Chronic suppurative otitis media (CSOM), usually accompanied by otorrhea, is a major cause of hearing impairment in children, especially in developing countries. Management is either unsatisfactory or very difficult [1]. Otorrhea typically occurs after perforation of the tympanic membrane as a complication of acute otitis media, with subsequent discharge of pus into the external ear canal. Acute otitis media is hard to diagnose in young infants due to non-specificity of presenting symptoms and the difficulty of visualizing the tympanic membrane [2]. Otorrhea, however, is easy to visualize and does not require any equipment [1].

The study was a retrospective folder search for otorrhea identified in the clinical notes and assumed due to a perforation of the tympanum, which is the most common cause. External otitis media was excluded clinically. Otitis media and otorrhea are World Health Organization (WHO) stage II events for HIV in children [3]. Both occur frequently in HIV-infected children [4–6]. Antiretroviral therapy (ART) was associated with a reduction in otitis media and CSOM in a retrospective cohort [7].

In a sub-analysis of a prospective randomized strategy trial, we compared otorrhea in young HIV-infected infants receiving early versus delayed ART. Infants aged 6 to 12 weeks of age with confirmed HIV infection and a CD4 percentage greater than or equal to 25% were randomized to deferred ART according to South African guidelines [8] (the deferred-therapy group) or immediate ART until 1 or 2 years of age (the early-therapy groups), followed by treatment interruption.

Eighteen infants developed at least one episode of otorrhea after baseline: 11 of 38 (29%) in the deferred-therapy group and 7 of 75 (9%) in the early-therapy groups (risk ratio 3.10, 95% CI, 1.31-7.36; p = 0.01). Evaluation for severe otorrhea showed a similar trend, but did not reach statistical significance. Seven (18.4%) in the deferred arm and 6 (8%) in the early ART arm had severe disease (risk ratio 1.74, 95% CI, 0.97-3.11; p = 0.10).

Three infants, one in the deferred-therapy group and two in the early-therapy groups, each had 2 separate events of otorrhea. In the deferred-therapy group, 40% of infants had not initiated ART at onset of otorrhea.

Eight infants had otorrhea fluid submitted for culture, 4 from the deferred and 4 on early ART. Two S. pneumonia isolates were from infants on the deferred arm. Of two Staphylococcus aureus isolates, one was methicillin sensitive and one was resistant. The other isolates included Pseudomonas, Proteus and Serrattia species and one of Kelbsiella pneumonia. Five of 7 organisms were resistant to TMP-SMX.

In this prospective randomized study, we showed that early initiation of ART provided significant protection against otorrhea compared to deferred ART, initiated on the basis of either CD4 depletion or progression of HIV disease. Infants in the deferred arm had 3 times the risk of developing otorrhea compared with those who started treatment immediately. Nevertheless, even in the early therapy group, the incidence was relatively high (9%). In a cross-sectional survey from Swaziland before the advent of HIV, the prevalence of otorrhoea in children attending grade 1 was only 1.3% [14].

Of note is that all infants entered the study being relatively asymptomatic and with high baseline CD4 percentages at a median of 7.4 weeks of age. We documented a high incidence of otorrhea in the first year of life. Our study provides additional evidence of the benefits of ART for infants. Preventing otorrhea in HIV-infected infants improves their quality of life, preserves hearing and reduces the burden on the health system. Interestingly, the CD4 counts were relatively high in both groups, likely due to careful monitoring and ready access to ART in a trial setting. We infer that ART must have a qualitative benefit on CD4 functioning.

Our results are similar to those from a retrospective cohort study from Brazil, which found that ART comprising three antiretrovirals was associated with a significantly lower prevalence of otorrhea (7.1%) than children receiving only one or two antiretrovirals (20.8%) (p = 0.02) [7]. The mean age in this study was 6.6 (± 2.5) years, far older than in our study. We now confirm that their observations are already noted in the first year of life. In an earlier study from our unit, the prevalence of otorrhea before the widespread availability of ART was 32%, close to the prevalence in our deferred-therapy group, and was linked to CD4 depletion [5]. Had we included infants with a CD4 percentage below 25% in CHER, it is likely that we would have documented a higher prevalence of otorrhea.

The majority of events in the delayed group occurred on therapy. Possible explanations may include immune reconstitution inflammatory syndrome as well as better qualitative antibody production with early antiretroviral therapy [11].

Our study is strengthened by the fact that it was a prospective randomized trial. In addition, the study population represents the broader infant population accessing ART through the public health system, making the results widely generalisable.

There are a number of limitations to our study. Infants were not evaluated by an otorrhinolaryngologist. Perforation of the tympanic membrane was not uniformly documented. Also, we did not consistently submit otorrhea specimens for bacterial culture or serotyping of S. pneumonia isolates. We were therefore unable to evaluate the effect of immunization with the conjugated pneumococcal vaccine. We have, however, demonstrated in the companion study to CHER that early initiation of ART is associated with improved qualitative function of specific antibodies to serotypes in vaccine [11]. An intriguing finding in the present study was that the S. pneumonia isolates were both from infants where ART was deferred.

Early ART in young HIV-infected infants significantly reduces the incidence of otorrhea, a condition associated with much morbidity.