In meta-analysis,we combine results of different clinical trials.One such tool is meta-regression.Consider trials of cholesterol lowering.Let x be the mean difference in cholesterol for a given trial, and y be the treatment-to-control log odds ratio for death.You have,say,10 such clinical trials,and you regress y on x.

Some people argue that cholesterol lowering drugs can do more harm than good if they are used in populations who do not achieve much cholesterol lowering.To see if this is true, what would you concentrate on, the slope or the intercept of the meta-regression equation?Explain.