Multiple secondary outcome analyses: precise interpretation is important.

Another problem is that if a secondary variable shows a clinically important effect or is highly significant (for example, with P < 0.0001), as one of the secondary outcomes in Baggot et al.

also in 132,142 other papers