Secondary Logo

Journal Logo

Original Studies

Pertussis is Common in Nonvaccinated Infants Hospitalized for Respiratory Syncytial Virus Infection

Korppi, Matti MD, PHD; Hiltunen, Johanna MS

Author Information
The Pediatric Infectious Disease Journal: April 2007 - Volume 26 - Issue 4 - p 316-318
doi: 10.1097/01.inf.0000258690.06349.91
  • Free

Abstract

Two studies, published nearly 20 years apart, have documented that respiratory syncytial virus (RSV) and Bordetella pertussis can cause respiratory infections concomitantly.1,2 In those studies, pertussis was identified in 16–20% of young infants hospitalized for lower respiratory tract infection, and a third to half were mixed infections with RSV.1,2 Laboratory reports indicate that infants still develop pertussis in our country, despite a good coverage (95.6% in 1999; 95% confidence interval 94.1–96.8%) of pertussis vaccinations given at 3, 5 and 12 months (at 3, 4, 5 and 24 months until 2004) of age (National Public Health Institute, Finland). Since 2005, booster vaccinations have been given also at 4 and 15 years of age.

In northern climates, severe RSV epidemics occur every second autumn and winter, alternating with mild spring epidemics.3 A nation-wide epidemic of RSV infections took place from November 2005 to January 2006 in our country.

This paper presents a retrospective analysis of infants, aged <6 months, who were treated in hospital for respiratory tract infection from October 2005 to February 2006, the period including an RSV epidemic. The aim of the study was to evaluate how often respiratory viral infections in infants, RSV infections in particular, are mixed infections with B. pertussis, and what is their clinical presentation.

MATERIALS AND METHODS

The population of <12-month-old infants is about 3000 in the province of Kuopio (Official Statistics of Finland). Our university hospital is responsible for the primary hospital care of all infants and young children in this area.

There was a nation-wide epidemic of RSV infections in Finland from November 2005 to January 2006. Because this epidemic was anticipated, we started, in September 2005, to monitor respiratory viral antigens in all infants treated as inpatients for respiratory tract infection in our hospital. The viral antigens were studied at the Department of Clinical Microbiology, Kuopio University Hospital, by a commercial direct immunofluorescence test (PathoDx Respiratory Virus Panel, Remel Inc., Lenexa, KS), including assays for RSV, parainfluenza 1, 2 and 3 virus, influenza A and B virus, and the adenovirus group. In addition, in-house polymerase chain reaction (PCR) to B. pertussis4 was studied at the Department of Medical Microbiology, University of Turku, if the infant suffered from cough.

We collected retrospectively from the patient records of our hospital clinical data of the infants <6 months of age who were hospitalized for respiratory tract infection from October 1, 2005 to February 28, 2006. The number of patients was 126, which is about 8% of the age-matched population. The information gathered included admission date, age (in months) on admission, gender, the presence of cough, the presence of respiratory distress, and the findings, observed and registered by the doctors on duty, on thoracal inspection (increased respiratory rate and chest indrawings) and auscultation (expiratory wheezes and inspiratory crackles). Oxygen saturation data on admission and at hospital were recorded. White blood cell (WBC) counts and serum C-reactive protein (CRP) were studied on admission. Chest radiographs were obtained during the hospital period, and pediatric radiologists interpreted the radiographs. The pattern of infiltration in chest radiograph, if present, was classified into alveolar or interstitial pneumonia, or into patchy atelectatic areas or perihilar inflammatory changes which were not regarded as diagnostic for pneumonia.

Based on parent-reported cough at home or nurse-reported cough in hospital, the infants were divided into 3 groups: Prominent cough, nasopharyngeal aspirate (NPA) samples for B. pertussis PCR were obtained and macrolides were started (group 1); Modest cough, NPA samples for B. pertussis were obtained but macrolides were not started (group 2); No or mild cough, NPA samples for B. pertussis PCR were not obtained and macrolides were not considered (group 3). The macrolide, when started, was clarithromycin, and the dosage was 15–20 mg/kg/d (divided into 2 doses) either orally or intravenously. The result of the PCR test was ready within 2–4 days, depending on week-ends when the tests were not done. If the result was negative, macrolide therapy was discontinued. If the result was positive, macrolide therapy was continued or started if not going on.

Viral antigens were studied in 117 (93%) of the 126 infants hospitalized for respiratory tract infection. PCR for B. pertussis was studied in 88 (75%) of these 117 infants. Among them, macrolide therapy was started in 40 and not started in 48 cases. Thus, based on the cough history, macrolide therapy was started for 28% of infants <6 months of age treated in hospital.

Student's t test and Fisher's exact test, with Bonferroni correction when appropriate, were used in the statistical analyses of the data.

RESULTS

RSV infection was documented in 91 (78%) infants, with a peak incidence in December 2005. Parainfluenza viruses (PIV) were detected in 2 cases. No other viruses were found. Eleven percent of the patients were <1 month old, 6% were >5 months old, and 18–24% belonged to each of the 4 other 1-month age groups. There were 65 (56%) boys and 52 girls. B. pertussis infection was documented in 9 (8%) infants. Six of them were <3 months old, and 2 were 3–4 months and 1 was 5–6 months old. Co-infection with RSV was present in 7 (78%) cases and with PIV type 3 in 1 case.

RSV and B. pertussis findings were analyzed in relation to cough history at home or in hospital (Table 1). B. pertussis was diagnosed in 10% of the infants with cough, but the classification of cough had no association with the presence of B. pertussis infection. PCR for pertussis was not studied in infants with no or mild cough.

TABLE 1
TABLE 1:
RSV and Bordetella pertussis Findings in Infants in Relation to the Presence of Cough

There were no clinical, nonspecific laboratory or radiologic characteristics to help the doctors on duty to screen pertussis or mixed RSV-pertussis cases (Table 2). Wheezing was rarely associated with pertussis and was present in 1 infant with RSV co-infection. On the other hand, no more than half of the RSV patients wheezed, which suggests that wheezing did not distinguish RSV and pertussis cases.

TABLE 2
TABLE 2:
Clinical Characteristics in the Infants With RSV and/or Bordetella pertussis Infection

The patient with pertussis alone was a 1-month-old girl, and the clinical, radiologic and nonspecific laboratory findings registered were all negative. The patient with mixed pertussis-PIV type 3 infection was a 3-month old girl, and there was mild hypoxemia on admission and hyperinflation in chest radiograph.

During hospitalization, a secondary presumably bacterial (other than pertussis) pneumonia was present in 4 cases, diagnosed by alveolar infiltrates in chest radiographs. In them, WBC counts ranged from 11.2 to 22.9 × 10E9/L and CRP from 13 to 61 mg/L. WBC count was >20 × 10E9/L in 1 and CRP >40 mg/L in 3 patients.

DISCUSSION

There are 3 main results in the present study carried out during an RSV epidemic. First, pertussis was surprisingly common (7%) in nonvaccinated or incompletely vaccinated infants hospitalized for respiratory tract infection. Second, nearly all cases were mixed infections with RSV. The findings are in agreement with recent observations from the United Kingdom,2 where 20% of infants requiring intensive care during an RSV epidemic had pertussis and 36% of the pertussis cases were mixed infections with RSV. Third, the RSV and pertussis cases could not be separated by clinical, nonspecific laboratory or radiologic findings, as also was found 20 years ago in a study from the United States.1 Thus, the typical characteristics of RSV bronchiolitis, or even the detection of RSV etiology, do not exclude pertussis.

Nelson et al studied upper respiratory secretions by culture and direct microbe detection using fluorescent antibody tests for RSV and B. pertussis in 180 children hospitalized for respiratory tract infection.1 Pertussis was diagnosed in 29 (15%) cases; 15 were mixed infections with RSV and 14 were single B. pertussis infections. These mixed and single cases were compared with 22 randomly selected sole RSV cases. The median age in all of the 3 groups was 2–3 months. Children with pertussis were more likely to have been premature than children with RSV infection alone. There were no other significant differences between infants with pertussis alone, RSV infection alone or mixed pertussis-RSV infection. The authors concluded, in accordance with the present results, that pertussis or simultaneous infection with pertussis should be considered in young children hospitalized for presumed viral respiratory illness.

Since 1980s, case presentations have confirmed the occurrence of mixed RSV-pertussis infections.5 In nonvaccinated infants, pertussis may be fatal, if not promptly diagnosed and treated by appropriate antibiotics.

Nearly 20 years later, Crowcroft et al2 studied B. pertussis culture and PCR in 126 infants aged <6 months requiring intensive care for lower respiratory tract infection, and among them, 25 (20%) had pertussis. Two infants with pertussis died, only 6 cases (24%) were culture positive, and 9 cases (36%) were mixed infections with RSV. On admission, pertussis was clinically suspected and antibiotics were started in less than a third of the cases. Infants with pertussis had cough, apneic spells and whooping episodes more often than infants without pertussis.2 The authors concluded that pertussis in infants is under-diagnosed, and that an RSV diagnosis, in accordance with the results of Nelson et al1 and the present results, does not exclude pertussis.

There are 2 recent studies with opposite results. In a study from the United States, only 1 of 166 children admitted to the hospital during RSV season was B. pertussis positive by PCR, suggesting that screening is not useful when pertussis prevalence is low, as was the case in the study population.6 Thurburn et al7 performed a prospective microbiologic analysis of lower airway secretions, obtained during intubation, in 165 RSV bronchiolitis patients requiring mechanical ventilation. B. pertussis was cultured in only 1 case, concomitantly with RSV.7

In the present study, pertussis was diagnosed only by PCR in upper respiratory samples. The PCR test has been positive in all cases diagnosed by B. pertussis culture,4 but the definite clinical applicability of the test is unresolved. Antigen detection in nasopharyngeal samples was used for RSV diagnosis, and this approach has proved to be both sensitive and specific in infants.3

REFERENCES

1. Nelson WL, Hopkins RS, Roe MH, Glode MP. Simultaneous infection with Bordetella pertussis and respiratory syncytial virus in hospitalized children. Pediatr Infect Dis. 1986;5:540–544.
2. Crowcroft NS, Booy R, Harrison T, et al. Severe and unrecognised: pertussis in UK infants. Arch Dis Child. 2003;88:802–806.
3. McIntosh K, Halonen P, Ruuskanen O. Report of a workshop on respiratory viral infections: epidemiology, diagnosis, treatment and prevention. Clin Infect Dis. 1993;16:151–164.
4. He Q, Mertsola J, Soini H, Skurnik M, Ruuskanen O, Viljanen M. Comparison of polymerase chain reaction with culture and enzyme immunoassay for diagnosis of pertussis. J Clin Microbiol. 1993;31:642–645.
5. Aoyama T, Ide Y, Watanabe J, Takeuchi Y, Imaizumi A. Respiratory failure caused by dual infection with Bordetella pertussis and respiratory syncytial virus. Acta Paediatr Jpn. 1996;38:282–285.
6. Siberry GK, Ross TL, Perl TM, Valsamakis A. Low prevalence of pertussis among children admitted with respiratory symptoms during respiratory syncytial virus season. Infect Control Hosp Epidemiol. 2006;27:95–97.
7. Thurburn K, Harigopal S, Reddy V, Taylor N, van Saene HK. High incidence of pulmonary bacterial co-infection in children with severe respiratory syncytial virus (RSV) bronchiolitis. Thorax. 2006;61:611–615.
Keywords:

RSV; pertussis; Bordetella pertussis; infant; mixed infection

© 2007 Lippincott Williams & Wilkins, Inc.