Every year influenza season rolls around.

Every year we need to discuss oseltamivir.

However, in 2025, it is no longer solely oseltamivir available for consideration and misuse.  We now have the next generation of influenza antivirals, including peramivir and baloxavir, not to mention potential antiviral candidates from around the world.  Unfortunately, despite the ubiquity and predictability of influenza season, the millions of prescriptions for these medications annually, and this parade of options, the evidence underpinning their use remains poor.

There have  been many individual studies published testing these influenza antivirals.  Further, there have been various systematic reviews, focussing primarily on oseltamivir.  Each of these types of evidence tells one specific type of story, but an interesting tool called a “network meta-analysis” provides scaffolding for a more expansive look.¹

The underlying principle behind a network meta-analysis allows it to perform indirect comparisons where direct, head-to-head trials may not have been performed.  Additionally, it can use information from these indirect comparisons to strengthen or attenuate findings from other, direct comparisons.  Such a tool can potentially integrate the entire landscape of influenza antivirals, assuming such antivirals were tested against each other, or placebo, in trials enrolling similar types of patients.

The network meta-analysis in question looks across the entire landscape of influenza antiviralsinvolving primarily, the aforementioned oseltamivir, peramivir, and baloxavir, but also gives mention to laninamivir, favipiravir, umifenovir, zanamivir, and amantadine.  The scope of this review involves only patients with “non-severe influenza” — i.e.  the ambulatory, outpatient population representing most emergency department patients, rather than those sick enough to require hopsital admission .

In approaching influenza, the United States Center for Disease Control and Prevention, echoing the World Health Organization, makes a distinction between “low risk” and “high risk” patients.² “High risk” patients are considered to be those who are pregnant, aged greater than 65 or younger than 2, those with lung disease, morbid obesity, and other types of serious chronic diseases.  Current recommendations consider treatment appropriate primarily in the “high risk” cohort, while leaving treatment in the “low risk” to individual clinician discretion.  In this network meta-analysis, where data were available, the authors perform the same mortality, hospitalization, and adverse effects comparisons for both these cohorts.

The total population for analysis consisted of 65 trials published between 1971 and 2023, comprised of 34,332 patients.  The largest body of evidence exists for oseltamivir, followed by zanamivir, and most comparisons were with placebo, rather than each other.  The first important descriptive finding these authors report is the risk of bias across the included trials.  Out of the 65 included, only 9 were at “low” or “probably low” risk of bias across all reported outcomes.  The most frequent threats to study validity were inadequate allocation concealment, lack of blinding, and in data analysis. 

The key starting place to consider which medications are efficacious is to recognize the bar for success for these antivirals is extraordinarily difficult to measure.  Absent antivirals, the chance of death from non-severe influenza was estimated at 1 in 5,000 in the low-risk cohort, and at 1 in 500 in the high-risk cohort.  Therefore, it is virtually impossible to measure a clinically meaningful effect on mortality.  Hospitalization rates are ten times greater however, and are the primary surrogate for efficacy relied upon in the included trials.

Getting straight to the main, practice-informing point of their findings: oseltamivir belongs on the junk heap.  Where these authors could generate evidence with high level of certainty, they were unable to find any signal oseltamivir reduced mortality or admission to hospital.  These findings were consistent across both low-risk and high-risk cohorts.  With respect to alleviation of symptoms, oseltamivir may decrease illness duration by a little more than half a day, but this outcome was uncertain due to serious risk of bias.  Likewise, oseltamivir may increase treatment-related non-serious adverse events, commonly nausea and vomiting, but this outcome is also affected by the same risk of bias.

The story is a bit different with baloxavir. Owing to the baseline low rate of mortality, baloxivir could not demonstrate a meaningful improvement in this outcome. Nevertheless, baloxivir did demonstrate a substantial effect on hospitalization and a full day’s worth of alleviation of symptoms.  The big caveat:  these outcomes are held with low and moderate certainty due to “serious imprecision”. This imprecision stems from a relative paucity of data on baloxavir, derived primarily from a mere three clinical trials sponsored by the manufacturer.

There is another important consideration for baloxavir:  resistance. In the studies included in this review, an estimated of ~10% of virus isolates demonstrated induced antiviral resistance following treatment with baloxavir.  While the controlled clinical trial evidence certainly favored baloxavir, widespread use during influenza season would likely attenuate this advantage and reduce its effectiveness during future seasons.³

The other antivirals mentioned are primarily used in overseas markets. In this review, the data regarding each was too sparse to suggest that any could yet be considered promising candidates to improve outpatient management of non-severe influenza.

So, what does “good prescribing” during influenza season look like?  In the vast majority of uncomplicated influenza managed as an outpatient, it ought to simply be recommendations for symptomatic self-care.  Oseltamivir has become ingrained in clinical practice, most likely as result of an initial favorable view arising from non-publication of early negative trials.^4,5 It is always more difficult to “de-adopt” practices than for them to take root.  On the other hand, baloxavir demonstrates obvious appeal.  However, baloxavir remains both costly from an individual standpoint, and damaging to public health from a population standpoint.  Where possible, baloxavir ought be prescribed conservatively in those at highest risk, as widespread use will ultimately become its downfall.

Dr. Radecki (@EMLITOFNOTE) is an emergency physician and informatician with Christchurch Hospital in Christchurch, New Zealand. He is the Annals of Emergency Medicine podcast co-host and Journal Club editor.

References

1. Gao Y, Zhao Y, Liu M, et al. Antiviral medications for treatment of nonsevere influenza: a systematic review and network meta-analysis. JAMA Intern Med. Published online January 13, 2025.
2. Treating Flu With Antiviral Drugs.  https://www.cdc.gov/flu/treatment/antiviral-drugs.html [accessed 16 Jan 2025]
3. Uehara T, Hayden FG, Kawaguchi K, et al. Treatment-emergent influenza variant viruses with reduced baloxavir susceptibility: impact on clinical and virologic outcomes in uncomplicated influenza. The Journal of Infectious Diseases. Published online July 16, 2019:jiz244.
4. Jefferson T, Jones M, Doshi P, Spencer EA, Onakpoya I, Heneghan CJ. Oseltamivir for influenza in adults and children: systematic review of clinical study reports and summary of regulatory comments. BMJ. 2014;348(apr09 2):g2545-g2545.
5. Baghdadi JD, Grady D, Morgan DJ. The limited role for antiviral therapy in influenza. JAMA Intern Med. Published online January 13, 2025.