Risk factors for positive mantoux tuberculin skin tests in children in SanDiego, California: Evidence for boosting and possible foodborne transmission

Objectives. Source case finding in San Diego, California, rarely detects th e source for children with tuberculosis (TB) infection or disease. One thir d of all pediatric TB isolates in San Diego are Mycobacterium bovis, a stra in associated with raw dairy products. This study was conducted to determin e risk factors for TB infection in San Diego. Design. Case-control study of children less than or equal to5 years old scr eened for TB as part of routine health care visit. Asymptomatic children wi th a positive (greater than or equal to 10 mm) Mantoux skin test (TST) were matched by age to 1 to 2 children with negative TST from the same clinic. We assessed risk factors for TB infection through parental interview and ch art review. Results. A total of 62 cases and 97 controls were enrolled. Eleven cases an d 25 controls were excluded from analysis because of previous positive skin tests. Compared with controls, cases were more likely to have received BCG vaccine (73% vs 7%, odds ratio [OR] 44), to be foreign born (35% vs 11%, O R 4.3), and to have eaten raw milk or cheese (21% vs 8%, OR 3.76). The medi an time between the most recent previous TST and the current test was 12 mo nths for cases and 25 months for controls. Other factors associated with a positive TST included foreign travel, staying in a home while out of the co untry, and having a relative with a positive TST. There was no association between contact with a known TB case. In a multivariable model, receipt of BCG, contact with a relative with a positive TST, and having a previous TST within the past year were independently associated with TB infection. Conclusions. We identified several new or reemerging associations with posi tive TST including cross border travel, staying in a foreign home, and eati ng raw dairy products. The strong associations with BCG receipt and more re cent previous TST may represent falsely positive reactions, booster phenome na, or may be markers for a population that is truly at greater risk for TB infection. Unlike studies conducted in nonborder areas, we found no associ ation between positive TB skin tests and contact with a TB case or a foreig n visitor. Efforts to control pediatric TB in San Diego need to address loc al risk factors including consumption of unpasteurized dairy products and c ross-border travel. The interpretation of a positive TST in a young child i n San Diego who has received BCG is problematic.