Neuraminidase inhibitors for influenza: what does the evidence really show?

This has been the subject of controversy. Here is an evidence summary published in a recent issue of the Annals of Energency Medicine. From the article:

This Cochrane review demonstrates that oseltamivir shortens symptom duration by 21 hours in patients who receive the drug within 48 hours of symptom onset. Unfortunately, there is little information that can be inferred from current available data about hospitalization rates and neuraminidase inhibitors’ effectiveness in decreasing transmission rates of the influenza virus. However, the CDC continues to recommend neuraminidase inhibitors, using information based on observational studies that showed decreased hospital stays and severe outcomes such as ICU admissions or death in patients treated with oseltamivir.3

In the past, Cochrane reviews examining the effect of neuraminidase inhibitors were based on published clinical studies conducted by drug manufacturers; however, these studies were found to reflect only a small portion of the trials conducted and bore discrepancies with their clinical study report counterparts. For the first time in Cochrane history, this review attempted to remove the reporting bias seen in the previous reviews by analyzing the unpublished regulatory data from clinical study reports rather than relying on the published trials.4 Unfortunately, despite major efforts by the systematic review authors, they were unable to obtain much of the data from trials sponsored by the drug manufacturers.

So we have high level data in support of NI based on soft outcomes (time to symptom relief) and low level data in support of them for hard outcomes (severe complications, death). The debate about this has been heated and overhyped. The final answer is that it depends on your evidence quality threshold, your preferences and your values.