Agreement in pathologic interpretation of liver biopsy specimens in posttransplant hepatitis C infection

Hepatitis C virus-related disease is rapidly becoming the most common indic ation for orthotopic liver transplant (OLT) in the United States. Although post-OLT hepatitis C viremia is universal, 40% to 60% of patients develop r ecurrent chronic hepatitis C. Distinguishing recurrent chronic hepatitis C infection from acute rejection may be difficult because of overlapping hist opathologic features. To improve our diagnostic accuracy we undertook a stu dy to determine interobserver and intraobserver agreement between pathologi sts examining post-OLT liver biopsy specimens in patients from our transpla nt database. Clinical data and microscopic sections from 26 patients with h epatitis C virus-related OLT were reviewed. Biopsy specimens were obtained because of abnormal liver enzymes (21/26) or routine post-OLT follow-up (5/ 26), representing both early (18 +/- 11 days) and late (252 +/- 206 days) p ost-OLT periods. Unidentified sections were examined by an experienced path ologist in a randomly assigned order and reexamined 6 weeks later in the sa me fashion by the initial reviewer and a second experienced pathologist. In terobserver and intraobserver agreement was calculated using kappa statisti c. The intraobserver agreement was 81% with a kappa coefficient of 0.67 (P = .001). The interobserver agreement was 78% with a kappa coefficient of 0. 60 (P < .001). The early post-OLT biopsy specimens (18 +/- 11 days) were th e most difficult to interpret.