> Spectacles that are marketed to filter out blue light probably make no difference to eye strain caused by computer use or to sleep quality, according to a rev

> Spectacles that are marketed to filter out blue light probably make no difference to eye strain caused by computer use or to sleep quality, according to a review of 17 randomized controlled trials of the best available evidence so far. This seems like a highly misleading summary of what the meta-analysis showed, or rather, didn't show. "There is no proof that it works" somehow transmutes into "this proves it doesn't work". They are not the same. The joke about Cochrane Review meta-analyses is that the answer is always, "more research is required", and this looks like another example: https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD013244.pub2/full https://www.cochranelibrary.com/cdsr/doi/10.1002/14651858.CD... First, they find too few studies to meta-analyze. So how can they show that it "probably makes no difference to sleep quality", as the headline screams? Let's go through their comments: Depression: the quoted CI is "‐33.95 to 53.47". You could drive a truck through that, and most of those values would be practically important. This is just uninformative. Critical flicker fusion: no one cares, so any numbers don't matter. Daytime alertness: too few studies / uninformative, so they provide no numbers. Sleep quality: 3 RCTs hit statistical-significance for benefits, 3 did not (did the 3 have point-estimates for benefit? unspecified), they decline to meta-analyze the 6 together to provide an actual number. ("Potential effects on sleep quality were also indeterminate, with included trials reporting mixed outcomes among heterogeneous study populations.") If this shows anything, it shows that blue-filtering spectacles probably make a difference to sleep quality... Adverse effects: some anecdotal evidence, no numbers. Various other metrics: not reported enough to discuss. So, this is largely completely uninformative, except for the one metric "sleep quality" where they explicitly contradict the headline. ('Indeterminate' does not mean 'probably makes no difference to sleep quality'!) Medicalxpress screwed this one up.