 |
 |
18th August 2009
An observational follow-up study to an RCT found that children prescribed amoxicillin for acute otitis media (AOM) were about 50% more likely to have recurrence of AOM over the following three years than those given placebo. There was no significant effect on rates of related referrals or ENT surgery.
Level of evidence:
Level 2 (limited-quality patient-oriented evidence) according to the SORT criteria.
Action
Health professionals should follow NICE guidance on management of upper respiratory tract infections. They should try to agree a no-antibiotic or delayed-antibiotic strategy for most children with AOM, and generally prescribe antibiotics only to children with AOM who are systemically very unwell or at high risk of serious complications because of pre-existing co-morbidity, or who appear unwell with symptoms and signs suggestive of mastoiditis. Depending on clinical assessment of severity, an immediate antibiotic antibiotic prescription could be considered for children with otorrhoea or children younger than two years with bilateral AOM, but it is important to weigh the likely benefits against the possible risk, including side effects. The NPC patient decision aids relating to AOM specifically and upper respiratory tract infections generally may be helpful in some consultations.
What is the background to this?
As discussed on the relevant section of NPC and in a MeReC bulletin in December 2006, a strategy of watchful waiting or delayed antibiotic prescribing is appropriate for most children with AOM. About two thirds of children recover within 24 hours, whether they receive antibiotics or not, and the number of children in whom antibiotics prevent pain at 2-7 days (about 7 in 100) is almost exactly the same as the number of extra children who suffer diarrhoea, vomiting or rash through antibiotic treatment (about 6 in 100). Some children are more likely to benefit from antibiotics than others (see action section above).
This study is a follow-up to an RCT which was published several years ago, in which 240 children aged 6 months to two years with a diagnosis of AOM presenting to 53 general practices in the Netherlands were randomised to receive amoxicillin or placebo, and which found results in terms of benefits and harms similar to those described above. Approximately three and half years after the start of the trial the researchers sent a questionnaire to parents of the participating children asking them about episodes of recurrent AOM, referral to secondary care (paediatrician or ENT surgeon) and ENT surgery.
What does this study claim?
Questionnaires were returned for 168 of the original 240 children (70%): this subgroup had similar baseline characteristics to the whole study population and the groups who had been given placebo or antibiotics were well matched. The authors say that 95% of parents were still unaware of the original treatment allocation.
Three years after randomisation, AOM had recurred in 47/75 (63%) children in the amoxicillin group compared with 37/86 (43%) children in the placebo group, relative risk 1.5 (95% confidence interval [CI] 1.1 to 2), odds ratio (OR) 2.22 (95%CI 1.2 to 4.2), absolute risk increase 20% (95%CI 5% to 35%), number needed to harm (NNH) 5 (95%CI 3 to 22). In other words, for every five children like those in the study given antibiotics for AOM, one developed AOM in subsequent years who would not have done had they all been given placebo.
Further analysis showed that patient sex, allergy, and history of recurrent AOM might be relevant confounding factors. Adjustment for these factors resulted in a slightly higher adjusted odds ratio of 2.5 (95%CI 1.2 to 5.0), suggesting an even greater risk of recurrence of AOM due to antibiotic treatment. There were no significant differences in the rates of referral or surgery.
So what?
This study has some limitations. In particular, the original sample size was not large, and this observational extension was smaller still. It also relies on parent-reported outcomes and is therefore at risk of recall bias, that is, inaccurate recollection of events. The authors say that this possibility should not invalidate the results (e.g. if parents of children prescribed antibiotics were more likely to recall and report later AOM infections), because most parents were still unaware of the original treatment allocation. However, it is not clear how they established this.
Despite these limitations, the findings of this study support NICE guidance that, generally speaking, antibiotics should be avoided for treatment of AOM, unless there are compelling reasons for treatment.
More information on AOM can be found on the common infections – respiratory section of NPC.
Study details
Patients: 168 of the original 240 children aged six months to two years presenting to GPs in the Netherlands with signs and symptoms of AOM.
Intervention and comparison: amoxicillin suspension 40 mg/kg/day in three divided doses for 10 days, or placebo suspension
Outcomes and results: Parents reported recurrence of AOM in 47/75 (63%) children in the amoxicillin group compared with 37/86 (43%) children in the placebo group, RR 1.5 (95%CI 1.1 to 2.0), OR 2.22 (95%CI 1.2 to 4.2), absolute risk increase 20% (95%CI 5% to 35%), NNH 5 (95%CI 3 to 22). Logistic regression analysis showed that patient sex, allergy, and history of recurrent AOM might be relevant confounding factors. Adjustment for these factors resulted in an adjusted OR of 2.5 (95%CI 1.2 to 5.0). Parents reported secondary care referral in 24/78 (31%) of children in the amoxicillin group compared to 27/89 (30%) in the placebo group (risk difference 0%, 95%CI –14% to 14%). Parents reported ENT surgery in 16/78 (21%) of children in the amoxicillin group compared to 27/90 (30%) in the placebo group (risk difference –9%, 95%CI –23% to 4%)
Sponsorship: Netherlands Organisation for Scientific Research (grant No 90458074) funded the original trial.
Feedback
Please comment on this blog in the NPC discussion rooms, or using our feedback form.
Make sure you are signed up to NPC Email updates — the free email alerting system that keeps you up to date with the NPC news and outputs relevant to you.