What are the criteria to compare vaccines?

India had a golden opportunity to evaluate every serious ‘adverse events following immunisation’, but in the absence of trained public health officers, the job was not done well, leading to unanswered questions. Lack of transparency has led to lack of vaccine confidence.

At a vaccination centre in Mumbai on April 16. In public health mode, when disease control is the goal, two special phenomena are important—‘herd immunity’ and ‘herd effect’.

If there was only one vaccine againsta disease, the choice is simple, yes or no. If the vaccine is safe and efficacious, and if the disease is posing a risk, yes is the sensible choice for individuals.

The government’s health management system must make a policy decision whether or not to roll it out for community-level application. Individual choice comes under ‘health care’ mode and programmed use comes under ‘public health’ mode.

When there are two (or more) vaccines against one disease, the choice becomes complicated. Two vaccines are available against whooping cough (pertussis) and also against polio. In the case of pertussis the two vaccines are whole-cell killed bacterial vaccine and ‘acellular’ vaccine without killed bacteria, but only some components. In India we use the former in the universal immunisation programme (UIP), and rightly so, whereas many rich countries use the latter. The biological qualities of safety and efficacy are the basic criteria for choice—but acellular vaccine is more costly, bringing one more factor to consider. In vaccine assessment, shades of grey exist, not always black or white.

The essential criteria for licensing a vaccine by the regulatory agency are, again, safety and efficacy. For the government to develop a policy to roll out a licensed vaccine in the community, in other words vaccination programme, there are three major elements to consider—epidemiology (is disease prevalent, how prevalent?), economics (cost per dose in health care, programme’s cost in public health) and ethics (biomedical ethics in health care, public health ethics in programme mode). Immunisation programme officers understand safety and efficacy, but not necessarily ethics, which is frequently violated with serious consequences.

Examples and principles of choice

Killed bacterial pertussis vaccine has higher and more durable efficacy; this is the reason for its preference in India in spite of more reactogenicity. Acellular vaccine is less reactogenic, the reason for its preference elsewhere.

Immune response is a silent physiological process, response being to the antigen(s) in the vaccine, the real stuff. Reactions (‘adverse events following immunisation’, AEFI) are mostly to other ingredients in the vaccine, put in there for keeping the antigen safe and bio-available, or as an ‘adjunct’, called ‘adjuvant’ in immunology, to provoke the immune system to respond vigorously.

Antigens can be separated from bacterial cells—that is how acellular vaccine is made. Bacterial cells are the main reason for AEFI to killed bacterial pertussis vaccine—pain, fever, and rarely prolonged inconsolable crying by the infant—none serious, all short, self-limited and inconsequential.

The lower efficacy of acellular vaccine can be overcome with additional booster doses during school age/adolescence. If so desired, the primary course in infancy may be with killed bacteria and boosters with acellular. Either vaccine can boost immune responses induced by the other.

Against polio there are the inactivated (killed) virus vaccine (IPV, Salk) and the live virus vaccine, as oral drops (OPV, Sabin). The choice between them is based on biological criteria of safety and efficacy, and additional practical issues of cost and ease of administration. IPV is injected, completely safe and exquisitely efficacious, no child ever having developed breakthrough polio after three doses, but it is about ten times costlier (per dose) than OPV.

The less expensive OPV is neither completely safe nor highly effective, but it is far cheaper than IPV, and very easily given to children anywhere, not only in clinics, by anybody, even untrained volunteers. While all rich countries use IPV exclusively, India’s UIP (like in most low-income countries) uses OPV. The regulatory agency has licensed IPV and it is available in the private market. Because of low per-dose efficacy, OPV has to be given repeatedly, some 10-15 times, but in UIP only five contacts are scheduled between health workers and children during infancy. So, additional annual pulsing is resorted to. Pulse campaigns are expensive—one OPV pulse campaign is roughly as costly as one IPV dose for all children. So the economic advantage was actually fallacious.

On the safety side: OPV causes, on rare occasions, polio, which the UIP counts only as AEFI, which it is by definition. Rich countries call the rare polio as ‘vaccine-associated paralytic polio’ (VAPP) and they all had switched to IPV to avoid VAPP, either prior to the year 2000 or soon thereafter since 2000 was the target year for eradication; any polio beyond 2000 was not acceptable. The continued programmatic use of OPV in low-income countries, when IPV is available, contravenes public health ethics.

When natural polio was prevalent, there was a different argument, that of benefit-risk assessment. In the 1980s, natural polio in India was highly prevalent, with 500 to 1,000 cases per day. The risk of natural wild virus polio was by far very much higher than that of VAPP; in other words, the benefit of averted wild virus polio was very high in comparison with the tiny risk of VAPP. Benefit/risk ratio was highly favourable. When natural polio was eliminated in 2011, benefit/risk ratio became highly unfavourable—benefit disappeared but risk remained. VAPP in the absence of natural polio is unethical, the reason why all rich countries abandoned OPV.

Now benefit/risk was no longer a criterion for choice. But we must compare the two vaccines IPV and OPV, and such comparison study is called in epidemiology, ‘non-inferiority’ trial. There has never been a non-inferiority vaccine trial between IPV and OPV, but we know what such a trial would have shown: the superiority of IPV over OPV for efficacy. So an unsafe and efficacy-wise inferior vaccine remains in use.

These kinds of problems illustrate why experts in public health ethics, health economics and theoretical/practical epidemiology are essential in national programmes—in UIP, tuberculosis control and COVID-19 pandemic, etc. But they had to be budgeted, pushing the health budget up. ‘Paisa wise, rupee foolish’ is the culture of poverty. To keep health budget minimal, we do not employ experts in the three disciplines, but make decisions that are weak on science and strong on polemics and rhetoric.

The purposes of vaccination

Vaccination can be purposed for individual benefit of preventing disease and death. This is what is meant by health-care mode, vaccination given as ‘preventive medicine’. There is high demand and huge market for vaccination of children in the private health-care market, parents paying for vaccines and service. Well-to-do families have one child or two children; protecting their health and life is taken seriously. Major vaccination centres are in hospitals, and they provide all licensed vaccines, those on the UIP list and those outside it. The vaccination clinic of CMC Vellore (genuinely not-for-profit) runs throughout the daily working hours, to keep up with demand—annually over one lakh children are vaccinated, 350-400 per day. During pandemic year 2020, the number fell to 70,000 and is now picking up.

The goal of UIP is much broader than that of health care: to “control” the burden (‘incidence rate’ in epidemiology) of selected childhood diseases. Disease control is not mere disease prevention at the individual level, but is a ‘public health’ goal and activity, defined as “ disease reduction to a pre-determined level by vaccination, within a pre-determined time period, and quantitatively monitored for success ”. No disease other than neonatal tetanus and polio is truly under control mode in India because the three highlighted fundamentals are missing for every other disease.

Eradication is the ultimate control, when zero case level is achieved and sustained globally. The time target for global polio eradication was the year 2000. India reached the goal only in 2011. At global level there are problems still—natural polio is present in Pakistan and Afghanistan, but OPV-caused polio is more widespread. VAPP is not counted anywhere, swept under the carpet. But ‘mutant variants’ derived from OPV viruses (called ‘circulating vaccine-derived polioviruses’, cVDPV) are also causing polio outbreaks in Pakistan and Afghanistan and, what is more disturbing, sin more than 20 African countries. In 2020, globally 90 per cent of all polio was cVDPV-caused and 10 per cent natural. Such is the embarrassing problem of the ‘culture of poverty’, choosing the cheaper vaccine and neglecting public health ethics, by too many low-income countries, all guided by rich country experts.

The other diseases under UIP’s control efforts in India (vaccines in parenthesis) include childhood tuberculosis (BCG), diphtheria, pertussis, tetanus (DPT), measles, rubella (MR), hepatitis B (HB vaccine), Haemophilus and Pneumococcal pneumonia and meningitis (Hib and Pneumo vaccines), rotavirus diarrhoea (rotavirus vaccine) and Japanese encephalitis (JE vaccine). All vaccines are offered free in government health-care centres, from sub-centres and primary health centres, right up to medical colleges. If you want non-UIP vaccines against mumps, chickenpox, viral hepatitis A, influenza or typhoid fever, or IPV or human papilloma virus vaccine to prevent cervical cancer, you have to go to private clinics. Public health is for equity, but how easy it is to create inequity in the name of public health, contrary to the spirit of public health.

Although UIP’s goal is control of all diseases under UIP list of vaccines, we do not practice case counting by public health surveillance and cannot quantitatively monitor disease reduction. Public health surveillance is when law demands that every doctor must report every case of any vaccine-preventable disease to a local public health office, and every district has a public health officer, supervised by the State public health officer. All these are avoided to keep expenditure at a minimum. So, UIP is essentially a vaccine delivery platform—and as a result, all vaccine-preventable diseases are still prevalent, some even causing outbreaks occasionally. We do not know if we are getting our money’s worth of benefits in UIP.

Obviously health care and public health objectives are not mutually exclusive but complementary. In public health mode, when disease control is the goal, two special phenomena are important—‘herd immunity’ and ‘herd effect’. The per cent of the total eligible population vaccinated (and protected) is herd immunity. When a majority is immunised, building herd immunity, the spread of infection slows down both in the vaccinated segment of the population and in the unvaccinated segment also—overall with much lower disease burden than before. The effect in the unvaccinated is called herd protective effect, or simply herd effect. Obviously, herd effect applies only to human-to-human transmitted diseases, spread educed by vaccine immunity. These concepts are relevant in today’s context of COVID-19 epidemic and vaccine-rollout, as we will see presently.

Fast forward to the pandemic

Public knowledge and understanding about the birthing process of vaccines have reached an advanced level, thanks to our ‘ring-side’ view of the unfolding events of COVID-19 vaccines, nationally and globally. In India at present we have two vaccines available, one created by Oxford University and manufactured (as Covishield) by Serum Institute of India, after technology transfer and contractual agreement. Without Oxford inviting SII, would we have made-in-India Covishield?

Bharat Biotech had experience with inactivated virus vaccine technology, and, for abundant caution for staff-safety, had built a Biosafety Level 3 unit, completed in 2019. When the SARS-CoV-2 pandemic hit India, BB ventured to make an inactivated virus vaccine, taking the risk of failure and financial loss. A unique adjuvant was added to enhance T-cell and antibody responses simultaneously. One investment step for a company, one giant leap for India.

India should/could have established a National Vaccine Task Force as soon as we knew about the pandemic, in February 2020. All players could have been networked and moulded for a national non-competitive but collaborative project, like the man-on-the moon mission— with the sole goal of making vaccines of different technologies, with common laboratory facilities to expedite standardised testing, one massive set of Phase I, II and III trials with one protocol and multiple trial sites distributed widely nationally—one arm for placebo against multiple arms for vaccine candidates. We should have shown India’s intellectual and scientific prowess to the world. India could have justified its reputation as the pharmacy of the world. We owed it to all developing countries less endowed and fortunate than India. But India did not.

No point of crying over spilled milk. What now? We have two vaccines. Are they equal and on par? How do we compare Covishield and Covaxin? Should we compare at all? On January 3, 2021, the regulatory agency gave ‘emergency use authorisation/approval’ for both vaccines. So the government became the vaccine-promoting agency, and ipso facto , the vaccine rollout became in public health mode with all its attendant principles described above.

Safety, efficacy, herd immunity, herd effect, vaccine effectiveness

During Phase II/III trials, the sponsoring companies were responsible for monitoring AEFI and vaccine efficacy. Now these responsibilities belong to the government. The government is equally responsible to all citizens. When vaccination is in public health mode, benefits and risks had to be clearly quantified and equally shared. ‘Effectiveness’ and safety had to be monitored by the government through its agencies.

The performance of each vaccine had to be measured—by quantifying vaccine effectiveness. Efficacy pertains to vaccination in trial participants, calculated from Phase III trial data. The overall disease reduction due to vaccination programme, due to herd immunity plus herd effect, is called effectiveness. Theoretically effectiveness is always higher than efficacy. The government ought to have data on the effectiveness of both vaccines—as all beneficiaries are computer documented with Aadhaar card numbers.

Application of public health ethics is important since neither vaccine is licensed for general use by the regulatory agency. In other words, safety and efficacy have not yet been certified, but are only provisional, to be carefully monitored to ensure no harm comes to any citizen. We have no experience in large-scale vaccination of adults, and no public health surveillance of even influenza that came as pandemic in 2009. As we saw earlier, even UIP does not have expertise in ethics, epidemiology and economics of vaccine rollout. So, the government had to have an ad hoc mechanism to evaluate serious AEFI—defined as “ health events within one month after vaccination that required hospitalisation; caused prolonged illness even if not hospitalised, particularly with any long-term disability; caused death ”. We had a golden opportunity to monitor every serious AEFI and evaluate safety, but in the absence of trained public health officers, the job was not done well, leading to many unanswered questions. Lack of transparency has led to lack of vaccine confidence.

There is a background frequency of deaths due to natural causes. However, many are sick for some time before demise. Vaccination is by self-selection and by healthy individuals. So deaths during the month after vaccination could not simply be declared unrelated to vaccination. Recently the National AEFI Committee reported some results. Although death resulted for 180 AEFI cases, only 124 had necessary information and 74 were related to blood supply problems to the heart (63) and brain (11). There were 111 deaths in the first week after vaccination as against 11 in the next three weeks (“ 180 AEFI deaths reported in India ”, The Hindu , April 9, 2021). This skewing, clustering towards the point of vaccination, is highly suspicious of a causal connection. Denmark has reported serious AEFI due to the Oxford vaccine with the frequency of one in 40,000 vaccinated persons. India is sitting on crucial data that the world needs now, but will we provide it?

Is AEFI distribution equal for both vaccines? Are we merely rolling out vaccines as in UIP, without carefully monitoring herd effect and vaccine effectiveness? If vaccine effectiveness is not quantified, and Phase III trial is interfered with (by vaccine rollout), how will the regulatory agency get enough data to licence either or both vaccines? Will India’s reputation as the pharmacy of the world be dented?

Epilogue

Just as COVID affects the body, organs and tissues several ways and cause complex pathology, the pandemic also affects the population, individuals, families, their livelihood, social, economic, educational, religious and recreational interactions in a very complex manner. The last thing we want is mistrust and suspicion that are bred by lack of relevant communications and lack of transparency on matters of concern, AEFI for example. Vaccination is the most effective, and cost-effective, tool of intervention we have against both sets of problems. Vaccine confidence must be built and reinforced and vaccination applied to mitigate both disease pathology and social pathology—government has an unprecedented challenge which is also a unique opportunity to learn by doing.

T. Jacob John is retired professor of clinical virology, CMC, Vellore.