Systematic bias in the reporting of angioplasty outcomes: accuracy of visual estimates of absolute lumen diameters.

To examine physician bias in reporting percutaneous transluminal coronary angioplasty (PTCA) results and analyze this potential source of errors, and to examine the ability to estimate absolute lumen diameters visually, the authors reviewed 56 successful PTCAs from their institution. Pre- and postprocedural cineangiograms were blindly reviewed by an experienced consensus panel (three members) and compared with the interventional cardiologist's reported outcome (percentage diameter stenosis) and quantitative coronary angiography (QCA) using the Cardiac Measurement System. Staff cardiologists significantly overestimated pre-PTCA stenosis severity (staff 83.7 versus panel 75.2%, P < 0.05) while underestimating residual narrowing (staff 18.4 versus panel 22.8%, P < 0.05), thus exaggerating overall angioplasty benefit (staff -65.3 versus panel -52.4%, P < 0.05). The cumulative error varied greatly among individual staff members (3.4 to 18.0%). Despite these findings, the consensus panel accurately identified pre-PTCA minimum lumen diameter, as measured by quantitative angiography (panel 0.66 versus QCA 0.67 mm, not significant), although they tended to overestimate absolute postprocedural luminal dimensions (panel 2.28 versus QCA 2.00 mm, P < 0.05) and thereby ultimate changes in minimum lumen diameter (panel 1.62 versus QCA 1.33 mm, P < 0.05). Therefore, substantial bias exists in the reporting of PTCA outcomes, which tends to magnify the perceived benefits of the procedure. Well-trained observers can accurately estimate pre-PTCA absolute lumen diameters, although difficulties occur in evaluating residual dimensions.