Systemic corticosteroids for improving symptoms in children with acute middle ear infection

Review question

We reviewed the evidence on the effects of corticosteroids given by mouth or injection for acute middle ear infection (acute otitis media (AOM)) in children, particularly in improving symptoms such as ear pain, fever, irritability, lack of sleep, and lack of appetite. We also looked at the side effects of corticosteroids.


Acute otitis media is common in children and causes ear pain and non-specific symptoms such as fever, irritability, and deafness. It is often treated with antibiotics, although ear pain generally resolves within two days, and antibiotics help symptoms only slightly. Other treatments (such as over-the-counter antihistamines and decongestants) do not help very much.

Corticosteroids are often prescribed to reduce inflammation in children for other illnesses, and so may also help symptoms of AOM, which is an inflammatory process. We investigated whether using corticosteroids was better or worse than nothing in improving AOM-related symptoms.

Search date

Our evidence is current to 20 February 2018.

Study characteristics

We included two studies involving 252 children with AOM, aged from three months to six years, receiving hospital ambulatory care. Children were treated with an antibiotic injection and either oral corticosteroid or a placebo (treatment with no effect). In one study, fluid from the middle ear was collected by inserting a needle through the eardrum to measure the level of inflammation.

Study funding sources

The National Institutes of Health (NIH) and the National Center for Research Resources, NIH, US Public Health Service funded both studies. Pharmaceutical companies provided the drug but did not contribute any other scientific or financial support.

Key results and quality of evidence

Corticosteroids did not make a significant difference in improving the symptoms and inflammation of the eardrum(s) at Day 5 and Day 14, but we are unsure of this effect due to the small numbers of children in the studies. There were no significant differences between the corticosteroid and placebo groups in terms of resolving fluid in children's middle ears (at 1, 2, and 3 months) and experiencing new episodes of AOM (at 1, 2, 3 months, and 4 and 6 months). Neither study reported a reduction in the duration of overall or specific symptoms, rupture of eardrum(s), the occurrence of middle ear inflammation in the other ear following the current ear infection, or serious complications. Only one study reported the overall side effects identified during the trial (e.g. drowsiness, dry mouth, diaper rash, nervousness).

We could not draw any conclusions regarding the effects of corticosteroids for AOM in children.

The quality of evidence included in this review was low to very low due to few children included in two small studies. We are uncertain about whether or not corticosteroids are useful in relieving pain from AOM.

Authors' conclusions: 

The evidence for the effect of systemic corticosteroids on AOM is of low to very low quality, meaning the effect of systemic corticosteroids on important clinical outcomes in AOM remains uncertain. Large, high-quality studies are required to resolve the question.

Read the full abstract...

Acute otitis media (AOM) is a common acute infection in children. Pain is its most prominent and distressing symptom. Antibiotics are commonly prescribed for AOM, although they have only a modest effect in reducing pain at two to three days. There is insufficient evidence for benefits of other treatment options, including systemic corticosteroids. However, systemic corticosteroids are potent anti-inflammatory drugs, and so theoretically could be effective, either alone or as an addition to antibiotics.


To assess the effects of systemic corticosteroids (oral or parenteral), with or without antibiotics, for AOM in children.

Search strategy: 

We searched the Cochrane Central Register of Controlled Trials (CENTRAL) which contains the Cochrane ARI Group's Specialised Register, MEDLINE (Ovid), Embase (Elsevier), CINAHL (EBSCO), Web of Science (Thomson Reuters), and LILACS (BIREME) for published studies, and and the World Health Organization (WHO) International Clinical Trials Registry Platform (ICTRP) for completed and ongoing studies, to 20 February 2018. We checked the reference lists of all primary studies and review articles for additional references and contacted experts in the field to identify additional unpublished materials.

Selection criteria: 

We included randomised controlled trials of children with AOM that compared any systemic corticosteroid (oral or parenteral) with placebo, either with antibiotics (corticosteroid plus antibiotic versus placebo plus antibiotic) or without antibiotics (corticosteroid versus placebo).

Data collection and analysis: 

Three review authors (EDS, RR, YP) independently screened the titles and abstracts and retrieved the full texts of potentially relevant studies. We independently extracted study characteristics and outcome data from the included studies, and assessed the risk of bias for each study using the criteria outlined in the Cochrane Handbook for Systematic Reviews of Interventions. We assessed study quality using the GRADE method.

Main results: 

We included two studies involving 252 children with AOM aged from three months to six years receiving hospital ambulatory care who were treated with intramuscular ceftriaxone, and who were then randomised to the corticosteroid group (corticosteroid and corticosteroid plus antihistamine) or the placebo group (antihistamine and double placebo). In one study, children also had a needle aspiration of middle ear fluid. Both studies were at unclear risk of bias for allocation concealment, and unclear to high risk of bias for selective reporting.

One study (N = 179) included pain as an outcome, but we were unable to derive the proportion of children with persistent pain at Day 5 and Day 14. Reduction of overall or specific symptoms was presented as improvement in clinical symptoms and resolution of inflamed tympanic membranes without the need for additional antibiotic treatment: at Day 5 (94% of children in the treatment group (N = 89) versus 89% in the placebo group (N = 90); risk ratio (RR) 1.06, 95% confidence interval (CI) 0.97 to 1.16) and Day 14 (91% versus 87%; RR 1.05, 95% CI 0.95 to 1.17). Low-quality evidence meant that we are uncertain of the effectiveness of corticosteroids for this outcome.

The second study (N = 73) reported a reduction of overall or specific symptoms without additional antibiotic treatment during the first two weeks as a favourable outcome. Children in the treatment group had more favourable outcomes (adjusted odds ratio 65.9, 95% CI 1.28 to 1000; P = 0.037), although the numbers were small. We were unable to pool the results with the other study because it did not report the proportion of children with this outcome by treatment group. Only one study reported adverse effects of corticosteroids (e.g. drowsiness, nappy rash), but did not quantify incidence, so we were unable to draw conclusions about adverse effects. Neither study reported a reduction in overall or specific symptom duration.