[Quality assurance in invasive cardiology. A prospective study for the evaluation of indications for coronary angiography and coronary dilatation using the "Rand Corporation" method].

Objective: To assess the indications for coronary angiography and percutaneous transluminal coronary angioplasty (PTCA) according to the criteria of the RAND Corporation's expert panel ratings; to compare the results with those already published in the literature; and to examine the method with respect to its appropriateness as a measure of quality control.

Patients And Methods: The parameters necessary for rating according to the RAND Corporation's published criteria were prospectively obtained in 116 consecutive patients (89 men, 27 women; mean age 59.4 +/- 10.7 years) undergoing coronary angiography and 138 patients (112 men, 26 women; mean age 61.5 +/- 9.4 years) undergoing PTCA.

Results: For coronary angiography the >>inappropriate rate << was 22.4%, for >> uncertain indications << 15.5%, >> appropriate indications << 24.1%, and for >> necessary indications << 37.9%. 38.6% of coronary angiographies, performed in the course of angiographic control after PTCA, were >> inappropriate <<, but only 12.5% of other coronary angiographies (P < 0.01). In 48.6% of >> appropriate << or >> necessary << procedures invasive treatment followed, compared with 19.2% of those rated >> inappropriate << (P < 0.05). In 10.1% of patients the indications for PTCA were judged >> inappropriate <<, in 32.6% as >> uncertain <<, in 13.0% as >> appropriate << and 44.2% as >> necessary <<. There was no correlation between ratings and the acute results of PTCA. These findings pertaining to indications for coronary angiography and PTCA correspond to those reported in the literature.

Conclusions: As the RAND criteria take inadequate account of individual peculiarities, they are not suitable for individual clinical decisions. But they are useful as screening method in a quality control project, because procedures for which indications have not been adequately proven are singled out by an unfavourable rating and can thus be thoroughly analysed in the individual case.