A recent meta-analysis in Britain was reported as showing that certain anti-depressants, namely SSRIs, are no better than a placebo in treating depression. STATS Maia Szalavitz explains on Scientific American why the coverage of this study misses the salient point: “when you are looking at aggregated data, huge individual differences can be washed out.”
Some people are strong responders to one drug –but give them another in the same class, and they become actively suicidal. Most people have a slight positive effect; some have a slight negative effect. In aggregate, a drug that is a home run for one person and potentially fatal to another looks inert.
More at Scientific American’s 60- Second Science.