Secondary Logo

Journal Logo

Brief Reports


Baptista, Paulo N. PhD*; Magalhães, Vera PhD; Rodrigues, Laura C. PhD; Rocha, MariaÂngela W. MsD*; Pimentel, Analíria M. MsD*

Author Information
The Pediatric Infectious Disease Journal: September 2006 - Volume 25 - Issue 9 - p 844-846
doi: 10.1097/01.inf.0000232642.25495.95
  • Free


In the last 2 decades, the number of confirmed cases of pertussis increased, even in countries with high vaccine coverage.1 Studies of pertussis vaccine effectiveness (VE) have found that 3 doses protect more than 80% of those vaccinated against clinical disease.1,2

In 2003, in the University Hospital Oswaldo Cruz in Recife, Brazil (HUOC), pertussis, as defined by the Centers for Disease Control and Prevention (CDC),3 was diagnosed in 158 subjects. In Brazil, until 2004, pertussis whole cell vaccine (DwPT) was given at 2, 4 and 6 months of age and a booster dose at 15 months of age. Between 1998 and 2003 in the study area, there was complete vaccine coverage for 3 doses of DwPT in children <5 years old.4 Pertussis VE has not been assessed in Brazilian households in which at least one case was bacteriologically confirmed. In households, exposure to infection is likely to be homogeneous; this reduces the likelihood of confounding, increases the power of the study and facilitates collection of information.5 This study estimated pertussis VE; vaccinated contacts had lower incidence of clinical disease, a low proportion of vaccinated cases had a positive culture and transmissibility from vaccinated cases was lower than from unvaccinated cases.


In 2003, a system was established for referral of children with symptoms consistent with pertussis to the HUOC. Once a pertussis case was confirmed, it was enrolled as an index case. The members of the index cases’ households were enrolled into the study: parents/guardians were interviewed and information was collected for each household member on recent history of cough (symptoms and date of onset), pertussis vaccination (by examination of the card; and by recall in 8 subjects, of which 3 aged <5 years) and age. Nasopharyngeal swabs were collected from every household member who had cough within the previous 21 days. Cultures were not collected from subjects with no reported symptoms. Clinical specimens were placed in Regan Lowe medium and incubated at 35°C to 37°C. Cultures were checked up to 12 days. Bordetella pertussis culture was confirmed by biochemistry test.6

The criteria for being a case was one of the following: 1) an illness with cough of any duration and positive culture for B. pertussis; 2) an illness with cough lasting at least 2 weeks and at least one of the following symptoms: paroxysm of coughing, inspiratory whooping, pos-tussive vomiting and a negative culture but a epidemiologic link to a culture confirmed case; or 3) an illness with cough lasting at least 14 days in a household with 2 or more pertussis cases where there was one culture confirmed case. Household members for whom information on recent cough was insufficient to confirm or exclude pertussis were excluded from analysis. The remaining was considered not cases.

Pertussis cases (including index cases) were classified into: primary case (first case in the household), coprimary case (case with onset of symptoms within 6 days after the onset in the primary case) or secondary case (case with onset 7 or more days after onset in the primary case). Households with coprimary cases were excluded from the analysis because contacts were likely to have been exposed to a higher bacterial challenge.

Household members formed 2 cohorts, 1 of vaccinated members (who had received at least 3 doses of DwPT) and 1 of unvaccinated members (unvaccinated or who had received <3 doses of DwPT). The first day of cough in the primary case was considered the first day at risk for the cohort.

Different measures of VE were estimated: VE against illness (attack rate in unvaccinated household contacts − attack rate in vaccinated household contacts) over attack rate in unvaccinated household contacts; VE against transmission (attack rate in contacts of unvaccinated primary cases − attack rate in contacts of vaccinated primary cases) over attack rate in contacts of unvaccinated primary cases; and VE against bacteriologic positivity in cases (proportion of unvaccinated cases with a positive culture − proportion of vaccinated cases with a positive culture) over proportion of unvaccinated cases with a positive culture.

All estimated VE were stratified by age and calculated using EpiInfo 6.04 software. Statistical significance was assessed using χ2 (Fisher exact test when the value of one of the cells in a table was <5). A confidence interval of 95% was calculated and probability was considered significant when ≤0.05. The study was approved by the ethics committee of HUOC. Parents/guardians signed a written consent form to participate.


Pertussis was confirmed by culture or epidemiologic link in 17.8% (51 of 287) of the suspected cases of pertussis. These 51 index cases of pertussis were members of 57 households. There were 349 subjects in these households. Among them, 158 were classified as having had clinical pertussis during the critical time (cases) and 169 were classified as not a case. Among the 158 cases, 52 were confirmed by culture and 106 by epidemiologic link. Twenty-two contacts were excluded because information about them was insufficient to classify them into cases or not cases.

Vaccine history of household primary cases was unknown in 43% (28 of 65), 15% (10 of 65) were unvaccinated or received <3 doses of DwPT, 31% (20 of 65) had received 3 to 4 doses of DwPT <10 years before the study and 11% (7 of 65) had received 3 to 4 doses >10 years before the study. Vaccine history of the contacts is shown in Table 1.

Secondary Attack Rates in Household Contacts by Age and Pertussis Vaccine Status of the Contacts

Vaccine Effectiveness Against Illness.

In the age group from 11 [1/2] to 19 years, the difference in the secondary attack rate in household contacts who were vaccinated <10 years ago (70%) and in those vaccinated >10 years ago (40%) was not statistically significant (P = 0.57). In the age group <6 years of age, VE in reducing clinical disease in contacts was 12.5% (95% confidence interval [CI] = −5.3 to 27.3) (P for Fisher exact test = 0.12). A less refined estimate was done comparing attack rates in vaccinated contacts who were <11 [1/2] years of age with attack rates in contacts with incomplete immunization who were <6 years of age. This was 36.4% (95% CI = 20.4–49.1) (P for Fisher exact test <0.001).

Vaccine Effectiveness Against Transmissibility.

The secondary attack rate in contacts of primary cases aged 7 months to 5 years and who had received 3 to 4 doses of DwPT was 16.7% (5 of 30); the secondary attack rate among contacts of primary cases of other age groups and of any vaccine status was 43% (63 of 145). The pertussis VE in reducing the transmissibility of a vaccinated primary case aged 7 months to 5 years (compared with all primary cases) was 61.6% (95% CI = 12.8–83.1%; P = 0.01).

Vaccine Effectiveness Against Bacteriologic Positivity in Cases.

Culture for B. pertussis was positive in 31% (13 of 42) of fully vaccinated cases and in 83.8% (31 of 37) of cases with incomplete vaccination (P < 0.001). VE in reducing bacteriologic positivity in cases was 63.1% (95% CI = 40.7–77.0; Fisher exact test P < 0.01).


The protective effect of vaccines can wane with time. Subjects >11[1/2] years who received DwPT in the first 2 years of life are believed to have low or no protection from pertussis.7 The fact that 84% of household contacts who did not become a case of pertussis were >11[1/2] years old can be explained, at least in part, by naturally acquired immunity.

In the present study, a low VE against clinical disease was found in fully vaccinated children. It is possible that this reflects low VE for a complete scheme; this is also consistent with high protection from an incomplete vaccination. Age can be a confounding factor2 and we would have liked to have been able to estimate VE in narrower age groups. Sample size and the high vaccine coverage did not allow this. Estimating VE within a wide age range can control incompletely for confounding; this would increase the probability of naturally acquired immunity to appear to enhance the estimated vaccine efficacy. We suggest that the effectiveness against clinical disease of the complete versus incomplete scheme could have been overestimated.

The proportion of cases confirmed by culture among unvaccinated cases was higher than in vaccinated cases. This suggests that vaccination reduces the bacterial load of cases. If the risk of disease increases with increases in bacterial challenge, then one would expect higher vaccine protection in households in which the primary cases were vaccinated.8,9 In this study, estimated VE in reducing the occurrence of bacteriologic-positive disease was 63%, very similar to the reduction in transmissibility from the vaccinated primary cases to the contacts.

It has been suggested that reduced transmissibility from cases contributes to indirect protection of vaccination.10 This suggests that the effect of the vaccine in reducing transmission would have a significant effect in reducing circulation of B. pertussis in the population.

In the study, individuals >11 [1/2] years were the primary cases for 79% of all secondary cases.4 Brazilian National Immunization Program has recently introduced a fifth dose of DwPT to be given between 4 and 6 years of age. Given that protection wanes with time and given the role of older children in transmission and the vaccine effect in reducing the ability of vaccinated cases to transmit infection, we suggest consideration of the fifth dose being given at age 11 years in Brazil.


The authors thank Dr. Gary Sanden, Pertussis Laboratory Program, Center for Disease Control and Prevention, Atlanta, GA.


1. Crowcroft NS, Britto J. Whooping cough a continuing problem. Pertussis has re-emerged in countries with high vaccination coverage and low mortality. BMJ. 2002;324:1537–1538.
2. Fine PEM, Clarkson JA. Reflections on the efficacy of pertussis vaccines. Rev Infect Dis. 1987;9:866–883.
3. Bisgard K. Definitions. In: Pertussis Guide. Atlanta: Centers for Disease Control and Prevention; 2000. (Available at): Accessed September 18, 2002.
4. Baptista PN, Magalhães V, Rodrigues LC, Rocha MAW, Pimentel AM. Source of infection in household transmission of culture-confirmed pertussis in Brazil. Pediatr Infect Dis J. 2005;24:1027–1028.
5. Fine PEM, Hall AJ, Glynn JR, Rodrigues LC, Smith PG. Measuring the transmissibility of infection. In: Course of Epidemiology and Control of Communicable Diseases—2004 London School of Hygiene and Tropical Medicine, Course Manual, Session 5. 2004.
6. Murphy T, Bisgard K, Sanden G. Diagnosis and laboratory methods. In: Pertussis Guide. Atlanta: Centers for Disease Control and Prevention; 2000. (Available at): Accessed September 18, 2002.
7. Jenkinson D. Duration of effectiveness of pertussis vaccine: Evidence from a 10 year community study. BMJ. 1988;296:612–614.
8. Taranger J, Trollfors B, Bergfors E, et al. Mass vaccination of children with pertussis toxoid decreased incidence in both vaccinated and nonvaccinated persons. Clin Infect Dis. 2001;33:1004–1010.
9. Fine PEM, Clarkson JA, Miller E. The efficacy of pertussis vaccines under conditions of household exposure. Int J Epidemiol. 1988;7:635–641.
10. Préziosi MP, Halloran ME. Effects of pertussis vaccination on transmission: Vaccine efficacy for infectiousness. Vaccine. 2003;21:1853–1861.

vaccine effectiveness; pertussis vaccine; transmissibility

© 2006 Lippincott Williams & Wilkins, Inc.