BACKGROUND: The British Society of Gastroenterology (BSG) suggests that a serum ferritin level <or=50 microg/L is still consistent with iron deficiency in the presence of coexistent pathology (inflammation, infection or malignancy), by implication excluding iron deficiency above this level. We aim to examine the validity of this cut-off level in three different groups of patients. METHODS: We used the soluble transferrin receptor/Log(10)ferritin ratio (sTfR-F Index or Index) as a determinant of body iron stores. If the Index was equal or more than 2, the patients were considered iron deficient. Patients were considered iron replete if Index was <or=1. The data was prospectively collected over a period of 3 years. All patients had normocytic anaemia. RESULTS: We collected data for 198 patients. Ninety-three had a sTfR-F Index >or=2 and 17 had Index <or=1. If a ferritin level <or=50 microg/L was used as the cut-off value for iron deficiency, the negative predictive value (NPV) of ferritin test was 22% and the positive predictive value (PPV) 100%; if the level is raised to 100 microg/L the NPV of ferritin test rose to 34.8% and the PPV was 97%. CONCLUSION: Patients with normocytic anaemia who have ferritin levels above 50 microg/L should not automatically be considered to have adequate iron stores. We suggest that the integration of sTfR-F Index in the diagnostic workup of these patients can improve patient care.
BACKGROUND: The British Society of Gastroenterology (BSG) suggests that a serum ferritin level <or=50 microg/L is still consistent with iron deficiency in the presence of coexistent pathology (inflammation, infection or malignancy), by implication excluding iron deficiency above this level. We aim to examine the validity of this cut-off level in three different groups of patients. METHODS: We used the soluble transferrin receptor/Log(10)ferritin ratio (sTfR-F Index or Index) as a determinant of body iron stores. If the Index was equal or more than 2, the patients were considered iron deficient. Patients were considered iron replete if Index was <or=1. The data was prospectively collected over a period of 3 years. All patients had normocytic anaemia. RESULTS: We collected data for 198 patients. Ninety-three had a sTfR-F Index >or=2 and 17 had Index <or=1. If a ferritin level <or=50 microg/L was used as the cut-off value for iron deficiency, the negative predictive value (NPV) of ferritin test was 22% and the positive predictive value (PPV) 100%; if the level is raised to 100 microg/L the NPV of ferritin test rose to 34.8% and the PPV was 97%. CONCLUSION:Patients with normocytic anaemia who have ferritin levels above 50 microg/L should not automatically be considered to have adequate iron stores. We suggest that the integration of sTfR-F Index in the diagnostic workup of these patients can improve patient care.