I trust the FDA and Cochrane Review here more than the CDC.
The small durational effect, high dependence on early-administration, and reliance on observational reports combine to add to my suspicion. Consider especially:
If patients are motivated to seek anything that might help, and they know that a prescription is only given in "first 48 hours", they might report their symptoms as having started later than they really did. Then, when their symptoms end at the usual time, it "observationally" looks like a shorter flu. Or maybe, other placebo/choice-supportive-bias effects cause them to rush to deem their symptoms 'over' – after all they went through the time and cost of acquiring the drug.
Thus a pronouncement like "works if rapidly administered" could become a self-fulfilling prophesy through social/psychological biases in reporting.
Similarly, the CDC is under political/social pressure to "do something", and have active modern recommendations, other than just generic 100-year-old patience/hygiene tips. This is especially the case when, as this year, their flu-strain guesses for the vaccine have been wrong.
Temperamentally and institutionally, the CDC needs to believe there's something else to do, that the stockpiles of antivirals they've bought at great cost are effective, and that these kinds of drugs (from specific companies they often rely upon) will help if something much worse than seasonal flu materializes. So they're going to err in the direction of faith-in-antivirals.
The small durational effect, high dependence on early-administration, and reliance on observational reports combine to add to my suspicion. Consider especially:
If patients are motivated to seek anything that might help, and they know that a prescription is only given in "first 48 hours", they might report their symptoms as having started later than they really did. Then, when their symptoms end at the usual time, it "observationally" looks like a shorter flu. Or maybe, other placebo/choice-supportive-bias effects cause them to rush to deem their symptoms 'over' – after all they went through the time and cost of acquiring the drug.
Thus a pronouncement like "works if rapidly administered" could become a self-fulfilling prophesy through social/psychological biases in reporting.
Similarly, the CDC is under political/social pressure to "do something", and have active modern recommendations, other than just generic 100-year-old patience/hygiene tips. This is especially the case when, as this year, their flu-strain guesses for the vaccine have been wrong.
Temperamentally and institutionally, the CDC needs to believe there's something else to do, that the stockpiles of antivirals they've bought at great cost are effective, and that these kinds of drugs (from specific companies they often rely upon) will help if something much worse than seasonal flu materializes. So they're going to err in the direction of faith-in-antivirals.