Acute OM aka "suppurative," "purulent":

I. Epidemiology/Clinical features

  1. 60% of children will have at least one episode of OM in first year of life; 17% will have 3 or more episodes
  2. Tends to peak in March
  3. If recurrent, ass'd with abundant nasopharyngeal lymphoid tissue, cleft palate, & trisomy 21
  4. Tend to be sick, febrile
  5. Physical findings: tympanic membrane bulging (LR 51) and TM redness (LR 8.4) (JAMA 304:2161, 2010-abst)
  6. 50% of kids will still have effusion 1mo post-treatment

II. Microbiology

  1. <1mo:
  1. Same as 1mo-6y, but also E. Coli
  1. Bugs 1mo-6y:
  1. Streptococcus pneumoniae (30-40%; us. abrupt onset)
  2. Nontypable (unencapsulated) H. Flu (20-30%; us. gradual onset)--more likely than S. pneumoniae to resolve spontaneously
  3. Branhamella Catarrhalis (8-18%)--more likely than S. pneumoniae to resolve spontaneously.
  4. Gp. A Strep (4-8%)
  5. Nonbacterial (25-30%)--e.g. RSV; acute viral illness may predispose to bacterial OM

III. Treatment--Antibiotics vs. placebo

A joint clinical practice guideline from AAFP & AAP: recommended antibiotics for all pts < 6mos old and for pts 6mos-2yo if dx is certain or illness is severe (mod-severe ear pain or temp > 39'C), and for pts > 2yo if dx is certain AND illness is severe.  Tx of choice is amoxicillin 80-90mg/kg/d unless illness is severe in which case should be amoxicillin-clavulanate (American Academy of Pediatrics and American Academy of Family Physicians. Subcommittee on Management of Acute Otitis Media. Clinical Practice Guideline. Diagnosis and Management of Acute Otitis Media--Peds 113:1451, 2004--JW)

The data on the effect of antibiotics in acute otitis media (AOM) in children come from a number of placebo-controlled randomized trials. The Cochrane collaboration has performed a meta-analysis of seven of these, published between 1968 and 2000. As of 2001, this is the most up-to-date meta-analysis that has been published on the question of difference between antibiotics and placebo for treatment of AOM in children; it was last updated in 4/2000 (1). The review details the methodology for finding and analyzing the studies which in typical Cochrane fashion, were quite thorough.

The Cochrane meta-analysis showed no difference with abx in pain at 24h (62% overall had no pain) but less likelihood of pain at followup (which occurred between two and seven days after initial evaluation--20.7% with placebo and 15.1% with antibiotics). The absolute risk reduction, thus, was 5.6%, giving a "number needed to treat" of about 17. Likelihood of developing contralateral AOM (in patients with unilateral AOM at intake) was statistically significantly less likely in antibiotic-treated patients (10.6% vs. 16.6%) with an absolute risk reduction of 6.0%. "Late recurrences" (not defined in the Cochrane review) almost identical in antibiotic and placebo groups.

As far as complications of AOM, The meta-analysis examined impact of antibiotics on subsequent hearing problems after AOM and found no statistically significant differences for tympanometry results at 1mo or 3mos, though they note that the data were sparse (only some studies reported this measure) and technique by which this test was performed may have varied. Tympanic membrane perforation was reported in two studies and again the meta-analysis showed no statistically significant difference between antibiotic and placebo-treated patients.

Serious complications were extremely uncommon in either the placebo or active-treatment groups in all the trials analyzed. These studies would not have sufficient power to rule out a benefit of AOM in a setting where complications of AOM are common. They do mention a "semi-randomized study" from 1954 (Acta Otolaryngol 1954;113(suppl):1-79) where the incidence of mastoiditis in untreated patients was 17% (and 0% in the treated patients)--Of interest, this study was excluded from the meta-analysis because of methodologic features that may not invalidate the dramatic findings. The conclude "In populations or sub-populations where mastoiditis is judged still a problem, such as in developing countries, antibiotic treatment would be strongly advised."

As far as complications of treatment, the Cochrane meta-analysis found that the likelihood of any vomiting, diarrhea, or rash during followup was significantly higher in antibiotic-treated children (16.5% vs. 10.8%) a 5.7% absolute risk increase in antibiotic-treated children.

The US Agency for Healthcare Research and Quality (AHRQ) conducted a review of initial management of AOM. This is included in the TRIP index. To date, only the summary has been published online (the full report is available in print but is on backorder so I was not able to review it). They conducted what sounds like a fairly thorough literature review, through 1999, and performed a meta-analysis of the effects of antibiotic treatment vs. placebo treatment in AOM and also the natural history of AOM not treated with antibiotics. The summary does not state how many trials were analyzed or give references. They found that the likelihood of "clinical failure"(the summary doesn't state how that was defined) was 22% at 4-7d after diagnosis, and found an absolute risk reduction for "clinical failure" of 12.3% at 2-7d with amoxicillin vs. placebo. They did NOT find a difference in pain at 24h, TM perforation, vomiting, diarrhea, rash, tympanometry findings at 1 month, or recurrent AOM. They note some limitations of the clinical trials analyzed, including variation in diagnostic criteria for AOM and poor methodologic quality of some studies.

Rosenfeld at al. published a meta-analysis of studies of antibiotics in AOM (3). This paper iincluded studies found through a search of Medline and Current Contents through 1992. They looked at 30 trials though only 4 included a placebo arm; the others compared different antibiotic regimens. They found the the likelihood of "successful primary control" (defined as resolution of all presenting signs or symptoms except for middle ear effusion at 7-14d after initiation of therapy) was 81% with placebo or no drug and 94% with antibiotics. Suppurative complications did not occur in any of the untreated children in the trials studied. All 4 of the studies comparing antibiotics vs. placebo were included inthe Cochrane meta-analysis. The main finding was that antibiotics vs. no antibiotics were associated with an absolute risk reduction of 16% of persistent signs or symptoms of AOM at 7-14 days after initiation of treatment. No mention is made of any difference in complication rates

Randomized trials published since these meta-analyses:

  • 240 children 6mo-2yo with OM randomized to amox 40mg/kg/d divided TID x 10d vs. placebo; both groups received oxymetazoline nasal TID. The amox group had sig. less incidence of persistent sx at 4d (59% vs. 72%) and sig. less use of analgesics, but no diff. in incidence of persistence of sx at day 11, recurrent OM, recurrent use of abx, middle ear infusion at 6wk f/u, or specialty referral over 6wk f/u. HOWEVER, 2 pts in in the placebo group had apparent complications: one admitted w/meningitis, another admitted because of "clinical deterioration." (4)

  • 283 children 6mos-12yo presenting to an ED with acute otitis media (but didn't appear toxic and had no systemic sx) randomized to standard antibiotic prescription vs. a "wait and see" prescription to use if sx not improving in 48h; all received ibuprofen and otic analgesic drops.  38% of the wait-and-see prescriptions were filled.  No sig. diff. between the groups in frequency of subsequent fever, otalgia, or unscheduled medical visits, but wait-and-see pts had sig. longer average duration of otalgia (2.4d vs. 2.0d) (JAMA 296:1235, 2006--abst)

In another meta-analysis of six trials involving 1,643 pts 6mo-12yo with acute otitis media ranodmized to early antibiotic therapy vs. placebo (or delayed treatment); incidence of "extended AOM" (pain, fever, or both 3-7d after starting treatment) was sig. lower in pts who received early treatment (RR 0.83); treatment effect was greater in pts < 2yo (RR 0.64) and pts with AOM & otorrhea (RR 0.52). (Lancet 368:1429, 2006--JW)

In another meta-analysis of 135 trials, short-term clinical success was higher for immediate tx (upon dx) of (ampicillin or amoxicillin) compared w/placebo (73% vs. 60%) but rate if rash or diarrhea was 3-5% higher (5).

1. Glasziou PP, Del Mar CB, Sanders, SL. Antibiotics for acute otitis media in children (Cochrane Review). In: The Cochrane Library, Issue 2, 2001. Oxford: Update Software. (NOTE--A meta-analysis with two of the same authors, that appears to be an initial version of this Cochrane review, was published as Del Mar, Christopher. Glasziou, Paul. Hayem, Mauricio. Are antibiotics indicated as initial treatment for children with acute otitis media? A meta-analysis. BMJ. 314(7093):1526-1529. Unlike the BMJ article, which was published in 1997, the Cochrane review incudes a randomized trial published in 2000: Damoiseaux RA, van Balen FA, Hoes AW, Verhiej 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 320:250-4).

2. Management of Acute Otitis Media. Summary, Evidence Report/Technology Assessment: Number 15, June 2000. Agency for Healthcare Quality and Research, Rockville, MD. http://www.ahrq.gov/clinic/otitisum.htm

3. Rosenfeld RM, Vertrees JE, Carr J, Cipolle RJ, Uden DL, Giebink GS, Canafax DM. Clinical efficacy of antimicrobial drugs for acute otitis media: meta-analysis of 5400 children from thirty-three randomized trials. J Ped 1994;124:355-67.

4. BMJ 320:350, 2000--AFP, JW

5. Diagnosis, microbial epidemiology, and antibiotic treatment of acute otitis media in children: A systematic review.  JAMA 2010;304:2161-2169

IV. Treatment--Other issues

  1. Duration of tx
    1. In a meta-analysis of 14 studies in 2,115 kids 4wk-18yo with acute OM randomized to tx with abx for either 5d or 8-10d, incidence of tx failure (lack of clinical resolution, or relapse within 1mo) was not sig. diff in the two groups, though nonsig. increased risk (OR 1.52) of relapse or reinfection at 20d (JAMA 279:1748, 1998--AFP)
  1. Treatment in children with tympanostomy tubes
    1. In a study in 80 children 6mo-12yo with tympanostomy tubes and acute otorrhea randomized to oral amoxicillin/clavulanate x 10d vs. topical (otic) ciprofloxacin/dexamethasone x 7d, the topical group had faster resolution of otorrhea (median time 4d vs. 7d) and sig. higher clinical cure rate at 18d (85% vs. 59%) (Peds. 118:e561, 2006--JW)
  2. Specific drugs:
  1. Amoxicillin
  1. Drug of choice as of 1999 per CDC; dose should be 80-90mg/kg/d divided TID, x 5-7d (rather than the traditional 20-40mg/kg/d, because of emergence of drug-resistance pneumococcus--such resistance, when it's to beta-lactams, generally can be overcome w/higher doses)--pts at "very low risk" for infection with drug-resistance S. pneumoniae (no recent abx, age > 2y, and no daycare attendance) can still be tx'd with amox 40-45mg/kg/d. (Ped. Inf. Dis. J. 18:1, 1999)
  1. If initial tx fails, CDC (op. cit.) recommends:
  1. Waiting at least 3d before declaring a tx failure
  2. Define tx failure as presence of signs/sx of persistent acute otitis media (ear pain, fever, redness or bulging of TM, and otorrhea) rather than persistent middle-ear effusion which occurs in 70% of pts anyway)
  3. Consider tympanocentesis for culture & sensitivity assays
  4. Empiric change of abx to one of the following (at day 3, iii and iv only if has had abx in the last month; at day 10-28, i, ii, and iii all OK if has had abx in the last month)
    1. Amoxicillin-Clavulanic acid (high dose--80-90mg/kg/d Amox, 6.4mg/kg/d Clavulanic acid; note this requires newer formulation of amox/clavulanate OR combination w/plain amox, because clavulanate dose must be < 10mg/kg/d)
    2. Cefuroxime axetil
    3. Ceftriaxone (50mg/kg IM QD x 3d--a single IM injection has a high failure rate when used after initial tx failure
    4. Clindamycin (only if HAS had abx in the last month--Clinda is not active against M. catarrhalis or H. flu)
  5. Note that resistance of S. pneumoniae against beta-lactams is often associated also with resistance to macrolides and sulfonamides and, unlike the case w/beta-lactams, this resistance is not amenable to simply increasing the dose of those meds; therefore, macrolides and sulfonamides are NOT recommended by CDC for use in cases of initial treatment failure.
  1. Because of resistance issues, initial tx with Penicillin, Macrolides, or Trimethoprim-Sulfamethoxazole are not recommended by CDC (op. cit.)
  2. Ceftriaxone 50mg/kg IM x 1 as effective as amox x 10d (Peds 91:23, 1993-JW) or TMP-SMZ (Peds 99:23, 1997-JW) for initial tx with no increase in side effects
  3. If penicillin-allergic, consider azithromycin (10mg/kg on day 1 then 5mg/kg QD on days 2-5), clarithromycin (15mg/kg/d divided BID), or clindamycin (10-30mg/kg/d divided TID)
  4. Treatment when tympanostomy tubes are present
    1. 599 children (mean age 2.5yo) with tympanostomy tubes and acute OM randomized to topical ciprofloxacin/dexamethasone (4 gtt BID x 7d) vs. ofloxacin (5 gtt BID x 10d); absence of or reduction in otorrhea on days 3, 11, and 18 was sig. more likely in ciprofloxacin/dexamethasone group (Peds. 113:e40, 2004--JW)
  5. Adjunctive treatments
    1. Neither prednisolone (2mg/kg/d) or chlorpheniramine (0.35mg/kg/d) were ass'd with any difference in clinical response compared with placebo in a randomized trial of 180 children (3mo-6yo) with acute OM and at least 2 prior AOM episodes (J. Peds. 143:377, 2003--JW)
    2. Analgesics-Including acetaminophen, ibuprofen, and Auralgan (antipyrene/benzocaine 2-4 gtt to ear canal Q2h PRN)

V. Prevention

  1. Xylitol--a natural sugar which inhibits growth of Streptococcus mutans
    1. 857 healthy children randomized to Xylitol gum 8.4g/d vs. control gum (very young kids got xylitol syrup vs. control syrup) 5x/d x 3mos; xylitol pts had sig. fewer episodes of acute otitis media (RR 0.6 for gum, 0.7 for syrup). No adverse f/x noted (Peds. 102:879, 1998--JW)
  2. Limiting pacifier use after 6mos of age was ass'd with sig. lower incidence of acute OM (RR 0.71) in a randomized trial of 484 pts < 18mo (Peds 106:483, 2000--JW)
  3. Influenza Immunization has been shown to reduce incidence--See link for details
  4. Pneumococcal Immunization has been shown to slightly reduce incidence--See link for details
  5. Prevention in pts with Recurrent OM
    1. Often defined as 3 episodes in 6 mos or 4 episodes in 12mos
    2. Natural history of recurrence
      1. 222 Finnish kids with recurrent OM (by that definition) followed without tx (no suppressive abx or myringotomy tubes) for 6 mos; 38% had no more OM; 28% had it x 1; 18% had it x 2; 12% had it x 3 or more; recurrences most common in kids >16mo; may not require suppressive Rx in those kids! (J. Fam. Prac. 43:258, 1996-JW)
    3. Myringotomy with tubes
    4. Chronic antibiotics
    5. Adenoidectomy or adeno-tonsillectomy in kids with recurrent OM but no tonsil-related indication for same
      1. Had little clinical benefit in terms of reduced OM recurrence c/w a control group not given surgery in a randomized trial of 304 pts (JAMA 282:945, 1999--JW)
      2. 180 children 10-24mo with h/o acute OM x 3 or more in prior 6mos randomized to adenoidectomy, sulfafurazole 50mg/kg QD x 6mos, or placebo.  At both 6mos and 2y, no sig. diff. in failure rate (defined as two AOM episodes in 2 months or three in 6 months, or middle ear effusion for 2 months ) (BMJ 328:487, 2004--JW)
      3. 217 children 12-48mos with AOM > 3x in previous 6mos, or chronic otitis media with effusion, all receiving myringotomy & tubes, randomized to adenoidectomy vs. no adenoidectomy.  Over 1y f/u, incidence of AOM was not sig. diff. between the two groups (Peds. 116:185, 2005--JW)
    6. Alpha-hemolytic Streptococci nasal spray--May protect against recurrence in pts w/multiple courses of abx for prior OM's
      1. Alpha-hemolytic streptococci interfere w/growth of pathogenic bacteria in vitro
      2. They are present in lower #'s in kids with h/o otitis media
      3. 108 children 6mo-6yo with h/o recurrent AOM (2x in 6mo or 5x in 1y) randomized to nasal alpha-hemolytic streptococcal nasal spray vs. placebo BID x 10d; given after 10d of abx for AOM.  The nasal spray tx was repeated 2mo later x 10d; recurrence over 3mo f/u was sig. lower in active tx goup (40% vs. 51%) (BMJ 322:210, 2001--JW)
    7. Osteopathic manipulation for prevention of recurrent AOM
      1. 57 children 6mo-6yo with recurrent acute otitis media (3 episodes in 6mos or 4 within 12mos) randomized to routine care or routine care + osteopathic manipulation.  Over 6mo f/u, manipulation recipients had sig. fewer AOM episodes per month than routine care group (0.19 vs. 0.27) and fewer myringotomy tube procedures (1 vs. 8 kids) but no diff. in audiologic findings.  No "sham" treatment in the control group and kids in the menopulation group were older on average (26 mos vs. 20mos), however (Arch. Pediat. Adol. Med. 157:861, 2003--JW)

V. OM with effusion aka "secretory," "nonsuppurative," "serous," "mucoid":

  1. Not too sick; often afebrile
  2. May be associated with hearing loss
  3. Theorized etiolgies include viruses, bacterial, allergies, and GERD
  4. Treatment options
    1. Antibiotics: optional but if use, cover for branhamella
    2. Decongestants don't help
    3. Corticosteroids 
      1. In a systematic review of 10 randomized trials seen to be ass'd with improvement in short-term (2wk) but not long-term (2mos) outcomes (Arch. Ped. Adol. Med. 155:641, 2001--JW)
    4. Tympanostomy with tubes
      1. Often used if doesn't resolve with watchful waiting but may not improve meaningful outcomes
      2. A randomized trial of immediate vs. delayed (9mos later) tympanostomy tubes in 429 kids 2mos-3yo with effusion x > 90d bilaterally of > 135d unilaterally showed that tubes were ass'd with sig. fewer days with effusion over 2y BUT no diff. in speech, language, cogniton, or development tests at 3y of age (NEJM 344:1179, 2001--JW)
      3. Simultaneous adenoidectomy or adenotonsillectomy may reduce risk of need for tube reinsertion in kids undergoing tympanostomy tube insertion (NEJM 344:1188, 2001--JW)
      4. In a meta-analysis of seven randomized studies, tube placement vs. no treatment for OME was associated with sig. better hearing at 6mos but no difference at 12-18mos after placement and also no diff.in language development (Arch. Dis. Child 90:480, 2005--AFP)
      5. In a study in 395 children < 3yo with bilateral middle ear effusions for > 90d (or unilateral effusion for > 135d) randomized to tympanostomy tubes vs. deferral of tube placement for 9mos and placement then if the effusion persisted. At age 6 there were no sig. diffs. between roups in assessments of cognitive, language, and psychosocial development (NEJM 353:576, 2005--JW)
        1. In a follow-up study with reassessment at ages 9-11, there were no sig. diffs between the groups in speech/language development intelligence, academic achievement, or behavior (NEJM 356:248, 2007--JW)

VI. Chronic OM--Tends to be suppurative & ass'd with mastoiditis or immunodeficiency