Skip navigation
Link to Back issues listing | Back Issue Listing with content Index | Subject Index

Antibiotics for acute otitis externa

Systematic review
Results
Comment

Acute otitis externa is an inflammation of the external ear canal, commonly known as swimmer's ear. It can be treated with systemic or topical antibiotics, and topical treatment can involve several different types of antibiotic or antiseptic, sometimes combined with cortiocosteroid. A systematic review reveals just how little we know about what works best [1].

Systematic review


Searching involved using different databases for studies in any language. Articles were limited to acute otitis externa, parallel group design, and comparing antimicrobial and placebo, antiseptic and antimicrobial, and steroid plus antimicrobial versus antimicrobial alone or steroid alone. Outcomes used included clinical cure (absence of all signs and symptoms) or improvement (partial or complete relief). Different end points over 3 to 21 days were examined, with the intention to combine the final results.

Results


Twenty trials described as randomised were found, 18 with the required information for data pooling. The median size was 79 patients (range 28 to 842), all but one including children and adults. Half did not explicitly define acute otitis externa, and only half were defined as double blind.

Using the Oxford quality scoring system for randomisation, blinding, and withdrawal description, only 10 scored 3 out of 5 points, associated with a relative lack of bias. Those that were adequate (3, 4, or 5 out of 5) did not improve with time (Table 1).


Table 1: Trial quality over five decades



Quality score
(range 1-5)
Decade
1 or 2
3-5
1960s
0
1
1970s
1
3
1980s
1
1
1990s
4
2
2000s
4
3



There were 13 meta-analyses of these 18 trials, without sensitivity analysis according to quality score. The effect of topical antibiotic (neomycin) plus corticosteroid compared with placebo was described in only two good quality studies, with only 89 patients. Cure rates were much higher for antibiotic plus steroid than with placebo at 3-10 days. The NNT calculated for these two trials was 2.2 (1.6 to 3.7).

The only other comparisons of at least two treatments with relatively unbiased trials was for antiseptic versus antibiotic in three trials, with identical cure rates of about 60% at 7-10 days and 80% at 14 to 28 days.

Comment


Another example where we have a paucity of data to guide therapy for a relatively common condition. Most trials were performed since 1990, yet most had poor quality scores indicating at best poor reporting quality, and at worse inadequate conduct. If anything, the quality of more recent trials was worse than those published earlier (Table 1). These are simple trials, and the influence of quality is well known. How can it be that trials of inadequate quality continue to be performed or reported? This shows a clear failure by ethics committees and journals, and a disservice to patients and professionals.

Reference:

  1. 1 RM Rosenfeld et al. Systematic review of topical antimicrobial therapy for acute otitis externa. Otolaryngology – Head and Neck Surgery 2006 134:S24-S48.

previous or next story