Deaths Reported after Pentavalent Vaccine Compared with Death Reported after Diphtheria-Tetanus-Pertussis vaccine - An Exploratory Analysis
Immunization is one of the most effective public health tools available to prevent death and disease. Serious adverse reactions following immunization (AEFI) are rare. However coinci-dental sudden-infant-death-syndrome (SIDS) deaths do occur temporally associated with vac-cination. In 2010 the Government of India (GoI) introduced a new standard operating proce-dure (SOP) to report AEFI.
There have been stray newspaper reports of deaths soon after the administration of the Penta-valent vaccine (PV) which was introduced by the GoI in December 2011. This study was con-ducted to examine if there is an epidemiological signal from the data collected passively under the new SOP.
Materials and Methods
We used data provided by the GoI on the number children who received 3 doses of Diphtheria-Tetanus-Pertussis vaccine (DPT), the number receiving PV and the number of deaths in the vac-cinated within 72 hours.
After PV was introduced in the states, 45 million infants received DPT vaccination and 25 mil-lion received PV. There were 217 deaths within 72 hours after DPT was administered and 237 following PV. There were 4.8 deaths per million vaccinated with DPT (95% CI 4.2-5.5) and 9.6 deaths (95% CI 8.4-10.8) per million vaccinated with PV(Odds ratio 1.98 (95% CI 1.65 to 2.38) There were 4.7 additional deaths (95% CI: 3.5-5.9), per million vaccinated with PV instead of DPT (p<0.0001)
Deaths following DPT vaccination would include the natural rate of deaths within that window period, plus deaths if any, caused by DPT. For purposes of this study we assumed that all the deaths associated with DPT are coincidental SIDS deaths. Taking that as the base rate of SIDS, we look for any increase in the death rate after PV.
This study demonstrated an increase in reports of sudden unexplained deaths within 72 hours of administering PV compared to DPT vaccine. Whether improvements in AEFI surveillance sys-tem or other factors contributed to this increase cannot be ascertained from this study.
These findings do not warrant deviation from current vaccination schedule but the differential death rates between DPT and PV do call for further rigorous prospective population-based in-vestigations.
Pentavalent vaccine (PV) (combined Haemophilus influenzae type B (Hib), hepatitis B, whoop-ing cough, tetanus, and diphtheria vaccine) was introduced in the national immunization pro-grammes of Sri Lanka and Bhutan before it was introduced in India. In both these countries, use of this vaccine was temporally associated with adverse events following immunization (AEFI) including unexplained deaths (1). These were investigated by the WHO and the deaths were declared as unlikely to be related to the vaccine. In India, starting 15 December 2011, PV was introduced into the country's immunization programme to replace the DPT vaccine (diph-theria, whooping cough and tetanus) in a staged manner with a view to add protection against Hib and Hepatitis B without increasing the number of injections given to infants. In India also, there have been sporadic news paper reports of deaths soon after administration of the new vaccine, including the unexplained death of twins, a day after vaccination (2).
Serious adverse reactions following vaccination are very rare. Miller and colleagues note how-ever that because a very large number are vaccinated, coincidental adverse events including deaths due to sudden-infant-death-syndrome (SIDS) that are temporally associated with vac-cination, do occur (3). It is said that the deaths associated with PV are merely coincidental SIDS deaths, associated temporally with immunization and they are unrelated to vaccination and the vaccine used (4).
Another factor that must be borne in mind is that the Government of India improved its sys-tems for surveillance of AEFI and developed a detailed standard operating procedure (SOP) manual for AEFI in 2010 (5). Details of this time-bound reporting system are quoted below.
Reporting and Investigation of AEFI
Serious AEFIs are defined as those that are life threatening and those that result in hospitaliza-tion , disability (or have the potential to result in disability) or death (5).
At the field level, Auxiliary Nurse Midwives (ANMs), Health Assistant (HAs) and other field lev-el health workers and Medical Officers (MO) of Primary Health Centers (PHC) are expected en-quire about and monitor the occurrence of AEFI. In case of a serious AEFI the MO (PHC) is ex-pected to be informed by telephone immediately. He/she has then to initiate an investigation personally to verify the facts, fill the first information report (FIR) and have it sent to the Dis-trict Immunization Officer (DIO) within 24 hours and also inform the DIO by telephone or fax immediately. The incriminated vial of vaccine and syringe used to administer the vaccine are collected and sent under cold chain requirements to DIO and finally to Central Research Insti-tute (CRI) Kasauli for laboratory investigation (5).
The DIO initiates an investigation and files a preliminary investigation report (PIR) and detailed investigation form (DIR). The FIR is sent to the Assistant Commissioner Universal Immuniza-tion Programme ( AC UIP) within 24 hrs, the PIR within 7 days and the DIR within 90 days. Deaths are investigated by the Regional Investigation Team (RIT) which team is informed through the State EPI Officer (SEPIO) by telephone or fax (5).
In the event of a death, the RIT is expected to make an onsite investigation and then file a pre-liminary report available to the SEPIO within 72 hours. The final report is readied within a rea-sonable time (3 months) period after completing necessary tests and detailed investigations (5).
State Expert Committee recommends cases for expert review and causality assessment. The Expert Review and Causality Assessment Team review individual serious and unusual AEFIs in order to assess a potential causal link between the event and the vaccine. The committee meets at least twice a year to review the serious and unusual AEFI (5).
Better reporting of AEFI through implementation of this new SOP could have contributed to the perception that there is an increased incidence of deaths after the introduction of PV. The SIDS rate in India is not known. For this study we assumed that all deaths within 72 hours of receiv-ing DPT are SIDS deaths. We hypothesize that if there is a significantly higher rate of deaths after PV compared to DPT administered to other children contemporaneously in the same state, the increased rate of deaths cannot be attributed to the natural rate of SIDS but may be related to PV. In each state we looked at the deaths following DPT only after PV was intro-duced in the state to ensure that all the AEFI deaths were being reported using the same SOP.
Materials and Methods
Data on AEFI deaths occurring within 72 hours after vaccine administration reported to the government surveillance system from April 2012 to 14 May 2016 and the numbers of infants (0 to 11 months old) vaccinated from April 2012 up to 29 March 2016 (as on 9 April 2016) were obtained from the Ministry of Health and Family Welfare (MoH&FW) of the Government of India under the Right to Information Act (RTI) 2005. The RTI reply is posted online (6). Data from the Health Information Management System (HIMS) of the MoH&FW on number of chil-dren vaccinated with DPT and PV were extracted to Microsoft Excel spreadsheets and the raw data used for the analysis is posted online (7). We utilized data on DPT and PV from states af-ter PV was introduced and as it was being phased-in, so some children in the state were re-ceiving the DPT and others were getting PV. If a state introduced PV in 2014, then data on DPT doses and deaths following vaccination were noted from that year on. We assume that within the state, the areas selected by the Government for early introduction of the vaccine was a matter of convenience and the underlying SIDS rate was the same in all areas of the state. There is no evidence to suggest that areas within the state with higher SIDS reporting were selected for early introduction of PV.
Statistical analysis was done with MedCalc 14 v14.8.1 (MedCalc Software bvba). The 95% CI was determined for rates. Comparison between groups of categorical data was carried out us-ing chi-square test. Correlation coefficient 'r' was examined for correlations and its p-value was determined for significance. A p-value <0.05 was considered as statistically significant.
Approximately 45 million infants (the actual figure 44936653 infants, was used in the analysis) received DPT vaccination and approximately 25 million (actual figures 24803770) were admin-istered 3 doses of PV. Two hundred seventeen infants died after DPT and 237 died following PV (Table 1). There were 4.8 deaths per million vaccinated with DPT (95% CI 4.2-5.5) and 9.6 deaths (95% CI 8.4-10.8) per million vaccinated with PV (Odds ratio 1.98 (95% CI 1.65 to 2.38). There were 4.7 additional deaths (95% CI: 3.5-5.9) within 72 hours, per million vaccinated with PV instead of DPT (p<0.0001).
There were wide differences between states in the AEFI death rates reported. States reporting higher rates of death with DPT were also the ones reporting more deaths with PV (correlation coefficient r= 0.65; p<0.0001 Spearman's rho 0.142 p= 0.146)). AEFI death rates with PV in the different states ranged from 0 going up to 430 deaths per million vaccinated. (Table 2)
It is often difficult to say whether a death soon after immunization is caused by the vaccine or is a coincidental event. To overcome this problem we have compared deaths following immun-ization with one or the other of the two vaccines given at the same age, in a large cohort of babies. The rate of coincidental deaths will be similar no matter what vaccine is given. The study assumes that all deaths following DPT are coincidental SIDS deaths and that even if no vaccine was given on that day; these children would have died anyway. We hypothesis further, that if PV does not result in deaths, there would be no increase in the death rates in children given this vaccine.
We have, in this analysis, looked at AEFI deaths within 72 hours of vaccination. Not all AEFI deaths occur within 72 hour and our calculations underestimate the total deaths from AEFI. However if the window period is enlarged there is an increased chance that more deaths unre-lated to vaccination would be included. We took the cue from the TOKEN study (8) looking at sudden infant deaths following Pentavalent vaccine which found significantly more deaths within the first 72 hours. It is a weakness of this analysis that we are not able to capture all AEFI deaths from the vaccine. It merely examines if there significant increase in deaths follow-ing vaccination with one of the vaccines within a small window period.
Extrapolating the data, using the mean values for the excess deaths with PV, we estimate that if the birth cohort of 26 million in India is vaccinated each year, there is would to be 122 addi-tional deaths (95% CI: 101-145) within 72 hours, due to the switch from DPT to PV.
There was great variation in the death rates reported from different states. This could reflect a real difference in susceptibility to AEFI in different states or it could be that some states record these deaths more meticulously. If we reject the conclusion that the large differences seen in the AEFI rate are related to differences in local susceptibility, we have to accept the possibility that it relates to poor recording of AEFI in some states. It follows that nation-wide projections must be made based on figures available from states with the best reporting. The analysis shows there is likely to be 7020 to 8190 deaths from the vaccine each year if data from states with the better reporting namely Manipur and Chandigarh are projected nationwide.
Most of these deaths have been reviewed by the Expert Review and Causality Assessment Team and none of the deaths were deemed as 'Consistent causal association to immunization: A1 Vaccine product-related reaction'. According to the revised WHO causality assessment manual of AEFI (9), only 'reactions for which there evidence in the literature that the vac-cine(s) may cause the reported event even if administered correctly' are classified as 'Con-sistent causal association to immunization: A1 Vaccine product-related reaction' (See Figure 1 for the algorithm and Figure 2 for the AEFI causality assessment classification for a single case) The product insert for DPT (10) and PV (11) do not report death as one of the adverse reactions and so it is not surprising that none of these deaths investigated are recorded as 'A1 Vaccine product-related reaction (As per published literature)'
Previous reports of Pentavalent AEFI Deaths
A pentavalent vaccine Quinvaxem (Crucell) was introduced in Sri Lanka on 1 January 2008 (1). This was followed by 3 deaths which were 'probably' caused by the vaccine where the WHO team of experts investigating the death reported that there was no alternate explanation for the events. The team however classified the deaths as 'unlikely' to be related to the vaccine (12).
The AEFI Manual Revised
Following this, in March 2013 the "User Manual for AEFI" was revised acknowledging that most of its concepts and definitions were adapted from 'Definitions and Application of Terms for Vaccine Pharmacovigilance Report of CIOMS/WHO Working Group on Pharmacovigilance' (13). The CIOMS/WHO document on page 170, 'Notes for Guidelines', states: "If there is ade-quate evidence that an event does not meet a case definition, such an event should be rejected and should be reported as 'Not a case of [AEFI]' (13).
Soon after the revised WHO AEFI causality assessment scheme was published, on May 4, 2013 (9), the Ministry of Health of Viet Nam suspended the use of Quinvaxem (Crucell) after it had caused 12 deaths (14). WHO experts investigated the Viet Nam deaths. This time they report-ed "Quinvaxem was prequalified by WHO. ... no fatal adverse event following immunisation (AEFI) has ever been associated with this vaccine" (15). This is the same brand of PV which was used in Sri Lanka where 3 death occurred (12). Using the revised AEFI causality assess-ment, the Sri Lankan deaths could be re-classified as 'Not a case of [AEFI]'. After that, the WHO 'Safety of Quinvaxem report' could state 'no fatal adverse event following immunisation (AEFI) has ever been associated with this vaccine'.
TOKEN Study of Deaths with Pentavalent Vaccine
An epidemiological study investigating PV link to unexplained sudden unexpected death (uSUD) of children between their 2 months and 2 years - is available in the TOKEN report (8). vonKries had previously found a statistically significantly increased standardised mortality ratio (SMR) within two days after vaccination with hexavalent vaccines (Hexavac®) (16) and the TOKEN study was done to confirm or refute the association of uSUD with PV and Hexavalent vaccines. This study, using exploratory analyses, indicated an elevated uSUD risk after PV vaccination but not after hexavalent vaccination. However, despite its rigorous methodology, the TOKEN study suffered from serious statistical and methodological limitations according to the authors and hence its findings should be interpreted with caution (8).
In spite of the data presented in this paper from a large cohort, it must be pointed out that the evidence is merely circumstantial and not conclusive. The precautionary principle states that "if an action or policy has a suspected risk of causing harm to the public, even if there is no scien-tific consensus, there is a social responsibility to protect the public from exposure to harm" (17) There are statutory obligations, for example under Article 2 European Convention on Hu-man Rights (Art 2 ECHR), "to establish framework of laws, precautions, and means of en-forcement which will, to the greatest extend reasonably practicable, protect life". A prospec-tive rigorous review of the deaths following PV is called for.
Strengths and Limitation of the study
The strength of our study is that it is based on a large, population-based cohort. Such an analy-sis has the potential to provide an accurate picture of AEFI. However, using this dataset also contributed to the limitations of the study. One important limitation of our data is that it is de-pendent on the reporting by health workers, many of whom have very limited education and literacy.
The government database does not also provide the exact ages when the infants were admin-istered the vaccinations. It is known that the SIDS rate is lower as the children become older. It is recommended that 3 doses of the vaccine are given at monthly intervals after 6 weeks of age. It is not known if there is any variation between states in compliance with this schedule and whether that is responsible for the differences seen but this seems unlikely. Further, the study looks only at a short term increase in deaths (within three days of vaccination) but it does not calculate the potential benefits on infant mortality, for example by protection against le-thal diseases like Haemophilus influenza.
In this study we consider PV as a single entity and it is not possible in this analysis to ascertain what component, whether vaccine antigen or additives or combination of these agents are re-sponsible for the deaths.
Vaccination Practices in Developed Countries
Most developed countries use the acellular pertusis vaccine in the DPT (DaPT) and in other combination vaccines. A Cochrane review of combined DTP-Hepatitis B-Hib vaccine found less immunological response to the combined vaccine than when they were administered separate-ly and there were more local reactions to the combined vaccine (18). Therefore this combina-tion vaccine is not used in the USA. The Vaccine Adverse Event Reporting System (VAERS) data in the USA suggests there have not been any serious adverse events with these vaccines given separately (19).
For developing countries too, it can be argued that the same benefit against lethal diseases can be achieved if the vaccines DPT, Hib and hepatitis B are given separately as it is done in the USA, and it may be safer.
This study has demonstrated a probable increase in sudden unexplained deaths within 72 hours of administering PV compared to DPT vaccine. Whether improvement in AEFI surveillance sys-tems or other hitherto unstudied/unrecognized factors contributed to this increase cannot be ascertained from this study. These findings do not warrant deviation from current vaccination schedule but the differential death rates with DPT and PV do call for further rigorous prospec-tive population-based investigation.
1. Global Advisory Committee on Vaccine Safety review of pentavalent vaccine safety concerns in four Asian countries. Report from the 12 June 2013 meeting. Available at http://www.who.int/vaccine_safety/committee/topics/hpv/GACVSstatement_pentavalent_June2013.pdf Accessed on 20 May 2016
2. Special Correspondent. Infant twins die a day after vaccination. The Hindu. December 5, 2015 available at
http://www.thehindu.com/news/cities/bangalore/infant-twins-die-a-day-after-vaccination/article7950650.ece Accessed on 1/8/16
3. Miller ER, Moro PL, Cano M, Shimabukuro TT. Deaths following vaccination: What does the evidence show? Vaccine. 2015;33:3288-92.
4. Deshmukh V, Lahariya C, Krishnamurthy S, Das MK, Pandey RM, Arora NK. Taken to health care provider or not, under-five children die of preventable causes: Findings from cross-sectional survey and social autopsy in rural India. Indian J Community Med. 2016;41:108-19.
5. Government of India. Standard operating procedure (SOP) for investigation of ad-verse events following immunization (AEFI). 2010. Available at searo.who.int/india/topics/routine_immunization/AEFI_standard_operating_procedures_SOPs_2010.pdf Accessed on 1/9/16
6. Ministry of Health and Family Welfare Government of India. RTI reply Z33013/01/2016-Imm(Pt) dated 3 May 2016. Available at http://bit.ly/RTIReply Ac-cessed on 5/8/16Ministry of Health and Family Welfare Government of India. Health Management Information System Reports. Available at https://nrhm-mis.nic.in/hmisreports/frmstandard_reports.aspx Accessed on 1/8/16
7. Kaur J, Puliyel A,Puliyel J. Data extracted in Excel sheets used for analysis. Available at http://bit.ly/DPTPentaData. Accessed on 5/8/16
8. Schlaud M, Poethko-Müller C, Kuhnert R, Hecker H. TOKEN Study on deaths in young children (2nd to 24th month of life) Study Report. Available at http://www.rki.de/EN/Content/Health_Monitoring/Projects/TOKEN_Study/Studyreport.pdf?__blob=publicationFile Accessed on 1/8/16
9. WHO. The Causality assessment of adverse event following immunization User manual for the revised WHO classification WHO/HIS/EMP/QSS. March 2013 http://www.who.int/vaccine_safety/publications/aevi_manual.pdf?ua=1
10. Serum Institute of India. Triple antigen. www.vaxinpoint.in/wp-content/uploads/2014/02/Triple-antigen.pdf Accessed on 14 May 2017
11. Sanofi Pasteur Ltd. Pentaxim Summary of product characteristic. s fda.moph.go.th/zone_search/files/2C_22_47_N.pdf accessed on 14 May 2017
12. Saxena KB, Banerji D, Qadeer I, Kurian NJ, Priya R, Shiva M, Puliyel J, Dabade G. "An-tivaccine lobby" replies to the BMJ. BMJ. 2010;341:c4001
13. Definitions and Application of Terms for Vaccine Pharmacovigilance CIOMS/WHO Definitions and Application of Terms for Vaccine Pharmacovigilance Report of CIOMS/WHO Working Group on Vaccine Pharmacovigi-lancehttp://www.who.int/vaccine_safety/initiative/tools/CIOMS_report_WG_vaccine.pdf
14. TuoiTre News. Vietnam suspends Quinvaxem vaccine following 9 deaths. TuoiTre News [Internet]. 2013 May 5 [cited 2013 Jun 22]. Available from: http://tuoitrenews.vn/society/9330/vietnam-suspends-quinvaxem-vaccine-following-9-deaths accessed on 1/11/15
15. WHO. Safety of Quinvaxem (DTwP-HepB-Hib) pentavalent vaccine. 10 May 2013. http://www.who.int/immunization_standards/vaccine_quality/quinvaxem_pqnote_may2013/en/ Accessed on 23 February 2017
16. Von Kries R, ToschkeAM, Strassburger K, Kundi M, Kalies H, Nennstiel U, Jorch G, Rosenbauer J, Giani G. Sudden and unexpected deaths after administration of hexa-valent vaccine (diphtheria, tetanus, pertussis, poliomyelitis,hepatitis, Haemophilus influenza type b): is there a signal? Eur J Pediatr. 2005;164:61-9.
17. Science and environmental health network. Precautionary Principle Wingspread Con-ference on the Precautionary Principle January 26, 1998 Available at http://www.sehn.org/wing.html Accessed on 23 May 2016
18. Bar-On ES, Goldberg E, Fraser A, Vidal L, Hellmann S, Leibovici L. Combined DTP-HBV-HIB vaccine versus separately administered DTP-HBV and HIB vaccines for pri-mary prevention of diphtheria, tetanus, pertussis, hepatitis B and Haemophilus influ-enzae B (HIB). Cochrane Database of Systematic Reviews 2009, Issue 3. [DOI: 10.1002/14651858.CD005530.pub2]
19. CDC. Safety Information About Specific Vaccines
Available at https://www.cdc.gov/vaccinesafety/vaccines/ Accessed on 25 May 2017