It is estimated that whooping cough caused 294,000 deaths globally in 2003, but even in countries with long-standing surveillance and immunization programs, official pertussis notifications recorded only a fraction of actual cases.1 Severity of disease and mortality is greatest in infants,2 but most industrialized countries do not recommend the first vaccine dose until 8 weeks of age.3 The World Health Organization (WHO) recommends the first dose be given at 6 weeks of age, and the vaccine is licensed for use from this age in Europe, North America, and Australia.3,4
Shinall et al5 recently modeled the effects of acceleration of the first dose of the primary series in infants from 8 weeks to 6 weeks of age in the United States, using aggregated data on notifications and their reported hospitalization status, assuming equal distribution of these over the first 2 months of life. Cortese et al,6 using US national data on hospitalizations coded as due to pertussis, found higher rates of hospitalization and a rate in the second and third months of life more than double than in the first 4 weeks of life, questioning the validity of the equal distribution assumption of Shinall et al.5 We used national line listed data for notifications and hospital days coded as due to pertussis during a 10-year period in Australia to evaluate the potential benefits of advancing the first vaccine dose in infants to 6 weeks of age in a setting, comparable to the United States, of an industrialized country with a long-standing pertussis immunization program.
Pertussis notifications with diagnosis date between January 1 1999 and December 31 2008 were extracted from the Australian National Notifiable Diseases Surveillance System. Hospitalizations coded in any field as whooping cough with an admission date between January 1 1997 and December 31 2006 were extracted from the Australian Institute of Health and Welfare Hospital Morbidity Database. Deaths were derived from notifications so coded in the National Notifiable Diseases Surveillance System.
Incidence rates were calculated using Australian Bureau of Statistics midyear population estimates7 in the denominator, with the assumption that the population for each month of age was one-twelfth that of the total population less than 1 year of age. Bed-days were calculated by summing total days in hospital for each age group.
Negative binomial regression was used to compare incidence rates across age groups using rate ratios. χ2 was used for trend comparison. Ninety-five percent confidence intervals (CI) and P values were calculated using Intercooled STATA version 9. P values less than 0.05 were considered statistically significant.
Following methodology established by Shinall et al,5 the effect of vaccine acceleration was estimated by using the average rate per month of pertussis notification or hospitalization multiplied by the number of live births in 2007 in Australia (285,213) to obtain the expected number of pertussis notifications or hospitalizations by month of age. Benefits were calculated using an incidence shift in whooping cough rates, hospital admissions, or bed-days of 2 weeks from the second to the fourth month. Benefits accrue because of the decrease in notifications and hospitalizations reported each month from 1 month to 3 months of age. Estimations assumed an equivalent immune response to DTPa vaccine at 6 and 8 weeks of age and that the timeliness of vaccine coverage under a 6- and 8-week first dose recommendation would be similar. The difference in rates of notifications and hospitalizations equaled the current rate minus the estimated rate. Similar calculations were applied for estimating bed-days saved. The primary series for pertussis containing vaccine in Australia is given at 2, 4, and 6 months of age, although 1 jurisdiction (New South Wales) has recommended commencing at 6 weeks since 2009.
Pertussis Notification and Hospitalization in Infants Less Than 1 Year Old.
For the period 1999 to 2008, the notification rate for pertussis in children less than 1 year of age was 2.6 times that of the general population. In the 10 years, between 1997 and 2006, infants aged less than 1 year represented 52.4% (3032 cases) of all recorded whooping cough hospital admissions. All the notifications and 41% (1228) of hospital admissions had Bordetella pertussis attributed as the causative agent.
When notification rates were examined by 1 month age brackets, those aged 1 and 2 months were 31/2 times more likely to be notified for pertussis than those aged 3 to less than 12 months (Table 1). Compared with infants aged between 3 and less than 12 months, the incidence rate ratio was 3.5 (95% CI: 2.7–4.6) for those aged 1 month, 3.4 (95% CI: 2.6–4.5) for those aged 2 months, and 2.0 (95% CI: 1.5–2.7) for infants in their first 4 weeks of life. Infants aged 3 and 4 months were 35% (P = 0.063) and 49% (P = 0.004) less likely to contract pertussis than those at 2 months. A similar pattern was seen for hospital admissions (34% and 57% reduction) and overall hospital bed-days (50% and 73% reduction) (Table 1). If notifications and total hospital admissions between 1999 and 2006 were compared, a larger proportion of infants aged 1, 2, and 3 months were hospitalized than other infants less than 1 year. Among hospitalized infants, there was a clear trend for decreasing length of hospital stay with increasing age. The proportion of infants spending more than 5 days in hospital declined from 52% for those less than 4 weeks of age to 23% at 6 months (P < 0.001, Table 1) and for more than 10 days in hospital from 29% to 5%, respectively.
Effect of Vaccine Acceleration.
It was estimated that accelerating the first dose of the DTPa vaccine from 2 months to 6 weeks of age would result in an annual reduction of 11 notifications (8%), 17 hospitalizations (9%), and 151 bed-days (12%). The number needed to vaccinate to achieve this reduction would be 25,465 for notifications and 16,679 for hospitalizations.
Whooping cough occurs in epidemic cycles, with a substantially greater burden and potential for public concern in an epidemic year. If calculated for 2001, an epidemic year common to both data sets, we find a greater reduction than the mean with 12% in notifications, 10% in hospitalizations and 12% in hospital bed-days.
Of the 11 deaths identified in the study period, 4 (36%) would have had the opportunity to be immunized at least 10 days before date of diagnosis of disease with 1 dose of vaccine if the first dose were to be advanced to 6 weeks of age.
Infants in their first 3 months of life are at significantly greater risk than any other age group of acquiring severe pertussis, as measured by hospitalization, length of hospital stay, and death. Potential measures to reduce the burden of early infant pertussis may be both indirect, such as maximizing the timeliness and population coverage of infant vaccination and vaccination of adults in contact with infants, including maternal vaccination,8 and direct such as earlier infant vaccination.
Our findings are in keeping with a recent analysis using aggregated US data5 which estimated that accelerating the first vaccine dose by 2 weeks reduced notified whooping cough cases by 8% and hospitalizations by 9%. Additionally, the longer hospital stay seen among the youngest infants could be reduced by 12% of hospital bed-days per year. We also estimated the potential benefits of an earlier first vaccine dose in epidemic years and found that the additional estimated benefit increased by 50% (from 8% to 12%) for notifications and by 11% (from 9% to 10%) for hospitalizations compared with an interepidemic period. As with the US study, the reduction in cases occurs mainly between the second and fourth month of life and depends on the assumption that some degree of protection against severe pertussis accrues after even 1 dose.9
A limitation of the databases used is that there is potential for under-ascertainment (if the most sensitive diagnostic test was not used) and for misclassification (if some cases coded as pertussis on clinical grounds without laboratory confirmation were hospitalized for other reasons). Furthermore, the estimated reduction in hospitalizations may be less than 12% as only 41% of hospital admissions were coded as being specifically due to B. pertussis.
The reductions between month 2 and 4 and then 4 and 6 in notifications (49% and 58%, respectively) and hospitalizations (57% and 82%, respectively) provide some evidence for the efficacy of the first and the second dose of a pertussis containing vaccine. This trend was also noted in the analysis of hospitalization rates by Cortese et al6 and in notification rate data by Shinall et al.5 Uncertainty regarding immune response is greater for acceleration of the third dose of vaccine, than for the first dose for postimmunization levels of pertussis, diphtheria, and tetanus antibodies.10 With respect to the first dose of vaccines including acellular pertussis antigens, administration from 6 weeks of age is approved by regulatory authorities in North America, Europe, New Zealand, and Australia, although little direct comparative data between first doses at 6 or 8 weeks are available.5 With respect to subsequent doses, some European countries have accelerated primary schedules with the third dose at 4 months of age, while others do not give the third dose, as a booster, until 12 months of age.3
There is evidence of lesser responses to both pertussis antigens10 and other antigens, particularly the protein conjugate polysaccharide vaccines,11 when the third dose is given at 4 months. In the light of these concerns, and the evidence of good protection against pertussis after 2 doses (when given at 2 and 4 months of age),10,11 should the first dose be accelerated to 6 weeks, caution would be required regarding a third dose any earlier than 6 months of age. With respect to the implementation of an earlier first dose, it is likely that if parents and vaccine providers were convinced about the benefits, this would be straightforward. This is because most industrialized countries have existing recommendations for a routine postnatal visit at approximately 6 weeks, which could be integrated with commencement of the infant vaccine schedule.
The authors thank the State and Territory Health Authorities and the Office of Health Protection for provision of data from the National Notifiable Diseases Database and the Australian Institute of Health and Welfare for provision of data from the Morbidity Database.
1. Crowcroft NS, Peabody RG. Recent developments in pertussis
2. de Greef SC, Mooi FR, Schellekens JF, et al. Impact of acellular pertussis
preschool booster vaccination on disease burden of pertussis
in The Netherlands. Pediatr Infect Dis J
5. Shinall MC, Peters TR, Zhu Y, et al. Potential impact of acceleration of the pertussis vaccine
primary series for infants. Pediatrics
6. Cortese MM, Baughman AL, Zhang R, et al. Pertussis
hospitalizations among infants in the United States, 1993 to 2004. Pediatrics
8. Rodriguez-Cobo I, Chen YF, Olowokure B, et al. Clinical and economic assessment of different general population strategies of pertussis vaccine
booster regarding number of doses and age of application for reducing whooping cough
disease burden: a systematic review. Vaccine
9. Juretzko P, von Kries R, Hermann M, et al. Effectiveness of acellular pertussis vaccine
assessed by hospital-based active surveillance in Germany. Clin Infect Dis
10. Booy R, Aitken SJ, Tudor-Williams G, et al. Immunogenicity of combined diphtheria, tetanus, and pertussis vaccine
given at 2, 3, and 4 months versus 3, 5, and 9 months of age. Lancet
11. McVernon J, Andrews N, Slack MP, et al. Risk of vaccine
failure after Haemophilus influenzae
type b (Hib) combination vaccines with acellular pertussis