Efficacy of oseltamivir treatment started within 5 days of symptom onset to reduce influenza illness duration and virus shedding in an urban setting in Bangladesh: a randomised placebo-controlled trial

Background Influenza causes substantial morbidity and mortality worldwide. Few data exist for the efficacy of neuraminidase inhibitors, which are the only readily available influenza treatment options, especially in low-income settings. We assessed the efficacy of treatment with the neuraminidase inhibitor oseltamivir to reduce patient illness and viral shedding in people with influenza, in whom treatment was started within 5 days of symptom onset, in an urban setting in Bangladesh. Methods We undertook a double-blind, randomised, controlled trial between May, 2008, and December, 2010. Patients with a positive rapid influenza test identified by surveillance of households in Kamalapur, Bangladesh were randomly allocated on a 1:1 basis to receive oseltamivir or placebo twice daily for 5 days. Randomisation lists for individuals enrolled less than 48 h and 48 h or longer since illness onset were generated with permuted blocks of variable length between two and eight. Participants and study staff were masked to treatment group. Participants provided nasal wash specimens at enrolment and 2,4, and 7 days later, and were visited daily to record symptoms. All specimens were tested for influenza with reverse-transcriptase PCR, and if the result was positive, we isolated the virus. The primary endpoints were duration of clinical illness and viral shedding in patients treated less than and more than 48 h since illness onset and the frequency of oseltamivir resistance during treatment. Analyses were intention to treat unless otherwise specified. Findings Overall, 1190 people with a median age of 5 years (IQR 2-9) were enrolled: 794 (67%) less than 48 h since symptom onset and 396 (33%) 48 h or longer since symptom onset. 592 participants were assigned to placebo and 598 to oseltamivir. The median duration of symptoms was shorter in the oseltamivir group (3 days, IQR 1-5) than in the placebo group (4 days, 1-6; p=0.01). When stratified by timing of treatment initiation, in participants enrolled 48 h or longer since illness onset, the median duration of symptoms was similar in both groups (oseltamivir 3 days [IQR 2-5], placebo 3 days [1-5]; p=0-04). The median duration of symptoms was reduced by 1 day in the group given oseltamivir who were enrolled less than 48 h since symptom onset compared with those given placebo, but this difference was not significant. In those with all swab specimens (n=1134), oseltamivir significantly reduced virus isolation on days 2 (placebo 374 [66%] vs oseltamivir 321 [56%]; difference 15.2%, 95% CI 9-5-20.8, p=0.0004), 4 (241 [43%] vs 174 [30%]; difference 30.2%, 95% CI 24.6-35.8, p<0.0001), and 7 (68 [12%] vs 36 [6%]; difference 47.5%, 95% CI 44.2-50.8, p=0.0009). In participants enrolled 48 h or longer since illness onset, oseltamivir treatment significantly reduced virus isolation on days 2 and 4, but not day 7. In participants enrolled less than 48 h since illness onset, oseltamivir treatment significantly reduced virus isolation on days 2,4, and 7. The emergency of resistance to oseltamivir during treatment was rare overall (<1%) and in influenza A H1N1pdm09 viruses (3.9%). Interpretation Oseltamivir treatment resulted in a modest reduction in the duration of symptoms and virus shedding in people with uncomplicated influenza infections, even when treatment was started 48 h or longer after illness onset.