Pertussis is a vaccine-preventable bacterial respiratory tract infection. Infants, often too young to have received their primary immunization series, have an increased risk of severe disease and mortality.1 Transmission from close contacts is the major source of infection in young children.2 In spite of high vaccination coverage, more cases have been reported in several European countries and in the United States over the last years, especially among adolescents.3 Norway reports a high rate of pertussis compared with other European countries, 85 cases per 100,000 population in 2012.4 The high number of cases in Norway is probably related to extensive laboratory testing and mandatory reporting of pertussis. In the period 1996–2010, the risk of pertussis infection among teenage children has increased, although the risk among infants did not significantly change.5
Worldwide, more than 1 in 10 babies are born preterm, that is, before 37 completed weeks of gestation.6 Infants born preterm could have a higher risk of pertussis than full-term infants because of incomplete transfer of maternal antibodies and an immature immune system.7 Preterm birth is correlated with low birth weight (LBW; <2500 g). Nevertheless, near full-term infants may have normal birth weight (NBW) and term infants with intra-uterine growth restrictions or other constitutional factors may have LBW. A few studies have found increased risk of pertussis in infants with LBW8,9 but one of these, a US cohort study, did not control for vaccination status, an important confounder since studies report delayed vaccinations in preterm infants.10,11 Furthermore, increased risk of pertussis hospitalization and pertussis-related mortality have been observed in preterm infants.12–14 However, no cohort studies assessing the risk of pertussis by preterm subcategories have, to our knowledge, been published.
We aimed to determine whether there is a higher risk of reported pertussis and pertussis hospitalizations among preterm than full-term infants during the first 2 years of life. We also wanted to estimate vaccine effectiveness (VE). This was done by linking data from the Medical Birth Registry of Norway (MBRN) to other national registries.
MATERIAL AND METHODS
We identified all live births in Norway between January 1, 1998, and December 31, 2010, through the MBRN. The MBRN contains extensive information about the pregnancy, the delivery and the infant.15 Gestational age (GA) is based on ultrasonography or the mother’s last menstrual period if ultrasonography was not recorded. The MBRN also contains date of death or date of emigration obtained through linkages with the Central Population Register on a regular basis. All Norwegian residents are assigned a unique, 11-digit identification number enabling linkage of data from MBRN with the Norwegian Surveillance system for Communicable Diseases (MSIS)16 and the Norwegian Immunisation Registry (SYSVAK).17 Each parent’s country of birth and level of education were obtained from Statistics Norway. Ethical clearance was obtained from the Regional Committee for Medical and Health Research Ethics, Southeast Norway.
Cases of pertussis were retrieved from MSIS.16 Pertussis has been mandatory notifiable since 1993. The notification includes data on disease and hospitalization. The case definition includes symptoms compatible with pertussis, and laboratory confirmation, or epidemiologic link to a laboratory-confirmed case. Laboratory confirmation is done by culture, serology or nucleic acid amplification test (NAAT). NAAT has been the most common method for diagnosis of pertussis in infants since 2002.
Since 1995, all vaccinations received within the Norwegian childhood immunization program are recorded in SYSVAK.17 Before this study (1952–1997), a whole cell vaccine was administered. From 1998, an acellular pertussis vaccine (Infanrix; GSK, Rixensart, Belgium) was used until it was replaced by Infanrix-polio+Hib (GSK, Rixensart, Belgium) in 2001. The recommended schedule was 3 doses at 3, 5 and 11–12 months of age regardless of GA.18 For the last decade, 90%–95% of infants have received 3 doses by 2 years of age,19 but about 20% receive the third dose later than recommended.20 In 2006, a booster vaccine (Tetravac; Sanofi Pasteur MSD, Diegem, Belgium) was introduced at 7–8 years of age.
All infants born alive in the period 1998–2010 who were residents of Norway were eligible for the study, in total 761,580 infants. We excluded infants who died within the first week of birth (n = 1328), multiple births (n = 26,640), infants with a GA less than 23 weeks 0 days (n = 59) or greater than or equal to 45 weeks 0 days (n = 191), unknown GA (n = 9592) or unknown birth weight (n = 504). Furthermore, infants with a sex- and GA-specific birthweight outside 3 standard deviations of sex- and GA-specific mean (n = 4566) were excluded, because we assumed that the recorded GA was incorrect.21,22 Finally, we excluded infants without information on country of birth for at least 1 parent (n = 299) and infants without information on education for at least 1 parent (n = 5235). A total of 713,166 infants were included in the analyses. The infants in the study sample were followed from birth until pertussis diagnosis, death, emigration or 730 days old, whichever occurred first.
Based on GA, the infants were categorized as full term (≥37 weeks) or preterm (23–36 weeks 6 days). Preterm infants were furthermore categorized as 35–36, 32–34, 28–31 and 23–27 weeks. The infants were also categorized according to birthweight with the following categories: ≥2500, 2000–2499, 1500–1999, 1000–1499 or <1000 g.
Each child’s number of siblings at birth was defined as the number of previous births to the mother and categorized as 0, 1, 2 or ≥3. Infants were classified as having an immigrant background if at least 1 parent was born outside Norway. Parental education was defined as mother’s attained education, categorized as compulsory education, secondary education, undergraduate level higher education or graduate level higher education. Father’s attained education was used for the 14,048 (2.0%) infants whose mothers did not have information on education.
We used a χ2 test to evaluate whether preterm and full-term cases differed in terms of sex, hospitalization status or diagnostic method. To compare age at diagnosis, we used a Wilcoxon rank-sum test. Crude incidence rates of reported pertussis and pertussis hospitalization for full-term and preterm infants were calculated as number of events divided by accrued person-time. For reported pertussis cases, we also calculated rates by the following age groups: 0 months (1–31 days), 1 month (32–61 days), 2 months (62–92 days), 3–5 months (93–183 days), 6–8 months (184–274 days), 9–11 months (275–365 days), 12–14 months (366–456 days), 15–17 months (457–547 days) and 18–23 months (548–730 days) and by calendar year. Since study subjects were born between 1998 and 2010, the age distributions of the person-time accrued in 1998–1999 and 2011–2012 were different from the overall age distribution. Therefore, we did not calculate rates for these years. We compared rates for 2000–2005 and 2007–2010, that is, before and after introduction of the booster dose at 7–8 years of age. Follow-up time for 2006, the year the booster dose was introduced, was excluded from this comparison. Incidence rate ratios (IRRs) and corresponding 95% confidence intervals (CIs) were estimated with Poisson regression. The multivariate model was adjusted for attained age (time scale) (1–31, 32–61, 62–92, 93–183, 184–365, 366–547 and 548–730 days), number of doses of pertussis vaccine (0, 1, 2 or 3), sex of child, mother’s education, immigrant background and number of siblings. Vaccination was modeled as a time-dependent variable with a lag of 14 days, that is, the number of received doses was updated 14 days after each dose.
We estimated the association between prematurity and pertussis risk for 1998–2005 and 2007–2012, respectively, by including interaction terms between number of vaccine doses and period. To evaluate whether VE differed among full-term and preterm infants, we included interaction terms between number of vaccine doses and a prematurity indicator variable. From this model, we estimated IRRs of 1, 2 and 3 vaccine doses compared with 0 doses for preterm and full-term infants, respectively. The VE for preterm and full-term infants was then calculated as (1 − IRR) × 100%. Tests of interaction were evaluated with a likelihood ratio test.
All tests were two sided, and P value less than 0.05 was considered statistically significant. The data were analyzed with Stata/SE 14.0 (StataCorp, College Station, TX).
Of the 713,166 infants included in our study, 36,913 (5.2%) were born preterm. Characteristics of infants in the study by GA are described in Table 1. We identified 999 reported pertussis cases. Of these, 968 cases (96.9%) were laboratory confirmed: 475 (47.5%) with NAAT, 327 (32.7%) with serology and 166 (16.6%) with culture only. Cases confirmed by serology were older than those confirmed using culture or NAAT, median age 290, 83.5 and 122 days, respectively, P < 0.001. Median age at diagnosis was 139 days, and 54.5% of the cases were boys. Of all reported cases, 460 (46.0%) were hospitalized with pertussis. The proportion of cases that were hospitalized differed with age (Fig. 1). Preterm cases had a significantly higher risk of hospitalization than full-term cases at 3–5 months of age (P = 0.01). We did not find statistically significant differences between the full-term and preterm cases in terms of diagnostic method, age at diagnosis or sex (P ≥ 0.28). According to the outcomes registered in MSIS, 1 case (with GA 35 weeks) died of pertussis. The incidence rate of reported pertussis in the study period was 67.9 (95% CI: 63.7–72.5) cases per 100,000 person-years for full-term infants and 115.2 (95% CI: 93.0–142.6) cases per 100,000 person-years for preterm infants. The incidence rate of pertussis with hospitalization was 30.7 (95% CI: 27.9–33.8) and 63.1 (95% CI: 47.2–84.2) cases per 100,000 person-years in full-term and preterm infants, respectively.
In both full-term and preterm infants, the highest incidence rate of reported pertussis was observed in 2004, 118.0 and 282.4 cases per 100,000 person-years, respectively. For the period 2000–2005, the incidence rate of reported pertussis was 67.5 (95% CI: 61.3–74.3) and 117.9 (95% CI 86.5–160.7) cases per 100,000 person-years in full-term and preterm infants, respectively. The rates were not significantly higher in the period 2007–2010, IRR = 0.90 (95% CI: 0.77–1.05) and 0.89 (95% CI: 0.52–1.52) for full-term and preterm infants, respectively.
A steep increase in the pertussis incidence rate from 0 months to 1 month of age was observed for both full-term and preterm infants (Fig. 2). The rate peaked at 2 months of age in full-term infants (233.6 cases per 100,000 person-years) and 1 month of age in preterm infants (431.1 cases per 100,000 person-years). The incidence rate remained low from 6 months of age in both full-term and preterm infants.
In the multivariable analysis, we observed a significantly higher rate of reported pertussis in preterm than in full-term infants, IRR = 1.65 (95% CI: 1.32–2.07). Furthermore, compared to full-term infants, a significantly higher rate was found for infants who were born at GA 35–36, 32–34 and 23–27 weeks, IRRs = 1.49 (95% CI: 1.11–2.01), 1.63 (95% CI: 1.06–2.51) and 4.49 (95% CI: 2.33–8.67), respectively (Table 2). The rate in infants born at GA 28–31 weeks was similar to that of the rate in infants born at GA 32–36 weeks. However, this was not significantly higher when compared with full-term infants, because of few events among the infants born at GA 28–31 weeks, IRR = 1.57 (95% CI: 0.78–3.15). A similar pattern of increased risk by low GA was also observed for pertussis hospitalizations, but the associations were slightly stronger, with IRR = 1.99 (95% CI: 1.47–2.71) when comparing preterm to full-term infants. When we restricted the analyses to NAAT and culture-confirmed cases, we still found significantly higher rates of pertussis in preterm infants when compared with full-term infants, IRR = 1.82 (95% CI: 1.40–2.37) for reported pertussis and IRR = 2.11 (95% CI: 1.50–2.97) for pertussis hospitalization.
We found that infants with LBW had a significantly higher risk of pertussis than NBW infants, IRR = 1.62 (95% CI: 1.22–2.14) for reported pertussis and IRR = 1.92 (95% CI: 1.31–2.82) for pertussis hospitalization. Compared to infants with NBW, we found significantly higher rates of both reported pertussis and pertussis hospitalizations among infants in the 2 lowest birth weight categories (1000–1499 or <1000 g) (Table 3). However, the rates were not significantly higher for infants with birth weight 2000–2499 or 1500–1999 g.
The associations between prematurity and pertussis risk before and after introduction of the booster dose in infants 7–8 years of age did not differ, P = 0.48, IRR = 1.84 (95% CI: 1.38–2.46) and 1.55 (95% CI: 1.04–2.29) for 1998–2005 and 2007–2012, respectively.
In full-term infants, the VE against reported pertussis for the third dose was 88.8% (95% CI: 84.3–92.0), whereas in preterm infants, the VE was 93.0% (95% CI: 85.8–96.5) (Table 4). The VE for the third dose against pertussis hospitalization was 91.1% (95% CI: 78.8–96.3) in full-term infants and 88.7% (95% CI: 54.3–97.2) in preterm infants. The VE did not differ significantly in full-term and preterm infants for either reported pertussis (P = 0.20) or pertussis hospitalization (P = 0.81).
In this national cohort, infants born at GA less than 37 weeks had increased risk of reported pertussis compared with those born at GA 37 weeks old or older. The risk was particularly high in infants born before GA 28 weeks, but an increased risk was also evident for infants born at GA 35–36 weeks. Furthermore, the risk of pertussis hospitalization was increased in preterm infants. The VE against pertussis disease and hospitalization was similar in preterm and full-term infants.
Importantly, the increased risk that we observed was not due to difference in vaccination status or socioeconomic factors among preterm and full-term infants. Although this is the first cohort study that found increased risk in infants with GA less than 37 weeks, an overrepresentation in preterm/LBW infants has also been described elsewhere.8,9,13,14 The increased risk could be related to incomplete transfer of maternal antibodies and an immature immune system. Immunoglobulin G antibodies are transferred from mother to fetus from 16 weeks of gestation, but an abundance of IgG is acquired during the last month of full-term pregnancy.23,24
We found that “late preterm” infants (GA: 35–36 weeks) had an increased risk of pertussis when compared with full-term infants. Although other studies have used LBW to identify infants with increased risk of pertussis, we would have underestimated the number of infants with increased risk by using LBW instead of GA, because many of the late preterm infants have NBW. Infants born close to term have similar size and weight as full-term infants, but they have increased morbidity and mortality.25 Immature lung structure and immune system put these infants at higher risk of infectious diseases.26 This is also the largest group of preterm infants in numbers, which is of public health importance. Hence, early and timely vaccination is important in infants born close to term.
We found a 4-fold increased risk of reported pertussis in infants born at GA less than 28 weeks compared with those born at GA 37 weeks old or older. The cohort study by Langkamp and Davis8 found a 4-fold increased risk in infants with birthweight less than 1500 g compared with those of NBW. Severe acute and chronic morbidities and mortality are high in infants born before 28 weeks or who have birth weight less than 1500 g. Infants born at GA 23–27 weeks will have few maternal antibodies present at birth. Although early second-trimester maternal immunization leads to neonatal antibodies in term infants,27 this has not been shown for (extreme) preterm neonates. To vaccinate all close contacts after birth (cocooning) could reduce the risk of pertussis transmission although alone, it is an insufficient strategy to prevent morbidity and mortality in new borns.28 Providing postexposure antimicrobial prophylaxis to infants that had contact with a pertussis case can prevent infection.29
The increased risk of pertussis hospitalizations in preterm/LBW infants described in the present study and elsewhere8,12 may indicate a prolonged susceptibility to severe disease in these infants as prematurity has been associated with more severe disease13 and mortality.14 This is supported by the persistent high hospitalization rate among cases, which was significantly higher than in full-term cases at 3–5 months of age. Although increased doctors’ awareness of preterm infants may impact pertussis diagnostics and hospitalization rates, this is probably less likely for “late preterm” infants than for extreme preterm infants. As young infants more often have severe disease compared with older infants, the age at disease onset and clinical presentation are probably more important hospitalization criteria than a history of preterm birth.
We estimated the VE to be similar in preterm and full-term infants. VE data on pertussis disease in preterm infants have not been published previously. In addition, we found that the VE against pertussis hospitalization in preterm infants was similar to that in full-term infants. Similar data for pertussis hospitalization were reported in a Danish study.12 Studies have shown that geometric mean titers to relevant pertussis antigens after 3 doses of pertussis-containing vaccines were similar in preterm and term infants.30,31 These data show the benefit of vaccinating preterm and full-term infants. Delaying vaccinations in preterm infants10,11 should be avoided as vaccinations reduce the burden of pertussis. Since 2013, Norwegian recommendations says that preterm infants less than GA 32 weeks and/or less than 1500 g should be vaccinated from 2 months chronologic age. To administer the first dose at 6 weeks and the second dose at 3 months could further decrease pertussis morbidity in preterms, although the VE for the first 2 doses could possibly be lower when administered early.
Among the strengths of this study are the cohort design and a large sample size that included all births in Norway and linkage to other national registries. Reporting of pertussis cases, vaccinations and birth data to national registries in Norway are mandatory. Almost all patients had a laboratory-confirmed diagnosis. We were able to control for socioeconomic factors and vaccination status, and we excluded multiple births to avoid overestimating the increased risk of preterm infants.
The limitations of our study include a lack of information about coinfections, comorbidity and severity of pertussis including duration of hospitalizations. Almost every case was laboratory confirmed; hence, it is very likely that reported pertussis underestimates the true disease incidence especially for mild cases. The severity of pertussis in young infants suggests that underestimation in young infants is less probable than in older age groups. In about one third of the cases, the diagnosis was confirmed by serology, either by seroconversion or increased antibody in 2 separate samples, or by high antibody level in a single sample. As antibodies detected can result from vaccination or maternal transfer as well as infection, serologic confirmation in infants is difficult without information about antibody kinetics. Maternal antibodies are probably of limited importance because the median age was 9.5 months in cases confirmed by serology. Previous vaccinations cannot be ruled out as a cause of positive antibody test. However, the increased risk in preterm compared with full-term infants was similar when serologically confirmed cases were omitted. Furthermore, the case definition also includes symptoms compatible with pertussis, and cases should be notified from both laboratory and clinician.
In summary, we provide novel evidence of increased risk of pertussis in preterm infants compared with full-term infants. The risk was particularly high in infants born before GA 28 weeks; however, an increased risk was also found in infants born at GA 35–36 weeks. The VE against pertussis disease and hospitalization was similar among preterm and full-term infants. Early and timely pediatric vaccinations as well as other strategies to prevent transmission to preterm infants are of utmost importance.
2. de Greeff SC, de Melker HE, Westerhof A, et al. Estimation of household transmission rates of pertussis and the effect of cocooning vaccination strategies on infant pertussis. Epidemiology. 2012;23:852–860.
3. Tan T, Dalby T, Forsyth K, et al. Pertussis across the globe: recent epidemiologic trends from 2000 to 2013. Pediatr Infect Dis J. 2015;34:e222–e232.
5. Lavine JS, Bjørnstad ON, de Blasio BF, et al. Short-lived immunity against pertussis, age-specific routes of transmission, and the utility of a teenage booster vaccine. Vaccine. 2012;30:544–551.
7. Bonhoeffer J, Siegrist CA, Heath PT. Immunisation of premature infants. Arch Dis Child. 2006;91:929–935.
8. Langkamp DL, Davis JP. Increased risk of reported pertussis and hospitalization associated with pertussis in low birth weight children. J Pediatr. 1996;128(5, pt 1):654–659.
9. Zamir CS, Dahan DB, Shoob H. Pertussis in infants under one year old: risk markers and vaccination status–a case-control study. Vaccine. 2015;33:2073–2078.
10. Tozzi AE, Piga S, Corchia C, et al. Timeliness of routine immunization in a population-based Italian cohort of very preterm infants: results of the ACTION follow-up project. Vaccine. 2014;32:793–799.
11. Ochoa TJ, Zea-Vera A, Bautista R, et al. Vaccine schedule compliance among very low birth weight infants in Lima, Peru. Vaccine. 2015;33:354–358.
12. Hviid A. Effectiveness of two pertussis vaccines in preterm Danish children. Vaccine. 2009;27:3035–3038.
13. Marshall H, Clarke M, Rasiah K, et al. Predictors of disease severity in children hospitalized for pertussis during an epidemic. Pediatr Infect Dis J. 2015;34:339–345.
14. Haberling DL, Holman RC, Paddock CD, et al. Infant and maternal risk factors for pertussis-related infant mortality in the United States, 1999 to 2004. Pediatr Infect Dis J. 2009;28:194–198.
15. Irgens LM. The Medical Birth Registry of Norway. Epidemiological research and surveillance throughout 30 years. Acta Obstet Gynecol Scand. 2000;79:435–439.
16. Norwegian Surveillance System for Communicable Diseases (MSIS). Available at: http://www.msis.no/
. Accessed June 8, 2016.
17. Trogstad L, Ung G, Hagerup-Jenssen M, et al. The Norwegian immunisation register–SYSVAK. Euro Surveill. 2012;17:20147.
18. Nøkleby H. [Vaccination of premature infants]. Tidsskr Nor Laegeforen. 1990;110:3781–3782.
19. World Health Organization. WHO Vaccine-Preventable Diseases: Monitoring System. 2016 Global Summary. May 27, 2016. Available at: http://apps.who.int/immunization—
monitoring/globalsummary. Accessed June 9, 2016.
20. Riise ØR, Laake I, Bergsaker MA, et al. Monitoring of timely and delayed vaccinations: a nation-wide registry-based study of Norwegian children aged < 2 years. BMC Pediatr. 2015;15:180.
21. Skjaerven R, Gjessing HK, Bakketeig LS. Birthweight by gestational age in Norway. Acta Obstet Gynecol Scand. 2000;79:440–449.
22. Moster D, Lie RT, Markestad T. Long-term medical and social consequences of preterm birth. N Engl J Med. 2008;359:262–273.
23. Simister NE. Placental transport of immunoglobulin G. Vaccine. 2003;21:3365–3369.
24. de Voer RM, van der Klis FR, Nooitgedagt JE, et al. Seroprevalence and placental transportation of maternal antibodies specific for Neisseria meningitidis serogroup C, Haemophilus influenzae
type B, diphtheria, tetanus, and pertussis. Clin Infect Dis. 2009;49:58–64.
25. Shapiro-Mendoza CK, Tomashek KM, Kotelchuck M, et al. Risk factors for neonatal morbidity and mortality among “healthy,” late preterm newborns. Semin Perinatol. 2006;30:54–60.
26. Engle WA, Tomashek KM, Wallman C; Committee on Fetus and Newborn, American Academy of Pediatrics. “Late-preterm” infants: a population at risk. Pediatrics. 2007;120:1390–1401.
27. Eberhardt CS, Blanchard-Rohner G, Lemaître B, et al. Maternal immunization earlier in pregnancy maximizes antibody transfer and expected infant seropositivity against pertussis. Clin Infect Dis. 2016;62:829–836.
28. Centers for Disease Control and Prevention (CDC). Updated recommendations for use of tetanus toxoid, reduced diphtheria toxoid and acellular pertussis vaccine (Tdap) in pregnant women and persons who have or anticipate having close contact with an infant aged <12 months—Advisory Committee on Immunization Practices (ACIP), 2011. MMWR Morb Mortal Wkly Rep. 2011;60:1424–1426.
29. Tiwari T, Murphy TV, Moran J; National Immunization Program, CDC. Recommended antimicrobial agents for the treatment and postexposure prophylaxis of pertussis: 2005 CDC Guidelines. MMWR Recomm Rep. 2005;54(RR-14):1–16.
30. Omeñaca F, Garcia-Sicilia J, García-Corbeira P, et al. Response of preterm newborns to immunization with a hexavalent diphtheria-tetanus-acellular pertussis-hepatitis B virus-inactivated polio and Haemophilus influenzae
type b vaccine: first experiences and solutions to a serious and sensitive issue. Pediatrics. 2005;116:1292–1298.
31. Slack MH, Cade S, Schapira D, et al. DT5aP-Hib-IPV and MCC vaccines: preterm infants’ response to accelerated immunisation. Arch Dis Child. 2005;90:338–341.
pertussis; preterm; gestational age; vaccine effectivenessCopyright © 2017 Wolters Kluwer Health, Inc. All rights reserved.