Characteristics of Hospitalized Cases of Pertussis in Catalonia and Navarra, Two Regions in the North of Spain

Abstract

Pertussis causes a large number of cases and hospitalizations in Catalonia and Navarra. We made a study of household cases of pertussis during 2012 and 2013 in order to identify risk factors for hospitalization in pertussis cases. Each primary case reported triggered the study of their contacts. Close contacts at home and people who were in contact for >2 hours during the transmission period of cases were included. The adjusted OR and 95% confidence intervals (CI) was calculated using logistic regression. A total of 1124 pertussis cases were detected, of which 14.9% were hospitalized. Inspiratory whoop (aOR: 1.64; CI: 1.02–2.65), apnoea (aOR: 2.47; CI: 1.51–4.03) and cyanosis (aOR: 15.51; CI: 1.87–128.09) were more common in hospitalized than in outpatient cases. Hospitalization occurred in 8.7% of correctly-vaccinated cases, 41.1% of non-vaccinated cases and 9.4% of partially-vaccinated cases. In conclusion, inspiratory whoop, apnoea and cyanosis were associated factors to hospitalization while vaccination reduced hospitalizations due to pertussis.

About the Authors:

Inma Crespo

* E-mail: [email protected]

Affiliations CIBER Epidemiology and Public Health. Barcelona, Spain, Public Health Department, University of Barcelona, Barcelona, Spain

Diana Toledo

Affiliations Public Health Department, University of Barcelona, Barcelona, Spain, Public Health Agency, Barcelona, Spain

Núria Soldevila

Affiliations CIBER Epidemiology and Public Health. Barcelona, Spain, Public Health Department, University of Barcelona, Barcelona, Spain

Iolanda Jordán

Contributed equally to this work with: Iolanda Jordán, Rubén Solano, Joan A. Caylà

Affiliation: Paediatric Intensive Care Medicine, Hospital Sant Joan de Déu, Barcelona, Spain

Rubén Solano

Contributed equally to this work with: Iolanda Jordán, Rubén Solano, Joan A. Caylà

Affiliations CIBER Epidemiology and Public Health. Barcelona, Spain, Public Health Department, University of Barcelona, Barcelona, Spain, Public Health Agency, Barcelona, Spain

Jesús Castilla

Affiliations CIBER Epidemiology and Public Health. Barcelona, Spain, Public Health Institute, Navarra Institute for Health Research, Pamplona, Spain

Joan A. Caylà

Contributed equally to this work with: Iolanda Jordán, Rubén Solano, Joan A. Caylà

Affiliations CIBER Epidemiology and Public Health. Barcelona, Spain, Public Health Agency, Barcelona, Spain

Pere Godoy

Affiliations CIBER Epidemiology and Public Health. Barcelona, Spain, Agency of Public Health of Catalonia, Barcelona, Spain

Carmen Muñoz-Almagro

Affiliation: Molecular Microbiology Department, Hospital Sant Joan de Déu, Barcelona, Spain

Ángela Domínguez

Affiliations CIBER Epidemiology and Public Health. Barcelona, Spain, Public Health Department, University of Barcelona, Barcelona, Spain

Working Group on the Transmission of Pertussis in Households

¶Membership of the Working Group on the Transmission of Pertussis in Households is listed in the Acknowledgments.

Introduction

Pertussis disease causes a high number of cases and hospitalizations even though it is a preventable disease with high vaccination coverages [1–3]. The vaccination schedule of pertussis in Spain includes the first series of vaccine at 2, 4, and 6 months and two booster doses at 18 months and 4–6 years [4,5]. Vaccination coverages in Spain were > 95% for the first series of vaccination, up to 90% for the first booster dose, and around 80% for the second booster dose in 2011 [6, 7]. In Spain, notification of pertussis disease is mandatory and must be reported by physicians to the Health Department. In recent years, the incidence of pertussis has increased in Catalonia and Navarra, in other regions in Spain, and worldwide [8], and an increase in hospitalizations has been observed. Reported reasons for the increasing incidence include improvements in diagnostic methods [9–10], the decrease in the protection provided by pertussis vaccination over time, and the fact that vaccination, or having the disease, does not provide life-long protection against new infection. Additional suggested reasons include waning immunity [11], the change from the whole cell to the acellular vaccine, vaccine failure, and changes in B. pertussis strains [12–14].

However, some reports suggest that vaccination with at least one dose reduces the risk of hospitalization [15–16]. Determination of risk factors for hospitalization could aid vaccination and decrease hospitalization due to pertussis and therefore reduce the disease burden for the public health system.

The aim of this study was to determine risk factors for hospitalization in cases of pertussis in two Spanish regions (Catalonia and Navarra).

Methods

Between January 2012 and December 2013, all confirmed pertussis cases and their contacts were reported to the surveillance system in Catalonia and Navarra, which have a population of 8,200,000 (17.4% of the Spanish population). Pertussis cases were defined as patients with ≥ 2 weeks of cough with other symptom such as paroxysms, inspiratory whoop, posttussive vomiting or apnoea, accompanied by one of the following: laboratory confirmation by real-time polymerase chain reaction (PCR) or culture or epidemiologic link with a laboratory-confirmed case. Bacterial culture were prepared with selective charcoal agar (5%) plates (Oxoid, Basingstoke, UK), defibrinated human blood (10%) and cephalexin (40mg/l) (Selective supplement Bordetella, Oxoid, Basingstoke, UK). The plates were incubated in a humid atmosphere at 37°C (±2°C) and were examined daily for 5 days. The real-time PCR target used in the study was insertion sequence IS481 of B. pertussis. This sequence presents a high copy number in B. pertussis (about 200 copies) and a low copy number in other Bordetella species as B. holmesii or B. bronchiseptica. A second real-time PCR was performed in samples with Ct values above 35, and these samples were also concentrated (x4) by easyMAG (BioMerieux Laboratories, Marcy l’Etoile, France). The target insertion sequence was IS1002, which allow discrimination between other Bordetella spp and B. pertussis, as low copy numbers are found in B. pertussis.

A modified definition was used for patients aged < 1 year, in whom coughing could last < 2 weeks, but had to be accompanied by paroxysm and another symptom. Household contacts were defined as usual residents of the house of the index case; other contacts were defined as people who spent > 2 hours in the house of the index case during the transmission period of the disease.

Each contact was surveyed about exposure to the case, respiratory symptoms, doses of vaccine received and chemoprophylaxis. Once the index case was detected, we identified all household and other contacts, who were surveyed immediately after identification. In addition, both cases and contacts were followed for 28 days after the date of report in order to determine complications and the appearance of secondary cases.

Contacts with cough or other symptom related to pertussis were laboratory tested to confirm or rule out the disease. Some contacts became cases after laboratory testing and were included as cases, not as contacts. Pertussis index cases were defined as the first case reported and diagnosed that generated the household study. After the household study, index cases were classified as primary case, co-primary case or secondary case. Primary cases were cases without an epidemiological link to a pertussis case previously detected in the house. Co-primary cases were cases epidemiologically linked to a pertussis case diagnosed in the household with onset of symptoms between 0 and 6 days after the other pertussis case. Secondary cases were cases with an epidemiological link with a pertussis case previously detected in the household with onset of symptoms between 7 and 28 days after the other pertussis case. Symptoms from cases were obtained from clinical reports, from the attending physicians and by interview with cases or their parents or tutors. The following variables were collected in pertussis cases: clinical features, age, length of hospitalization, laboratory test used, vaccination status, date of onset of symptoms, date of diagnosis, vaccination status and type of case (primary or secondary). The following variables were collected in contacts: exposure to the case, respiratory symptoms, doses of vaccine received and chemoprophylaxis.

Diagnostic delay was calculated as the days between the onset of symptoms and the day of diagnosis. Cases were classified according to the vaccination status as follows: Correctly-vaccinated cases were cases that had received the number of doses corresponding to their age (and had received at least one dose). Non-vaccinated cases were cases which had not received any dose of vaccine. Partially-vaccinated cases were cases which had received some but not all vaccinations corresponding to their age. The Chi square test was used to compare the vaccination status.

The odds ratio (OR) and 95% confidence intervals (CI) between hospitalized and outpatient cases were calculated to determine the variables influencing the risk of hospitalization. To control for possible confounding variables, the adjusted OR was calculated using logistic regression, including other variables, such as symptoms and vaccination status, in the model. We also stratified the OR by age.

Population figures were taken from the National Institute of Statistics (INE) [17]. The statistical analysis was performed using the SPSS v.18 package. Statistical significance was established as α = 0.05.

The study was conducted according to the principles expressed in the Declaration of Helsinki. All cases included in the study received detailed information about the aims of the study. We first obtained verbal consent to participate and then written consent from all cases. When cases were minors, consent was obtained from relatives, carers or guardian. The study was approved by the Ethics Committee of Hospital Sant Joan de Deu in Barcelona, Spain.

Results

In the study period, 668 pertussis cases were reported to surveillance units and included in the study. These generated 2793 household contacts, 456 of whom became pertussis cases and generated 59 new household contacts. Therefore, 1124 pertussis cases occurred during the study, of which 536 (47.7%) were male and 588 (52.3%) female.

Seven hundred and thirty-two cases were determined by PCR, 5 by culture, 32 by both PCR and culture, and 355 by epidemiological link. Seven hundred and thirty eight (65.7%) cases were classified as primary or co-primary and 386 (34.3%) as secondary.

Table 1 shows the median age, duration of cough and diagnostic delay in hospitalized and outpatient cases.

[Figure omitted. See PDF.]

Table 1. Age, duration of cough and diagnostic delay in pertussis cases.

https://doi.org/10.1371/journal.pone.0139993.t001

One hundred and sixty-seven (14.9%) cases were hospitalized, 936 (83.3%) were outpatient cases and no information was available in 21 cases. The hospitalization rate was 1.01 per 100,000 inhabitants. Eighty-four hospitalized cases (50.3%) were male and 83 (49.7%) female (p = 0.511). Eighty-two (49.1%) cases were hospitalized in 2012 and 85 (50.9%) in 2013 (p = 0.221).

Of the 167 hospitalized cases, 163 (97.6%) were index cases, of which 100 (61.3%) were primary or co-primary cases and 63 secondary cases. Of the 4 cases that were not index cases, 3 (75.0%) were primary cases and 1 was a secondary case. Secondary hospitalized cases acquired the infection most frequently from the mother (26.5%), siblings (23.4%), the father (20.3%), other relatives (20.3%) and grandparents (9.5%).

One hundred and fifty four hospitalized cases (92.2%) were confirmed by PCR, 8 (4.8%) by both PCR and a positive culture test, 1 (0.6%) by a positive culture and 4 cases (2.4%) were epidemiologically-linked to a previous laboratory-confirmed case.

Differences in age group between outpatient and hospitalized cases are shown in Fig 1. Of hospitalized cases, 91% were aged < 6 months, while 34% of outpatient cases were aged > 15 years.

[Figure omitted. See PDF.]

Fig 1. Number of hospitalized and outpatient pertussis cases according to age group.

mo. months; y. years.

https://doi.org/10.1371/journal.pone.0139993.g001

Table 2 shows the symptoms in hospitalized and outpatient cases. Cough ≥ 2 weeks was excluded from the table, as it was mandatory in reporting a case. The adjusted OR for symptoms shows that inspiratory whoop, apnoea and cyanosis were associated with hospitalization. In pertussis cases aged 0 to 6 months, apnoea was associated with hospitalization (S1 Table).

[Figure omitted. See PDF.]

Table 2. Predictive symptoms in hospitalized and outpatient pertussis cases.

https://doi.org/10.1371/journal.pone.0139993.t002

The vaccination status was known in 905 (81.5%) cases. Table 3 shows the vaccination status in hospitalized and outpatient cases.

[Figure omitted. See PDF.]

Table 3. Vaccination status in hospitalized and outpatient cases.

https://doi.org/10.1371/journal.pone.0139993.t003

There were fewer hospitalizations in correctly-vaccinated cases than in non-vaccinated cases (p<0.001), and fewer hospitalizations in partially-vaccinated cases than in non-vaccinated cases (p<0.001). Of hospitalized cases, 56 had received some dose of vaccine: 41 had received 1 dose (73.2%), 7 cases 2 doses (12.5%), 3 cases 3 doses (5.3%), 3 cases 4 doses (5.3%) and 2 cases had received 5 doses (3.7%).

Discussion

A large number of pertussis cases were detected in 2012 and 2013 in Catalonia and Navarra, of which approximately 15% were hospitalized (> 90% were aged <1 year). More than half the cases detected were primary or co-primary cases, an important factor for both early treatment and breaking the transmission of the disease.

Our results show that some symptoms were more frequent in hospitalized than in outpatient cases, including inspiratory whoop, apnoea and cyanosis (which is not a symptom but a complication of the disease). A New Zealand study found that paroxysmal cough and cyanosis were predictive factors for not discharging a case from hospital, in order to avoid readmission [18]. A Swiss study found that hospitalized cases more frequently had cyanosis [19].

We found a reduction in hospitalizations when patients were correctly vaccinated or partially vaccinated. This reinforces the importance of vaccination and suggests that vaccination reduces disease severity and protects against hospitalization. Similar conclusions have been reported by other authors, and in some studies one dose of vaccine was sufficient to produce this reduction [15–16, 19–21].

Some studies have reported that whole cell pertussis vaccine offers greater protection than the acellular vaccine. We could not analyse this factor, as all hospitalized cases in our study were aged <10 years (except 1 case aged <45 years with unknown vaccination status) and all were vaccinated with the acellular vaccine, which has been available in Catalonia since 2002 [22].

Previous studies in Catalonia found a large percentage of pertussis cases in infants [23–25]. In our study, which included active surveillance, a large number of cases in patients aged 15–44 years were detected. Normally a diagnosis of pertussis is not made in this age group as they have less-severe disease than infants [26–27]. However, several studies in different countries have found that the age pattern of pertussis has changed and the incidence of cases in adolescents and adults has increased [28–33]. This may be because the vaccination coverage in older people is lower than in infants and also because of a loss of immunity [34].

Several studies have found that unreported and misdiagnosed cases are a public health problem [35–36]. Our study avoided this problem because we made active surveillance. For this reason, we found a higher incidence in adults in Catalonia and Navarra than previously described [6, 8].

A limitation of the study is the large number of epidemiologists involved. Although all epidemiologists applied the same protocol, we cannot rule out differences in the interpretation of some variables. However, we believe this did not affect the results, as the Working Group made a considerable effort to correct possible differences in data interpretation.

Regarding PCR diagnosis, B. bronchiseptica strains that contain a copy of IS481 (it is estimated that IS481 is present in only 1% of B. bronchiseptica) could by erroneously identified as B. pertussis [37].

In conclusion, we found that inspiratory whoop, apnoea and cyanosis were good predictors of hospitalization in pertussis cases. Vaccination reduces the disease severity and hospitalizations.

Supporting Information

[Figure omitted. See PDF.]

S1 Table. Predictive symptoms in hospitalized and outpatient pertussis cases by age group.

https://doi.org/10.1371/journal.pone.0139993.s001

(DOCX)

Acknowledgments

The members of the Working Group on the Transmission of Pertussis in Households (PI11/02557) are:

Miquel Alsedà, Josep Alvarez, Cesar Arias, Irene Barrabeig, Neus Camps, Glòria Carmona, Mónica Carol, Maria Company, Joaquim Ferràs, Glòria Ferrús, Mireia Jané, Sofia Minguell, Raquel Rodríguez; María-Rosa Sala, Roser Torra (Agència de Salut Pública de Catalunya), Pere Godoy, Pedro Plans (Agència de Salut Pública de Catalunya and CIBERESP), Inma Crespo, Diana Toledo, Ángela Domínguez and Rubén Solano (Public Health Department, University of Barcelona and CIBERESP), Joan A. Caylà, Sara Lafuente and Cristina Rius (Agència de Salut Pública de Barcelona and CIBERESP), Manuel García Cenoz, Rosana Burgui, Jesús Castilla (Instituto de Salud Pública de Navarra and CIBERESP), Eva del Amo, Iolanda Jordán, Héctor de Paz and Carmen Muñoz-Almagro (Hospital de Sant Joan de Dèu, Barcelona).

Author Contributions

Conceived and designed the experiments: IC AD PG JC JAC. Performed the experiments: DT PG CMA. Analyzed the data: IC DT NS AD CMA. Contributed reagents/materials/analysis tools: CMA. Wrote the paper: IC DT NS RS JC JAC PG AD CMA IJ.

Citation: Crespo I, Toledo D, Soldevila N, Jordán I, Solano R, Castilla J, et al. (2015) Characteristics of Hospitalized Cases of Pertussis in Catalonia and Navarra, Two Regions in the North of Spain. PLoS ONE 10(10): e0139993. https://doi.org/10.1371/journal.pone.0139993

References

1. Waters V, Halperin SA. Bordetella pertussis. In: Principles and Practice of Infectious Diseases. Edited by Bennett JE, Dolin R, Blaser MJ. 8th ed. Philadelphia: Elsevier; 2015. pp. 2619–2628.

2. Edwards K, Decker MD. Whooping cough vaccine. In: Plotkin SA, Orenstein WA, Offit PA, editors. Vaccines. 6th ed. Philadelphia: Elsevier; 2013. Pp. 447–492.

3. Feigin RD, Cherry J, Harrison N, Kaplan M. Pertussis and other Bordetella infections. In: Feigin RD, ed. Feigin and Cherry's Textbook of Pediatric Infectious Diseases. 6th ed. Philadelphia: Saunders Elsevier; 2009. Pp. 1683–1706.

4. Departament de Salut. Generalitat de Catalunya. Vaccination Schedule 2014. Available: http://www20.gencat.cat/portal/site/canalsalut/menuitem.af261f715269a25d48af8968b0c0e1a0/?vgnextoid=b7a3c118ab2da210VgnVCM1000008d0c1e0aRCRD&vgnextchannel=b7a3c118ab2da210VgnVCM1000008d0c1e0aRCRD&vgnextfmt=default#div_02

5. Servicio de Epidemiología, Prevención y Promoción de la Salud. Instituto de Salud Pública y Laboral de Navarra. Vaccination Schedule 2014. Available: http://www.Navarra.es/home_es/Gobierno+de+Navarra/Organigrama/Los+departamentos/Salud/Organigrama/Estructura+Organica/Instituto+Navarro+de+Salud+Publica/Publicaciones/Publicaciones+profesionales/Epidemiologia/Vacunaciones.htm

6. Grupo de trabajo de tos ferina 2012. Ponencia de Programa y Registro de vacunaciones. Revisión del Programa de Vacunación frente a tos ferina en España. Enero 2013. Ministerio de Sanidad Servicios Sociales e Igualdad.

7. Crespo I, Domínguez A. Tos ferina: cambios en la epidemiología. España 2000–2011. Vacunas. 2013; 14:155–1561.

8. Centro Nacional de Epidemiología. Situación Epidemiologica de la tos ferina en España. September 2012.

9. Hochwald O, Bamberger E, Srugo I: The return of pertussis: who is responsible? What can be done? IMA. 2006; 8:301–307.

10. DeVincenzo JP, Guyton C, Rea H, Elmore E, Patel S, Wynn L, et al. Molecular detection and quantification of pertussis and correlation with clinical outcomes in children. 2013; 76: 10–15.

11. Klein NP, Bartlett JRowhani-Rahbar A, Fireman B, Baxter R. Waning Protection after Fifth Dose of Acellular Pertussis Vaccine in Children. N Engl J Med. 2012; 367:1012–1019. pmid:22970945

12. van Gent M, Bart MJ, van der Heide HG, Heuvelman KJ, Mooi FR. Small mutations in Bordetella pertussis are associated with selective sweeps. PLoSOne. 2012;7(9):e46407.

13. Fingermann M, Fernández J, Sisti F, Rodríguez ME, Gatti B, Bottero D et al. Differences of circulating Bordetella pertussis population in Argentina from the strain used in vaccine production. Vaccine. 2006; 24: 3513–3521. pmid:16545509

14. Sheridan SL, McCall BJ, Davis CA, Robson JM, Hull BP, Selvey CE, et al. Acellular pertussis vaccine effectiveness for children during the 2009–2010 pertussis epidemic in Queensland. Med J Aust. 2014; 200:334–338. pmid:24702091

15. Cortese MM, Baughman AL, Zhang R, Srivastava PU, Wallace GS. Pertussis hospitalizations among infants in the United States, 1993 to 2004. Pediatrics. 2008;121:484–492. pmid:18310196

16. Barlow RS, Reynolds LE, Cieslak PR, Sullivan AD. Vaccinated children and adolescents with pertussis infections experience reduced illness severity and duration, Oregon, 2010–2012. Clin Infect Dis. 2014;58:1523–1529. pmid:24633685

17. National Institute of Statistics. INE. Available: http://www.ine.es/inebmenu/mnu_padron.htm.

18. Lurie G, Reed PW, Grant CC. When to discharge children hospitalized with pertussis? Acad Pediatr. 2009; 9:118–122. pmid:19329103

19. Wymann MN, Richard JL, Vidondo B, Heininger U. Prospective pertussis surveillance in Switzerland, 1991–2006. Vaccine. 2011;29:2058–2065. pmid:21251904

20. Briand V, Bonmarin I, Lévy-Bruhl D. Study of the risk factors for severe childhood pertussis based on hospital surveillance data. Vaccine. 2007;25:7224–7232. pmid:17707962

21. Juretzko P, von Kries R, Hermann M, Wirsing von König CH, Weil J, Giani G. Effectiveness of acellular pertussis vaccine assessed by hospital-based active surveillance in Germany. Clin Infect Dis. 2002 Jul 15; 35: 162–167. pmid:12087522

22. Generalitat de Catalunya. Departament de Sanitat i Seguretat Social. Manual de vacunacions. Barcelona: Generalitat de Catalunya; 2006.

23. Cardeñosa N, Romero M, Quesada M, Oviedo M, Carmona G, Codina G, et al. Is the vaccination coverage established enough to control pertussis, or it is a re-emerging disease? Vaccine. 2009; 27:3489–3491. pmid:19200822

24. Crespo I, Cardeñosa N, Godoy P, Carmona G, Sala MR, Barrabeig I, et al. Epidemiology of pertussis in a country with high vaccination coverage. Vaccine. 2011;29:4244–4248. pmid:21496465

25. Crespo Fernández I, Soldevila N, Carmona G, Sala MR, Godoy P, Domínguez A. Surveillance of hospitalized and outpatient cases of pertussis in Catalonia from 2003 to 2009. Hum Vaccin Immunother. 2013;9:667–670. pmid:23302866

26. Miyashita N, Akaike H, Teranishi H, Kawai Y, Ouchi K, Kato T, et al. Diagnostic value of symptoms and laboratory data for pertussis in adolescent and adult patients. BMC Infect Dis. 2013;13:129. pmid:23496900

27. von König CH, Halperin S, Riffelmann M, Guiso N. Pertussis of adults and infants. Lancet Infect Dis. 2002;2:744–750. pmid:12467690

28. de Greeff SC, de Melker HE, van Gageldonk PG, Schellekens JF, van der Klis FR, Mollema L, et al. Seroprevalence of pertussis in The Netherlands: evidence for increased circulation of Bordetella pertussis. PLoS One. 2010;5:e14183. pmid:21152071

29. Cherry JD. Epidemic pertussis in 2012—the resurgence of a vaccine-preventable disease. N Engl J Med. 2012; 367:785–787. pmid:22894554

30. Shapiro ED. Acellular vaccines and resurgence of pertussis. JAMA. 2012; 308: 2149–2150. pmid:23188034

31. Lavine JS, Bjørnstad ON, de Blasio BF, Storsaeter J. Shortlived immunity against pertussis, age-specific routes of transmission, and the utility of a teenage booster vaccine. Vaccine. 2012; 30: 544–551. pmid:22119924

32. Amirthalingam G, Gupta S, Campbell H. Pertussis immunisation and control in England and Wales, 1957 to 2012: a historical review. Euro Surveill. 2013; 18: pii = 20587.

33. Skowronski DM, Janjua NZ, Tsafack EP, Ouakki M, Hoang L, De Serres G. The number needed to vaccinate to prevent infant pertussis hospitalization and death through parent cocoon immunization. Clin Infect Dis. 2012; 54: 318–327. pmid:22156850

34. Chiappini E, Stival A, Galli L, de Martino M. Pertussis re-emergence in the post-vaccination era. BMC Infect Dis. 2013; 26;13:151. pmid:23530907

35. Fernández-Cano MI, Armadans L, Martínez X, Campins M. Incidence of whooping cough in Spain (1997–2010): an underreported disease. Eur J Pedirat. 2014;173:721–726.

36. van den Brink G, Wishaupt JO, Douma J, Hartwig N Versteegh F. Bordetella pertussis: an underreported pathogen in pediatric respiratory infections, a prospective cohort study. BMC Infect Dis. 2014; 14:526. pmid:25267437

37. Roorda L, Buitenwerf J, Ossewaarde JM, van der Zee A. A real-time PCR assay with improved specificity for detection and discrimination of all clinically relevant Bordetella species by the presence and distribution of three insertion sequence elements. BMC Res Notes. 2011; 4:11. pmid:21255383