Screening for tuberculosis infection in urban children
C. Christy, M. L. Pulcino, B. P. Lanphear and K. M. McConnochie
Department of Pediatrics, University of Rochester School of Medicine and Dentistry, NY, USA.
OBJECTIVES: To determine the proportion of children who are at high risk
for tuberculosis (TB) as defined by the American Academy of Pediatrics
(AAP) criteria, the rate of compliance with visits for tuberculin skin test
(TST) interpretation, and the prevalence of TB infection. DESIGN: A
cross-sectional study of 401 children, 12 months to 18 years of age, who
attended a hospital-based, urban pediatric clinic for well-child visits was
undertaken from April 13, 1994, through August 30, 1994. Respondents
completed a self-administered questionnaire, an intradermal TST was
applied, and an appointment was scheduled for skin test interpretation in
48 to 72 hours. SETTING: Hospital-based, pediatric primary care center in
Rochester, NY, serving children of low to moderate income (67% were
receiving Medicaid). RESULTS: Of the 401 children, 342 (85%) had at least 1
risk factor for TB identified: 96 (24%) reported contact with persons who
were considered to be at high risk for TB; 170 (42%) had at least 1 parent
who was born in a high prevalence country; and 269 (67%) reported a
household income of less than $15,500. Of the 401 children, 300 returned
for TST interpretation, 257 (64%) by 48 to 72 hours and an additional 43
(11%) by 96 hours. Four (1.3%) of the 300 children had a positive TST (ie,
induration > or = 10 mm). All 4 of the children who were TST positive
had at least 1 parent from a high-risk country and were identified using
AAP-defined risk criteria. The mean age of children who were TST positive
was 15.3 years (range 13-17 years) compared with 8.1 years for those who
were TST negative (P < .01). The positive predictive value of the
questionnaire, which included income as a risk factor for TB, was only 1.5
(95% confidence interval = 0.5-4.0); when household income was not
considered a risk factor, the positive predictive value was 2.0 (95%
confidence interval = 0.7, 5.5). The estimated cost per child who was TST
positive ranged from $430 for those who had contact with an incarcerated
adult to $855 per child who was TST positive identified by using
AAP-defined criteria. CONCLUSIONS: The overall sensitivity of the
AAP-defined criteria and having at least 1 parent from a TB-endemic country
were high. However, because of the low prevalence of TB infection, the
positive predictive value of these criteria was very low. These data
support AAP recommendations only to skin test children who are at high risk
for TB, but they also suggest that annual testing may not be cost-effective
for many communities in the United States.
