The accuracy of a lead questionnaire in predicting elevated pediatric blood lead levels
E. K. France, B. A. Gitterman, P. Melinkovich and R. A. Wright
Department of Community Health Services, Denver Health and Hospitals, Colo, USA.
OBJECTIVES: To determine the prevalence of elevated blood lead levels and
to evaluate the accuracy of a lead screening questionnaire in a western
United States urban inner-city pediatric population. DESIGN: A convenience
sample of children between the ages of 6 months and 6 years seen for a
well-child visit were enrolled. Venous blood lead levels were measured and
a lead screening questionnaire was completed. SETTING: The primary care
clinics of the 10 community health centers of the city and county of
Denver, Colorado. Approximately 85% of children receiving services are
below the 150% poverty level and 54% are insured through the state's
Medicaid program. SUBJECTS: A total of 2978 children seen for a well-child
visit from February 1993 to January 1994. MAIN OUTCOME MEASURES: The
prevalence of elevated blood lead levels and the operating characteristics
of both the Centers for Disease Control and Prevention lead screening
questionnaire and the complete questionnaire used in Denver, using venous
blood lead levels as the criterion standard. RESULTS: The mean blood lead
level was 0.20 mumol/L (4.19 micrograms/dL). Eighty-five children had blood
lead levels of 0.48 mumol/L (10 micrograms/dL), representing 2.9% of the
study group (95% confidence interval [CI], 2.3-3.5). Only 0.3% of the
cohort had blood lead levels greater than 0.96 mumol/L (20 micrograms/dL).
The sensitivity, specificity, and positive predictive value of the Centers
for Disease Control and Prevention questionnaire was 57%, 51%, and 3%,
respectively. The sensitivity, specificity, and positive predictive value
of the complete questionnaire was 59.7%, 36%, and 2.6%, respectively. The
marginal cost of identifying a child with a blood lead level greater than
0.96 mumol/L (20 micrograms/dL) was $4925. CONCLUSIONS: Few of the
low-income children in this study had blood lead levels greater than 0.48
mumol/L (10 micrograms/dL). The questionnaire did little better than chance
at predicting the presence or absence of elevated blood lead levels and
cannot replace a blood lead level test for childhood lead screening in this
community.
