Effectiveness of Prenatal Tetanus, Diphtheria, and Acellular Pertussis Vaccination on Pertussis Severity in Infants

(See the Major Article by Winter et al on pages 3–8.)

Young infants are at greatest risk of severe disease and death from pertussis [1–8]. From 1994 to 2004, 58% of infant pertussis deaths in the United States occurred in infants <2 months of age [9] and in 2010 in California, 10 deaths from pertussis occurred in infants <8 weeks of age [10].

Passively acquired maternal antibodies against many infectious diseases offer protection during the first months of life [11]. In the United States, women are recommended to receive tetanus, diphtheria, and acellular pertussis (Tdap) vaccine at 27–36 weeks gestation during each pregnancy to protect infants through transplacental transfer of antibodies [12]. Data from the United Kingdom show that this strategy reduces the risk of pertussis among infants <2 months of age [13, 14]. However, to our knowledge, no studies have evaluated the impact of this strategy on the outcomes of infected infants. Studies have shown that infants who develop pertussis after receiving the first dose of diphtheria and tetanus toxoids and acellular pertussis (DTaP) vaccine have lower risks of hospitalization and death than unvaccinated infants [8, 15]. It is possible that, although still infected, these infants receive sufficient antibody to pertussis toxin to prevent development of severe leukocytosis, which can lead to irreversible pulmonary hypertension and death [16, 17].

In the baboon model, infant baboons born to mothers vaccinated against pertussis in the third trimester of pregnancy were protected against leukocytosis after a direct challenge with Bordetella pertussis [18]. Similarly, infants born to women who receive Tdap vaccine during pregnancy who are infected with B. pertussis may obtain sufficient antibody to pertussis toxin and other antigens to prevent severe disease and death [19, 20]. In the current study, we retrospectively evaluate a cohort of infants with pertussis to determine whether those whose mothers received Tdap vaccine during pregnancy had less severe pertussis illness than those whose mothers were unvaccinated during pregnancy.

METHODS

By California statute, laboratories and clinicians must report suspected cases of pertussis to local public health departments; all cases with an acute cough illness meeting the Council of State and Territorial Epidemiologists clinical case definition [21] or with laboratory detection of B. pertussis by culture or polymerase chain reaction are subsequently reported to the California Department of Public Health (CDPH). Local health department investigators complete case report forms containing demographic, clinical, laboratory, and epidemiologic data through patient and provider interviews and medical record reviews. In addition, medical records are requested and reviewed by CDPH staff to ensure accuracy of the surveillance data for hospitalized cases occurring in young infants. We evaluated the cohort of infants reported in California with pertussis onset occurring at <63 days of age and born from 1 January 2011 through 31 December 2015. We used an age cutoff of 63 days because most infants receive the first dose of DTaP vaccine at about 2 months of age, and we sought to minimize potential confounding effects of protection from routine active immunization.

CDPH pertussis surveillance reports were linked to birth certificate records from the CDPH Center for Health Statistics for the mother and infant and California Immunization Registry (CAIR) Tdap immunization records for the mother using a matching algorithm consisting of combinations of the following variables: infant's last name, infant's first name, infant's date of birth, infant's sex, mother's last name, mother's first name, mother's date of birth, father's last name, and city of residence.

Symptoms, clinical course, DTaP vaccination, and outcomes were reported on surveillance reports and verified through medical record review. Length of hospital stay was calculated as the number of days between admission and discharge dates and summed across hospital stays for multiple admissions or transfers. The infant's gestational age and birth weight, and maternal age, country of birth, date of last menses, and number of prenatal visits were reported on the birth certificate. Maternal Tdap vaccination histories were self-reported on public health case investigation reports or verified by the provider through documentation in the mother or infant's medical record or in CAIR.

Mothers were considered to have had prenatal Tdap if they self-reported or had documented receipt of Tdap vaccine in their medical record or in CAIR at any point during the pregnancy for the infected infant. Where available, the trimester and date of Tdap receipt were recorded; week of gestation for the dose administration was calculated as the number of weeks between the date of last menses reported on the birth certificate and Tdap dose.

Mothers were considered unvaccinated if they self-reported or had with documentation in their medical record of not receiving Tdap vaccine during pregnancy (eg, vaccine refusal). In addition, mothers who had documentation in their medical record or CAIR of receipt of Tdap within 2 weeks after the delivery were assumed to have been unvaccinated during the pregnancy, because postpartum Tdap is recommended only for unvaccinated mothers. If no Tdap dose was recorded on the surveillance report, medical records, or CAIR and the mother did not recall Tdap receipt during or immediately after pregnancy, prenatal Tdap was considered unknown.

Bivariate comparisons of demographics, symptoms, clinical course and outcomes between infants born to vaccinated and unvaccinated mothers were calculated using χ² and Fisher exact tests; associated relative risk (RR) and 95% confidence intervals (CIs) were constructed. Two-sided P values were calculated using analysis of variance or the Wilcoxon-Mann–Whitney test for continuous variables and the Cochran-Armitage test to evaluate trends, as appropriate.

To avoid misclassification of Tdap vaccination, infants were excluded from analyses if their mother's vaccination status was unknown. To assess potential bias due to missing data, bivariate comparisons of demographics, birth and clinical characteristics of infant pertussis cases with and without known exposure data were evaluated and associated odds ratios (ORs) were calculated. In addition, birth certificate records were unavailable for infants born in 2015 or born outside California and were excluded from estimations requiring maternal age, maternal country of birth, number of prenatal visits, gestational age, and birth weight.

The primary outcomes of interest were hospitalization for ≥24 hours, length of hospital stay, admission to an intensive care unit (ICU) for any length of time, intubation, and death from pertussis. Using multivariate logistic regression, we calculate ORs and 95% CIs to estimate the effect of prenatal Tdap vaccination on risk of (1) hospitalization and (2) ICU admission, while simultaneously adjusting for the important covariates of infant's chronological age, gestational age, and receipt of DTaP vaccine.

The vaccine effectiveness (VE) of prenatal Tdap in preventing hospitalization among infants with pertussis was calculated as (1 – OR) × 100% [22]. VE estimates adjusted for chronological and gestational age of the infant and receipt of DTaP were also calculated. A subanalysis included only infants whose mothers received Tdap vaccine during the third trimester.

RESULTS

A total of 752 infants born from 1 January 2011 through 31 December 2015 were reported with pertussis at <63 days of age. Among those with complete information, 495 (68%) were hospitalized, 25% of whom required intensive care. Forty-four infants (7%) required intubation, and 6 (1%) died. Four (1%) received the first dose of DTaP vaccine ≥14 days before onset of pertussis illness. Three infants were born outside California, and 116 were born in 2015 and did not have available birth data. Of the remaining 633 infants, 607 (96%) were successfully linked to their birth certificate record.

A total of 420 (56%) infants had known maternal vaccination status. Comparisons of infants with and without complete data are presented in Table 1. Characteristics were similar between these 2 groups, but infants with missing exposure data were significantly more likely to be preterm, have apnea, be admitted to the ICU, have seizures, or die (Table 1).

Of the 420 infants with complete maternal vaccination data, 49 (12%) had mothers who received Tdap during pregnancy, whereas the remaining mothers were unvaccinated or vaccinated post partum. Among those vaccinated during pregnancy, 8 (16%) received Tdap during the first or second trimester, 37 (76%) received Tdap during the third trimester, and 4 (8%) received Tdap at an unknown point during the pregnancy. Among the 18 mothers vaccinated during the third trimester with documented administration dates, 14 (78%) received Tdap during the recommended window of 27–36 weeks gestation, and 4 (22%) received Tdap after 36 weeks gestation; ≥5 infants (10%), including all 4 whose mothers were vaccinated after 36 weeks gestation, were born <14 days after maternal vaccination.

Vaccinated mothers were less likely to have Medicaid insurance (RR, 0.6; 95% CI, .4–.9). There was no difference in age, country of birth, or number of prenatal visits between vaccinated and unvaccinated mothers (Table 2). Infants born to vaccinated mothers were older when they developed pertussis (P = .03) but did not differ with regard to race or ethnicity, sex, gestational age, birth weight, or DTaP vaccination (Table 2). Clinical symptoms differed between the groups, and infants born to vaccinated mothers were less likely to have the classic pertussis symptoms of paroxysmal cough (RR, 0.41; 95% CI, .25–.68), apnea (0.66; .47–.91), cyanosis (0.53; .39–.73), or whoop (0.78; .62–.99), although the frequency of posttussive vomiting was similar between groups (Table 2).

Infants whose mothers received prenatal Tdap had significantly lower risk of hospitalization (RR, 0.5; 95% CI, .4–.6) or ICU admission (0.8; .7–.9). Among hospitalized infants, those whose mothers were vaccinated had shorter hospitalization stays (median 3 vs 6 days; P = .02). No infants born to vaccinated mothers had seizures, required intubation, or died (Table 2).

In the first multivariate regression model (model 1), infants whose mothers received Tdap vaccine during pregnancy were significantly less likely to have been hospitalized (OR, 0.4; 95% CI, .2–.9) after adjustment for the infant's chronological and gestational age and receipt of DTaP (Table 3). In model 2, no infants who received DTaP were admitted to an ICU; this was dropped from the model to allow for convergence. After adjustment for the infant's chronological and gestational age, prenatal Tdap was associated with a lower odds of ICU admission (OR, 0.5; 95% CI, .2–1.2), but the CI included the null value (Table 3).

The overall VE for preventing hospitalization among infants with pertussis was 72% (95% CI, 49%–85%); after adjustment for the infant's chronological and gestational age and receipt of DTaP vaccine, it was 58% (15%–80%) (Table 4). VE estimates remained similar when restricted to infants whose mothers were known to have received Tdap vaccine during the third trimester.

DISCUSSION

This is the first known study demonstrating that prenatal Tdap vaccination of mothers reduces the severity of disease in infants who are infected with pertussis. Infected infants whose mothers received Tdap during pregnancy had a lower risk of hospitalization and ICU admission and shorter hospital stays than infants whose mothers were unvaccinated. No infected infants born to vaccinated mothers required intubation or died. Classic pertussis symptoms were observed less frequently among infants born to vaccinated mothers, suggesting that they had more mild disease.

Among the women vaccinated with Tdap during pregnancy, ≥24% were vaccinated outside the recommended window of 27–36 weeks gestation, and ≥5 (10%) infants were born <14 days after vaccination; it is possible that optimal timing of vaccination could have prevented pertussis infection for these infants altogether. Recently, high pertussis antibody titers were observed in infants born to mothers vaccinated late in the second or early in the third trimester [23], stressing the importance of prompt vaccination. All pregnant women should be offered Tdap vaccination during routine prenatal visits at the earliest opportunity, starting at 27 weeks gestation.

This study has several important limitations. Foremost, misclassification of Tdap immunization could have occurred because maternal vaccination histories were not exclusively provider verified, and there may have been recall bias if mothers of severely ill infants reported Tdap vaccination differently from mothers of infants with more mild disease. However, the use of several overlapping data sources, including those with provider verification to obtain maternal vaccination histories, reduced the risk of recall bias and exposure misclassification.

Complete maternal Tdap immunization data were available for only 56% of infants with pertussis, and some differences were observed between infants with and those without complete data. Infants with known maternal Tdap status were more likely than with missing data to be admitted to the ICU, have seizures, or die. It is possible that public health investigators were more persistent in obtaining maternal vaccination histories for more severe cases. Hospital admission records for infants were available as an additional data source for more severe cases. There may have been differential completeness of the exposure data because documentation of nonreceipt of vaccine (eg, notation of refusal or postpartum vaccination) was necessary to determine nonexposure. Therefore, it is possible that more infants with unknown maternal Tdap status were born to unvaccinated mothers. Most characteristics were similar between our infants with and those without maternal Tdap data, so it unclear whether incomplete data had a meaningful impact on our study results.

The cohort of infected infants in this study includes only those with reported pertussis, so some cases may have been missed. However, because pertussis reporting is mandated by both laboratories and clinicians and ill infants in this age group are often brought to medical attention, underreporting, particularly of hospitalized infants, is probably lessened. In addition, routine surveillance activities include reviews of hospital discharge and death certificate databases to identify possible infant pertussis cases that may not have been initially reported.

Finally, our VE is only a partial estimate of the true impact of prenatal Tdap vaccination. All the infants in our cohort were infected with pertussis; therefore, we are unable to estimate the effect of prenatal Tdap on preventing pertussis illness in infants. Many mothers received postpartum Tdap but information on breastfeeding status was not collected, so we could not evaluate possible protective effects of postpartum Tdap or antibodies transferred via breast milk, which may have lowered the VE estimates presented here.

California is a large state with a diverse population; therefore our findings should be generalizable to other populations. Other countries experiencing a high incidence of pertussis should consider recommending routine Tdap vaccination of pregnant women. These findings should be communicated to both prenatal care providers and pregnant women to encourage uptake of Tdap vaccine during each pregnancy.

In conclusion, prenatal Tdap vaccination was 58% effective in preventing hospitalizations in infants infected with pertussis. Infected infants whose mothers did not receive Tdap during pregnancy had a higher risk of hospitalization and ICU admission and longer hospital stays. No infants born to vaccinated mothers required intubation or died of pertussis. Prenatal Tdap vaccination of mothers is a critical strategy for reducing morbidity and mortality rates associated with pertussis.

Notes

Acknowledgments. We are grateful for Rosie Glenn-Finer from CDPH for her assistance with data collection and to Arthur Reingold from the University of California, Berkeley, School of Public Health for his critical review of this manuscript. We are indebted to the 61 local health jurisdictions for their ongoing pertussis surveillance and vaccination efforts.

Potential conflicts of interest. J. D. C. has served as a consultant for GlaxoSmithKline and as a speaker for Sanofi Pasteur, outside the submitted work. All other authors report no potential conflicts. All authors have submitted the ICMJE Form for Disclosure of Potential Conflicts of Interest. Conflicts that the editors consider relevant to the content of the manuscript have been disclosed.

References