Usefulness, usability, and quality criteria for noninvasive methods in cardiovascular clinical pharmacology

Validation of study methods is a prerequisite for their usability. Empirical quality criteria based on test-theoretical principles are useful for this purpose. These criteria are discussed for several noninvasive methods used in cardiovascular clinical pharmacology: electrocardiography, systolic time intervals, blood pressure, and estimates of stroke volume. Several of these methods are highly reliable and sensitive to drug effects. However, they are inherently low in validity because they are based on oversimplified algorithms and yield estimates rather than measurements of function and changes therein. These estimates are method-specific and are likely to be subject to method*subject*effect interaction. Highlighting these constraints and the limitations of these methods need not preclude their useful contribution to the early evaluation of drug action in humans.