The Vaccine Conundrum

by

Preventive or Promotive?

Prevention in public health is defined as a “call for action in advance, based on knowledge of natural history and the social context of disease occurrence in order to make it improbable that the disease will progress subsequently,” according to Leavell and Clarck (qtd in Czeresnia 1999: 705). The social context of disease occurrence and the risk of contracting the disease when interventions are non-existent are the two important and related characteristics inherent in the concept of prevention. Interventions that prevent specific diseases during epidemic situations by changing the behaviour of people through organised community efforts are deemed to be “preventive” interventions.

  • Historically, John Snow’s classical intervention of preventing cholera during the epidemic by altering the source of drinking water was a preventive intervention, which in the context of the non-endemic situation has transformed into a health promotion activity, as is the case of ensuring safe drinking water now.
  • Health “promotion” is defined as “measures that are not directed to a given disease or disorder, but serve to increase overall health and well-being” (Czeresnia 1999: 705). What is implied in the definition of both these concepts is the importance of context, namely the prevalence of the problem against which interventions are developed and not merely whether the interventions are targeted at one or more diseases at a time. The classic experiences in public health demonstrate this and hence the case of vaccination needs further deliberation.
  • The context of the introduction of newer vaccines reveals that they are meant for preventing only one sub-category of a major disease, whose proportionate prevalence and the case fatality are not very high, thanks to the overall social development and improved coverage and advancement in medical therapies.
  • Newer vaccines intend to safeguard populations from one subcategory of a disease caused by a specific infectious agent. This is a clear departure from the vaccines historically introduced to target specific diseases. For instance, BCG (Bacillus Calmette–Guérin) vaccine for TB, DPT vaccine against diphtheria, pertussis, and tetanus and measles and polio vaccines for those diseases.
  • The current logic of newer vaccines is to reduce susceptibility towards specific strain of a virus or a bacterium, the infectious agent, which can protect from one specific type of the “parent”3 disease. For instance, the Hib (Haemophilus influenzae type b) vaccine can protect only from Hib induced influenza, though all forms of influenza leading to meningitis and pneumonia are usually projected as the one targeted by the vaccine.
  • The important aspect to note here is that not all influenza is caused due to Hib bacterium and not all Hib induced influenza leads to meningitis (Bajpai and Saraya 2012; DHR and ICMR 2010). In other words, it becomes extremely difficult to identify those infections which can be exclusively attributed to Hib vaccine in a situation where several forms of influenza exist in the population with similar clinical presentations.
  • A more technology-driven diagnostic mechanism is needed even to identify the specific types. This could possibly be an extension of “laboratory medicine” in public health, which in medicine is characterised by the domination of laboratory parameters in every facet of medical care.
  • The same is true for rotavirus vaccine, as not all diarrhoeas among children can be attributed to the rotavirus, posing a serious limitation in evaluating the impact of newer vaccines. In other words, for each newer vaccine introduced, there exist a “parent” disease of which only one subcategory will be prevented through the vaccine.
  • This also raises another challenge, that is, mere rise and fall of any “parent” disease, say, diarrhoea or influenza cannot be attributed to the success or failure of a vaccine, as the very fact that only one variant of the “parent” disease could be attributed to a vaccine whose proportionate contribution to the “parent” disease becomes significant.
  • This was evident in one of the rotavirus vaccine trial carried out in Niger, which reported an increase in number of cases of diarrhoea among the cases than the control group, a contradictory finding, which has spurt several controversies on the capacity of rotavirus to reduce diarrhoea among children (Isanaka et al 2017; Puliyel 2017).
  • There are criticisms that population prevalence of those diseases for which newer vaccines are introduced are either unknown or indicates a very low prevalence as compared to several other diseases. For instance, the population prevalence of Hib influenza is estimated to be around 0.007% (Gupta and Puliyel 2009).
  • Similar is the case with rotavirus-induced diarrhoea, as there is serious disagreement among experts on the actual population prevalence and the deaths caused due to the same (Bhan et al 2014; Puliyel 2014). This is partly due to the error in some of the estimates that attribute all forms of severe diarrhoea among children that got admitted in hospitals to rotavirus.
  • These estimates range from 18% to 39% among children (Bajpai and Saraya 2012; Banerjee et al 2006; Bhan et al 2014). Thus, for those diseases against which newer vaccines are introduced, we neither have adequate prevalence data at the population level nor an estimate of their proportionate contribution to the “parent” diseases.
  • In other words, there was never an adequate effort to examine the population prevalence of infections like Hib influenza out of total influenza, or the proportion of pneumococcal infections out of all forms of pneumonia or the proportion of those diarrhoeas that is attributable to rotavirus and so on.
  • Instead, most of the estimates rely on specific hospital-based data to represent the population parameter, which is a serious methodological error in public health, especially in a context where the utilisation of healthcare service is low and random. For instance, according to an estimated prevalence of rotavirus diarrhoea to total diarrhoea cases, the prevalence was 7% from a community-based study, whereas it was 27% from hospital-based data (Banerjee et al 2006).
  • It is a well-known fact that community-based prevalence of any disease will be higher than the hospital-based prevalence as not all cases from the population will get reported in healthcare facilities as the latter depends on the extent of utilisation. Hence, the population-based prevalence of diseases is used for efficient programme planning.

Herd Immunity

  1. It is the inadequacies of population-based prevalence that poses a major challenge while evaluating mass immunisation programmes. As mentioned earlier, it is the prospect of herd immunity that qualifies vaccination as a public health intervention. Herd immunity is dependent upon three major factors:
  2. the reproduction of disease in a population (R0 or the basic reproduction number), which is a product of the prevalence of the disease in a population and its infective rate, the latter further depends on the context of potential human interaction possible in any society;
  3. the vaccine efficacy and the population covered through mass vaccination drives; and (iii) the extent of “natural immunity” prevalent in a population towards the said disease (Fine et al 2011).
  4. This has also resulted in serious controversies4 in deciding the threshold coverage necessary for attaining herd immunity for various diseases; it is considered high for measles characterised by its high prevalence and greater infectivity (Fine 1993; Fox 1983). This threshold is generally considered greater in the case of airborne infections as compared to vector-borne infections.
  5. Another important aspect which is not given adequate consideration is the extent of natural immunity that populations acquire by getting exposed to the same microbes or closely similar species of microbes during their life course. For instance, a study in Kerala population among antenatal women shows that the prevalence of antibody5 of rubella among unvaccinated women was 94.3% (Jayakrishnan 2016a) and among adolescent girls, it was reported around 68.3% (Jayakrishanan 2016b).
  6. What are the potential inferences possible? The inference drawn by the study was that this is a marker of prior exposure to rubella
    virus of the pregnant women and hence posed a risk of getting congenital rubella syndrome (CRS) by their children.
  7. It is also possible to argue that if there are already antibodies developed against rubella by un-vaccinated women, is there not a possibility of “natural immunity” that exists in the society due to prior exposure or due to their living conditions? As the study also reported that all the respondents were free of any specific clinical symptoms significant to rubella, it indicates that the population was free of the disease.
  8. A similar situation has been identified for Hib disease where there is a possibility of “natural immunity” that exists among population due to prior infections with bacteria with cross-reacting antigen (Puliyel et al 2001). One of the possible reasons for the low prevalence of Hib influenza is attributed to this feature.
  9. This feature of natural immunity that exists among populations needs serious investigations, as this could be another important factor that shall decide the need for newer vaccines at a population level. It is necessary to examine this component of herd immunity more closely and critically. Instead, using data on the high prevalence of antibodies among the unvaccinated population as a justification for introducing rubella vaccine by claiming prior exposure to the disease is unethical, as the outcome expected to achieve after immunisation is also a form of immunity (acquired).
  10. Additionally, scholars have also cautioned based on the rubella vaccine experience in the United States and the United Kingdom that a vaccination coverage lower than the threshold coverage necessary for rubella control can lead to an increase in the cases rather than its reduction. This threshold level for high prevalent African regions is estimated to be around 90% as the safe limit so that the cases will not increase further due to vaccination (Fine 1993).
  11. This can only be monitored when the baseline information indicating the current population prevalence of rubella disease is available. There is a gross lack of evidence in this regard. Any future attempt to evaluate the efficacy of mass immunisation programmes need to rely on this information. Hence, it becomes the responsibility of the governments to generate baseline information on the population prevalence of those specific diseases against which newer vaccines are being introduced.

Conclusions

  1. Introduction of newer vaccines at a time when the population prevalence of those specific diseases is low, transforms vaccination intervention from a “preventive” intervention to a “promotive” one.
  2. The transformation is fuelled by the fact that disease-specific prevention vanishes and is replaced by the protection of individuals from “strain-specific infections,” and the latter calls for sophisticated laboratory support. This becomes a challenge in a context when even basic laboratory services are lacking in the healthcare system.
  3. This further constraints any attempt to estimate the exact prevalence of the infectious agent and case fatality rates due to specific causal agents, which in turn restrict the prospective evaluation of immunisation outcomes.
  4. The success or failure of public health interventions like immunisation needs to be evaluated based on an interdisciplinary approach guided by the principles of public health, namely social justice, population, and prevention. This calls for a critical engagement with the logic of introduction of vaccines from biomedical, public health, economic and ethical perspective.
  5. A critical inquiry and engagement that address different realms need to be considered before introducing any vaccine as a mass immunisation programme for the nation. Instead, the current challenge is that there is an inherent assumption among policymakers that biomedical logic and public health logic are similar and any critical inquiry towards vaccine from the latter perspective is generally dismissed as if it is triggered by the anti-vaccine lobby based on misconceptions towards vaccines.

Notes

1 Threshold coverage is the minimum proportion of people to be vaccinated for a specific disease to attain herd immunity for the entire population. This ranges from 70% to 99% depending on the type of causative agent, rate of infection and so on.

2 Newer vaccines imply the set of vaccines introduced post 2005, including hepatitis B, Hib influenza, rotavirus, pneumococcal vaccine and most recently measles-rubella (MR). This is in addition to the older vaccines that were part of the UIP, namely DPT, measles and polio.

3 “Parent disease” is used in this article to make a distinction between the major diseases popularly understood, namely influenza, diarrhoea, pneumonia, from the only one of its type caused by one among several of the infectious agents (microbe) as in the case of Hib, rotavirus and pneumococcal types.

4 Threshold coverage required for developing herd immunity to measles for a population was estimated based on several studies in history across several populations and was estimated at 70% to 96%, which resulted in a search for a more accurate value across all population along with controversies that argued for heterogeneity of populations.

5 Antibodies are produced in humans as a response to an exposure to a specific infectious agent (microbe). The presence of antibody in an individual can be a prior exposure to the infectious agent and does not necessarily progress to a disease earlier, instead can also lead to a protection towards that disease by acquiring immunity to that specific infectious agent.

Leave a Comment