A prime example of the harm that comes from exploratory analyses that are represented as confirmatory is one of the results from the Multiple Risk Factor Intervention Trial (MRFIT)  study. Published in 1982, it was designed to demonstrate that reductions in the risk factors associated with atherosclerotic cardiovascular disease would be translated into reduction in clinical events, e.g., myocardial infarction and stroke. Patients in the intervention group received treatment for elevated blood pressure, joined cigarette smoking cessation programs, reduced their weight, and lowered their serum lipid levels; patients in the control group followed their usual accepted standard of living. At the conclusion of the study, the investigators found and reported that there was no difference in clinical outcome between those patients who received risk factor intervention and those who did not.
These null findings were a disappointment to the risk factor interventionists, who then poured over the data to identify if any effect could be found in a fraction of the patients that might explain the null overall effect. They found one, and it was a bombshell. When the researches ignored the results in the entire randomized cohort (i.e., all randomized patients), and concentrated on men who were hypertensive and had resting electrocardiograph (ECG) abnormalities at baseline, they dis covered that these patients had a worse outcome when randomized to anti-hypertensive therapy than those who received no such therapy.
This result was published, and had a major impact on the momentum to treat essential hypertension. At this time in clinical medicine, the importance of identification and treatment of hypertension galvanized physicians. Screening programs were well underway. New therapies (e.g., hydrochlorothiazide, alphamethyl-dopa, and clonidine) for the hypertensive patient became available. All of the necessary forces for a war on undiagnosed and untreated hypertension were maneuvering into position when the MRFIT analyses were released. This finding slowed the momentum for the treatment of hypertension by raising disturbing, and ultimately unhelpful, questions e.g., "Maybe not all hypertension was bad after all?", or "Maybe hypertensive disease itself was bad, but the treatment was worse?"
The real question however, was, "Is it just a distorted treatment effect?" For years after this finding, clinical trials in hypertension were forced to address this unusual result. None of the major studies ever found that hypertensive men with resting ECG abnormalities were better off when their hypertension remained unchecked. Nevertheless, an exploratory analysis, dressed as a confirmatory one, produced an important interruption in the treatment of a deadly cardiovascular disease.
Was this article helpful?
Do You Suffer From High Blood Pressure? Do You Feel Like This Silent Killer Might Be Stalking You? Have you been diagnosed or pre-hypertension and hypertension? Then JOIN THE CROWD Nearly 1 in 3 adults in the United States suffer from High Blood Pressure and only 1 in 3 adults are actually aware that they have it.