Elizabeth D. Morrison, MD; David J. Brandhagen, MD; Pradyumna D. Phatak, MD; James C. Barton, MD; Edward L. Krawitt, MD; Hashem B. El-Serag, MD, MPH; Stuart C. Gordon, MD; Mark V. Galan, MD; Bruce Y. Tung, MD; George N. Ioannou, MD, MS; Kris V. Kowdley, MD
*The odds ratio form of Bayes' theorem states that the post-test odds = pretest odds likelihood ratio. The likelihood ratio is the frequency of a finding in patients with a disease divided by its frequency in patients who don't have the disease.
Grant Support: In part by grants DK02957 and DK 38215 from the National Institutes of Health. Dr. Ioannou is a Fellow at the Health Services Research and Development program of the Veterans Affairs Puget Sound Health Care system.
Potential Financial Conflicts of Interest: None disclosed.
Requests for Single Reprints: Kris V. Kowdley, MD, Division of Gastroenterology/Hepatology, University of Washington, 1959 NE Pacific Street, Box 356174, Seattle, WA 98195; e-mail, firstname.lastname@example.org.
Current Author Addresses: Drs. Morrison, Tung, Ioannou, and Kowdley: Division of Gastroenterology/Hepatology, University of Washington, 1959 NE Pacific Street, Seattle, WA 98195.
Dr. Brandhagen: Mayo Clinic, 200 First Street SW, Rochester, MN 55905.
Dr. Phatak: Department of Medicine, Rochester General Hospital, 1425 Portland Avenue, Rochester, NY 14621.
Dr. Barton: G-105, 2022 Brookwood Medical Center Drive, Birmingham, AL 35209.
Dr. Krawitt: Department of Medicine, University of Vermont College of Medicine, Given Building, Burlington, VT 05405.
Dr. El-Serag: Houston Veterans Administration Medical Center (152), 2002 Holcombe Boulevard, Houston, TX 77030.
Drs. Gordon and Galan: William Beaumont Hospital, Division of Gastroenterology and Hepatology, 3601 West 13 Mile Road Royal Oak, MI 48073.
Author Contributions: Conception and design: E.D. Morrison, D.J. Brandhagen, P.D. Phatak, E.L. Krawitt, K.V. Kowdley.
Analysis and interpretation of the data: E.D. Morrison, D.J. Brandhagen, H.B. El-Serag, B.Y. Tung, G.N. Ioannou, K.V. Kowdley.
Drafting of the article: E.D. Morrison, E.L. Krawitt, H.B. El-Serag, K.V. Kowdley.
Critical revision of the article for important intellectual content: E.D. Morrison, D.J. Brandhagen, P.D. Phatak, J.C. Barton, H.B. El-Serag, S.C. Gordon, M.V. Galan, G.N. Ioannou, K.V. Kowdley.
Final approval of the article: D.J. Brandhagen, J.C. Barton, H.B. El-Serag, S.C. Gordon, B.Y. Tung, K.V. Kowdley.
Provision of study materials or patients: E.D. Morrison, D.J. Brandhagen, P.D. Phatak, J.C. Barton, E.L. Krawitt, S.C. Gordon, B.Y. Tung, K.V. Kowdley.
Statistical expertise: J.C. Barton, H.B. El-Serag, G.N. Ioannou.
Obtaining of funding: K.V. Kowdley.
Administrative, technical, or logistic support: K.V. Kowdley.
Collection and assembly of data: E.D. Morrison, D.J. Brandhagen, P.D. Phatak, M.V. Galan, K.V. Kowdley.
DNA-based HFE gene testing can confirm hereditary hemochromatosis in most people of Northern European descent. However, liver biopsy is important to detect cirrhosis.
To develop noninvasive criteria to predict the presence or absence of advanced hepatic fibrosis or cirrhosis in Americans with hemochromatosis.
Six tertiary care referral clinics.
182 patients with phenotypically defined hemochromatosis.
Liver histopathology and serum ferritin, aspartate aminotransferase, and alanine aminotransferase levels. Multivariate logistic regression analysis was used to examine factors associated with cirrhosis (defined as bridging fibrosis or unequivocal cirrhosis on biopsy).
Cirrhosis was present in 40 of 182 (22%) patients in the overall group and in 35 of 147 (24%) of C282Y homozygotes. Only 1 of 93 patients with a serum ferritin level less than 1000 g/L had cirrhosis compared with 39 of 89 patients with serum ferritin levels greater than 1000 g/L (P < 0.001). No C282Y homozygotes or C282Y/H63D compound heterozygotes with serum ferritin levels less than 1000 g/L had cirrhosis. Elevated serum aminotransferase levels (P = 0.001) and serum ferritin levels greater than 1000 g/L (P = 0.001), but not age older than 40 years (P = 0.2), were independently associated with cirrhosis. In a multivariate model, the probability of cirrhosis was 7.4% among patients with serum ferritin levels less than 1000 g/L compared with 72% among patients with serum ferritin levels greater than 1000 g/L after adjustment for age and elevated serum liver enzyme levels.
Patients with hemochromatosis and serum ferritin levels less than 1000 g/L are unlikely to have cirrhosis. Liver biopsy to screen for cirrhosis may be unnecessary in such patients, regardless of age or serum liver enzyme levels.
Morrison ED, Brandhagen DJ, Phatak PD, Barton JC, Krawitt EL, El-Serag HB, et al. Serum Ferritin Level Predicts Advanced Hepatic Fibrosis among U.S. Patients with Phenotypic Hemochromatosis. Ann Intern Med. ;138:627–633. doi: 10.7326/0003-4819-138-8-200304150-00008
Download citation file:
Published: Ann Intern Med. 2003;138(8):627-633.
Gastroenterology/Hepatology, Liver Disease.
Results provided by:
Copyright © 2018 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use