Objective. We investigated the prevalence and the etiology of acute otitis media (AOM) in children with bronchiolitis to determine whether AOM in such children is due entirely or mainly to respiratory syncytial virus (RSV), in which case routine antimicrobial treatment would not be appropriate.
Methods. The study group consisted of children aged 2 to 24 months with bronchiolitis. In patients with AOM at entry, nasal washings for RSV enzyme-linked immunosorbent assay were obtained, and Gram-stained smear, bacterial culture, and reverse transcriptase polymerase chain reaction to detect the presence of RSV were performed on middle-ear aspirates. Patients without AOM were reevaluated at 48 to 72 hours, 8 to 10 days, and 18 to 22 days.
Results. Forty-two children with bronchiolitis were enrolled. Sixty-two percent had AOM at entry or developed AOM within 10 days. An additional 24% had or eventually developed otitis media with effusion. Only 14% remained free of both AOM and otitis media with effusion throughout the 3-week observation period. All patients with AOM had 1 or more bacterial pathogens isolated from one or both middle-ear aspirates. Of 33 middle-ear aspirates, Streptococcus pneumoniae was isolated in 15, Haemophilus influenzaein 8, Moraxella catarrhalis in 8, and Staphylococcus aureus in 2. Two middle-ear aspirates yielded 2 pathogens each; 2 aspirates had no growth. RSV was identified in 17 (71%) of 24 patients with AOM.
Conclusion. Bacterial AOM is a complication in most children with bronchiolitis. Accordingly, in patients with bronchiolitis and associated AOM, antimicrobial treatment is indicated.
- AOM =
- acute otitis media •
- RSV =
- respiratory syncytial virus •
- MEE =
- middle-ear effusion •
- TM =
- tympanic membrane •
- OME =
- otitis media with effusion •
- RT-PCR =
- reverse transcriptase polymerase chain reaction •
- ELISA =
- enzyme-linked immunosorbent assay •
- MIC =
- minimum inhibitory concentration
Peaks in the occurrence of acute otitis media (AOM) have been associated with epidemic peaks of respiratory infection attributable to rhinovirus, influenza A virus, parainfluenza virus, adenovirus, and respiratory syncytial virus (RSV).1-5Of these viral infections, RSV infection seems to be the one most commonly accompanied by AOM6; in retrospective studies AOM has been reported to occur in 57% to 67% of children with RSV infection.6 7 Similarly, in children presenting with AOM, RSV or RSV antigen has been found in specimens obtained from the nasopharynx8 9 and, more often than any other virus or virus antigen, also in middle-ear aspirates—either alone or, more often, in association with pathogenic bacteria.8-16
In children with bronchiolitis specifically, the occurrence of AOM has not, to our knowledge, been systematically studied. Having frequently encountered AOM in such children, we wondered whether AOM is a usual accompaniment or complication of bronchiolitis, and whether AOM occurring in children with bronchiolitis might be due entirely or mainly to RSV, rather than to one or another of the bacterial pathogens usually found in the middle ear. Should that be the case, routine antimicrobial treatment of AOM in children with bronchiolitis would not be appropriate. To address this question we undertook a prospective study.
Eligible participants were children aged 2 to 24 months who presented to the Primary Care Center or the Emergency Department of the Children's Hospital of Pittsburgh from December 1995 to January 1997 and who were diagnosed with bronchiolitis. The study was approved by the hospital's Human Rights Committee.
For the purpose of the study, bronchiolitis was defined as an acute lower respiratory tract illness characterized by high-pitched cough, tachypnea (respiratory rate >40/minute), signs of acute respiratory distress, and prolonged expiratory phase, with or without wheezing and/or crackles.17
Criteria for the diagnosis of AOM included either: 1) the presence of purulent otorrhea of less than 24 hours duration, not caused by otitis externa; or 2) evidence of middle-ear effusion (MEE) in addition to at least one indicator of acute inflammation. Evidence of MEE consisted of the presence, on pneumatic otoscopy, of at least two of the following abnormalities of the tympanic membrane (TM): decreased or absent mobility; yellow, white, amber, or blue discoloration; opacification other than because of scarring; and air-fluid interfaces. Indicators of acute inflammation consisted of ear pain within 24 hours (including unaccustomed tugging or rubbing of the ear), marked redness of the TM, and distinct fullness or bulging of the TM. Ears with effusion but without indicators of acute inflammation were considered to have otitis media with effusion (OME; secretory otitis media).
Excluded from the study were children whose parents refused to agree to possible tympanocentesis if their child had AOM or were to develop AOM during the follow-up period. Also excluded were children who had or met any of the following conditions: more than two previous episodes of wheezing; severe respiratory distress rendering tympanocentesis problematic; presence of tympanostomy tubes; antibiotic treatment during the previous 7 days; perforation of the TM with drainage for more than 24 hours; craniofacial abnormalities associated with prolonged MEE, such as cleft palate or Down syndrome; and chronic underlying disease.
On each enrolled child we obtained a standardized history and performed a complete physical examination, including pneumatic otoscopy.
In patients with AOM, tympanocentesis was performed 15 to 20 minutes after administering acetaminophen combined with codeine (1 mg/kg of codeine). Patients were restrained using a papoose board. Under direct vision through a surgical otoscope head, the external auditory canal was cleaned using a Buck curet (No. 0) as necessary, and a Farrell applicator with its tip wrapped with alcohol-soaked cotton to create a wet mop. To avoid inadvertently sterilizing the middle-ear specimen, no alcohol was instilled in the ear canal. To obtain and collect the specimen, we used a disposable tympanocentesis aspirator (Xomed Surgical Products, Jacksonville, FL; No. 91–19010), connected to wall suction. The TM was punctured at the point of maximum bulge in the posteroinferior or anteroinferior quadrant.18Gram-stained smear and bacterial culture were performed on each middle-ear aspirate, and when possible, reverse transcriptase polymerase chain reaction (RT-PCR) to detect the presence of RSV also was performed, along with nasal washings for RSV-enzyme-linked immunosorbent assay (ELISA). Specimens obtained from the middle ear were inoculated onto 5% sheep blood agar, chocolate agar, CNA agar, and McConkey agar media for culture and incubated at 35°C in a 5% CO2 incubator for 48 hours. Penicillin susceptibility ofStreptococcus pneumoniae isolates was determined using the Sensititer HP (Accumed, Westlake, OH) panel. Isolates with minimum inhibitory concentrations (MICs) ≤0.01 μg/mL were considered susceptible, isolates with MICs of 0.1 to 1.0 μg/mL intermediately resistant, and isolates with MICs ≥2 μg/mL highly resistant. Beta-lactamase production by Haemophilus influenzae andMoraxella catarrhalis isolates was assessed using the cefinase disc method.
RT-PCR was performed in 19 of 26 patients with AOM. Samples from middle-ear aspirates were frozen at −80°C until RNA extraction was performed using TRIzol LS reagent (Gibco-BRL Life Technologies, Inc, Gaithersburg, MD). RT-PCR was performed using MMLV-Reverse Transcriptase (Gibco-BRL) to produce complementary DNA of the RSV. The RT product was then amplified via PCR using Taq Polymerase (Perkin-Elmer Cetus, Norwalk, CT) according to the manufacturer's instructions. The PCR product was then detected by liquid hybridization-gel retardation analysis. A positive PCR result with the experimental specimen was determined by the presence of a retarded probe band comigrating with amplified pure RSV control complementary DNA.19 20 Nasal washings for RSV-ELISA were performed in 24 of 26 patients with AOM using Testpack RSV (Abbott, Abbott Park, IL) for rapid viral detection.
After tympanocentesis all patients were initially treated with standard doses of amoxicillin, pending the outcome of middle-ear aspirate culture and sensitivity testing. Parents were contacted by telephone 48 to 72 hours after study entry, and follow-up visits arranged as necessary. All patients with AOM at entry were reevaluated at days 11 to 13. Patients without AOM at entry were reevaluated at 48 to 72 hours, at 8 to 10 days, and at 18 to 22 days. Parents were instructed to contact the investigators if their children developed any symptoms suggestive of AOM during the follow-up period.
Forty-two children with bronchiolitis were enrolled. Their mean age was 6.75 months, and the mean duration of their rhinorrhea was 4.0 days. Thirty-nine (93%) of the patients had no previous episodes of wheezing, and 3 (7%) had one episode each. Other selected characteristics of the patients are shown in Table1.
The Figure shows the middle-ear status of the 42 patients during the course of the study period. At entry, 21 children (50%) had evidence of AOM. During the first 10 days after entry, 5 additional patients (12%) developed AOM, 4 of whom had OME at entry in the affected ear(s). By the end of the 3-week follow-up period, a total of 26 patients (62%) had or developed AOM (18 unilateral and 8 bilateral), 10 additional patients (24%) had or developed OME, and only 6 patients (14%) maintained a normal middle-ear status throughout. Of the children with AOM, 96% had distinct fullness or bulging of the TM, 72% had marked redness, and 68% had both. In no case did the diagnosis depend on the presence of middle-ear effusion plus a history of ear pain exclusively.
Tympanocentesis was performed on 33 ears from 25 patients with AOM (the parents of one patient refused tympanocentesis). Each of the 25 patients had at least 1 bacterial pathogen isolated from one or both middle-ear aspirates. Bacteriologic findings are summarized in Table2. Three of the 15 S pneumoniae isolates were resistant to penicillin (1 intermediate and 2 highly resistant); all 8 of the M catarrhalis isolates and 3 of the 8 H influenzae isolates produced β-lactamase. In each of 2 patients with bilateral AOM, 1 of the aspirates showed no growth. In one of these patients, in the aspirate with no growth the smear showed Gram-negative rods, and the aspirate from the contralateral ear yielded H influenzae β-lactamase negative. In the other patient, the aspirate from the contralateral ear yielded resistant S pneumoniae.
RSV was identified in 17 (71%) of 24 patients with AOM—in 7 both by ELISA in nasal washings and by PCR in middle-ear aspirates, in 9 by ELISA only, and in 1 by PCR only. The results are summarized in Table 3.
No reports to our knowledge have described either the prevalence or the etiology of AOM in infants and young children with bronchiolitis. In the present study of such patients, we found that most either had AOM at entry or developed AOM soon afterward. From the middle-ear aspirates of all patients with AOM who underwent tympanocentesis we recovered usual middle-ear bacterial pathogens. These findings suggest that RSV is rarely, if ever, the sole cause of AOM in patients with bronchiolitis.
In a previous study of patients with OME, RSV was identified by RT-PCR and nested-PCR in the middle-ear aspirates of 9 (75%) of 12 patients who had RSV isolated by culture from the nasopharynx.21 None of the patients in that study had AOM. In our study, RSV was found in 54% of the middle-ear aspirates of patients in whom RSV was isolated from the nasopharynx. Possible explanations for not finding evidence of RSV in middle-ear aspirates of patients with RSV isolated from the nasopharynx include actual absence of RSV from the middle-ear cavity and methodologic limitations leading to false-negative PCR results.
We conclude that AOM in patients with bronchiolitis is usually caused by S pneumoniae, H influenzae, or M catarrhalis, and that the finding of AOM in patients with bronchiolitis calls for antimicrobial therapy. Curtailment of antimicrobial therapy on an assumption that AOM associated with bronchiolitis is viral or mainly viral in etiology is not warranted. On the other hand, as in patients generally, stringent diagnostic criteria are necessary to distinguish AOM from OME to avoid use of inappropriate antimicrobial therapy.22
This project was supported in part by the Biomedical Research Support Grant Program, Division of Research Resources from the National Institutes of Health, Bethesda, MD.
We thank Kenneth D. Rogers, MD, for his advice on design and analysis; Diana Kearney, RN, for her unstinting efforts in caring for patients and her assistance in all aspects of the study; and J. Christopher Post, MD, and Garth D. Erlich, PhD, for their laboratory determinations. We also thank the Children's Hospital of Pittsburgh house staff and nurses for their invaluable assistance.
- Received June 23, 1997.
- Accepted September 4, 1997.
Reprint requests to (A.H.) Children's Hospital of Pittsburgh, 3705 Fifth Ave, Pittsburgh, PA 15213-2583.
- Arola M,
- Ruuskanen O,
- Ziegler T,
- et al.
- Chonmaitree T,
- Howie VM,
- Truant AL
- ↵Ackerman VL, Salva PS. Bronchiolitis. In: Loughlin GM, Eigen H, eds. Respiratory Disease in Children: Diagnosis and Management. Baltimore, MD: Williams & Wilkins; 1994:291–300
- Saiki R,
- Gelfand DH,
- Stoffel S,
- et al.
- ↵Erlich GD. PCR-based laboratory methods for the detection of the human retroviridae and Hepadnaviridae. In: Erlich GD, Greenberg SJ, eds. PCR-Based Diagnostics for Infectious Disease. Boston, MA: Blackwell Scientific Publications; 1993:415–446
- Okamoto Y,
- Kudo K,
- Shirotori K,
- et al.
- Paradise JL
- Copyright © 1998 American Academy of Pediatrics