Acellular Pertussis Vaccine

Effectiveness Over Time

Ousseny Zerbo, PhD, Joan Bartlett, MPH, MPP, Kristin Goddard, MPH, Bruce Fireman, MA, Edwin Lewis, MPH,
Nicola P. Klein, MD, PhD

To determine pertussis risk by diphtheria-tetanus-acellular pertussis (DTaP)

OBJECTIVES:
vaccination status and time since last DTaP dose.
abstract
METHODS: Children born at Kaiser Permanente Northern California between 1999 and 2016
were followed from 3 months of age until they tested positive for pertussis; disenrolled from
Kaiser Permanente Northern California; received the tetanus toxoid, reduced diphtheria
toxoid, and acellular pertussis, adsorbed vaccine; turned 11 years of age, or the end of the
study period. DTaP vaccination status was categorized on the basis of the number of doses
received in relation to the number of doses expected according to the Advisory Committee on
Immunization Practice–recommended ages.
RESULTS: Among 469 982 children ages 3 months to 11 years, we identified 738 pertussis cases.
A total of 99 cases were unvaccinated, 36 were undervaccinated, 515 were fully vaccinated,
and 88 were fully vaccinated plus 1 dose. Pertussis risk was 13 times higher among
unvaccinated (adjusted hazard ratio [aHR] = 13.53; 95% confidence interval [CI] 10.64–17.21)
compared with fully vaccinated children and 1.9 times higher (aHR = 1.86; 95% CI 1.32–2.63)
among undervaccinated children. Among vaccinated children ages 19 to ,84 months,
pertussis risk was 5 times higher (aHR = 5.04; 95% CI 1.84–13.80) $3 years vs ,1 year after
vaccination. Among children ages 84 to 132 months, risk was 2 times higher (aHR = 2.32; 95%
CI 0.97–5.59) $6 years vs ,3 years after vaccination.
CONCLUSIONS:Undervaccinated and especially unvaccinated children were at greater risk of
pertussis. However, most pertussis cases occurred among children age-appropriately
vaccinated who were further away from their last DTaP dose, suggesting that suboptimal
vaccine effectiveness played a major role in recent pertussis epidemics.

Division of Research, Vaccine Study Center, Kaiser Permanente Northern California, Oakland, California WHAT’S KNOWN ON THIS SUBJECT: Children who are not
age-appropriately vaccinated are at increased risk of
Drs Zerbo and Klein, Mr Lewis, and Mr Fireman conceptualized and designed the study; Ms Bartlett pertussis. However, age-appropriately vaccinated children
conducted all statistical analyses; and all authors critically reviewed the manuscript and approved are also at risk because of the waning diphtheria-tetanus-
the final manuscript as submitted. acellular pertussis vaccine immunity.
DOI: https://doi.org/10.1542/peds.2018-3466
WHAT THIS STUDY ADDS: Compared with children fully
Accepted for publication Mar 19, 2019 diphtheria-tetanus-acellular pertussis vaccinated,
Address correspondence to Ousseny Zerbo, PhD, Division of Research, Vaccine Study Center, Kaiser unvaccinated and undervaccinated children were at
Permanente Northern California, 1 Kaiser Plaza, 16th Floor, Oakland, CA 94612. E-mail: a greater risk of pertussis. However, most pertussis
ousseny.x.zerbo@kp.org cases occurred among age-appropriately vaccinated
PEDIATRICS (ISSN Numbers: Print, 0031-4005; Online, 1098-4275). children, suggesting that suboptimal vaccine
effectiveness played a major role in recent pertussis
Copyright © 2019 by the American Academy of Pediatrics epidemics.
FINANCIAL DISCLOSURE: Dr Klein has received research grant support from Sanofi Pasteur,
Novartis, GlaxoSmithKline, Merck, MedImmune, Pfizer, Protein Sciences (now Sanofi Pasteur), and To cite: Zerbo O, Bartlett J, Goddard K, et al. Acellular
Dynavax for unrelated studies; the other authors have indicated they have no financial relationships Pertussis Vaccine Effectiveness Over Time. Pediatrics.
relevant to this article to disclose. 2019;144(1):e20183466

Downloaded from www.aappublications.org/news by guest on November 11, 2020

PEDIATRICS Volume 144, number 1, July 2019:e20183466 ARTICLE
Pertussis, or whooping cough, is the ACIP recommendation and The current study included children
a highly contagious vaccine- waning of DTaP vaccine immunity who were born between 1999 and
preventable disease caused by over time. In recent years, there has 2016 and were members of KPNC at
Bordetella pertussis.1 This disease can been an increase in parental 2 months of age. Children born before
occur at any age but is most severe in hesitancy to vaccinate children, 1999 were excluded to restrict the
infants, for which most fatal cases leading to an increase in the number study population to those receiving
occur.2 Before the availability of of children unvaccinated or only acellular pertussis vaccines. We
vaccines, pertussis was a common undervaccinated,13–19 which has been also excluded children with culture-
cause of morbidity and mortality reported to cluster geographically.20, or polymerase chain reaction (PCR)-
21
among children. After the But despite vaccine hesitancy, confirmed pertussis before study
introduction of a whole-cell pertussis DTaP vaccine coverage in the United entry. Children had to be
vaccine, annual rates of the disease States from 2012 to 2016 was continuously enrolled in KPNC
dropped considerably from 150 of between 93.7% and 95.0% for $3 starting at 2 months of age to ensure
100 000 of the population in the doses,8 suggesting that waning of accurate ascertainment of DTaP
1940s to 1 of 100 000 in the 1970s in vaccine immunity may play a strong vaccination status.
the United States.3 role.22–27
Study Design
The whole-cell pertussis vaccine was A previous study has revealed that
children who are not age- The study was a retrospective cohort
associated with systemic and local
appropriately vaccinated are at study. The outcome was pertussis
side effects, and in the 1990s,
increased risk of pertussis.15 identified by a real-time PCR test for
a combined diphtheria-tetanus-
However, studies have also shown B pertussis. Since 2006, nearly all
acellular pertussis (DTaP) vaccine
that children who are age- pertussis testing is conducted in
was introduced and eventually
appropriately vaccinated are also at a centralized laboratory by using PCR
replaced the whole-cell pertussis
risk partly because of waning DTaP on nasopharyngeal swabs. Eligible
vaccine in many developed
immunity.22,23 The objective of this children were followed starting on
countries.4,5 The US Advisory
study was to determine pertussis risk January 1, 2006, or when they were
Committee on Immunization Practice
by DTaP vaccination status and time 3 months of age, whichever was later,
(ACIP) recommends a total of 5 doses
since last DTaP dose. and continued until they tested PCR-
of DTaP: 1 dose each at ages 2, 4, and
positive for pertussis, disenrolled
6 months, 1 dose between 12 and
from KPNC, received their Tdap
18 months of age, and 1 dose METHODS vaccine, turned 11 years of age, or the
between ages 4 and 6 years.6 In 2006,
end of the study period on June
the ACIP recommended a booster Study Population
30, 2017.
dose of tetanus toxoid, reduced The study setting was Kaiser
diphtheria toxoid, and reduced The main exposure of interest was
Permanente Northern California
acellular pertussis (Tdap) vaccine for DTaP vaccination status, a time-
(KPNC), an integrated health care
adolescents between 11 and 12 years varying covariate with 4 levels
delivery organization that provides
of age.7 comprehensive care to ∼4 million
(unvaccinated, undervaccinated, fully
vaccinated, or fully vaccinated plus 1
Despite robust vaccination members. Members receive almost all
dose). We categorized DTaP
schedules and high vaccination medical care at KPNC-owned
vaccination status on the basis of the
rates,8 many countries, including facilities, including clinics, hospitals,
number of DTaP doses received in
the United States, have seen large pharmacies, and laboratories. KPNC
relation to the number of DTaP doses
pertussis outbreaks since the databases are used to capture
expected according to the ACIP-
replacement of the whole-cell detailed information on all medical
recommended ages for the 5-dose
pertussis vaccine with the acellular services, including vaccinations and
series (2, 4, 6, and 12–18 months and
pertussis vaccine.9–13 In 2010 and laboratory tests, as well as on
4–6 years). ACIP-recommended ages
2014, California experienced 2 large enrollment and demographics. KPNC
for DTaP did not change during our
pertussis outbreaks with .9000 members are similar to the broad
study period.
cases each.11,12 catchment population in Northern
California in terms of We classified children as fully
The reasons for these outbreaks are sociodemographic characteristics, vaccinated if they received the
certainly due to multiple factors, except the extremes of income expected number of DTaP doses by
including suboptimal vaccine distribution are underrepresented.28 1 month after the ACIP-recommended
coverage in some children not age- Members receive all their routine age (ie, 1 dose by 3 months, 2 doses
appropriately vaccinated according to vaccinations free of charge. by 5 months, 3 doses by 7 months, 4

Downloaded from www.aappublications.org/news by guest on November 11, 2020

2 ZERBO et al
doses by 19 months, and 5 doses by models. The models included a 4- pertussis in relation to the number of
84 months of age). Children with level vaccination indicator variable to years since the last DTaP dose,
fewer doses than expected for their estimate the pertussis hazard ratios adjusting for DTaP vaccination status.
age were undervaccinated, and (HRs) for children who were Years since last DTaP dose was
children with 1 more dose than unvaccinated, undervaccinated, or a time-dependent covariate that was
expected were fully vaccinated plus 1 fully vaccinated plus 1 dose divided into categories and updated
dose. We created the fully vaccinated compared with children who were as children got further from their last
plus 1 dose category so that children fully vaccinated. The Cox models DTaP dose or received a new DTaP
who had all their recommended were specified with a calendar time dose. For children ages 19 to ,84
doses were not combined with those line and stratified by the year and months, there were 4 categories:
who had more than the month of birth, which, together, adjust ,1 year (reference), 1 to ,2 years, 2
recommended doses because their for the month of age. We included to ,3 years, and $3 years. For
risk of pertussis may be different covariates to adjust for sex, race or children ages 84 to 132 months, there
(Table 1). Children are considered ethnicity, type of insurance (Medicaid were 5 categories: ,3 years
age-appropriately vaccinated if they or other subsidized insurance versus (reference), 3 to ,4 years, 4 to ,5
are in either the fully vaccinated or commercial insurance), and the years, 5 to ,6 years, and $6 years.
the fully vaccinated plus 1 dose medical clinic where children We estimated the pertussis HR for
category. received most of their care. The last 2 each category for years since
covariates were time-varying and vaccination in relation to the
We counted DTaP doses starting
updated monthly (insurance type) or reference category with a Cox model
8 days after receipt to allow time for
quarterly (medical clinic). We like those described above. We used
the vaccine to take effect.29 We
estimated HRs for the entire study a calendar time line stratified by the
censored children who had 2 more
population over the entire follow-up year and month of birth and included
doses than expected (eg, 4 doses by
period. We also estimated HRs for the the same covariates. We also
5 months of age) or children
$7 years of age with 6 doses. following age groups: 3 to ,5, 5 to conducted analyses in which we did
,7, 7 to ,19, 19 to ,84, and 84 to not control for the year and month of
We updated DTaP vaccination status 132 months. Each of these age groups birth because years since last DTaP
during follow-up as children aged or has a different number of DTaP doses dose and age are closely correlated.
received additional DTaP doses. For (ranging from 1 to 5) corresponding
example, at ages 3 to ,5 months, We conducted all analyses with SAS
to fully vaccinated status. For the
children with 1 DTaP dose were fully version 9.3 (SAS Institute, Inc, Cary,
youngest and oldest age groups, the
vaccinated, and those with 2 DTaP NC). We used the Lexis macro to
model included a 3-level rather than
doses were fully vaccinated plus 1 partition person-time (http://
a 4-level vaccination indicator
dose; at ages 5 to ,7 months, bendixcarstensen.com/Lexis/Lexis.
variable (Table 1). We estimated
children with 1 dose were sas).
vaccine effectiveness (VE) as 1 minus
undervaccinated, those with 2 doses the adjusted hazard ratios (aHRs) The study was approved by the KPNC
were fully vaccinated, and those with using these same models, except that Institutional Review Board with
3 doses were fully vaccinated plus 1 unvaccinated children constituted the a waiver of written informed consent
dose. Children with no DTaP doses reference group. because the study had no direct
were classified as unvaccinated at all contact with study participants.
ages (Table 1). In each of the 2 oldest age groups of
children, we examined waning of
Statistical Analysis vaccine immunity over time. We RESULTS
We examined the risk of pertussis in restricted the analyses to children The study included 469 982 children
relation to DTaP vaccination status who had received at least 1 dose of born between 1999 and 2016 and
using a series of Cox regression DTaP and estimated the risk of followed them from 2006 to 2017.

TABLE 1 Number of DTaP Doses by Vaccination Status and Age at KPNC, January 2006–June 2017
DTaP Vaccination Status Ages 3 to Ages 5 to Ages 7 to Ages 19 to Ages
,5 mo ,7 mo ,19 mo ,84 mo 84–132 mo
Unvaccinated 0 0 0 0 0
Undervaccinated N/A 1 2 3 4
Fully vaccinated 1 2 3 4 5
Fully vaccinated plus 1 dose 2 3 4 5 N/A
N/A, not applicable.

Downloaded from www.aappublications.org/news by guest on November 11, 2020

PEDIATRICS Volume 144, number 1, July 2019 3
The average duration of follow-up vaccinated, and 88 (12%) fully children across all follow-up in all age
was 4.6 years per child. Most vaccinated plus 1 dose. groups. Children who were fully
children were born at a KPNC vaccinated plus 1 dose had a lower
Overall incidence rates of pertussis
hospital (89%) and entered the risk of pertussis compared with fully
among children in the study
study at 3 months of age (73%) vaccinated children (aHR = 0.48; 95%
population varied from 389 out of
(Table 2). During follow-up, the CI 0.34–0.68; Table 4).
100 000 P-Ys for unvaccinated
study population contributed children to 16 out of 100 000 P-Ys for Risk of pertussis according to DTaP
2 138 835 person-years (P-Ys) fully vaccinated plus 1 dose children. vaccination status varied somewhat
distributed as follows by DTaP Incidence rates of pertussis also by age. In each age group,
vaccination status: 1% unvaccinated, varied by vaccination status within unvaccinated children were at
3% undervaccinated, 70% fully each age category (Table 3). a significant increased risk for
vaccinated, and 26% fully vaccinated pertussis compared with fully
After adjustment for covariates,
plus 1 dose (Table 3). We identified vaccinated children, ranging from
pertussis risk was 13 times higher
738 PCR-confirmed pertussis cases 4 times higher (aHR = 4.54; 95% CI
among unvaccinated children (aHR =
classified into the following 1.62–12.69) for children ages 3 to
13.53; 95% confidence interval [CI]
categories according to their ,5 months to 23 times higher (aHR =
10.64–17.21) and 1.9 times higher
vaccination status: 99 (13%) 23.62; 95% CI 16.32–34.17) for
among undervaccinated children
children ages 19 to ,84 months.
unvaccinated, 36 (5%) (aHR = 1.86; 95% CI 1.32–2.63)
Undervaccinated children ages 5 to
undervaccinated, 515 (70%) fully compared with fully vaccinated
,7 and 19 to ,84 months were also
at significantly increased risk for
pertussis compared with fully
TABLE 2 Characteristics of the Study Population at KPNC, January 2006–June 2017 vaccinated children, and children who
Study Population (N = were fully vaccinated plus 1 dose at
469 982), n (%) ages 7 to ,84 months were at
Age at study entrya significantly reduced risk compared
3 mo 343 712 (73) with the fully vaccinated reference
.3 mo to ,1 y 21 036 (4) group (Table 4).
1 to ,2 y 24 323 (5)
2 to ,3 y 21 597 (5) Across all follow-up and all age
3 to ,4 y 17 461 (4)
4 to ,5 y 15 543 (3)
groups, VE was 86% (95% CI
5 to ,6 y 14 140 (3) 80%–91%) for undervaccinated
6 to ,7 y 12 146 (3) children compared with unvaccinated
7y 24 (0.0) children. VE was even higher for fully
Born at KPNC hospital 419 823 (89) vaccinated children and for those
Year of birth
1999–2001 41 816 (9)
who were fully vaccinated plus 1
2002–2004 63 366 (14) dose. VE = 93% (95% CI: 91%–94%)
2005–2007 91 594 (19) for fully vaccinated children and
2008–2010 91 512 (19) VE = 96% (95% CI: 95%–97%) for
2011–2013 88 911 (19) children who were fully vaccinated
2014–2016 92 783 (20)
Sex
plus 1 dose (Supplemental Table 6).
Female 229 154 (49)
Male 240 828 (51)
Vaccinated children who were further
Race or ethnicityb away from their last DTaP dose were
White 182 938 (39) at increased risk of pertussis. Among
Asian or Pacific Islander 112 142 (24) children ages 19 to ,84 months, the
Hispanic (regardless of race) 111 165 (24) crude incidence rate of pertussis rose
Black 28 673 (6)
American Indian or Alaskan native 2260 (0)
from 11 to 32 out of 100 000 P-Ys
Missing 32 804 (7) when time since last DTaP dose
Pertussis status increased from ,1 to $3 years.
PCR-positive for pertussis 738 (0.2) Adjusting for vaccination status and
a Children born in 1999 entered the study on January 1, 2006, at 6 years of age, except for the few children born on other covariates, including the child’s
January 1, 1999, who entered the study at 7 years of age; children born in 2000 entered the study on January 1, 2006, at age, risk of pertussis was 5 times
5 years of age, except for the few born on January 1, 2000, who entered the study at 6 years of age; and so on. Children
born in or after October 2005 entered the study on their 3-month birthday. higher for children whose last DTaP
b The categories are mutually exclusive. dose was $3 years before compared

Downloaded from www.aappublications.org/news by guest on November 11, 2020

4 ZERBO et al
 with those whose last dose was

Crude Incidence
,1 year before (aHR = 5.04; 95% CI
Children Ages 84–132 mo
Expected DTaPa = 5 (N =

Rates
1.84–13.80; Table 5). Adjusting for

505

—
50
63

65
the same covariates except the child’s
172 760)

age, risk of pertussis was .2 times

12 031 (2) higher for children whose last DTaP
2573 (1)

484 290

498 894
P-Y (%)

dose was $3 vs ,1 year before (aHR

(100)
(97)
—
= 2.58; 95% CI 1.49–4.46; Table 5).
Crude Incidence
Children Ages 19 to ,84 mo

Similarly, among fully vaccinated

Expected DTaPa = 4 (N =

children ages 84 to 132 months,

TABLE 3 P-Y of Follow-up and Incidence Rate of Pertussis per 100 000 P-Y by DTaP Vaccination Status, Overall and by Age Group at KPNC (January 2006–June 2017)

Rates
436
51
16

14

21

crude incidence rates of pertussis

381 578)

rose from 24 to 133 out of

100 000 P-Ys when time since last
702 155 (58)

465 299 (38)

12 166 (1)
37 331 (3)

DTaP dose increased from ,3 to $6

1 216 951
P-Y (%)

(100)

years. After adjusting for covariates,

including age, the risk of pertussis
was .2 times higher $6 years after
Crude Incidence
Children Ages 7 to ,19 mo
Expected DTaPa = 3 (N =

the last DTaP vaccination compared

Rates
351

with ,3 years (aHR = 2.32; 95% CI

57
32

15

35
357 699)

0.97–5.59; Table 5). After adjusting

for all covariates except age, the risk
68 462 (22)

of pertussis was .4 times higher

13 916 (5)
6264 (2)

221 617

310 259
P-Y (%)

(100)
(71)

$6 years after the last DTaP

vaccination compared with ,3 years
(aHR = 4.66; 95% CI 2.81–7.71;
Crude Incidence
Children Ages 5 to ,7 mo
Expected DTaPa = 2 (N =

Table 5).
Rates
362.6
102
26

24

39
343 099)

DISCUSSION
1655 (3)
2928 (5)
P-Y (%)

39 109

12 248

55 939
(100)

Children ages 3 months to 11 years

(70)

(22)

who did not receive any DTaP vaccine

or who were undervaccinated
Crude Incidence
Children Ages 3 to , 5 mo
Expected DTaPa = 1 (N =

according to the ACIP-recommended

Rates

schedule were at increased risk of

179
—
44

51

53
348 968)

pertussis compared with children

who were fully vaccinated. Although
noncompliance with the vaccination
2786 (5)
P-Y (%)

38 454

15 552

56 792
(100)
(68)

(27)
—

schedule is an important public

a The expected number of DTaP doses corresponds to fully vaccinated.

health problem that leads to

increased pertussis risk, most
Crude Incidence
Expected DTaPa = 1–5 (N =
Children Ages 3–132 mo

children in our study received all

Rates
389
54
35

16

34

their recommended DTaP doses.

469 982)

Unvaccinated and undervaccinated

children accounted for only 4% of
Fully vaccinated plus 1 561 560 (26)

P-Ys of follow-up in our study

25 443 (1)
66 206 (3)
1 485 625

2 138 835
P-Y (%)

(100)
(69)

population. Furthermore, most

undervaccinated children had only 1
fewer DTaP dose than recommended
for their age so that ,2% of follow-
up was spent missing $2 doses. Thus,
DTaP Vaccination

Undervaccinated
Fully vaccinated

—, not applicable.
Unvaccinated

despite the pertussis risk being

13 times higher among unvaccinated
dose
Status

Total

children and the risk 1.9 times higher

among undervaccinated children,

Downloaded from www.aappublications.org/news by guest on November 11, 2020

PEDIATRICS Volume 144, number 1, July 2019 5
 these 2 groups together comprised

5.00–15.71
(Children 498 894 P-Y)

0.38–1.94
DTaP Doses); 172 760 ,20% of pertussis cases.
Ages 84–132 mo (5

95% CI
With 323 Pertussis

—
The majority of pertussis cases
Cases (.80%) in our study population
aHRb occurred among children who had
8.87
0.86
—
received all their recommended DTaP
doses. Children who were further
away from their last DTaP dose were

b aHR adjusted for sex, race or ethnicity, type of insurance, medical clinic, and for age and calendar date because risk sets were defined on a calendar time line and stratified by year and month of birth.
16.32–34.17
(Children 1 216 951 P-Y)
Ages 19 to ,84 mo (4

1.78–4.88
0.34–0.87
DTaP Doses); 381 578

at increased risk of pertussis. Our

95% CI
With 253 Pertussis

results reveal that waning of DTaP

Cases

immunity was an important cause of

pertussis in children .18 months of
age who have a longer interval
23.62
aHRb
Age at Follow-up (No. DTaP Doses Corresponding to Fully Vaccinated Status)

2.94
0.54

between recommended doses. By age

19 months, nearly all children had
received at least 1 DTaP dose, and
7.23–20.15
Children (310 259 P-Y)
Ages 7 to ,19 mo (3

0.91–4.02
0.19–0.80
DTaP Doses); 357 699

95% CI

children who were age-appropriately

With 110 Pertussis

vaccinated accounted for .95% of

Cases

follow-up.
TABLE 4 Risk of Pertussis in Children by DTaP Vaccination Status, Overall and by Age Group at KPNC (January 2006–June 2017)

Although undervaccinated children

12.07
aHRb

1.91
0.39

were at a much lower risk than

unvaccinated children, their risk was
twice that of fully vaccinated children.
5.02–45.49
1.20–17.15
0.18–3.98
Children (55 939 P-Y)
Ages 5 to ,7 mo (2

These results are broadly consistent

DTaP Doses); 43 099

95% CI
With 22 Pertussis

with 2 previous case-control studies.

Cases

Authors of 1 study, which included 72

children PCR-positive for pertussis
and 288 matched controls, found
15.11
aHRb

4.54
0.85

a significant increased risk for

pertussis in undervaccinated children
aged 3 to 36 months,15 whereas
a Children fully vaccinated according to the ACIP-derived age cut points served as the reference group.
1.62–12.69

0.54–4.36
Children (56 792 P-Y)
Ages 3 to ,5 mo (1
DTaP dose); 348 968

95% CI
With 30 Pertussis

authors of another study evaluated

—

145 cases and 2900 controls and

Cases

found undervaccination associated

with a more than twofold increased
aHRb
4.54

1.53

risk of pertussis among children 3 to

—

35 months old.30 In both studies,

undervaccination was defined on the
Overall: Ages 3–132 mo;

10.64–17.21
(2 138 835 P-Y) With 738

basis of the number of missing DTaP

1.32–2.63
0.34–0.68
95% CI
469 982 Children

Pertussis Cases

doses in relation to the number of

recommended doses. Children who
were age-appropriately vaccinated
were the comparison group. The
13.53
aHRb

1.86
0.48

results for waning of DTaP immunity

in this study are consistent with our 2
previous studies, although we use
Fully vaccinated plus 1 dose

different methods and measures of

DTaP Vaccination Statusa

waning.22,23 Our results are also

consistent with those of a meta-
Undervaccinated

analysis of 11 studies in which

—, not applicable.
Unvaccinated

authors found the odds of pertussis

increased by 33% for every
additional year after the third or the
fifth DTaP dose.24

Downloaded from www.aappublications.org/news by guest on November 11, 2020

6 ZERBO et al
TABLE 5 Risk of Pertussis in Vaccinated Children by Years Since Last DTaP Dose, by Age Group at KPNC, January 2006–June 2017
Age at Follow-up (No. DTaP Doses Corresponding to Fully Vaccinated Status)
Ages 19 to ,84 mo (4 DTaP Doses); 377 030 Children Ages 84–132 mo (5 DTaP Doses); 171 701 Children (496 321
(1 204 786 P-Y) With 200 Pertussis Cases P-Y) With 310 Pertussis Cases
HR (95% CI)a HR (95% CI)b HR (95% CI)a HR (95% CI)b
Years since last DTaP dose — —
(versus ,1 y)
1 to ,2 y 1.40 (0.86–2.27) 1.36 (0.93–1.97) — —
2 to ,3 y 2.01 (1.05–3.85) 1.96 (1.34–2.87) — —
3+ y 5.04 (1.84–13.80) 2.58 (1.49–4.46) — —
Years since last DTaP dose — — — —
(versus ,3 y)
3 to ,4 y — — 1.55 (0.88–2.71) 1.90 (1.17–3.11)
4 to ,5 y — — 1.37 (0.68–2.76) 2.14 (1.31–3.50)
5 to ,6 y — — 2.03 (0.92–4.52) 3.42 (2.13–5.50)
6+ y — — 2.32 (0.97–5.59) 4.66 (2.81–7.71)
—, not applicable.
a HR estimates are adjusted for sex, race or ethnicity, type of insurance, medical clinic, DTaP vaccination status, and for age and calendar date because risk sets were defined on

a calendar time line and stratified by year and month of birth.

b Same as footnote “a,” except HR estimates are not adjusted for age (ie, risk sets are not stratified by year and month of birth).

The study strengths include its large, addition, PCR testing may have Within our study population, .80%
racially diverse population and misclassified pertussis status for of pertussis cases occurred among
precise data on the number and a few individuals; however, such age-appropriately vaccinated
timing of DTaP doses. The large misclassification was unlikely to be children. Children who were further
population allowed us to analyze related to time since vaccination or to away from their last DTaP dose were
pertussis risk relative to vaccination vaccination status. at increased risk of pertussis, even
status separately by age group for after controlling for undervaccination.
Although waning immunity is clearly
each level of recommended DTaP Our results suggest that in this
an important factor driving pertussis
doses. We also carefully adjusted for population, possibly in conjunction
epidemics in recent years, other
calendar date and age as we with other factors not addressed in
factors that we did not evaluate in
examined pertussis risk in relation to this study, suboptimal VE and waning
this study might also contribute to
vaccination status. Because we used played a major role in recent
pertussis epidemics individually or in
Cox models on a calendar time line, pertussis epidemics.
synergy. Results from studies in
pertussis cases were only compared
baboons suggest that the acellular
with other children at risk on the
pertussis vaccines are unable to
same day who have nearly the ABBREVIATIONS
prevent colonization, carriage, and
same age.
transmission.32–34 If this is also true ACIP: Advisory Committee on
Our study has some limitations. We for humans, this could contribute to Immunization Practice
could not fully disentangle the effects pertussis epidemics. The causes of aHR: adjusted hazard ratio
on risk of pertussis of age and time recent pertussis epidemics are CI: confidence interval
since last DTaP dose because these 2 complex, and we were only able to DTaP: diphtheria-tetanus-acellular
factors are highly correlated. In our address some aspects in our study. pertussis
primary analyses, we closely adjusted HR: hazard ratio
for months of age by comparing KPNC: Kaiser Permanente
children who were the same age in CONCLUSIONS Northern California
months, which resulted in minimal Compared with fully vaccinated PCR: polymerase chain reaction
variability in time since last DTaP children, pertussis risk was 13 times P-Y: person-year
doses and reduced precision of our higher among unvaccinated children Tdap: tetanus toxoid, reduced
estimates. Although we likely and 1.9 times higher among diphtheria toxoid, and
captured most pertussis cases in our undervaccinated children. Although acellular pertussis,
study, we may have missed mild cases undervaccinated and especially adsorbed
that did not come to medical unvaccinated children were at greater VE: vaccine effectiveness
attention or cases that presented risk of pertussis, they represented
.2 weeks after cough onset.31 In a small fraction of all pertussis cases.

Downloaded from www.aappublications.org/news by guest on November 11, 2020

PEDIATRICS Volume 144, number 1, July 2019 7
FUNDING: Funded by Kaiser Permanente Northern California and in part by grant 1K01AI139275-01 from the National Institute of Allergy and Infectious Diseases to
Dr Zerbo. Funded by the National Institutes of Health (NIH).
POTENTIAL CONFLICT OF INTEREST: Dr Klein reports potential conflicts of interest relevant to this article: the pertussis vaccines purchased by Kaiser Permanente
Northern California, which are the focus of this study, were manufactured by GlaxoSmithKline and Sanofi Pasteur; the others authors have indicated they have no
potential conflicts of interest to disclose.

REFERENCES
1. Hegerle N, Guiso N. Epidemiology of 10. Amirthalingam G. Strategies to control immunization requirements: secular
whooping cough & typing of Bordetella pertussis in infants. Arch Dis Child. trends and association of state policies
pertussis. Future Microbiol. 2013;8(11): 2013;98(7):552–555 with pertussis incidence. JAMA. 2006;
1391–1403 11. Winter K, Glaser C, Watt J, Harriman K; 296(14):1757–1763
2. Kilgore PE, Salim AM, Zervos MJ, Centers for Disease Control and 20. Atwell JE, Van Otterloo J, Zipprich J,
Schmitt HJ. Pertussis: microbiology, Prevention (CDC). Pertussis et al. Nonmedical vaccine exemptions
disease, treatment, and prevention. Clin epidemic–California, 2014. MMWR Morb and pertussis in California, 2010.
Microbiol Rev. 2016;29(3):449–486 Mortal Wkly Rep. 2014;63(48):1129–1132 Pediatrics. 2013;132(4):624–630
3. Cherry JD. Epidemic pertussis in 12. Winter K, Harriman K, Zipprich J, et al. 21. Omer SB, Enger KS, Moulton LH, Halsey
2012–the resurgence of a vaccine- California pertussis epidemic, 2010. NA, Stokley S, Salmon DA. Geographic
preventable disease. N Engl J Med. J Pediatr. 2012;161(6):1091–1096 clustering of nonmedical exemptions to
2012;367(9):785–787 13. Jackson DW, Rohani P. Perplexities of school immunization requirements and
pertussis: recent global associations with geographic clustering
4. Klein NP. Licensed pertussis vaccines in
epidemiological trends and their of pertussis. Am J Epidemiol. 2008;
the United States. History and current
potential causes. Epidemiol Infect. 2014; 168(12):1389–1396
state. Hum Vaccin Immunother. 2014;
10(9):2684–2690 142(4):672–684 22. Klein NP, Bartlett J, Fireman B, et al.
14. Dempsey AF, Schaffer S, Singer D, Waning protection following 5 doses of
5. Cody CL, Baraff LJ, Cherry JD, Marcy SM, a 3-component diphtheria, tetanus, and
Manclark CR. Nature and rates of Butchart A, Davis M, Freed GL.
Alternative vaccination schedule acellular pertussis vaccine. Vaccine.
adverse reactions associated with DTP 2017;35(26):3395–3400
and DT immunizations in infants and preferences among parents of young
children. Pediatrics. 1981;68(5):650–660 children. Pediatrics. 2011;128(5): 23. Klein NP, Bartlett J, Rowhani-Rahbar A,
848–856 Fireman B, Baxter R. Waning protection
6. Savoy M. ACIP releases 2017 childhood after fifth dose of acellular pertussis
immunization recommendations. Am 15. Glanz JM, Narwaney KJ, Newcomer SR,
et al. Association between vaccine in children. N Engl J Med. 2012;
Fam Physician. 2017;95(4):260–261 367(11):1012–1019
undervaccination with diphtheria,
7. Broder KR, Cortese MM, Iskander JK, tetanus toxoids, and acellular pertussis 24. McGirr A, Fisman DN. Duration of
et al; Advisory Committee on (DTaP) vaccine and risk of pertussis pertussis immunity after DTaP
Immunization Practices (ACIP). infection in children 3 to 36 months of immunization: a meta-analysis.
Preventing tetanus, diphtheria, and age. JAMA Pediatr. 2013;167(11): Pediatrics. 2015;135(2):331–343
pertussis among adolescents: use of 1060–1064
tetanus toxoid, reduced diphtheria 25. Quinn HE, Snelling TL, Macartney KK,
toxoid and acellular pertussis vaccines 16. Glanz JM, Newcomer SR, Narwaney KJ, McIntyre PB. Duration of protection
recommendations of the Advisory et al. A population-based cohort study after first dose of acellular pertussis
Committee on Immunization Practices of undervaccination in 8 managed care vaccine in infants. Pediatrics. 2014;
(ACIP). MMWR Recomm Rep. 2006; organizations across the United States. 133(3). Available at: www.pediatrics.
55(RR-3):1–34 JAMA Pediatr. 2013;167(3):274–281 org/cgi/content/full/133/3/e513

8. Hill HA, Elam-Evans LD, Yankey D, 17. Luman ET, Barker LE, McCauley MM, 26. Tartof SY, Lewis M, Kenyon C, et al.
Singleton JA, Kang Y. Vaccination Drews-Botsch C. Timeliness of Waning immunity to pertussis following
coverage among children aged 19- childhood immunizations: a state- 5 doses of DTaP. Pediatrics. 2013;131(4).
35 months - United States, 2016. MMWR specific analysis. Am J Public Health. Available at: www.pediatrics.org/cgi/
Morb Mortal Wkly Rep. 2017;66(43): 2005;95(8):1367–1374 content/full/131/4/e1047
1171–1177 18. Luman ET, McCauley MM, Stokley S, Chu 27. Schwartz KL, Kwong JC, Deeks SL, et al.
9. Choi YH, Campbell H, Amirthalingam G, SY, Pickering LK. Timeliness of Effectiveness of pertussis vaccination
van Hoek AJ, Miller E. Investigating the childhood immunizations. Pediatrics. and duration of immunity. CMAJ. 2016;
pertussis resurgence in England and 2002;110(5):935–939 188(16):E399–E406
Wales, and options for future control. 19. Omer SB, Pan WK, Halsey NA, et al. 28. Krieger N. Overcoming the absence of
BMC Med. 2016;14(1):121 Nonmedical exemptions to school socioeconomic data in medical records:

Downloaded from www.aappublications.org/news by guest on November 11, 2020

8 ZERBO et al
 validation and application of a census- Hum Vaccin Immunother. 2017;13(4): insights from the baboon model of
based methodology. Am J Public Health. 757–761 pertussis. J Infect. 2017;74(suppl 1):
1992;82(5):703–710 31. Lee AD, Cassiday PK, Pawloski LC, et al; S114–S119
29. Klein NP, Bartlett J, Fireman B, Baxter R. Clinical Validation Study Group. Clinical 33. Trainor EA, Nicholson TL, Merkel TJ.
Waning Tdap effectiveness in evaluation and validation of laboratory Bordetella pertussis transmission.
adolescents. Pediatrics. 2016;137(3): methods for the diagnosis of Bordetella Pathog Dis. 2015;73(8):ftv068
e20153326 pertussis infection: culture, polymerase 34. Warfel JM, Zimmerman LI, Merkel TJ.
30. Huang WT, Lin HC, Yang CH. chain reaction (PCR) and anti-pertussis Acellular pertussis vaccines protect
Undervaccination with diphtheria, toxin IgG serology (IgG-PT). PLoS One. against disease but fail to prevent
tetanus, and pertussis vaccine: 2018;13(4):e0195979 infection and transmission in
national trends and association 32. Pinto MV, Merkel TJ. Pertussis disease a nonhuman primate model. Proc Natl
with pertussis risk in young children. and transmission and host responses: Acad Sci U S A. 2014;111(2):787–792

Downloaded from www.aappublications.org/news by guest on November 11, 2020

PEDIATRICS Volume 144, number 1, July 2019 9
 Acellular Pertussis Vaccine Effectiveness Over Time
Ousseny Zerbo, Joan Bartlett, Kristin Goddard, Bruce Fireman, Edwin Lewis and
Nicola P. Klein
Pediatrics 2019;144;
DOI: 10.1542/peds.2018-3466 originally published online June 10, 2019;

Updated Information & including high resolution figures, can be found at:
Services http://pediatrics.aappublications.org/content/144/1/e20183466
References This article cites 34 articles, 11 of which you can access for free at:
http://pediatrics.aappublications.org/content/144/1/e20183466#BIBL
Subspecialty Collections This article, along with others on similar topics, appears in the
following collection(s):
Infectious Disease
http://www.aappublications.org/cgi/collection/infectious_diseases_su
b
Epidemiology
http://www.aappublications.org/cgi/collection/epidemiology_sub
Vaccine/Immunization
http://www.aappublications.org/cgi/collection/vaccine:immunization
_sub
Permissions & Licensing Information about reproducing this article in parts (figures, tables) or
in its entirety can be found online at:
http://www.aappublications.org/site/misc/Permissions.xhtml
Reprints Information about ordering reprints can be found online:
http://www.aappublications.org/site/misc/reprints.xhtml

Downloaded from www.aappublications.org/news by guest on November 11, 2020

 Acellular Pertussis Vaccine Effectiveness Over Time
Ousseny Zerbo, Joan Bartlett, Kristin Goddard, Bruce Fireman, Edwin Lewis and
Nicola P. Klein
Pediatrics 2019;144;
DOI: 10.1542/peds.2018-3466 originally published online June 10, 2019;

The online version of this article, along with updated information and services, is
located on the World Wide Web at:
http://pediatrics.aappublications.org/content/144/1/e20183466

Data Supplement at:

http://pediatrics.aappublications.org/content/suppl/2019/06/07/peds.2018-3466.DCSupplemental

Pediatrics is the official journal of the American Academy of Pediatrics. A monthly publication, it
has been published continuously since 1948. Pediatrics is owned, published, and trademarked by
the American Academy of Pediatrics, 345 Park Avenue, Itasca, Illinois, 60143. Copyright © 2019
by the American Academy of Pediatrics. All rights reserved. Print ISSN: 1073-0397.

Downloaded from www.aappublications.org/news by guest on November 11, 2020