Many randomized trials have been carried out which attempted to assess
the effect of drug, dietary, or surgery intervention for cholesterol reduction 
on the incidence of ischemic heart disease (IHD). These data are from a 
meta-analysis originally carried out and reported on in Law, Thompson,  
and Wald (1993) and Law, Wald, and Thompson (1993).

25 randomized trials are summarized. The data are IHD events (IHD 
fatalities and non-fatal myocardial infarction combined) in each of a 
treated and control group as well as the total number of subjects in each 
of the treated and control groups. Diagnostic criteria for the IHD
events were similar (but not identical) across the 25 trials and were
recorded by observers blinded to an individual's treatment. 

The treatment in each trial is classified as either drug, diet, or 
surgery. Trials with more than two treatments being simultaneously 
compared are exluded. The patient population in each trial is classified 
as either primary (if the subjects were free of IHD at baseline) or 
secondary (if the subjects had all had some non-fatal IHD event prior 
to baseline) or a mixture. Also reported are the trial-specific average 
duration of follow-up (in years) and the average post-treatment difference 
in cholesterol levels (control group minus treated group, so a positive 
value means the control group had higher cholesterol levels than the 
treated group).  Note that the cholesterol values do not reflect 
AVERAGE CHANGE OVER TIME (did cholesterols go down) but instead reflect 
AVERAGE DIFFERENCE DUE TO TREATMENT GROUP. Cholesterol changes are given 
as mmol/L.

Previous analyses have analyzed these data on the odds ratio scale, not
on the log odds ratio scale.

Data source: Thompson SG (1993). "Controversies in meta-analysis: the
case of the trials of serum cholesterol reduction," Statistical Methods
in Medical Research, 2: 173-192.

Law MR, Wald NJ, and Thompson SG (1994). "Serum cholesterol reduction 
and health: by how much and how quickly is the risk of ischaemic heart
disease lowered?" British Medical Journal, 308: 367-372. 

Law MR, Thompson SG, and Wald NJ (1994). "Serum cholesterol reduction
and health: assessing possible hazards," British Medical Journal, 308:
373-379.