Cover Image
close this bookImmunization Policy, 1996 (WHO - OMS, 1996, 63 p.)
View the document(introduction...)
View the document1. Introduction
View the document2. Vaccines used in the EPI
View the document3. Basic immunization schedules and strategies
View the document4. The expected effect of immunization on disease epidemiology
View the document5. Additional schedules and strategies
View the document6. HIV infection and immunization
View the document7. Reactions following immunization
View the document8. Other vaccines that can be used as a part of EPI
View the document9. References

4. The expected effect of immunization on disease epidemiology

The epidemiology of vaccine-preventable diseases changes after immunization programmes are well established and high coverage is attained. The degree of change depends on the mechanism of action of the vaccine, the level of coverage achieved, the presence of non-human hosts for the organism, and characteristics of the infectious organism.

An important issue is whether vaccines protect against infection of humans by a bacteria or virus, or whether they only reduce the severity of disease among persons who are infected by the organism. Live viral vaccines such as measles and polio vaccines, for example, confer resistance to infection (and thereby prevent disease); this means that transmission of infection is also reduced by immunization. For these vaccines, the concept of “herd immunity” is relevant, that is the indirect action of the vaccine, once a high proportion of the community is immunized, on reducing transmission of the infectious agent and thereby making it less likely that persons (including unvaccinated persons) will be exposed to the agent.

Vaccines such as pertussis, however, appear to protect the individual against severe disease, but do not confer complete protection against infection and hence have less effect on transmission of the organism. BCG vaccine is thought to have very little effect on reducing transmission of TB because it confers such variable protection against pulmonary TB.

Toxoid vaccines such as TT confer antibodies against the toxin produced by certain organisms and hence protect against disease, but would not normally be expected to protect against infection by the organism. Interestingly, diphtheria toxoid immunization has resulted in a dramatic decline in both clinical disease and carriage rates. It is thought that the vaccine may reduce transmission because infected vaccinees are usually asymptomatic, and therefore they contribute less to airborne spread of the toxigenic organism than do unvaccinated infected individuals who have a membrane and cough. Pappenheimer’s work on the molecular biology of diphtheria suggested that widespread immunization led to a decreased prevalence of the toxigenic strain of diphtheria, suggesting a mechanism for the “herd immunity” phenomenon (Pappenheimer et al 1983).

The presence of a non-human host or reservoir for the organism, such as monkeys for the forest pattern of yellow fever, means that immunization of humans may have little effect on transmission of the agent.

Vaccines that protect against infection have two major effects on the epidemiology of disease. These have perhaps been most extensively documented for measles but similar changes have also occurred with poliomyelitis and diphtheria:

· immunization changes the relative age distribution of cases, with a shift to older ages;
· outbreaks are likely to occur after some years of low incidence.

In addition, with all vaccines the following changes are likely:

· the proportion of cases of disease that occur in immunized individuals increases as coverage increases.

· antibody levels among immunized persons are often lower than among persons who acquired immunity through natural infection (an exception is tetanus immunization). This in turn will lead to lower levels of antibody transferred from mothers to their infants, with implications for the age at immunization of infants.

These main changes are discussed briefly below. There are of course other possible changes but those mentioned above have been of major programmatic importance for the EPI vaccines and are therefore discussed here.

Changing age distribution of cases. For vaccines that protect against infection and are administered in infancy, immunization of a large proportion of the community reduces transmission of the agent and reduces the chance of susceptible persons being exposed to the agent. Unimmunized children are therefore likely to reach an older age before they are exposed to the infectious agent, leading to an increase in the average age of infection. Though the proportion of cases in older children increases, the absolute number of such cases ultimately falls, due to the reduction in the overall incidence rate of the disease. The proportion of cases which occur among children below the recommended age for immunization may also increase, since this age group does not benefit directly from immunization. The number of cases will fall when very high coverage is reached, however. A shift in the age distribution of diphtheria has been seen in several developing countries recently, sometimes associated with a temporary increase in incidence in older age groups. In Yogyakarta, Indonesia, the incidence rate decreased markedly among children under five years of age after vaccine was introduced in 1977, but increased slightly in children aged 5-9 years from 1978-82 (Kim-Farley et al 1987).

The implications of changes in the age at infection depend on age-related changes in the outcome of infection. For measles, severity of disease is highest in children under 3 years of age. For rubella, which is a mild disease in children, the consequences of infection are much more serious in the childbearing years because of the risk of congenital rubella if a woman is infected during pregnancy. For poliovirus, the risk of paralysis increases with age at infection. Depending on the consequences of infection in older persons, countries may need to consider immunizing persons outside the primary target age group of the EPI once their programmes are well established (see section 5). For hepatitis B, infection during early childhood is almost always asymptomatic but leads to development of the chronic carrier state in many infants. Conversely, adult infection is often symptomatic but is less likely to progress to chronic carriage of the virus.

Outbreaks. In recent years, many outbreaks of EPI diseases have been reported in countries that have well established immunization programmes, for several reasons. As discussed above, if immunization protects against infection, it slows the rate of accumulation of susceptibles so that there may be many years of low incidence followed by a large outbreak. Measles outbreaks were reported in 1988-9 in countries or areas with coverage between 64% and 85%, such as Harare, Zimbabwe (Kambarami et al 1991), Muyinga health sector, Burundi, (Chen et al 1994) and many Latin American and Caribbean countries, after several years of low incidence. These outbreaks may involve a large proportion of older children and adults, including unimmunized persons and immunized persons who did not respond to the vaccine or whose immunity waned. There have been a number of polio outbreaks in countries with relatively high immunization coverage (64% - 87%) with 3 or more doses of OPV in the routine immunization schedule (Deming et al. 1992, Kim-Farley et al. 1984, Otten et al. 1992, PAHO 1986, Schoub et al. 1992, Sutter et al. 1991). Low seroconversion rates to the primary series of three or four doses of OPV in hot climates appeared to contribute to some of these outbreaks.

Second, there may be pockets of low coverage, which are likely to occur in certain geographic areas, such as urban slums, remote rural areas or islands, or in certain population groups, such as ethnic and racial minorities (Hersh et al. 1991), nomadic peoples (Loutan et al. 1992), or persons with religious or philosophical objections to immunization (Novotny et al. 1988). Outbreaks in such pockets have been documented for measles and polio.

At the beginning of the 1990s there were several outbreaks of diphtheria. In industrialized countries, falling levels of immunity in adults has contributed to these outbreaks, perhaps because of reduced boosting of antibody levels from declining exposure to C. diphtheriae as circulation of the wild organism is reduced. However, other factors such as declining coverage among young children, and high population movement, have also been important in Russia, the Ukraine and some Newly Independent States of the former Soviet Union (EPI 1993h, 1994a, Galazka et al. 1995a). Outbreaks in developing countries such as Algeria, China, Ecuador, Jordan, Lesotho, Sudan and Yemen (Galazka et al. 1995b), highlight the need to maintain immunity against diphtheria in all populations, and to monitor the epidemiology of diphtheria in developing countries.

Proportion of cases in immunized children. As immunization coverage increases, a higher proportion of cases occurs among immunized children as illustrated in table 8. The closer to 100% is the coverage, the more likely it is that a case will be a “vaccine failure”, ie a child who was vaccinated but whom the vaccine failed to protect. Among all cases, then, as more are due to vaccine failure and fewer are due to non-immunization, the absolute number of cases decreases whilst the proportion of immunized cases increases.

Table 8: Illustration of changes in the proportion of cases which occur in immunized children at different levels, for a hypothetical population of 100 000 children

Coverage 40%

Coverage 80%

Total number of children

100 000

100 000

Number of unvaccinated children

60 000

20 000

Number of cases in unvaccinated children

30 000

10 000

Number of vaccinated children

40 000

80 000

Number of cases in vaccinated children

2 000

4 000

Total number of cases

32 000

14 000

Proportion of total cases which are in vaccinated children



Disease incidence among unimmunized children: 50% per year
Disease incidence among immunized children: 5% per year

Antibody levels in immunized versus naturally infected adults. As discussed above, the lower antibody levels induced by immunization compared to natural infection have implications for the duration of immunity in immunized populations. As transmission of the agent decreases, antibody levels in immunized persons are less likely to be boosted by exposure to the agent, and immunity may be lost.

Another consequence of lower antibody levels in adults who were immunized as children, as compared to adults who were infected with the wild virus, is that mothers transfer less antibody to their infants, and infants therefore lose protective antibody sooner. For measles, for example, it may be possible to immunize children at an earlier age once most women of childbearing age have acquired immunity through immunization rather than disease. For most developing countries, this time has not yet arrived, since high measles immunization coverage has only been achieved relatively recently.

In tetanus, where infection with Cl. tetani does not induce immunity, the immunity level in a given age group depends only on immunization coverage. Thus, as TT immunization coverage increases in women of childbearing age, a higher proportion of neonates will have tetanus antitoxin. While this may reduce the response to the First and second doses of DPT, antibody levels after the third dose have been shown to be equal in infants whose mothers were or were not immunized against tetanus (Sangpetchsong et al. 1985).