There is no generally agreeable – objective – way to define the scope of clinical research in the general framework of medical research, commonly taken to encompass ‘basic’ and ‘applied’ medical research. The prevailing textbook-definitions of clinical research are, accordingly, highly variable; and they are in various other ways untenable besides [16].

But just as in the framework of epidemiological research, it is possible to identify the most important, quintessentially ‘applied’ genre of clinical research. This is research to advance the knowledge-base of clinical medicine, of scientific clinical medicine. It thus is research to develop the general (abstract-general) knowledge-base for setting gnostic – dia-, etio-, and prognostic – probabilities. Needed to this end is research on gnostic probability functions [16].

Preeminent in the needed research is study of diagnostic probabilities as functions (descriptive) of diagnostic indicators, and of prognostic probabilities as functions of prognostic indicators (descriptively) jointly with choice of intervention (causally) and prognostic time. Etiognostic clinical research is less eminent a topic, with probability functions for iatrogenesis of illness (or mere sickness) deserving the highest priority in this. Remarkably, advancement of the knowledge-base of clinical etiognosis (about iatrogenesis) has not yet emerged as the principal mission in epidemiologists’ ‘pharmaco-epidemiology.’

The needed diagnostic research is challenging because of the need to ascertain, in each of the person-moments in the study series/base, the truth about the presence/absence of the illness at issue. The prognostic research, by contrast, can be based on data now routinely collected in randomized trials on interventions [9, 16].