In trachoma control programmes, azithromycin is distributed to treat the strains of chlamydia that cause ocular disease. We aimed to compare the effect of annual versus twice-yearly distribution of azithromycin on infection with these strains.

We did a cluster-randomised trial in 24 subdistricts in northern Ethiopia, which we randomly assigned to receive annual or twice-yearly treatment for all residents of all ages. Random assignment was done with the RANDOM and SORT functions of Microsoft Excel. All individuals were offered their assigned treatment of a single, directly observed, oral dose of azithromycin. A 6 week course of topical 1% tetracycline ointment, applied twice daily to both eyes but not directly observed, was offered as an alternative to azithromycin in patients younger than 12 months, and in patients with self-reported pregnancy, with allergy, or who refused azithromycin. Our primary, prespecified outcome was the prevalence of ocular chlamydial infection in a random sample of children aged 0-9 years at baseline and every 6 months for a total of 42 months within sentinel villages. Our analysis was by intention to treat. This study is registered with ClinicalTrials.gov, number NCT00322972.

Antibiotic coverage of children aged 1-9 years was greater than 80% (range 80·9 to 93·0) at all study visits. In the groups treated annually, the prevalence of infection in children aged 0-9 years was reduced from a mean 41·9% (95% CI 31·5 to 52·2) at baseline to 1·9% (0·3 to 3·5) at 42 months. In the groups treated twice yearly, the prevalence of infection was reduced from a mean 38·3% (29·0 to 47·6) at baseline to 3·2 % (0·0 to 6·5) at 42 months. The prevalence of ocular chlamydial infection in children aged 0-9 years in groups treated annually was not different from that of the groups treated twice yearly at 18, 30, and 42 months (pooled regression p>0·99, 95 % CI -0·06 to 0·06). The mean elimination time in the twice-yearly treatment group was 7·5 months earlier (2·3 to 17·3) than that of the annual group (p=0·10, Cox proportional hazards model).

After 42 months of treatment, the prevalence of ocular infection with chlamydia was similar in the groups treated annually and twice yearly. However, elimination of infection might have been more rapid in the groups of villages that received treatment twice yearly.

National Institutes of Health (NEI U10 EY016214).