Surveillance for tuberculosis (TB) in children can be conducted with either universal or targeted screening. Depending on epidemiologic risk, screening practices have varied between these two strategies. Current recommendations, coinciding with a decreased prevalence of TB, call for targeted screening of high-risk children, but these risk assessment questions have not been formally tested. Ozuah and colleagues conducted this prospective study to determine the sensitivity, specificity and predictive validity of the New York City Department of Health (NYCDOH) TB screening questionnaire.

At the time of this study, the TB case rate for the population in question (children living in the South Bronx, New York City) was 38.6 per 100,000. The only exclusion criterion was a previous positive purified protein derivative (PPD) test result. At health maintenance visits, children and their caretakers were asked the following NYCDOH risk assessment questions: (1) has your child had any contact with a person who has TB; (2) was any household member, including your child, born in or traveling in areas where TB is common (e.g., Africa, Asia, Latin America or the Caribbean); (3) does your child have regular (e.g., daily) contact with adults who are at high risk for TB (i.e., the homeless, incarcerated persons or illicit drug users, or those who have human immunodeficiency virus (HIV) infection); and (4) does your child have HIV infection.

Any “yes” response was considered a positive risk factor for TB. All children received a Mantoux tuberculin test. The clinic nurse or physician read test results between 48 and 72 hours later. Self-report and parental reports were not accepted. A PPD test was considered positive if there was induration of greater than 10 mm. All children who tested positive were given a chest radiograph.

There were 3,093 children eligible, of whom 2,920 were available for analysis. The NYCDOH screening questionnaire identified 14 percent of the children as being at high risk for TB. Of these children, 5.6 percent (n = 23) had positive PPD tests. Of the low-risk children, 0.16 percent (n = 4) had positive PPD tests; 75 percent of these children were older than 11 years. None of the children had evidence of active disease on chest radiograph. Children having a positive screen (“yes” to any one of the above questions) were 35 times more likely to have an induration greater than 10 mm than children with a negative screen (odds ratio: 35.2).

The authors conclude that the NYCDOH screening questionnaire is able to identify children who are at much higher risk of having a positive PPD test, especially if the child is older than 11 years. In a given population, it will be necessary to be aware of the prevalence of positive skin tests and of TB, but targeted screening may be appropriate. This strategy is less costly and more efficient than universal screening.