Background and aims: The aim of this study was to noninvasively assess the severity of chronic hepatitis C virus (HCV) in large patient populations. It would be helpful if fibrosis scores could be calculated solely on the basis of data contained in the patients' electronic medical records (EMR). We performed a pilot study to identify all HCV-infected patients in a large health care system, and predict their fibrosis stage on the basis of demographic and laboratory data using common data from their EMR.
Materials and methods: HCV-infected patients were identified using the EMR. The liver biopsies of 191 HCV patients were graded using the Ishak and Metavir scoring systems. Demographic and laboratory data were extracted from the EMR and used to calculate the aminotransferase to platelet ratio index, Fib-4, Fibrosis Index, Forns, Göteborg University Cirrhosis Index, Lok Index, and Vira-HepC.
Results: In total, 869 HCV-infected patients were identified from a population of over 1 million. In the subgroup of patients with liver biopsies, all 7 algorithms were significantly correlated with the fibrosis stage. The degree of correlation was moderate, with correlation coefficients ranging from 0.22 to 0.60. For the detection of advanced fibrosis (Metavir 3 or 4), the areas under the receiver operating characteristic curve ranged from 0.71 to 0.84, with no significant differences between the individual scores. Sensitivities, specificities, and positive and negative predictive values were within the previously reported range. All scores tended to perform better for higher fibrosis stages.
Conclusions: Our study demonstrates that HCV-infected patients can be identified and their fibrosis staged using commonly available EMR-based algorithms.