The purpose of this study was to determine intraobserver and interobserver variations in scoring of liver biopsies of patients with chronic hepatitis in order to Knodell Scoring System. Five pathologists and two residents were contributed to this study. 80 patients which diagnosed as chronic active hepatitis in order to the old classification were included in the study. The slides which stained with hematoxylen&eosin, Masson's trichrome, Van Gieson, Reticulin, periodic acid Schiff with and without diastase were reviewed. Piece-meal necrosis, portal inflammation, lobular inflammation and fibrosis were scored in order to Knodell Scoring System. . These cases were reevaluated by the same observers a month later. Concordance and discordance between the evaluation of different observers and the evaluation of each observer in the first and second time were statistically calculated.

Consistency was moderate among pathologists with respect to the evaluation of piece-meal necrosis (r=0.52-0.63). Correlation was strong among observers from the same center while it was poor among observers from different centers in the evaluation of portal inflammation (r=0.40-0.51 ). Only two pathologists from the same center showed good concordance with respect to the evaluation of lobular inflammation. The strongest consistency was found in the evaluation of fibrosis in this study (r=0.53-0.84 in pathologists and r=0.49-0.59 in residents. Consistency between the first and the second evaluation of each observer (intraobserver) was also strong (r=0.62-0.92).

Intraobserver and interobserver variations in chronic hepatitis. The purpose of this study was to determine intraobserver and interobserver variations in scoring of liver biopsies of patients with chronic hepatitis in order to Knodell Scoring System. Five pathologists and two residents were contributed to this study. 80 patients which diagnosed as chronic active hepatitis in order to the old classification were included in the study. The slides which stained with hematoxylen&eosin, Masson's trichrome, Van Gieson, Reticulin, periodic acid Schiff with and without diastase were reviewed. Piece-meal necrosis, portal inflammation, lobular inflammation and fibrosis were scored in order to Knodell Scoring System. . These cases were reevaluated by the same observers a month later. Concordance and discordance between the evaluation of different observers and the evaluation of each observer in the first and second time were statistically calculated.

Consistency was moderate among pathologists with respect to the evaluation of piece-meal necrosis (r=0.52-0.63). Correlation was strong among observers from the same center while it was poor among observers from different centers in the evaluation of portal inflammation (r=0.40-0.51 ). Only two pathologists from the same center showed good concordance with respect to the evaluation of lobular inflammation. The strongest consistency was found in the evaluation of fibrosis in this study (r=0.53-0.84 in pathologists and r=0.49-0.59 in residents. Consistency between the first and the second evaluation of each observer (intraobserver) was also strong (r=0.62-0.92).