The Limitations of Transferring Evidence from Clinical Trials to the Care of Individual Patients

Abstract

The randomised clinical trial has, justifiably, become the premier research technique by which the efficacy and, to a lesser extent, safety of hypertensive therapy are established. Through blinding and randomisation, the influence of potentially confounding factors is minimised. In these circumstances, a comparison of outcomes with an identical control group permits an assessment of the impact of the intervention. Useful as this strategy has been in establishing a scientific basis for blood pressure reduction, the process has limitations — particularly in the comparison of different antihypertensive drugs. It may not always be appropriate to apply the result of a clinical trial to the individual patient. Since most participants in these trials do not experience an endpoint, the effect of the intervention probably will not be experienced by most patients who fit the entry criteria of the study. Moreover, the risk of cardiovascular events is unevenly distributed among trial participants even though they have similar blood pressures and, thus, the potential for actual benefit likewise varies depending upon absolute cardiovascular risk. Finally, unrecognised mechanistic differences may produce different results, particularly when comparing mechanistically different drugs, among patients heterogeneous in terms of pathophysiology of blood pressure control. For these reasons, and others, it is likely that the net results of a trial mask substantial outcome differences hidden within the overall study group. In short, the average result may not apply to many patients, even when they meet the criteria for participation in the study. This is not an argument against clinical trials but rather a recognition that clinical practice must not be guided exclusively by epidemiological data but, in addition, by all the information that derives from pathophysiology and pharmacology.