["Soft" and "hard" end points in the example of hypertension].
Hypertension is a disease with a long latency. Those afflicted will, if untreated, die earlier than an age-matched cohort within the general population. In a first phase of intervention trials it was demonstrated that anti-hypertensive treatment, in comparison with placebo, results in a reduction of so-called hard end points (death, stroke, myocardial infarction). In a second phase of clinical trials, which is still continuing, active treatment strategies are being compared. The problem in this phase is that differences in these hard end points are smaller than in placebo-controlled trials. For this reason signs of organ damage that occur in the time between the diagnosis of high blood pressure and the occurrence of hard end points (e.g. left ventricular hypertrophy, microalbuminuria, intima/media thickness) are used as additional end points. According to epidemiological data, these intermediate or surrogate end points are more or less predictive of hard end points. Systematic inclusion of such surrogates in large trials has been tried for some years in an attempt to raise the predictive power of these parameters. Another patient-relevant endpoint of antihypertensive treatment is the new occurrence of diabetes mellitus.