Marzouk HA, Nathawad R, Hammerschlag MR, Weedon J, Bachman D, Goldstein NA. Methicillin-Resistant Staphylococcus aureus Colonization in Otitis-Prone Children. Arch Otolaryngol Head Neck Surg. 2011;137(12):1217-1222. doi:10.1001/archoto.2011.192
Author Affiliations: Divisions of Pediatric Otolaryngology (Drs Marzouk and Goldstein and Mr Bachman) and Pediatric Infectious Disease (Drs Nathawad and Hammerschlag) and Scientific Computing Center (Dr Weedon), State University of New York Downstate Medical Center and Long Island College Hospital (Drs Marzouk and Goldstein), Brooklyn.
Objectives To examine the prevalence of methicillin-resistant Staphylococcus aureus (MRSA) colonization among children undergoing bilateral myringotomy and tube insertion with or without adenoidectomy for chronic otitis media with effusion or recurrent acute otitis media, as well as to examine the occurrence of postoperative otorrhea in children who have vs do not have MRSA colonization.
Design Prospective cohort study.
Setting Hospital-based pediatric otolaryngology practice in a metropolitan area.
Patients Seventy-six children (51 boys and 25 girls), with a mean (SD) age of 3.6 (1.8) years.
Interventions Cultures for S aureus from the nasopharynx, external auditory canals, middle ears, and adenoid were obtained at the time of surgery, as well as middle ear cultures for bacteriologic culture and sensitivity. Patients were followed up for the development of otorrhea.
Main Outcome Measures Prevalence of MRSA colonization and predictors of subsequent otorrhea.
Results The prevalence of S aureus colonization at the time of bilateral myringotomy and tube insertion was 7.9% (95% CI, 3.0%-16.4%), and the prevalence of MRSA colonization was 3.9% (95% CI, 0.8%-11.1%). All MRSA-positive specimens were resistant to erythromycin, and 2 were resistant to clindamycin. The mean (SD) follow-up period was 11.6 (3.6) months. Twenty-seven patients (35.5% [95% CI, 25.1%-46.9%]) developed at least 1 episode of otorrhea. One of 3 patients with MRSA colonization had subsequent otorrhea. The only predictor of otorrhea was younger age.
Conclusions The prevalence of MRSA colonization among otitis-prone children was similar to rates reported among the general pediatric community. Methicillin-resistant S aureus colonization at the time of bilateral myringotomy and tube insertion was not predictive of subsequent otorrhea.
Bilateral myringotomy and tube insertion (BMT) has become a mainstay of treatment for recurrent acute otitis media (RAOM) and chronic otitis media with effusion (COME) among children in the United States. It is the most frequently performed surgical procedure among US children.1 Otorrhea is the most common complication associated with BMT, with rates ranging between 26% and 83%.2- 5 It is typically a consequence of infection within the middle ear space, with pathogens likely from the nasopharynx or from external sources. Approximately 24% of otorrhea cases following BMT result from Staphylococcus aureus infection.3
Of growing concern within the medical community is the emergence of methicillin-resistant S aureus (MRSA) and its increasing incidence within the community, particularly among the pediatric population. Despite what seems to be a constant colonization rate of S aureus within the community, the prevalence of MRSA colonization has been steadily increasing.6,7 In Brooklyn, New York, 44% of S aureus isolates collected were found to be MRSA.8 Moreover, increased prevalence of more virulent strains of MRSA was associated with lower socioeconomic regions and with overcrowding.9
In a 2005 study,10 the prevalence of MRSA colonization among the pediatric community was found to be about 9.2%. There has been a significant nationwide increase in pediatric MRSA head and neck infections.11 Methicillin-resistant S aureus infections are notoriously refractory to standard antibiotic therapy and cause more severe infections than methicillin-susceptible S aureus (MSSA), particularly certain strains with specific virulence factors, such as Panton-Valentine leukocidin.10 Methicillin-resistant S aureus has been noted to be a source for COME and RAOM, with rising incidence.12 The strain has also been found in post-BMT otorrhea, although most of the evidence comes from retrospective reviews of few cases.13
Our objectives herein were (1) to examine the prevalence of MRSA colonization within the middle ear and nasopharynx among children in Brooklyn undergoing BMT with or without adenoidectomy in a prospective manner and (2) to determine if there was any association between MRSA colonization and the incidence of postoperative otorrhea in children who have vs do not have MRSA colonization.
We conducted a prospective cohort study of patients at the State University of New York Downstate Medical Center and the Long Island College Hospital, Brooklyn, recruited between July 30, 2009, and February 17, 2010. Patients (age range, 6 months to 18 years) undergoing BMT with or without adenoidectomy for treatment of COME or RAOM were included. Recurrent acute otitis media was defined as 3 or more episodes of acute otitis media in the preceding 6 months or 4 or more episodes in the preceding 12 months. Chronic otitis media with effusion was defined as a minimum of 3 months of bilateral middle ear effusion or 6 months of unilateral effusion. Patients who were immunocompromised, with diagnoses including but not limited to immunodeficiency syndromes, diabetes mellitus, current chemotherapy, or chronic corticosteroid dependence, were excluded from the study. Patients with craniofacial syndromes, history of cleft palate, or prior ear surgery other than BMT were also excluded from the study. The study was approved by the institutional review boards of the State University of New York Downstate Medical Center and the Long Island College Hospital, informed written consent was obtained from the parent or legal guardian, and assent was obtained from patients older than 7 years. A convenience sample was recruited based on the availability of one of us (H.A.M.) to attend the surgical procedures and to collect the specimens.
Information was collected before the procedure about the frequency of infections and the number of antibiotic courses received by patients over the past year, any complications of otitis media, the number of previous BMTs, the patient's daily environment (home, day care, or school), exposure to passive cigarette smoking, the number of members of the household working in a health care environment, type of health insurance, and education level of 1 or both parents. All patients underwent audiologic testing before surgery. All BMTs were performed using general mask anesthesia, and all BMTs with adenoidectomy were performed using general endotracheal anesthesia.
At the time of surgery, the following cultures were obtained: (1) A nasopharyngeal swab was obtained by swabbing the nasopharynx with a small calcium alginate swab via direct vision if the patient was undergoing adenoidectomy or transnasally if the patient was undergoing BMT only. (2) Bilateral external auditory canal swabs were obtained by swabbing the canals with small calcium alginate swabs. One milliliter of 70% isopropyl alcohol was placed in each external auditory canal for 1 minute and was suctioned clear. (3) If middle ear fluid was found in either ear, middle ear cultures were collected by attaching a sputum trap to the end of the Frazier suction used to evacuate the middle ear fluid after myringotomy. (4) For patients undergoing adenoidectomy, a portion of the adenoid specimen was collected.
After surgery, patients were followed up for 1 year through office visits or by telephone to determine if there had been any episodes of otorrhea since their last visit or last telephone call and, if so, how many. Information about tube patency or extrusion was also obtained.
All cultures obtained were evaluated for the presence of MRSA. Swabs were incubated in trypticase soy broth for 24 hours and were then plated on mannitol salt agar and incubated for 24 to 48 hours. Mannitol-fermenting colonies from each morphotype (on the mannitol salt agar plate) were subcultured on trypticase soy agar plus 5% sheep blood agar plates and were incubated for 24 hours at 37°C. Cultures on the trypticase soy agar plates were confirmed to be S aureus using latex agglutination and catalase tests. Confirmed S aureus isolates were screened for oxacillin resistance using the Clinical and Laboratory Standards Institute disk diffusion test. Bacterial suspension was prepared in Mueller-Hinton broth to the 0.5 McFarland standard and was plated on Mueller-Hinton agar with a 1-μg oxacillin sodium disk within the inoculum. Zone diameters were measured and recorded after 24-hour incubation at 35 ± 2°C (≥13 mm was considered susceptible S aureus, 11-12 mm intermediate, and ≤10 mm resistant).
Isolates determined to be resistant to oxacillin (MRSA) by disk diffusion were tested against common antibiotics using a commercially available test (Etest; bioMérieux SA, Marcy l’Etoile, France). Assessment of antibiotic sensitivity included the following: oxacillin, trimethoprim-sulfamethoxazole, tetracycline, clindamycin, erythromycin, ciprofloxacin, vancomycin, linezolid, and rifampin. Double-disk diffusion test was performed on all isolates that were resistant to erythromycin but sensitive to clindamycin.
Middle ear cultures were sent for routine culture and sensitivity. They were processed using standard techniques.
The sample size estimation was based on the assumption of a 9% MRSA colonization rate in children as documented in prior investigations.10 As a result, a minimum of 72 patients was required assuming a 90% CI with a width of 10 (ie, 9% [90% CI, 4%-14%]). This sample size estimation was determined using statistical software (NCSS/PASS; NCSS, Kaysville, Utah).
The prevalence (95% CI) of MRSA colonization at the time of BMT was determined. Poisson regression analysis was used to predict the number of otorrhea episodes with patent tubes during the follow-up period. A scaling variable derived from the deviation statistic was introduced where necessary to correct for mild overdispersion. Predictors investigated were the following: age, sex, allergies, race/ethnicity, history of previous BMT, gastroesophageal reflux disease, adenoidectomy at the time of BMT, exposure to passive cigarette smoking, family members working in health care, MRSA colonization at the time of BMT, positive middle ear culture at the time of BMT, the number of antibiotic courses in the year before BMT, patient's daily environment (day care vs home or school), type of health insurance (commercial or military vs public or none), indication for BMT (COME vs RAOM or both RAOM and COME), the number of episodes of acute otitis media in the year before BMT, middle ear effusion at the time of BMT (none vs unilateral vs bilateral), and maternal education level (high school education or less vs greater than a high school education). Univariate analyses were conducted. Exact Poisson regression analysis was performed for significant predictors on univariate analysis. Otorrhea incidence rates are reported as the number of events per person-year (EPPY). P < .05 was considered statistically significant.
Eighty-five patients were offered participation in the study, 6 families refused, and 79 patients were recruited. However, 3 patients were excluded from analysis because of technical error when intraoperative cultures were acquired, so 76 patients comprised the study sample. Patient demographics, baseline characteristics, and operative findings are given in Table 1. The mean (SD) age of patients was 3.6 (1.8) years, 51 patients (67.1%) were male, and 51 patients (67.1%) were of white race/ethnicity. Tubes were placed for RAOM in 12 patients (15.8%), for COME in 39 patients (51.3%), and for both in 25 patients (32.9%). The mean (SD) number of episodes of acute otitis media was 3.0 (3.0) in the year before enrollment, and the mean (SD) number of antibiotic courses was 3.7 (2.8). Sixty-two patients (81.6%) were undergoing their first tube placement. Adenoidectomy was performed in 37 patients (48.7%). No middle ear effusion was found in 8 patients (10.5%), unilateral effusion was found in 18 patients (23.7%), and bilateral effusion was found in 50 patients (65.8%). Thirty patients (26.3%) had family members who worked in a health care setting.
Two nasopharyngeal cultures were positive for MSSA, and 2 were positive for MRSA. In 1 patient with a nasopharyngeal culture positive for MSSA, his adenoid specimen was also positive for MSSA. No other adenoid specimen demonstrated MSSA or MRSA. One external auditory canal culture was positive for MRSA, and 1 middle ear culture was positive for MSSA. No other external auditory canal or middle ear cultures were positive. Therefore, the prevalence of S aureus colonization among our patient population was 7.9% (95% CI, 3.0%-16.4%), and the prevalence of MRSA colonization was 3.9% (95% CI, 0.8%-11.1%). All MRSA-positive specimens were resistant to erythromycin, 2 were resistant to clindamycin, and 1 was resistant to tetracycline. All were sensitive to vancomycin, linezolid, ciprofloxacin, rifampin, and trimethoprim-sulfamethoxazole.
Middle ear culture results are given in Table 2. Eighty-three specimens (72.2%) yielded no growth. Haemophilus influenzae and coagulase-negative Staphylococcus were the most prevalent organisms. Of 3 patients with MRSA colonization, all had bilateral middle ear effusion, 2 had middle ear cultures that grew H influenzae, and 1 had no growth on his middle ear cultures. Because only 3 patients had MRSA colonization, evaluation of potential predictors of MRSA colonization could not be performed.
The mean (SD) follow-up period, including office visit or telephone contact, was 11.6 (3.6) months, with 63 (82.9%) patients having at least 12 months of follow-up care. Fifty-two patients (68.4%) had office visit follow-up care, and the mean (SD) duration of tube patency for at least 1 ear in this group was 12.0 (2.8) months. One patient had tubes reinserted 8 months after study entry, so his time with patent tubes was considered 8 months. Assuming that parents contacted by telephone were accurate historians about the patency of the tubes, the mean (SD) tube patency for at least 1 ear for the entire study sample was 11.3 (3.7) months.
Twenty-seven patients (35.5% [95% CI, 25.1%-46.9%]) developed at least 1 episode of otorrhea during the follow-up period. Nineteen patients had 1 episode, 3 had 2 episodes, 3 had 3 episodes, and 2 had 4 episodes. The overall otorrhea incidence rate was 0.62 EPPY (95% CI, 0.44-0.87 EPPY). Three of 27 patients (11.1%) developed immediate postoperative otorrhea; the remainder of the patients had delayed otorrhea. One of 3 patients with MRSA colonization (in his external auditory canal) had an episode of delayed otorrhea. On univariate analysis, the following were predictive of otorrhea among patients with patent tubes during the follow-up period: no previous BMT (estimated incidence ratio of EPPY vs previous BMT, 4.7 [95% CI, 1.0-25.0]; P = .01), day care attendance (EPPY vs home care or school, 2.9 [95% CI, 1.5-5.5]; P = .01), gastroesophageal reflux disease (EPPY vs no gastroesophageal reflux disease, 2.6 [95% CI, 1.3-5.1]; P = .008), and younger age (no episodes of otorrhea were observed in patients older than 5 years vs an EPPY of 0.7 for younger patients, P < .001). Exact Poisson regression analysis that included all 4 predictors showed significant results for younger age only (estimated incidence ratio of EPPY, 7.0 [95% CI, 1.2 to infinity]; exact P = .02).
Community prevalence rates of MSSA and MRSA colonization vary depending on the population studied and the geographic region. Creech et al10 reported prevalences of MSSA colonization of 36.4% and MRSA colonization of 9.2% among children attending health maintenance visits in Nashville, Tennessee; the rate of MRSA colonization increased from 0.8% in 2001 to 9.2% in 2004. Fritz et al14 reported a mean MRSA prevalence of 2.6% among children seen for sick and well-child care visits in metropolitan St Louis, Missouri, with rates ranging from 0% to 9% depending on the pediatric practice. In 2003-2004 vs 2006-2007, Lee et al15 reported stable prevalences of MSSA (14.6% vs 14.1%) and MRSA (0.2% vs 0.9%) among children attending sick and well-child care visits in 8 Massachusetts communities. The prevalence rates of MRSA among children in day care settings were observed to be 6.7% in Galveston, Texas,16 and 1.3% in North Carolina and Virginia.17 Vegunta et al18 found prevalences of MSSA of 36.6% and MRSA of 28% on routine screening cultures among children undergoing elective pediatric general surgical procedures. Risk factors for MRSA colonization vary across studies but include minority race/ethnicity,9,14 low socioeconomic status,9,14 family member working in health care,10,14 asthma medication use,16 and exposure to antibiotics.16
To our knowledge, this is one of the first studies to evaluate MRSA and MSSA colonization in otitis-prone children. Although not specifically evaluating for S aureus, Jacobs et al19 performed nasopharyngeal cultures on 201 children undergoing BMT in 2004 and 2005 in Ohio and Pennsylvania and found no isolate of MRSA or MSSA. Herein, we expected a higher rate of colonization compared with that among healthy children, as children with otitis media have been shown to have higher rates of nasopharyngeal colonization of organisms typical of otitis media, including Streptococcus pneumoniae, nontypable H influenzae, and Moraxella catarrhalis.20 Antibiotic exposure may also be a risk factor for colonization, and our patients received a mean of 3.7 antibiotic courses in the year before enrollment. In addition, given our urban inner-city setting, with much overcrowding, we anticipated that the MRSA colonization rate among our patient population would be higher than the rates quoted in other studies.
However, the MRSA colonization rate in our cohort (3.9%) is consistent with the rates reported in investigations of healthy children and is lower than those reported in several investigations. Of S aureus isolated from our cultures, about one-half of those specimens were MRSA. This is consistent with previous findings from our community that showed 44% of S aureus isolated was MRSA.8 The resistance pattern of the MRSA isolates is consistent with that of community-acquired MRSA, with sensitivity to trimethoprim-sulfamethoxazole and variable sensitivity to clindamycin. Because of the few MRSA isolates, we could not evaluate potential predictors of MRSA colonization.
The most common pathogenic organism identified from middle ear cultures was H influenzae, and there was 1 isolate each of S pneumoniae and MSSA. Coagulase-negative Staphylococcus is considered nonpathogenic but is frequently isolated from middle ear effusions. Our findings are in agreement with prior evidence in terms of overall percentages of positive cultures and a high percentage of H influenzae and nonpathogenic Staphylococcus, but we found no M catarrhalis, a common organism usually identified in about 10% of middle ear effusions.21,22 Our low rate of S pneumoniae is in agreement with studies22,23 that demonstrate lower prevalence since the introduction of the pneumococcal 7-valent conjugate vaccine. Jung et al24 identified MRSA among 4.1% and MSSA among 3.8% of 289 children undergoing BMT in Seoul, Korea, between 2004 and 2008, a 5-fold higher percentage of S aureus than that obtained in our study.
Our rate of postoperative otorrhea (35.5% [27 of 76 patients]) is in agreement with previous studies.2- 5 We did not assess the bacteriologic culture of postoperative otorrhea in our patients, as most of them receive treatment for postoperative otorrhea from their pediatricians and we did not have the resources for the patients to be seen in the otolaryngology offices. Although MRSA colonization at the time of tube insertion was not predictive of postoperative otorrhea, the number of patients with MRSA colonization was few and may have precluded our obtaining statistically significant results. The only predictor of postoperative otorrhea on multivariate analysis was younger age, which has been found to be a risk factor in previous investigations.25 Although low socioeconomic status has been found to be predictive of otorrhea,2 we did not find this association.
The strengths of our study are its prospective nature, the adequate sample size, an 89.4% (76 of 85) participation rate of patients offered entry, and the blinding of the investigators performing the cultures to the patients' clinical information. Weaknesses include the lack of cultures during episodes of postoperative otorrhea, a reliance on parental history about tube patency and otorrhea history for 24 patients (31.5%), and the absence of a control group of healthy children without otitis to determine the local prevalence of S aureus among the Brooklyn pediatric community.
In conclusion, we found a low 3.9% prevalence of MRSA colonization among children undergoing BMT for treatment of RAOM and COME. Half of the S aureus isolates were MRSA. Methicillin-resistant S aureus colonization was not predictive of post-BMT otorrhea.
Correspondence: Nira A. Goldstein, MD, MPH, Division of Pediatric Otolaryngology, State University of New York Downstate Medical Center, 450 Clarkson Ave, Campus Box 126, Brooklyn, NY 11203 (email@example.com).
Submitted for Publication: June 3, 2011; accepted July 31, 2011.
Author Contributions: Drs Marzouk, Nathawad, Hammerschlag, and Goldstein had full access to all the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis. Study concept and design: Hammerschlag and Goldstein. Acquisition of data: Marzouk, Nathawad, and Bachman. Analysis and interpretation of data: Marzouk, Nathawad, Hammerschlag, Weedon, and Goldstein. Drafting of the manuscript: Marzouk and Goldstein. Critical revision of the manuscript for important intellectual content: Nathawad, Hammerschlag, Weedon, and Bachman. Statistical analysis: Weedon. Obtained funding: Goldstein. Administrative, technical, and material support: Bachman. Study supervision: Hammerschlag and Goldstein.
Financial Disclosure: None reported.
Funding/Support: This study was supported by a State University of New York Downstate College of Medicine 2009 Research Investment Initiative Program grant (Dr Goldstein).
Previous Presentation: This study was presented as a poster at the American Society of Pediatric Otolaryngology meeting; April 29, 2011; Chicago, Illinois.
Additional Contributions: Ari J. Goldsmith, MD, and Richard M. Rosenfeld, MD, MPH, allowed us to approach their patients for study recruitment.