ISSN:
1433-8726
Source:
Springer Online Journal Archives 1860-2000
Topics:
Medicine
Notes:
Summary In order to investigate the reproducibility of grading systems for prostatic carcinoma currently in use, a comparative histological grading study was done. These studies were carried out on tissue sections from radical prostatectomy specimens (N=50) stained with hematoxylin and eosin. Five pathologists with varying professional experience participated in the study, using five different grading systems: those of Broders, Brawn, Gleason (for statistical compilation the modified version), Mostofi, and a modified Mostofi grading method recently described by Schroeder and Mostofi. Weighted kappa coefficients ranged from 0.21 to 0.52. None of the systems investigated demonstrated a high degree of reproducibility (k〉0.70). Reproducibility of the systems described by Broders and Brawn was reasonably good (k=0.52 and 0.41, respectively). With the modified Gleason method (rearrangement of Gleason scores into 3 grades), a considerable difference was noted between the numerical agreement score (among at least 3 observers) and the measured kappa value (100% and 0.30, respectively). The methods described by Mostofi and Schroeder-Mostofi revealed only limited reproducibility (k=0.21 and 0.34, respectively).
Type of Medium:
Electronic Resource
URL:
http://dx.doi.org/10.1007/BF00327011
Permalink