine was withdrawn in 1999 when it was known that it increased the incidence of intussusception by 1 per 10,000 kids vaccinated. Intussusception has a mortality rate of about 1 per 100 cases. With Pentavalent vaccine in Kerala in the first 6 months of 2012 there has been 1 death per 10,000 children vaccinated but it is business as usual with people saying ‘causation is not
    proven.’

    In the investigation of deaths following Pentavalent vaccine in Sri Lanka the WHO committee could not find any alternate cause for the deaths. The deaths were also most probably
    due to the vaccine.

    I will quote below from a letter of concern written to the Government of India about these deaths.

    The deaths from Pentavalent are sporadic, meaning that thousands receive the vaccine with no reactions, but a few have died. All hypersensitivity reactions are like that –
    only a very few people react adversely. However when a drug is known to cause this form of fatal reactions, it is no longer permissible to give the medication to anyone without first testing to see if the person will react adversely. We do that with penicillin skin test for example, and doctors are liable for negligence if they administer penicillin without testing.

    Pentavalent deaths have now been reported in different countries with use of Pentavalent vaccine from different manufacturers. It cannot therefore be explained on defects in some specific batch of the vaccine, nor can it be blamed on some programmatic error – because of incorrect storage or administration. There is no test available to identify the child who is going to react and die. We as public health experts
    will be liable for these deaths, if we fail to recognize this pattern of reactions seen with this vaccine.

    It is important that we must make every one of these deaths count no matter where in the world they have occurred and no matter how economically poor the country is.

    Jacob Puliyel,
    Pediatrician, St Stephens Hospital, Delhi

    Author reply to
    response from Drs Phadke and Pramila

    6 February 2013

    We thank the doctors Phadke and Pramila for their response to our article.

    They raise 4 points which I will address individually

    a) Disease burden in the population.

    We have stated that the Minz study has found that the under five incidence of Hib meningitis is 7.1/100,000 children under 5.

    We said that if we immunize the birth cohort in one year (of 25 million babies) we can expect to prevent about 1750 cases of Hib meningitis each year. (Total 8750 in 5 years.) In this we assume deaths are prevented uniformly over the 5 years. However this is not strictly true because more cases are prevented in the first year. We have not bothered to differentiate the deaths in each year because we were looking at the final number of lives saved over 5 years and total lives saved each year by vaccinating the birth cohort in one year.

    The correspondents acknowledge that we have quoted the Minz study accurately but they then go on to confuse the issue saying that the incidence of Hib meningitis is 32/100,000
    in the first year and less in subsequent years. To prevent an additional 32/100,000 in the next year they need to vaccinate a second birth cohort. In 5 years they have to vaccinate 100 million babies.

    When 25 million are vaccinated 7/100,000 cases of Hib meningitis are prevented each year. If 100 million are vaccinated we can expect to prevent 7 X 5 = 35/100,000 cases of Hib meningitis. This is akin to what they report 32/100,000 cases prevented. We hope the correspondents will appreciate that to get to the figure they quote, they have a five fold increase in numbers of babies vaccinated and resultant costs.

    Later in the paragraph they state “If we consider all cases of possible meningitis, the annual incidence for below 1 year is reported to be 357 per 100,000″; as if Hib vaccine can prevent all forms of meningitis. There is little evidence for this in literature.

    The correspondents point out correctly that we did not consider deaths from Hib pneumonia prevented. The incidence of pneumonia is some 10 times higher than the incidence of meningitis but the mortality from pneumonia is very much lower. According to this study (http://pdf.usaid.gov/pdf_docs/PNADM961.pdf) pneumonia deaths in Vellore was only 0.77%. The total deaths (pneumonia and meningitis) may be double the figure for meningitis alone but not much higher. This is why most people rely on more easily confirmed figures of Hib meningitis.

    Finally the authors point out again that Hib is difficult to culture and so data we have from cultures is unreliable. This was the very argument used to do the ‘Probe studies'(which did
    not need culture proof) in Indonesia and Bangladesh.
    It will be remembered that the vaccine did not significantly reduce incidence of meningitis or pneumonia there.

    b) Safety and efficacy of the vaccine.

    We used information obtained from the Government under the Right to Information act 2005, that 5 children had died in the first 6 months when about 40,000 children were vaccinated. The correspondent reports the number of doses of vaccine used, without reporting how many children were vaccinated in the whole year. This is statistically less useful as a denominator. Also they suggest that 12 children have died. However this report says the Government thinks there were at least 15 deaths not 12. (http://timesofindia.indiatimes.com/city/thiruvananthapuram/Central-team-…).

    c) Cost effectiveness and sustainability.

    The correspondent states that the NTAGI did not mean to look at the deaths from the vaccine before rolling out to other states. The correspondents are wrong in this. The intention of the NTAGI is clearly stated in the minutes of the meeting and so it needs no further elaboration.

    d) Logistic and administrative issues

    The authors have suggested that the cost effectiveness model used by Gupta et al must be accepted. As we have not made any reference to this cost calculation nor have we read this paper, we are not in a position to comment.

    In summary

    Vaccines are given to a large number of health children to prevent disease and deaths. A vaccine that itself causes deaths is not acceptable (especially given that Hib infection responds to antibiotic treatment). Pentavalent vaccine, as a
    combined vaccine (DPT + Hepatitis B + Hib) was introduced to as a means to increase the uptake of Hib and Hepatitis B vaccine by piggy backing it on the routinely used DPT vaccine. (http://www.one.org/c/us/policybrief/3260/)

    As such, it is being pushed in developing countries.

    It seems to cause an unacceptably high incidence of hypersensitivity reactions and deaths. The deaths in Vietnam
    are said to be at least 6 in 2 months. (http://www.whale.to/v/vaccine_death77.html).

    The correspondents state that it is important to note that causality could not be established in many of these deaths.

    Often when such hypersensitivity deaths occur, the vials are sent for laboratory testing and if no contaminant is found, it is reported that the vaccine was not at fault. This is a negation of the Brighton Classification of adverse events following immunization (AEFI). If a death follows immunization and there is no other adequate reason for the death, it must be presumed that the death was probably due to the vaccine.

    I am sure that the correspondents will want these deaths in India and neighboring countries investigated using the Brighton Classification and we must continue the immunization programme only if it is safe.

    Jacob Puliyel
    Pediatrician St Stephens
    Hospital Delhi

You can follow any responses to this entry through the RSS 2.0 feed.
BMJ blogs homepage

The BMJ

Helping doctors make better decisions. Visit site



Creative Comms logo

Latest from The BMJ

Latest from The BMJ

Latest from BMJ podcasts

Latest from BMJ podcasts

Blogs linking here

Blogs linking here