From Infectious Disease Special Edition

Why Do Vaccine-Preventable Disease Outbreaks Occur in the US?

James D. Cherry MD, MSc; Kathleen H. Harriman PhD, MPH

In This Article


Individuals who are infected by or vaccinated against Bordetella pertussis only acquire short-term immunity. In addition to the new birth cohort of susceptible infants, there is a gradually increasing cohort of both older children and adults who once again become susceptible to pertussis. Despite high pediatric immunization rates in the United States, pertussis infections continue to occur in unvaccinated, susceptible children as well as in previously vaccinated older children, adolescents, and adults, due to primary vaccine failure and waning immunity.

Pertussis is a markedly different disease than measles and mumps, and it is far more complex.[4,5] Although it has been known for more than 2 decades that pertussis circulates in adults, this knowledge often has been overlooked. The most important fact is that reported cases of pertussis only represent the tip of the iceberg—B. pertussis causes infection and illness in all age groups, from birth through age 90 years or older. Currently, the majority of pertussis cases in adults are not diagnosed despite the fact that they are an important cause of infections in infants.[6,7]

During the 2010 epidemic in California, more than 9,000 cases of pertussis were reported, which was the highest number of cases reported in the state in more than 60 years. However, it’s possible that during the previous epidemic in 2005, when approximately 3,000 cases were reported, many more cases occurred but were undetected. Reasons for the higher number of cases detected in 2010 may include an increased awareness by the public and clinicians, along with the nearly universal use of polymerase chain reaction (PCR) testing, a diagnostic method that is far more sensitive than traditional culture. In 2004, PCR became the predominant diagnostic method for pertussis in California; in 2005, 66% of laboratory-confirmed cases were diagnosed by this method and by 2010, 94% were confirmed using PCR.[8]

As with any outbreak of a vaccine-preventable disease, the 2010 outbreak in California was caused in part by unvaccinated children; however, the contributions of this group were modest compared with the effects of vaccine failure and waning vaccine-derived immunity. Vaccine failure and waning immunity have always occurred with pertussis, but the effects likely were exaggerated in this epidemic because acellular pertussis vaccines, which have been recommended for the entire 5-dose series since 1997, are not as efficacious as the whole-cell vaccines that they replaced and the first generation of children who received only acellular vaccines had reached adolescence. The efficacy of current pertussis vaccines ranges from 60% to 70%, and this protection decreases over time following vaccination.[9] Data from the 2010 California epidemic suggest that this immunity may be waning more quickly than experts previously believed.[10]

Pertussis infection is endemic in both adolescents and adults. Although not widely recognized, pertussis occurred in adults even in the pre-vaccine era. The 2- to 5-year cycles of pertussis epidemics are caused by an accumulation of susceptible children in the population in whom pertussis is more easily recognized. The clinical manifestations of pertussis in older children, adolescents, and adults depend on the time interval since their last exposure to the pathogen. If they are exposed on an annual basis, then they will likely have asymptomatic infections.[11] If a longer period of time has elapsed since their last exposure, then this will lead to waning antibody levels and, therefore, they will have symptomatic pertussis with varying degrees of severity.

Solving the pertussis problem (ie, obtaining sufficient community immunity to interrupt transmission) is complex and currently not possible. Unlike measles, pertussis vaccination or infection does not confer lifelong immunity.[12] Because the R0for pertussis is similar to that for measles, a similar level of community immunity (ie, >90%) would be necessary to control its transmission.[1,2,3] And because pertussis immunity is constantly waning, even if immunization rates greater than 90% could be achieved, it would not be possible to sustain high levels of immunity without regular revaccination. However, revaccinating all adults, even at 10-year intervals, is extremely unlikely because the universal immunization of adults against influenza and pneumococcal disease has not been successful.[13]

A more realistic goal may be to focus on reducing the number of infant infections, because young infants have the most severe disease and comprise most fatal cases. During the 2010 pertussis epidemic, the primary goal of the California Department of Public Health was to prevent infant deaths. Several approaches can be used to reduce the incidence of infections in infants.

Since 2006, “cocooning” has been the recommended approach for the prevention of pertussis in young infants. This strategy involves the vaccination of anyone who will have contact with infants who are not yet old enough to be vaccinated. However, logistically it is difficult to vaccinate all people who might have contact with an infant, and information on the efficacy of this strategy is currently lacking.[14]

One alternate and promising approach is the vaccination of pregnant women, because this method has the potential to provide direct protection to infants. Pertussis antibodies are transplacentally transmitted to the infant and may offer protection until the infant is old enough to be vaccinated.[15] Although it is clear that women who are vaccinated during pregnancy transmit pertussis antibodies to their infants, it has not yet been proven that these antibodies will in fact protect infants from infection. A third strategy is the vaccination of infants at birth.[16] This method offers the potential for direct protection of the infant, but newborns still would be at risk for pertussis infections until they produced adequate levels of antibodies. More research is needed on the efficacy of this approach as well.

The current DTaP (diphtheria, tetanus, and acellular pertussis) and Tdap (tetanus, diphtheria, and pertussis) vaccines are less than optimal, so research should be directed toward new vaccines with more antigens and a better balance of those currently included. Alternately, we could return to whole-cell DTwP (diphtheria, tetanus, and pertussis) vaccines in which lipopolysaccharide (LPS) has been attenuated. LPS was the cause of most adverse reactions following DTwP vaccination, but LPS antibodies are likely important for protection. DTaP vaccines are less reactogenic than DTwP vaccines because virtually all LPS has been removed.