Antiviral medications for influenza

Clinical question

Do the neuraminidase inhibitors (NIs) oseltamivir or zanamivir improve clinical outcomes in influenza?

Bottom line

Biased, poor-quality, mostly unpublished evidence suggests that oseltamivir and zanamivir shorten the duration of influenza symptoms by 0.6 to 0.7 of a day. Pneumonia and hospitalizations are not decreased.

Evidence

Three 2014 systematic reviews of placebo-controlled RCTs (including > 160 000 pages of previously unreleased clinical study reports)^1–3 found the following when treating otherwise healthy adults with influenza or influenzalike illness with oseltamivir (11 RCTs)^1,2 or zanamivir (14 RCTs)^1,3:

• Time to symptom improvement was 0.6 to 0.7 days (about 10%) better^1–3; the benefit of zanamivir was similar to “relief medications” (like acetaminophen).^1,3

• There was no benefit for pneumonia (x-ray scan confirmed)^1–3 and hospitalizations were not reported^1,3 or there was no benefit.^1,2

• Adverse events included the following:

  • -For oseltamivir,^1,2 the number need to harm was 28 for nausea and 22 for vomiting.

  • -Postmarketing surveillance reports (frequency unknown) identified bronchospasm with zanamivir⁴ and delirium and self-injury with oseltamivir.⁵

A 2015 systematic review⁶ concluded adults receiving oseltamivir had faster symptom alleviation, and fewer lower respiratory tract complications and hospitalizations.

• The review used similar studies^1,2 but the conclusion was based on a subgroup with documented influenza.

• The review was funded by, and 2 authors had pre-existing financial affiliations with, the manufacturer of oseltamivir.

From 26 systematic reviews,⁷ authors with financial conflicts of interest were 5 times more likely to report benefits of NI use (this includes a systematic review of cohort studies from the 2009–2010 pandemic suggesting that NIs decreased mortality in hospitalized patients⁸) and less likely to report on publication bias and the quality of included studies.

• Other concerns¹: unpublished protocols; inconsistent outcome definition; “placebos” with potential adverse effects; and incomplete reporting (eg, missing symptom cards).

Context

• Oseltamivir sales are > $18 billion, half from government and company stockpiling. Most have not been used.⁹

• The NIs are not recommended if symptoms have lasted longer than 48 hours.^4,5 Zanamivir is contraindicated in asthma and COPD.⁴ There is limited evidence for oseltamivir in underlying cardiac or respiratory disease.⁵

• Limited data suggest NIs are likely safe in pregnancy,¹⁰ but the manufacturers do not recommend (zanamivir)⁴ or conclude that there are insufficient data and to use them only when the potential benefit justifies the potential risk to the fetus (oseltamivir).⁵

Implementation

Headache or muscle aches do not reliably differentiate influenza from other respiratory infections. Cough with fever might have the greatest diagnostic value,¹¹ but knowledge of current local influenza rates is more important. In outbreaks, 79% of patients with fever and cough have influenza,¹² but influenza accounts for only 10% of identified respiratory pathogens in a typical winter.¹³ Trials of NIs generally exclude the very young, the old, and those with comorbidity. Targeting those at greatest risk of complications during influenza outbreaks might provide greater utility. High-quality trials enrolling such patients are needed.