University of Michigan
Department of Pediatrics
Evidence-Based Pediatrics Web Site

Question

• In infants and toddlers, does a hemoglobin level, compared to complete iron studies, adequately screen for iron deficiency?

Clinical Bottom Lines

1. Most anemic toddlers are not iron deficient, and most iron deficient toddlers in the U.S. are not anemic.
2. Screening for iron deficiency cannot be adequately performed by obtaining only a hemoglobin level, as indicated by its sensitivity of 30%. The test has a positive likelihood ratio of 3.7, and in a selected population, the predictive value was 28%.

Summary of Key Evidence

1. The study was based on NHANES III data (1988-1994), in the form of a validation cohort study.1
2. 1289 children, 12-35 months old, were analyzed, all of whom had CBCs, ferritin, transferrin saturation, and free erythrocyte protoporphyrin (EPP) levels available to evaluate for anemia and the presence of iron deficiency.
3. Anemia was defined as Hgb <11, and iron deficiency was diagnosed if at least two of the following were present: ferritin <10 g/L, transferrin saturation <10%, or EPP >1.42 mol/L.
4. Subjects were stratified according to ethnicity and economic status.
5. Prevalence of all anemias was found to be 9%, prevalence of iron deficiency was 9%, and prevalence of iron deficiency anemia was 3%.
6. Children with iron deficiency were more likely to have a statistically lower Hgb, were more likely to be below the poverty line, and were more likely to have Mexican-American parents (all p-values <0.02).
7. Sensitivity of Hgb <11 for iron deficiency was 30% and the positive predictive value was 28%. Negative predictive value was 92%, and specificity was 91%. Likelihood ratio of a positive test was 3.7.
8. Positive predictive value increased as the threshold of a diagnosis of anemia was decreased. PPV was 38% when the cutoff for a diagnosis of anemia was Hgb <10.7.
9. Most subsets involving ethnic and economic factors did not influence the sensitivity or PPV.

Additional Comments

• The goal of this study was to assess the adequacy of hemoglobin screening for iron deficiency, a condition which, if left untreated, could result in poor cognitive outcomes.
• The study stresses that by using a hemoglobin screen, many actual iron deficient children will be missed, and many anemic children will be treated inappropriately.
• It was unclear whether all children in this age group evaluated by the NHANES study had this same complete laboratory investigation, or if this was just a subset of a much larger group.
• The article did not indicate if those performing the study were unaware of the patients' iron deficiency status when analyzing the hemoglobin levels; therefore, diagnostic uncertainty was potentially not experienced.
• Other studies confirm the findings in this study; however, most of the others do not use the same criteria for iron deficiency as is stipulated in this analysis.
• The study also stresses primary prevention of iron deficiency with fortified formula and cereal; however, this is only applicable mainly to the developed world.
• Treatment for iron deficiency is fairly benign and can possibly prevent an untoward outcome, but screening also involves lab draws and possible misdiagnosis, both of which are unsettling to children and parents.
• A study by Ullrich et al. recently indicated that screening for iron deficiency with reticulocyte hemoglobin content (CHr) is more effective than Hgb; however, the study was much smaller (202 patients), and was an initial study examining levels at which CHr predicts iron deficiency.2 Further evaluation of this is warranted.
  

Citation

1. White KC. Anemia is a poor predictor of iron deficiency among toddlers in the United States: for heme the bell tolls. Pediatrics 2005;115:315-20.
2. Ullrich C, Wu A, Armsby C, et al. Screening healthy infants for iron deficiency using reticulocyte hemoglobin content. JAMA 2005;294:924-30.

CAT Author: Jeff Knipstein, MD

CAT Appraisers: Robert Schumacher, MD

Date appraised: September 28, 2005