As part of their annual meeting, the Pediatric Academic Societies (PAS) and the Pediatric Infectious Diseases Society cosponsored a Hot Topic session where a number of speakers reviewed diagnosis and treatment of acute otitis media (AOM). The session began with a review of management guidelines and included descriptions of the most recent data related to current treatment approaches.
Review of the AOM Guidelines
Guidelines published in 2004 by the Subcommittee on Management of Acute Otitis Media of the American Academy of Pediatrics (AAP) were reviewed by Dr. Jerome O. Kline, Boston University School of Medicine, Boston, Massachusetts. He primarily focused on 4 areas: (1) the definition of AOM; (2) the recommendation to consider watchful waiting with certain patients; (3) recommendations for initial therapy and second-line treatment; and (4) possible prevention of AOM.
According to the AAP guidelines, the following criteria are essential for the diagnosis of AOM:
- Recent onset of signs and symptoms of middle ear inflammation
- Presence of middle ear effusion, with either bulging of the tympanic membrane, decreased mobility, or an air-fluid level
- Signs and symptoms of middle ear inflammation such as erythema or otalgia (ear pain)
Dr. Kline stressed that determination of some of the criteria may necessitate use of pneumatic otoscopy (ie, to demonstrate decrease in tympanic membrane mobility).
To successfully implement these guidelines, a clinician must be able to determine whether a patient has severe AOM or nonsevere AOM. The challenges of determining severe or nonsevere AOM were also reviewed in more detail later in the session. A general definition of severe AOM includes a temperature greater than 39� C in association with severe otalgia.
Perhaps the most controversial aspect of the 2004 guidelines is the suggestion that selected patients may be managed with watchful waiting instead of being treated with antibiotics at the time of diagnosis. For the purposes of the recommendations, the guidelines divided patients with AOM into 2 categories, based on the clarity of the history and the physical examination findings -- a certain and an uncertain group. The guidelines suggest children younger than 6 months of age should be treated with antibiotics regardless of the clarity of AOM diagnosis, so that even uncertain cases should be treated with antibiotics. Children aged 6 to 24 months with certain AOM should also receive antibiotics. However, if the diagnosis is uncertain and the symptoms are not severe, the practitioner could consider watchful waiting, with reassessment within 72 hours.
For children older than 24 months, those with certain AOM can be observed if the symptoms are not severe, and all with uncertain diagnoses should be observed. Dr. Kline maintained that it is too early to know whether these suggestions for watchful waiting have met with widespread acceptance by clinicians. Additional data are needed before this can be determined.
The importance of the recommended first- and second-line therapies put forth in the guidelines was reiterated. A first-line drug would be prescribed if treatment was initiated at diagnosis or after symptoms continue or worsen during a period of observation. It is Dr. Kline's opinion that the recommended drug therapy components of the guidelines have not been adequately followed by many clinicians.
For nonsevere episodes of AOM, amoxicillin remains the first-line drug of choice, whether prescribed at time of diagnosis or after failed observation. For severe episodes, the guidelines recommend the use of amoxicillin/clavulanic acid, again, whether prescribed at time of diagnosis or after failed observation.
Dr. Kline stressed that cephalosporins should only be considered as acceptable first-line treatment for patients with penicillin allergy. The drug with the most universal acceptance as a second-line drug in a penicillin-allergic patient is ceftriaxone, given intramuscularly for 1 to 3 days; ceftriaxone intramuscularly was the consensus recommendation for treating severe AOM, regardless of whether treatment is initial, after observation, or after failed treatment with another antibiotic.
Dr. Kline concluded with a review of statistics about diagnosis of and prescribing for AOM. In 2003, Finklestein and colleagues demonstrated a general decline in antibiotic prescribing from 1996-2000. In particular, the investigators demonstrated that the decrease in prescribing for AOM accounted for 59% of the decrease in overall antibiotic prescribing rates, and this drop was attributed to fewer diagnoses of AOM over the time period.
Implementing Watchful Waiting
Dr. David P. McCormick, University of Texas Medical Branch, Galveston, Texas, discussed whether the concept of watchful waiting can work for children diagnosed with AOM. He reviewed the spectrum of AOM presentations and described a tool used to identify candidates for watchful waiting.
In an investigation by Siegel and colleagues, almost 200 children with AOM were managed by providing a safety net prescription and treating pain. The parents were told to fill the safety net prescription if symptoms were not improved within 72 hours. This approach resulted in a 69% decrease in antibiotic prescribing -- only 31% of parents filled the prescription. However, this investigation was not a randomized study.
McCormick and colleagues designed and completed a randomized study, published in 2005. This investigation evaluated the safety, efficacy, and acceptability of a watchful waiting approach in children with nonsevere AOM. The 266 subjects had clinical symptoms and evidence of AOM seen on otoscopic examination. Nasopharyngeal cultures from each participant were obtained at enrollment and at Day 10.
The severity of participants' AOM symptoms were evaluated at enrollment and at Days 12 and 30. The classification of nonsevere involved symptom scores and physical finding scores previously developed by the authors. Patients' AOM was considered nonsevere if they were in the lower 50% of the symptom/discomfort reporting scale.
Children in the treatment arm received amoxicillin at 90 mg/kg/day (divided twice daily) for 10 days. If a treatment-group subject experienced treatment failure, he or she was treated with amoxicillin/clavulanate (same dose as amoxicillin). Watchful waiting subjects initiated the amoxicillin-only treatment in the case of treatment failure of watchful waiting. No routine pain control regimen was prescribed. Parental satisfaction was measured by survey at Days 12 and 30. A participant was considered a treatment failure if he or she returned during Days 1-12 with a symptom severity index higher than their severity index at enrollment, or if they returned with physical findings worse than at the initial visit
This study demonstrated that a watchful waiting approach could reduce overall antibiotic use by 66%: two thirds of the parents of children assigned to the watchful waiting arm did not fill the prescription for antibiotics during the 12-day observation period. However, the outcomes were slightly different for patients older or younger than 2 years of age. In children older than age 2, 77% were better within the 12-day observation period if they were treated with antibiotic compared with 76% of the patients with watchful waiting, a difference that was neither statistically nor clinically different.
In children younger than age 2, the study findings favored treatment. Among children younger than 2 years of age, 77% treated immediately with amoxicillin were better within the 12-day observation period compared with only 56% of those randomized to the watchful waiting arm. Antibiotic resistance of nasopharyngeal organisms was higher at Day 12 in the treatment group compared with the watchful waiting group.
Parental satisfaction scores were almost identical between the treatment and waiting groups at both the 12-day and 30-day visits. However, symptom diaries kept by parents documented quicker improvement for participants treated with amoxicillin, and this pattern held in the subgroup of children younger than 2 years of age. But this translated to a relatively small medication effect, because treatment group parents gave only an average of 4.3 fewer doses of pain medication compared with parents of watchful waiting subjects. There was also no difference in visit behavior, and no differences in days of daycare or work missed, despite the fact that children younger than 2 years of age did seem to have more failures with watchful waiting.
On the basis of these data, the overall conclusion was that children older than 2 years of age with nonsevere AOM seem to be excellent candidates for watchful waiting. Children younger than 2 years of age may not be good candidates for this approach.
A grading system developed by Dr. McCormick and colleagues for classifying AOM was described. Although specific scales can be used for AOM symptom grading, these primarily focus on symptoms reported by the patient and/or parent. By contrast, the scale developed by McCormick and colleagues focuses on the appearance of the tympanic membrane (TM). It goes from grade 0, a normal tympanic membrane, to grade 7, which includes acute inflammation, opacity, a bulging TM, as well as a bulla on the surface of the TM. An Internet link to this scale is provided at the conclusion of this summary.
Use of Antihistamines or Steroids as Adjuncts for Treatment of AOM
Data on the potential use of antihistamines and steroids in treating AOM were presented by Dr. Tasnee Chonmaitree, University of Texas Medical Branch, Galveston, Texas. She began by reviewing some important concepts related to the use of adjunct treatment in AOM, including a reminder that viruses alone can cause AOM. Therefore, AOM is not just a pure bacterial disease. Because of this, previous investigators and clinicians have wondered whether adjunctive therapy would be useful in improving symptoms or other outcomes for patients with AOM.
Several studies that evaluated microbial results of middle ear fluid obtained from patients with AOM were described. In one, research demonstrated that bacteria alone caused 55% of the cases of AOM, with another 19% of cases caused by bacteria and viruses combined. Viruses were the sole agents of AOM in relatively few subjects. The possibility of a mixed infection sheds interesting light on why some patients may fail initial antibiotic treatment: if a patient does not improve after beginning antibiotics for AOM, is it necessarily the case that the patient suffers from a resistant bacterium? In fact, the antibiotic may have been very effective in treating the bacterial component but not the viral coinfection, resulting in limited symptom improvement until the illness has run its course.
Data on use of antihistamines in treating AOM were also described. The basic science suggesting that treatment with antihistamines might be effective comes in part from Dr. Chonmaitree's own work, where she and colleagues demonstrated that the middle ear fluid obtained from patients with bacterial AOM contained increased levels of histamine. They proceeded to test whether antihistamines would help as an adjunctive therapy for patients with AOM.
Two studies have helped clarify the potential use of adjunctive agents. The first, involving 80 patients, focused on the effect of antibiotic treatment on inflammatory mediators in the middle ear fluid. Participants had myringotomy and fluid was obtained at enrollment as well as at Day 5 of treatment. In addition to antibiotic treatment, the patients were randomized to receive antihistamine for 5 days, prenisolone for 5 days, or placebo. At the end of the treatment period, there was no difference in histamine or leukotriene levels in the middle ear fluid of participants, regardless of treatment group.
The second study involved 179 patients, but did not include the double tympanocentesis. All participants had suffered at least 2 previous episodes of AOM and were followed for 6 months after initial enrollment. They received antibiotic treatment and were randomized to receive antihistamine for 5 days, prenisolone for 5 days, or placebo.
This study again showed no difference in the clinical failure rate across treatment groups. A surprising finding was that participants who were treated with antihistamines had 73 days' average duration of middle ear effusion vs an average duration of 20 to 30 days for those treated with other methods. Therefore, the overall conclusion of the investigators was that the use of antihistamines not only did not improve histamine levels in the middle ear or clinical outcomes, but actually served to prolong middle ear effusion in these groups.
Dr. Chonmaitree concluded by providing the results of a Cochrane Database of Systematic Reviews on the question of whether antihistamine or decongestant use was beneficial for adjunct treatment of AOM. The review contained 15 studies involving a total of more than 2600 pediatric patients; all were randomized controlled trials. The primary outcome of the studies was whether there was resolution of the AOM, with secondary outcomes being whether treatment provided symptom control, produced more medication side effects, or prevented complications of AOM.
The collective data showed that treatment with antihistamines or steroids did increase the rate of medication side effects, but neither produced benefit in treating AOM. Therefore, Dr. Chonmaitree concluded that, on the basis of her research findings and those analyzed in the Cochrane review, no benefits of steroid or antihistamine use in treatment of AOM have been identified.
Proper Antibiotic Use in AOM
The final presenter, Dr. Ron Dagan, Soroka University Medical Center, Beer Sheva, Israel, began by emphasizing the following points:
- Antibiotic use for AOM will reduce bacterial load in children under 2 years of age;
- If bacterial eradication does not occur by Day 3 of treatment, this raises the risk of treatment failure or complication;
- 20% to 40% of children with AOM will have a recurrent episode within 2 months; and
- Treating for AOM with antibiotics selects for more resistant pathogens in the nasopharynx.
These points raise interesting questions, including the issue of whether to select a different drug to treat AOM in a patient who has recently been treated for a different episode of AOM. Data to answer that question come in part from 108 cases of recurrent AOM studies by Liebovitz and colleagues. These investigators demonstrated that 54% of recurrent AOM episodes were caused by organisms different from those causing the index episode, with another 28% being due to the same organism. This 28% represents the true relapse rate and is notably high.
Dr. Dagan's own work, published in 2001, demonstrated that although nasopharyngeal organisms may not be present in the middle ear at the time of diagnosis, the nasopharyngeal organisms will colonize the middle ear during the course of treatment. These nasopharyngeal organisms may also be more resistant overall than those involved initially in the AOM episode. This raises the possibility that a treatment failure -- when a patient is not improved after 72 hours of therapy -- may result from adequate treatment of the initial causative organism, followed by opportunistic involvement by a resistant one.
This phenomenon was also demonstrated in almost 500 children studied by Libson and colleagues. In this investigation, 208 children with sterile middle ear fluid had a positive nasopharyngeal culture. Within 3 weeks, 32% of those children experienced an episode of AOM. Among the 286 children with sterile middle ear fluid and sterile nasopharynx, only 22% experienced an episode of AOM within the subsequent 3 weeks. In addition, McCormick's previously reviewed study of watchful waiting demonstrated a similar phenomenon of more resistant bacteria being recovered from the nasopharynx after antibiotic treatment of AOM.
Dr. Dagan's primary reason for reviewing these studies was to emphasize that although antibiotics have proven short-term benefits, they also have notable short-term complications, particularly with regard to selection for resistant organisms. The reviewed findings support the overall approach of providing more selective treatment of older children with AOM.
Although the presenters in the AOM sessions did not necessarily offer a lot of new information, these sessions served as a well-planned, focused review of the current state of the evidence and practice guidelines for diagnosing and treating this very common pediatric problem. There were 2 very important take-home reminders from these sessions. The first is to keep antibiotic treatment of AOM simple: high-dose amoxicillin at 80-90 mg/kg/day for first- line, nonsevere infections; amoxicillin/clavulanic acid for first-line severe infections. After that, ceftriaxone remains the second-line drug of choice. Oral cephalosporins should be reserved for penicillin-allergic patients. The second important take-home message is that watchful waiting seems to be associated with a high success rate when used with older children and can prevent selecting for resistant organisms.
Grading System Acute Otitis Media. Available at: http://www.utmb.edu/pedi_ed/AOM-Otitis/grading.htm. Accessed June 27, 2006.