Save Yourself Earache – Antibiotics and Otitis Media in the ED

Save yourselfearache

When I was a medical student a friend of mine completed a special study module in ENT which involved a case-based, evidence-based essay on a clinical topic of his choice. He chose acute otitis media. “How exciting!” you might (not) think.

Fast forward to today and in PED I continue to be amazed by the willingness of our junior doctors to prescribe antibiotics for acute otitis media. I’m not sure why the learning points from someone else’s essay have stayed with me for more than ten years (yes, I am that old) but I have to remind myself that not everyone read his essay. But apparently they haven’t read the NICE Clinical Knowledge Summary either.

Otitis Media in PED

Although we do see acute otitis media in the adult ED it’s far more common in paediatrics. And it’s pretty common. Why? Well, if you’ve ever had an acute otitis media you’ll know – it really hurts. Typically the patients we see present in the evening or overnight, sometimes with fever and sometimes without, complaining of pain in an ear if they are old enough to localise it and communicate it or crying/pulling their ear if they aren’t.

Sometimes we see red tympanic membranes as part of the assessment of a child with coryzal symptoms or a fever. Here’s where I think we get foxed – a red tympanic membrane does not acute otitis media make (nor antibiotics necessitate). It’s difficult to even diagnose otitis media: ear rubbing and ear pain are suggestive but their absence does not help in excluding the diagnosis (positive likelihood ratios but not strongly negative likelihood ratios). On otoscopy, a bulging or red (haemorrhagic or strongly/moderately red) tympanic membrane is the most useful sign with strongly positive and strongly negative likelihood ratios but this is something we commonly see in conjunction with a fever, cough and snotty nose.

The CKS: How should we manage these patients?

The Clinical Knowledge summary produced by NICE is quite helpful. It reminds us that for most people antibiotics are not required and that, without antibiotics, the average clinical course of the illness is just four days.

Pain and fever in children should be treated with analgesia/antipyretics.

There are a couple of outlined “antibiotic strategies”;

No antibiotics

Most people with acute otitis media do not require antibiotics. It is often viral and self limiting in origin. Yes, it’s painful and there is moderate evidence that antibiotic administration reduces pain – but it seems crazy to give antibiotics to treat pain. If we give antibiotics we risk causing harm (increased antibiotic resistance in the population, diarrhoea for the patient) – in fact, this recently published summary at suggests while 1 in 20 patients in the <12 years group had reduced pain, 1 in 9 experienced diarrhoea. The treatment harms seem to outweigh a small benefit.

If you are not giving antibiotics, counsel the patient/parent that they are not needed. They will not change the clinical illness, may cause additional symptoms (vomiting, diarrhoea, rashes…) and will contribute to antibiotic resistance. Reiterate that it is likely things will have settled within four days and that if this is not the case it might be pertinent to follow-up with a GP. As ever, safety netting is important. Routine follow up is not necessary.

Delayed antibiotics

This is a strategy adopted by some GPs. Advice is given as above but with a prescription to be “cashed in” if things aren’t better at four days (perhaps post-dated). I’m not sure this works well in the Emergency Department though as a return to the hospital pharmacy doesn’t add any additional convenience – although that might work in ensuring the prescription isn’t started before you want it to be!

Consider antibiotics

Consider antibiotics where there is visible perforation (or associated pus/discharge in the canal).  In children under the age of two with bilateral otitis media antibiotics might also be sensible. There are other caveats below.

Important Caveats

Sick kids

Of course, children who are systemically unwell (that is, beyond an associated fever) should be considered for antibiotic therapy. Recognising sick kids is tricky in itself, not least because children with a fever tend to also be tachycardic, tachypnoeic and miserable. I think the key here is that if you would be discharging them with a diagnosis of a “viral upper respiratory tract infection” had you not spotted that red ear, they probably don’t need antibiotics – but speak to a senior for advice if you are unsure.

Children <3 months of age

Otitis media is unusual in this group. Evidence is limited in the <6 month age group and as with all babies under 3 months be very careful about dismissing “simple” infections and have a low threshold for admitting and treating these patients.

Patients with complications

In the presence of complications of otitis media – meningitis, mastoiditis or facial nerve palsy – specialist ENT review is indicated.

Children with cochlear implants, craniofacial abnormalities and other underlying conditions

Children with cochlear implants are a special case because in these patients the risk of bacterial meningitis is higher, particularly in the first two months post-op. This great post form Pediatric EM Morsels covers otitis media in the child with cochlear implants. These children generally require intravenous antibiotics (especially in the first two months post-operatively) and our ENT colleagues usually admit them.

Patients with craniofacial abnormalities such as cleft palate may be at additional risk but there is limited evidence in this area.

In summary…

Think! Antibiotics usually aren’t needed; your patient will most likely get better within four days and antibiotics are more likely to cause them diarrhoea than improve their pain. Don’t do it!


Cite this article as: Natalie May, "Save Yourself Earache – Antibiotics and Otitis Media in the ED," in St.Emlyn's, February 18, 2015,

4 thoughts on “Save Yourself Earache – Antibiotics and Otitis Media in the ED”

  1. I always wonder if the high NNT quoted in studies for AOM is simply the fact that those studies are including a population of non-AOM.

    How many clinicians either don’t have access to or utilise pneumatic otoscopy?

    What would the true NNT be lower if treatment was only begun if the entire triad of redness, bulging and decreased TM movement was required for diagnosis.

    Any treatment has a high NNT if the diagnostic criteria of actual disease is not well defined.

  2. Pingback: LITFL Review #170 - LITFL

  3. There are 2 groups that do need antibiotics. Those kids from poor crowded dwellings and Aborigines. In 1971 in Bourke, 24% of Aboriginal children < 15years had at least one perforated ear drum.16% had one or two suppurative OM. Equivalent figures for white kids were
    perforations 5% with hardly any SOM.
    Training up 3 matriarchs with an auriscope and a pocket full of ampicillin reduced the perforation rate to 7% buy 1974.

  4. Pingback: LITFL Review 170 • LITFL Medical Blog • FOAMed Review

Thanks so much for following. Viva la #FOAMed

Scroll to Top