PEDIATRICS Vol. 108 No. 3 September 2001, pp. 823

Screening for Iron Deficiency by Dietary History in a High-Risk Population

To the Editor.

We appreciate the soundness and validity of the article by Bogen et al¹ and their interest in including a dietary survey in the context of well-child care. We submit the following comments to further stimulate research and discussion into the complex matter of nutritional assessment.

Our study,² largely quoted in this article, was based on a retrospective review of patient records. The data of interest came from a brief nutritional questionnaire that was part of a broader systematic review of health and psychosocial issues incorporated into all of the health supervision visits of the lead author (M.B.). Bogen's article, in contrast, was constructed as a prospective study in which the dietary data were obtained by a research assistant from patients recruited solely for the purpose of the study. Thus, there were significant differences in the nature of the sample and in the manner in which the data were collected.

The outcomes of the 2 studies diverge most markedly in the proportion of parents who responded positively to the screening questionnaire. This difference would account for the marked difference in specificity (15% vs our 79%), as many more truly noniron-deficient patients screened positive in Bogen's study.

Nutritional histories are known to be highly unreliable,^3-6 possibly because nutrition and nurture are closely related and emotionally loaded. The dramatic difference in specificity may be attributed to the manner and context of the interview. It is also possible that a research format created a self-selection bias weighed towards high-reporting parents in Bogen's study.

Many practices make routine use of standard history-taking questionnaires on various routine as well as sensitive issues. Additional research, comparing history-taking methodology, and better defining screening questions may be worthwhile to answer the following question: How do clinical interviews by primary care physicians compare with written questionnaires obtained by research assistants outside of the physician-patient relationship?

We support the recommendation that pediatric primary care should emphasize nutritional history and promote consumption of food items rich in readily-bioavailable iron sources as well as in other micronutrients. This is simply good, common-sense practice. More clinical practice-based research is needed to determine whether structured history-taking can reduce the need for universal hematologic screening for iron deficiency.

Mireille Boutry, MD
Robert Needlman, MD
Division of General Academic Pediatrics
Rainbow babies and Childrens Hospital
Mather House Mail Stop 6019
Cleveland, OH 44106

REFERENCES

1. Bogen DL, Duggan AK, Dorer GJ, Wilson MH Screening for iron deficiency anemia by dietary history in a high-risk population. Pediatrics. 2000; 105:1254-1259 [Abstract/Free Full Text]
2. Boutry M, Needlman R Use of diet history in the screening of iron deficiency anemia. Pediatrics. 1996; 98:1138-1142 [Abstract/Free Full Text]
3. Yip R. Pediatric surveillance systemUS, 1980-1990. MMWR Morb Mortal Wkly Rep. 1992
4. Kristal AR, Beresford SAA, Lazovich DA Assessing change in diet-assessment research. Am J Clin Nutr. 1994; 59:185S-189S [Abstract/Free Full Text]
5. Willett WC, Reynolds RD, Cottrell-Hoener S, Browne MI Validation of a semi-quantitative food-frequency questionnaire comparison with a 1-year diet research. J Am Diet Assoc. 1987; 87:43-47 [Medline]
6. Armstrong H Nutritional assessment of black pre-school children in Mississipi. J Am Diet Assoc. 1975; 66:488-493 [Medline]

In Reply.

We would like to thank Drs Boutry and Needlman for their comments, in particular for raising the issue of the possible differences between interviews conducted by physicians with their patients, interviews conducted by research assistants, and patient- completed surveys. For our study, we selected the survey format because we had hoped that a standard questionnaire could be used in either doctors' offices or WIC clinics to identify high-risk children in need of blood testing for iron-deficiency anemia. There are both limitations and advantages to the use of surveys as opposed to interviews in the primary care setting, and these vary by the nature of the questions. We are unaware of any published studies evaluating the difference between these methods for health screening.

The difference in specificity for the two studies could be attributable to selection bias, as suggested. Another possible explanation is that we did not exclude children with recent infection, premature birth, failure-to-thrive, and elevated lead, as was done in the Boutry and Needlman study. In addition, 25% of our sample had not received WIC in the past 12 months. In an attempt to mimic the broad use of the questionnaire in the community, we approached all children whom the primary care physician was going to screen for anemia at the well child visit.

We agree that there is a need for future work in the area of office-based structured assessment of nutritional risk and promotion of healthy diets for children.

Debra Bogen, MD
Division of General Academic Pediatrics
Children's Hospital of Pittsburgh
Pittsburg, PA 15213

Anne Duggan, ScD
George Dover, MD
Department of Pediatrics
Johns Hopkins University School of Medicine
Baltimore, MD 21205

Modena Wilson, MD, MPH
American Academy of Pediatrics
Elk Grove Village, IL 60007