Objective.The role of routine antimicrobial treatment of acute middle-ear infections is under debate, because the efficacy of antimicrobials in the resolution of middle-ear fluid has not been unambiguously proven. Acute tube otorrhea is regarded as evidence of acute otitis media, and for methodologic reasons it was chosen to provide objectivity for diagnostics and outcome assessment. The objective of this study was to assess whether amoxicillin-clavulanate accelerates the resolution of acute tube otorrhea.
Design and setting.Randomized, double-blind, placebo-controlled study in outpatient setting.
Patients.Volunteer sample of basically healthy 6- to 72-month-old children with a tympanostomy tube. Eligibility required having acute tube otorrhea of <48 hours’ of duration and no prior treatment within the last 2 weeks. The mean age of the participants was 25 months; they had a history of 3 episodes of acute otitis media (median), and 99% had manifestations of a concomitant respiratory infection. Of 79 randomized patients, 7 were withdrawn because of adverse events; 66 patients completed the study.
Interventions.Amoxicillin-clavulanate (N = 34; 45 mg/kg/d) or matching placebo (N = 32) for 7 days and daily suction of middle-ear fluid through tympanostomy tube.
Main outcome measures.Duration of acute tube otorrhea and duration of bacterial growth in middle-ear fluid.
Results.The median duration of tube otorrhea was significantly shorter in amoxicillin-clavulanate than in the placebo group (3 vs 8 days). At the end of the 7-day medication period, tube otorrhea was resolved in 28 of 34 children receiving amoxicillin-clavulanate compared with 13 of 32 children on placebo (treatment-control difference 41%; 95% confidence interval, 20%–63%; number needed to treat, 2.4). The median duration of bacterial growth in middle-ear fluid was shorter in amoxicillin-clavulanate than in the placebo group (1 vs 8 days).
Conclusions.Oral antibiotic treatment significantly accelerates the resolution of acute tube otorrhea by reducing bacterial growth in middle-ear fluid.
The use of antibiotics as a standard treatment of acute otitis media has become a subject of increasing controversy. Although many experts still favor the use of antibiotics,1–3 several others advocate withholding the routine use of antibiotics in most cases of acute otitis media.4–6 The current debate on the value of antibiotics is understandable, because the efficacy of antibiotic treatment on the resolution of middle-ear fluid has not been convincingly demonstrated in previous randomized, double-blind, placebo-controlled studies.7–16 However, considering the well-known difficulties of determining the presence of middle-ear fluid in ears with intact tympanic membranes, the effect of antibiotics might have remained undetected.
In children with tympanostomy tubes, acute otitis media can be reliably determined by noticeable acute otorrhea in addition to acute symptoms.17 The causative pathogens in young children with acute otorrhea are mainly the same as in acute otitis media with intact tympanic membrane.18 Furthermore, the presence of middle-ear fluid, the cornerstone of the diagnosis of acute otitis media, can be objectively verified by suctioning middle-ear fluid through a tympanostomy tube, which also provides samples for microbiologic detection. Despite these advantages, no previous randomized placebo-controlled study has been published of the treatment of acute tube otorrhea, although it is a common pediatric problem.19
We hypothesized that antibiotics have a substantial impact on the outcome of middle-ear infections by shortening the duration of bacterial growth in middle-ear fluid. For methodologic reasons, we decided to study the therapeutic effects of antimicrobials in children with tympanostomy tubes. We conducted a randomized, double-blind, placebo-controlled study to determine by daily follow-up whether treatment with amoxicillin-clavulanate shortens the duration of acute tube otorrhea.
The study participants were recruited between September 1998 and June 1999 by giving personal written information to all families whose children had undergone tympanostomy tube placement at 3 local public hospitals. Furthermore, to also reach those using private health care, information was distributed through daycare centers, family daycare, health centers, well-infant clinics, a large private pediatric office, and local media within the area of Turku, Finland. In all children, the decision of tympanostomy tube placement had been made independently by their own physicians who were not involved in this study. Children aged 6 months to 6 years qualified for enrollment if they had acute otorrhea through a tympanostomy tube that had started within the preceding 48 hours. The exclusion criteria were: tympanostomy tube insertion or any systemic or local antibiotic or steroid therapy within the preceding 2 weeks; tube otorrhea within the preceding 4 weeks; allergy to penicillin or amoxicillin; immunodeficiency; Down syndrome; cleft palate; or granulation or polyp in the tympanic membrane. All participants were primary care patients, although the study clinic was located at Turku University Hospital.
The study protocol was approved by the Ethics Committee of Turku University Hospital. Written informed consent was obtained from the parents of all children. All visits were free of charge; no compensation for participation was given.
This was a randomized, double-blind, placebo-controlled study and its objective was to assess the efficacy of amoxicillin-clavulanate in the treatment of acute tube otorrhea. At enrollment, each child underwent a thorough clinical examination. All otologic examinations and procedures were done using an otomicroscope, and in 94% of cases by the same investigator (A.R.). The ear canal was carefully cleaned by suction without disinfectants. Using a separate sterile suction trap, middle-ear fluid was suctioned through a tympanostomy tube, which confirmed the diagnosis of acute tube otorrhea. The middle-ear fluid obtained was diluted with 1.0 mL of saline, and 10 μL of the solution was inoculated on prewarmed blood-agar and chocolate-agar plates for bacterial culture by routine methods. The sensitivity of the bacterial isolates to amoxicillin-clavulanate was tested by disk diffusion, and the results were interpreted using National Committee for Clinical Laboratory Standards.20 The physicians responsible for the clinical examinations were unaware of the bacterial findings that were evaluated daily by one of us uninvolved in the patient examinations.
The children were randomly allocated to receive either amoxicillin-clavulanate (45 mg/kg/d of amoxicillin divided into 2 daily doses; amoxicillin to clavulanate ratio, 7:1) or placebo for 7 days (both drugs provided by SmithKline Beecham Pharmaceuticals, Bristol, TN). During the 7-day medication period, the use of otic drops or any other drugs, except paracetamol (acetaminophen), was not permitted. The parents were given a dosage card for recording daily the doses of medications given. The children were reexamined at the study clinic every day during the medication period: middle-ear fluid was suctioned and a specimen for bacterial culture was obtained as described above. When no more middle-ear fluid was obtained by suction, these reexaminations were discontinued.
If acute tube otorrhea persisted at the end of the 7-day medication period, the child was treated openly with antibiotics. The predetermined criteria to stop the study medication and to provide the study patient with an open antibiotic treatment were the following: bacteria resistant to amoxicillin-clavulanate, occlusion of the tympanostomy tube, extrusion of the tympanostomy tube combined with closure of perforation, migration of the tympanostomy tube into tympanum, development of granulation or polyp in the tympanic membrane, and an adverse reaction possibly caused by the study drug. At 1 week after the end of the study medication, all children were reexamined and the dosage cards and the medication bottles were returned.
The random allocation sequence was done in blocks of 4 according to a computerized scheme by the Department of Biostatistics, University of Turku, and the medication bottles were labeled and provided for us by the pharmacy of Turku University Hospital. The randomization list included numbers from 1 to 100, and the study physicians allocated each enrolled individual in consecutive order of study entry to receive the medication bottles with the next, ie, the lowest number in the list. The sealed envelopes containing the codes for each medication were provided for us together with the medication bottles by the pharmacy and kept under lock in the study clinic. None of the envelopes were opened before the completion of the study. The masking of the study drugs was ensured by the identical appearance, smell, and taste of the syrups.
Outcomes and Definitions
The diagnosis of acute tube otorrhea was based on ear discharge that had started within 48 hours before enrollment and that was verified by suction of middle-ear fluid through a tympanostomy tube. The primary outcome was the duration of tube otorrhea. Otorrhea was defined as resolved on the day when the study physician could not obtain any more middle-ear fluid by suction through the tympanostomy tube. The secondary outcome was the duration of bacterial growth in middle-ear fluid. Bacterial growth was defined as having continued until the day when no more bacteria could be cultured from middle-ear fluid or no more middle-ear fluid was obtained by suction. These outcomes were measured as whole days and a part of a 24-hour period was defined as 1 day; the day of enrollment was day 0. If tube otorrhea or bacterial growth were still detected on day 7, these outcomes were defined as 8 days in the analyses. In the case of bilateral otorrhea, the ear that drained longer was included in the analyses. The primary and secondary outcomes were analyzed both on a per-protocol and on an intention-to-treat basis. In the intention-to-treat analyses, the worst-case principle was obeyed, meaning that if the actual duration of tube otorrhea was not known it was determined as 8 days.
Noncompliance with the study medication was defined as 2 or more consecutive missed doses, or a total of 3 or more missed doses recorded in the dosage card. Compliance with the medication was additionally evaluated by measuring the amount of drug left in the bottles.
Sample size estimation was based on the assumption that the duration of tube otorrhea would be reduced by at least 2 days, and on the expectation that the standard deviation for the mean duration of tube otorrhea would be 2 days.21 To achieve 95% power with α = .05, the minimum required number of children in each group was 26.
The baseline characteristics were compared between the amoxicillin-clavulanate and the placebo group using the Student t test for continuous variables, the Mann-Whitney U test for nonparametric continuous data, and the χ2 test for categorical variables. The differences in the primary outcome, the duration of tube otorrhea, between the 2 groups were assessed by comparing the medians by using the Mann-Whitney U test as well as using a log-rank test to compare the distributions of the primary outcome. The percentages of patients cured after the first week were compared using the χ2 test. The treatment-control rate difference and its 95% confidence intervals were calculated and the number needed to treat was derived from the rate difference. The results of the secondary end point, the duration of bacterial growth, were first evaluated by comparing the medians between the groups using the Mann-Whitney U test. Furthermore, the differences between the 2 groups on each study day in the proportions of patients with bacterial growth were compared by the χ2 test. All statistical tests were 2-sided and a P value of < .05 was considered statistically significant. The statistical analyses were done using SPSS software (SPSS, Chicago, IL).
Demographics and Compliance
Altogether, 259 families contacted the study personnel by phone, 135 children met the exclusion criteria (the main reasons were antibiotic treatment within the last 2 weeks, tube otorrhea within the last 4 weeks, no otorrhea, or allergy to the study medication), 27 families declined, and 97 visited the study clinic. The details of the participant flow are presented in Fig 1. Of the 66 children who acceptably completed the study, 34 received amoxicillin-clavulanate and 32 placebo. The baseline characteristics of the groups were comparable (Table 1). Compliance to the medication was good; 6 children receiving amoxicillin-clavulanate missed 1 dose, and 7 and 2 children in the placebo group missed 1 and 2 doses, respectively. According to the amount of the drug left in the bottles, none of the children received less drug than recorded in the dosage card. The bottles of 4 children were washed before returning, and those of 3 children were not returned.
Duration of Tube Otorrhea
Table 2 illustrates the main outcomes of the study. Overall, amoxicillin-clavulanate significantly reduced the duration of acute tube otorrhea, as shown by the Kaplan-Meier estimates (Fig 2). The median duration of tube otorrhea was 5 days shorter in children treated with amoxicillin-clavulanate than in the placebo recipients (3 vs 8 days; P = .002), and the proportion of children cured during the study medication period was 2 times higher in the amoxicillin-clavulanate group than in the placebo group (82% vs 41%; P < .001). The absolute treatment-control rate difference was 41% (95% confidence interval, 20%–63%; number needed to treat, 2.4).
The bacterial findings in middle-ear fluids are summarized in Table 3. At enrollment, pathogenic bacteria were cultured in the middle-ear fluids of 27 children (79%) receiving amoxicillin-clavulanate and of 25 children (78%) on placebo. In 5 placebo recipients, positive bacterial cultures were only obtained from day 1 onwards. In 5 other placebo recipients a second pathogen appeared in the middle-ear fluid, whereas in the amoxicillin-clavulanate group a second pathogen appeared in the middle-ear fluid of 3 children during the study medication period. All pathogens were sensitive to amoxicillin-clavulanate; 3 strains of Streptococcus pneumoniae were intermediately resistant to penicillin. All except 1 strain of Moraxella catarrhalis and 14% of Haemophilus influenzae strains produced β-lactamase.
After the onset of the study medication, the proportion of children with bacterial growth in middle-ear fluid was each day significantly lower in the amoxicillin-clavulanate group than in the placebo group (P = .001; Fig 3). The median duration of bacterial growth in middle-ear fluid was significantly shorter in the amoxicillin-clavulanate group compared with the placebo group (1 day vs 8 days; P < .001; Table 2).
Relationship Between Duration of Tube Otorrhea and Bacterial Growth
Of the 66 children, 25 still had tube otorrhea at the end of the study medication, and bacterial cultures of the middle-ear fluids remained positive in 20 (80%) of these children (Fig 4). Acute tube otorrhea resolved during the medication period in altogether 41 children, 33 of whom had had bacteria in middle-ear fluid. Otorrhea resolved within 1 day of the first negative bacterial culture in 19 (58%) of these 33 children, and within 2 days in 29 (88%) of 33 children. Of these 33 children, 22 were treated with amoxicillin-clavulanate, and in 20 (91%) of them bacteria were eradicated before the resolution of tube otorrhea, whereas in the placebo group this occurred in 4 (36%) of 11 children in whom otorrhea ceased during the first week. The duration of tube otorrhea according to the causative bacterium is shown in Fig 5 for patients who had a single causative agent in the middle-ear fluid at study entry. Because the numbers of observations were small, no statistical analyses were done, but in the placebo group, tube otorrhea tended to be longest if M catarrhalis was found in primary cultures.
Because of an adverse event, the investigators withdrew 2 children on amoxicillin-clavulanate (extrusion of the tympanostomy tube, and vomiting) and 5 children on placebo (contralateral acute otitis media with perforation of the tympanic membrane, eczema, development of granulation, and extrusion of tympanostomy tube in 2 children). Additional mild adverse events (mainly gastrointestinal) that had no effect on compliance occurred in 8 amoxicillin-clavulanate recipients and in 1 placebo recipient.
Our results provide clear evidence that antibiotic treatment is superior to placebo in the treatment of acute tube otorrhea in young children. The beneficial effect of amoxicillin-clavulanate was observed both on the duration of tube otorrhea as well as on the duration of bacterial growth in middle-ear fluid. The aim of antibiotic therapy is to eradicate bacteria at the site of the infection and thus help the host to recover. This was demonstrated in this study by the close connection between the duration of bacterial growth and the duration of tube otorrhea. In most cases, the eradication of bacteria resulted in rapid resolution of middle-ear fluid, whereas bacterial growth persisted in most children in whom tube otorrhea continued after the medication period. Our results are in agreement with previous findings on the clinical effect of bacterial eradication in children with acute otitis media. In studies where the presence of bacteria in middle-ear fluid was documented by double tympanocentesis, clinical symptoms persisted for 3 to 6 days after onset of antibiotic treatment significantly more often in children who still had bacteria in middle-ear fluid than in children with negative bacterial cultures at the second tympanocentesis.22,23 The complexity of middle-ear infections is highlighted by our finding that otorrhea continued after the medication period also in some children receiving amoxicillin-clavulanate, although the bacteria isolated in middle-ear fluids were sensitive to the drug. It is probable that the inflammatory reaction in the middle-ear mucosa plays a role in the prolongation of the middle-ear infection. This is supported by our recent study demonstrating that the combination of oral prednisolone and amoxicillin-clavulanate provides significantly faster resolution of acute tube otorrhea than does amoxicillin-clavulanate alone.21 Nevertheless, since bacteria are the main evokers of the inflammatory reaction, bacterial eradication seems to be the basis for a recovery process in the middle ear.
Tube otorrhea ceased during the first week in 41% of placebo recipients, demonstrating that tube otorrhea may resolve also without systemic antibiotic treatment. However, careful daily suction of middle-ear fluid through the tympanostomy tube may have boosted recovery. So far, our knowledge about the spontaneous eradication of bacteria in middle-ear fluid has been based on a single study that evaluated placebo and 4 different antimicrobial agents in the treatment of acute otitis media.7 Our results do not seem to support the concept derived from that study that spontaneous eradication of H influenzae or M catarrhalis would occur more frequently than that of S pneumoniae. However, it must be acknowledged that the numbers of children with each organism were small in our study. In agreement with our data, the previous study showed that children on placebo were able to eradicate bacteria from middle-ear fluid, but this occurred significantly less often in them than in children treated by antibiotics. Another finding comparable to ours was that new pathogens appeared in middle-ear fluid significantly more often in the placebo than in the antibiotic groups. Because the nasopharynx has been verified to be the source of the middle-ear pathogens,24,25 it may be that in children without oral antimicrobials, new pathogens from the nasopharynx enter the middle-ear and contribute to the ongoing infection and inflammatory process. Therefore, a primarily negative culture of middle-ear fluid does not seem to be a good reason to withhold antimicrobial treatment of acute middle-ear infections.
Acute tube otorrhea is a common problem18,19,26 that mainly coincides with upper respiratory infection.27 The lack of consensus about its best treatment appears to arise from a lack of randomized controlled studies in well-defined populations.28–31 The recommendations favor systemic antibiotic treatment in children with concomitant respiratory infection.30,32 On the other hand, topical otic solutions, especially fluoroquinolones, are often recommended for the treatment of tube otorrhea. Topical fluoroquinolones have few adverse effects and they are regarded not to provoke bacterial resistance.33,34 However, the penetration of otic drops through a tympanostomy tube into the middle ear is doubtful, especially while treating a draining ear.35–37 In the study of Goldblatt et al,33 no difference was found between ofloxacin otic solution and amoxicillin-clavulanate for treatment of tube otorrhea. However, their study population was heterogeneous: tube otorrhea could have lasted for up to 3 weeks, prior therapy was permitted, the age distribution was wide (from 1 to 12 years), and patients having Pseudomonas aeruginosa were included although amoxicillin-clavulanate could not be considered an effective treatment for those. Furthermore, the clinical outcome was evaluated after 3 weeks from the beginning of treatment, although such a long interval has a high tendency for spontaneous recovery. Because otorrhea may restrain daycare attendance or other social activities and because prolonged tube otorrhea has clearly been shown to impair the quality of life,38 the duration of tube otorrhea should be restricted to be as short as possible. The present study demonstrates that in young children with concomitant respiratory infection, acute tube otorrhea can be cured with oral antimicrobials in most cases within 3 days, provided that the causative bacteria are sensitive to the drug.
All of our study participants had a patent tympanostomy tube, and thus the conditions in the middle ear were different from those in children with intact tympanic membranes, which is a limitation for direct generalization of our results. However, the characteristics of our study population parallel with those of children suffering acute otitis media with intact tympanic membranes.39 Virtually all cases of acute tube otorrhea developed during a viral type upper respiratory infection, which is the usual sequence of events in most children with acute otitis media.39,40 None of the participants had signs of infection in their ear canals. The bacteria isolated, as well as the age of the study population, were typical of children with acute otitis media.41–43 Finally and perhaps most importantly, although the study children had undergone tympanostomy tube placement, the occurrence of acute otitis media in these children agreed with the general incidence in this young age group.43,44 All of these features indicate that the study children did not represent a small fraction of children with exceptionally severe otitis problems, but rather that they belonged to the majority of children in whom acute middle-ear infections and the persistence of middle-ear fluid are frequent problems. Therefore, the present study attempts to support the view that antibiotics are beneficial in the treatment of acute otitis media with an intact tympanic membrane,45 although the previous randomized, double-blind, placebo-controlled studies have failed to show any consistent beneficial effect on the duration of middle-ear effusion.7–16 We believe that the most important reason for the difference between the present and the previous studies is our study design that enabled objective detection of the presence of middle-ear fluid. The majority of the previous studies have not used objective methods in diagnostics and outcome assessment, and none of them have followed the resolution of middle-ear fluid on a daily basis, both of which may have diluted the potential differences between antimicrobial and placebo groups.
This study provides evidence for the efficacy of antibiotics in the treatment of acute middle-ear infection. The superiority of antimicrobial therapy to placebo in the eradication of bacteria from middle-ear fluid and in shortening the period with middle-ear fluid suggests that antibiotics may have a greater impact in the treatment of acute otitis media than earlier studies have indicated.
This work was supported by the Foundation of Pediatric Research, the Jenny and Antti Wihuri Foundation, and the Academy of Finland.
We are indebted to Kirsi-Maija Suomela, RN, for excellent assistance in the conduct of this study, and to Timo Kattelus for producing the graphs.
- Received May 21, 2002.
- Accepted September 18, 2002.
- Address correspondence to Aino Ruohola, MD, Department of Pediatrics, Turku University Hospital, PL 52, FIN-20521 Turku, Finland. E-mail:
This work was presented in part at the Fourth Extraordinary Symposium on Recent Advances in Otitis Media; April 2001; Sendai, Japan.
- ↵van Buchem FL, Peeters MF, van’t Hof MA. Acute otitis media: a new treatment strategy. BMJ.1985;290 :1033– 1037
- ↵Del Mar C, Glasziou P, Hayem M. Are antibiotics indicated as initial treatment for children with acute otitis media? A meta-analysis. BMJ.1997;314 :1526– 1529
- ↵Howie VM, Ploussard JH. Efficacy of fixed combination antibiotics versus separate components in otitis media: effectiveness of erythromycin estrolate, triple sulfonamide, ampicillin, erythromycin estolate-triple sulfonamide, and placebo in 280 patients with acute otitis media under two and one-half years of age. Clin Pediatr (Phila).1972;11 :205– 214
- Thalin A, Densert O, Larsson A, Lyden E, Ripa T. Is penicillin necessary in the treatment of acute otitis media? Proceedings of the International Conference on Acute and Secretory Otitis Media, Part 1, Jerusalem, Israel, November, 1985:441–446
- Kaleida PH, Casselbrant ML, Rockette HE, et al. Amoxicillin or myringotomy or both for acute otitis media: results of a randomized clinical trial. Pediatrics.1991;87 :466– 474
- Burke P, Bain J, Robinson D, Dunleavey J. Acute red ear in children: controlled trial of non-antibiotic treatment in general practice. BMJ.1991;303 :558– 562
- Appelman CL, Claessen JQ, Touw-Otten FW, Hordijk GJ, De Melker RA. Co-amoxiclav in recurrent acute otitis media: placebo controlled study. BMJ.1991;303 :1450– 1452
- ↵Damoiseaux RA, van Balen FA, Hoes AW, Verheij TJ, De Melker RA. Primary care based randomised, double blind trial of amoxicillin versus placebo for acute otitis media in children aged under 2 years. BMJ.2000;320 :350– 354
- ↵Bluestone CD, Klein JO. Otitis Media in Infants and Children. 3rd ed. Philadelphia, PA: WB Saunders; 2001:1–15
- ↵Ah-Tye C, Paradise JL, Colborn DK. Otorrhea in young children after tympanostomy-tube placement for persistent middle-ear effusion: prevalence, incidence, and duration. Pediatrics.2001;107 :1251– 1258
- ↵National Committee for Clinical Laboratory Standards. 1997. Performance standards for antimicrobial disc susceptibility testing. Approved standard M2–A6. National Committee for Clinical Laboratory Standards, Wayne, PA. 27
- ↵Murphy TF, Bernstein JM, Dryja DM, Campagnari AA, Apicella MA. Outer membrane protein and lipooligosaccharide analysis of paired nasopharyngeal and middle ear isolates in otitis media due to nontypable Haemophilus influenzae: pathogenetic and epidemiological observations. J Infect Dis.1987;156 :723– 731
- Isaacson G, Rosenfeld RM. Care of the child with tympanostomy tubes: a visual guide for the pediatrician. Pediatrics.1994;93 :924– 929
- ↵Rosenfeld RM, Isaacson G. Tympanostomy tube care and consequences. In: Rosenfeld RM, Bluestone CD, eds. Evidence-Based Otitis Media. Hamilton, Canada: BC Decker Inc; 1999:315–336
- ↵Arola M, Ruuskanen O, Ziegler T, et al. Clinical role of respiratory virus infection in acute otitis media. Pediatrics.1990;86 :848– 855
- Copyright © 2003 by the American Academy of Pediatrics