Inability of caliper measurement to enhance observer agreement in the interpretation of coronary cineangiograms.

Important clinical decisions are made on the basis of the interpretation of the degree of stenosis in the coronary cineangiogram. Thus, minimization of observer variation (both intra and inter) is crucial. The purpose of this investigation was to determine if measurement of stenotic lesions by caliper (CM) as compared with the usual eyeball method (EM) would reduce observer variation. Five observers, using the AHA scoring system, estimated maximal luminal narrowing in 14 segments on 10 films providing 2 reports on each film by EM and 2 by CM, thus making observations on 2800 segments. This allowed analysis of intraobserver variation for each reporter for method EM and CM, and for interobserver variation by each method. Since chance agreement is an important component in an observational study, a correction for this was made using the Kappa statistic. The results revealed that the mean Kappa by EM was 0.71 compared to 0.63 by CM (1 being perfect agreement and 0 no agreement). Agreement was much greater in proximal rather than distal segments. Interobserver agreement was consistently higher by EM K = 0.62 than CM K = 0.53 (p = 0.0007). A delineation of the types of disagreement and their clinical significance are important considerations in the evaluation of observer variation. It is concluded that the use of a caliper does not enhance either intra or interobserver agreement.