OBJECTIVE: To evaluate the predictive value of a set of laboratory markers for the assessment of liver fibrosis in chronic viral hepatitis patients. STUDY DESIGN: Cross-sectional study. PLACE AND DURATION OF STUDY: Baqai Medical University, Combined Military Hospital, Malir, Karachi, from November 2006 to May 2008. METHODOLOGY: Twenty laboratory parameters were measured in 100 treatment-naïve chronic viral hepatitis patients who also had liver biopsy performed. Descriptive statistics, areas under the ROC's curves, and multivariate logistic regression analysis identified a fibrosis panel, a set of five most useful markers, for the assessment of stages of fibrosis, stage 0 to stage 4. The fibrosis index, FibroScore, consisted of bilirubin, Gamma glutamyl transferase, Hyaluronic acid, alpha 2 macroglobulin, and platelets evaluation. RESULTS: A score of ≥ 0.5 predicted stages 2, 3 and 4, with a sensitivity of 82%, and specificity of 92%. A score ≥ 0.5 for stages 3 and 4 had a sensitivity of 85%, and specificity of 89%. At a score of > 0.80, for stages 3 and 4, the sensitivity was 70%, specificity was 97%, and PPV 87% (there was ≥ 85% possibility of presence of stage 3 or 4). A score of ≤ 0.20 predicted the absence of stages 2, 3, and 4 with a sensitivity of 91%, specificity of 86%, and NPV of 96%. Scores from 0.00 to 0.10 almost certainly ruled out the presence of stages 2-4 (NPV=98%). The areas under the ROC curve were: 0.808 for stage 2; 0.938 for stage 3; and 0.959 for stage 4. CONCLUSION: A combination of 5 markers is very useful in predicting various stages of liver fibrosis, and is helpful in the non-invasive assessment of liver fibrosis in chronic viral hepatitis patients.
OBJECTIVE: To evaluate the predictive value of a set of laboratory markers for the assessment of liver fibrosis in chronic viral hepatitispatients. STUDY DESIGN: Cross-sectional study. PLACE AND DURATION OF STUDY: Baqai Medical University, Combined Military Hospital, Malir, Karachi, from November 2006 to May 2008. METHODOLOGY: Twenty laboratory parameters were measured in 100 treatment-naïve chronic viral hepatitispatients who also had liver biopsy performed. Descriptive statistics, areas under the ROC's curves, and multivariate logistic regression analysis identified a fibrosis panel, a set of five most useful markers, for the assessment of stages of fibrosis, stage 0 to stage 4. The fibrosis index, FibroScore, consisted of bilirubin, Gamma glutamyl transferase, Hyaluronic acid, alpha 2 macroglobulin, and platelets evaluation. RESULTS: A score of ≥ 0.5 predicted stages 2, 3 and 4, with a sensitivity of 82%, and specificity of 92%. A score ≥ 0.5 for stages 3 and 4 had a sensitivity of 85%, and specificity of 89%. At a score of > 0.80, for stages 3 and 4, the sensitivity was 70%, specificity was 97%, and PPV 87% (there was ≥ 85% possibility of presence of stage 3 or 4). A score of ≤ 0.20 predicted the absence of stages 2, 3, and 4 with a sensitivity of 91%, specificity of 86%, and NPV of 96%. Scores from 0.00 to 0.10 almost certainly ruled out the presence of stages 2-4 (NPV=98%). The areas under the ROC curve were: 0.808 for stage 2; 0.938 for stage 3; and 0.959 for stage 4. CONCLUSION: A combination of 5 markers is very useful in predicting various stages of liver fibrosis, and is helpful in the non-invasive assessment of liver fibrosis in chronic viral hepatitispatients.