By the age of 3, roughly five in six US children have had at least one ear infection. Most of those visits ended with a prescription — and most of those prescriptions probably did nothing the child's immune system would not have done on its own. Here is what the AAP, two landmark NEJM trials, and the 2023 Cochrane review actually say a pediatrician — and a parent — should do.
Acute otitis media is a clinical diagnosis: a middle-ear infection with rapid onset, signs of middle-ear inflammation, and middle-ear effusion. The AAP 2013 guideline tightened the diagnostic criteria to reduce the over-treatment that had become the norm in US pediatric practice in the prior two decades1:
The single most important distinction in modern pediatric ENT is between AOM and otitis media with effusion (OME). OME is the persistence of middle-ear fluid without the acute infection signs. OME is common after AOM resolution, often persists for weeks, and — critically — does not warrant antibiotics. The same red-pink tympanic membrane that a busy clinic might call "ear infection" is, in many children, simply lingering effusion from a viral upper-respiratory illness. Treating it as AOM gives the child a course of antibiotics they do not need2.
The clinical takeaway: in any child whose ear exam is borderline, ask for a clinician who is comfortable with pneumatic otoscopy (or tympanometry). The distinction between AOM and OME is the highest-leverage diagnostic decision in the entire pediatric primary-care antibiotic stewardship landscape.
In the late 1990s, US pediatricians wrote roughly 30 million AOM prescriptions per year — more than one prescription per US child per year6. The CDC and AAP both concluded by the mid-2000s that the practice was driving avoidable antibiotic resistance, avoidable side effects (the most common being antibiotic-associated diarrhea, which affects roughly 10–15% of treated children), and avoidable cost — without proportional clinical benefit for the majority.
The 2013 AAP guideline formalized the observation option for milder cases, paired with a safety net: a "wait-and-see prescription" the family fills only if symptoms have not improved by 48–72 hours, or have worsened at any point. Multiple randomized and observational studies have shown that this approach reduces antibiotic use by roughly 30–50% without measurable harm to children16.
The AAP 2013 guideline operationalized the decision tree as follows1. Pediatricians can — and should — share the same framework with families during the visit.
Severity is operationalized in the guideline as moderate-to-severe otalgia, otalgia of ≥48 hours, or temperature ≥39.0°C (102.2°F) in the past 48 hours1. The fever threshold is important: a "low-grade" fever (under 39°C) plus mild ear pain does not, by itself, push a child into the severe category.
Two parallel placebo-controlled trials published in the same January 2011 issue of NEJM remain the highest-quality evidence on antibiotic effect in children with stringently-diagnosed AOM. Hoberman et al., from Pittsburgh, enrolled 291 children aged 6–23 months with strict AOM criteria and randomized them to amoxicillin-clavulanate versus placebo for 10 days. Symptomatic resolution was modestly faster with the antibiotic (35% vs 28% by day 2; 80% vs 74% by day 7), and treatment failure (symptoms or exam not improved by day 4–5) was substantially less common in the antibiotic group (16% vs 51%)3.
Tähtinen et al., from Finland, enrolled 319 children aged 6–35 months and randomized them to amoxicillin-clavulanate versus placebo for 7 days. Treatment failure occurred in 19% of the antibiotic group versus 45% of the placebo group4.
Both trials reported significantly more diarrhea and diaper-area rash in the antibiotic arm. The NNT (number needed to treat) for one additional cured child by day 7 was approximately 4 in this strictly-diagnosed AOM cohort — meaningful, and the reason antibiotics remain on the table for the indicated bands above. But the corresponding NNH (number needed to harm) for one additional case of antibiotic-associated diarrhea or rash was roughly 5, which is why observation in the lower-severity bands is not just a stewardship preference but a clinically reasonable equipoise.
The 2023 Cochrane review on antibiotics for AOM in children synthesized 13 randomized trials (n=3,401) and concluded that antibiotics reduce ear pain at 2–3 days from 23% to 17%, with an NNT of 13 for symptom benefit at 2–3 days and no measurable reduction in rare complications such as mastoiditis at the population level5. The Cochrane review and the AAP guideline are mutually consistent: antibiotics work, the magnitude is modest, and selective use is the evidence-based path.
The "safety-net" or "wait-and-see" prescription is the operational backbone of AAP-aligned observation. A pediatrician hands the family a real, signed prescription, with explicit verbal and written instructions:
Multiple studies of safety-net prescribing show that roughly 60–70% of these prescriptions go unfilled, with no detectable increase in complications1. The approach respects parental judgment, sidesteps the "I drove an hour and got nothing" frustration, and aligns with the principle that the goal is the right amount of antibiotic — not the maximum amount.
When the decision is for treatment, the AAP's first-line recommendation remains high-dose amoxicillin (80–90 mg/kg/day divided in two doses)1, on the grounds that:
Amoxicillin-clavulanate (Augmentin) is appropriate as second-line in children who have received amoxicillin in the prior 30 days, who have concurrent purulent conjunctivitis (suggesting H. influenzae), or who have failed amoxicillin. For penicillin-allergic children with a non-Type-I (non-anaphylactic) history, second-generation cephalosporins (cefdinir, cefuroxime) are reasonable. For Type-I allergy, azithromycin or clindamycin are the standard substitutes, with the caveat that macrolide resistance in pneumococcus has risen in recent surveillance data6.
Duration: 10 days remains standard for children <2 years and for severe disease. For children ≥2 years with non-severe disease, 5–7 day courses have shown non-inferiority in multiple trials and are an acceptable alternative1.
The clinically relevant fact for parents is that antibiotics do not provide meaningful pain relief in the first 24 hours. Pain control is what matters for the child during that window, and it is independent of the antibiotic decision.
For the broader fever-management framework parents need around the AOM visit, see Wermom's companion explainer on AAP-aligned home care during febrile illness and the parent guide on distinguishing viral URIs from bacterial complications.
Recurrent AOM — typically defined as ≥3 episodes in 6 months or ≥4 episodes in 12 months with at least one in the past 6 months — is the most common indication for tympanostomy tube placement in children. The AAO-HNS and AAP joint guidelines support tube referral after meeting that definition, particularly in children with co-occurring developmental concerns, persistent OME ≥3 months, or hearing loss documented on audiometry2.
Two recent shifts deserve mention. First, tympanostomy tubes do not eliminate AOM — they redirect drainage and shorten episodes, but children with tubes still get ear infections (often presenting as painless drainage rather than the classic pain syndrome). Second, the recent TOPIC trial and subsequent reviews have raised the threshold for tube placement in non-high-risk children, on the grounds that the long-term outcome differences with conservative management are smaller than once believed. The conversation between pediatrician, ENT, and family should integrate the child's hearing function, language development trajectory, and family-life burden — not just the episode count.
Three prevention measures have the strongest evidence behind them.
Annual influenza vaccination reduces AOM incidence by approximately 30–55% during the influenza season in children aged 6 months and older1. The biological rationale is straightforward: influenza is one of the major viral predecessors of bacterial AOM superinfection.
Pneumococcal conjugate vaccination has reduced both vaccine-serotype AOM and the rate of severe AOM requiring tube placement. The trajectory has been one of the clearest public-health wins in pediatric ENT over the past 20 years16.
Breastfeeding for ≥6 months is associated with roughly a 50% reduction in AOM in the first year of life in cohort data, with the effect attenuating but persisting through age 22. The biological mechanism is multifactorial — secretory IgA, breastfeeding posture reducing eustachian-tube reflux, and microbiome differences all contribute.
Beyond these, household smoke exposure reduction, pacifier weaning by 6–12 months (pacifier use beyond infancy is associated with somewhat higher AOM rates), and minimizing prolonged supine bottle feeding all have modest supporting evidence.
Three practical questions to ask, in the order that matters:
And one practical note about the family-life side: in Wermom's parent-question corpus across the 2025 calendar year, more than 1 in 3 questions logged under "ear infection" were actually questions about OME or post-AOM residual fluid — not active AOM. Parents and pediatricians who jointly default to the distinction between active infection and residual effusion reduce unnecessary visits, unnecessary antibiotics, and the anxiety of the "another ear infection" loop.
The Wermom App's symptom and vaccine tracker captures the data parents and pediatricians actually need — without manual paper trails.
Explore the Wermom App