Oseltamivir is a neuraminidase inhibitor that prevents replication of the influenza virus. The evidence supporting its use for influenza is primarily from clinical trials in outpatients, which demonstrate a reduction in symptom duration of 16-24 hours.

Based on its antiviral activity and some observational data suggesting efficacy in more unwell patients, international guidelines have recommended oseltamivir for the treatment of critically ill patients with influenza.

Before the REMAP-CAP trial, there had never been a randomised trial of oseltamivir in the critically ill.

REMAP-CAP is an adaptive platform trial, which randomises patients with severe pneumonia and influenza to receive various treatments. The results being reported relate to critically ill patients (aged 12 or older) with influenza who were randomised to receive oseltamivir or to receive no antiviral treatment. 

Unexpectedly, the trial found that in patients with influenza who were in ICU receiving organ support, those receiving oseltamivir had a higher death rate. In 319 patients with severe influenza who received oseltamivir, 62 (19.4%) died, but in the 124 patients who didn't receive antivirals, 17 (13.7%) died.

REMAP-CAP differs from traditional trials in that it uses Bayesian analysis methods. In this framework, the results are expressed as a probability of oseltamivir being effective or harmful based on the odds of mortality compared to receiving no antiviral (rather than the frequentist statistical approach of hypothesis testing). The trial found that the probability of oseltamivir being effective was less than 2%, and the probability that it is harmful is greater than 98%.

These results ONLY apply to critically ill patients aged 12 or older with confirmed influenza. They DO NOT apply to patients with influenza who are in the community or hospitalised but not critically ill, and do not apply to paediatric patients less than 12 years old.

It is clear, from randomised trials that enrolled thousands of patients in the community with influenza, that oseltamivir is safe and shortens the duration of symptoms (by ~18 hours), but with little or no effect on admission to hospital or mortality in this group. There are no large randomised trials of oseltamivir in hospitalised patients who are not critically ill and the efficacy of oseltamivir in that group is not clear. The role of oseltamivir in influenza outbreak settings, such as post-exposure prophylaxis, is also not affected by these results.  

REMAP-CAP continues to evaluate the safety and effectiveness of oseltamivir for hospitalised non-critically ill patients with influenza.

The REMAP-CAP investigators recommend that clinicians should stop routinely using oseltamivir in critically ill patients aged 12 years or older with influenza. 

These results do not relate to the use of oseltamivir in other populations (e.g., non-hospitalised patients, non-critically ill hospitalised patients, and children) with influenza. The use of oseltamivir in these patients should follow local guidelines.

This result relates to patients aged 12 years and older who were critically ill with influenza in an ICU. REMAP-CAP continues to evaluate the safety and effectiveness of oseltamivir for hospitalised children younger than 12 years; however, at this point in time, we do not have any results relating to oseltamivir in this patient population. 

The mechanism of the observed increase in mortality with oseltamivir is not known. Further work will be required to understand the mechanism that underpins the REMAP-CAP findings. Possible mechanisms include the effects of oseltamivir on the patient’s neuraminidase enzymes, which may predispose patients to worse outcomes (e.g., due to deranged haemostasis or an increased risk of secondary infections). However, this is just one hypothesis that requires further investigation. 

Yes, the data suggests that patients who were more unwell with shock, patients who had bacterial infection in addition to influenza virus infection, patients who had been symptomatic with influenza infection for five days or more, and patients who were immunocompromised had a greater increase in mortality with oseltamivir treatment than patients without those characteristics. 

REMAP-CAP is an adaptive platform conclusion that utilises Bayesian statistical methods. No pre-specified sample size is set, but instead statistical stopping rules are pre-specified. Regular analyses are conducted as data accumulates, and recruitment stops when a statistical trigger is reached. This avoids indeterminate results, prevents recruiting too few or too many participants, and facilitates sharing results as soon as there is sufficient statistical confidence that the results have a very low probability of occurring by chance alone. In this case, the independent Data and Safety Monitoring Board advised the trial steering committee that recruitment to the 5-day and 10-day oseltamivir arms should cease as the stopping rule for inferiority of those interventions had been reached (<0.2% probability of being the most effective intervention).

One of the major determinants of the number of necessary trial participants is the size of the difference in outcome between control and intervention. In the full analysis after recently recruited participants had been followed-up, there was a 1.7% and 1.1% probability that oseltamivir is effective and a 98.7% and 98.9% probability that oseltamivir is harmful, for the 5-day and 10-day treatments respectively. This indicates that, despite a moderate sample size, there is a high statistical confidence in these results.

The trial team considered the results to be of significant public health importance and so they were presented at Critical Care Reviews on 10 June, prior to a peer-reviewed publication.

The manuscript reporting the comparison of oseltamivir to no influenza antiviral from this domain of REMAP-CAP is being prepared for publication. This process includes collecting day 90 follow-up data for the small proportion of patients who were recruited only recently, prior to closure of the domain, which may mean that there are small changes in the mortality rates in the final publication. Once the manuscript has undergone peer review and is published, it will be available via the REMAP-CAP website. 

Alternative antiviral medications for influenza (e.g., Baloxavir) remain under investigation within the REMAP-CAP platform. There are currently no antivirals that have randomised evidence of benefit for critically ill patients with influenza.

These results highlight the importance of trials such as REMAP-CAP for evaluating the safety and efficacy of interventions for the treatment of respiratory tract infection, including interventions that are already in widespread use. 

The REMAP-CAP investigators would like to acknowledge the important contribution of our participating patients and sites for helping to generate important evidence to improve the care of future patients.

If you would like to participate in REMAP-CAP, please contact info@remapcap.org, or the regional coordinating centre in your region.