The interpretation bias and trap of multicenter clinical research