Date of Award

Spring 5-2019

Degree Name

Master of Public Health (MPH)


Lu-Yu Hwang, MD

Second Advisor

Jeffrey Starke, MD


Background More than 300,000 refugees arrived in the United States (U.S.) from 2010-2015, and Texas accepts the 2nd highest number of refugees. Texas also accepts large numbers of asylees, parolees, and special immigrant visa holders. Additionally, a large proportion of trafficked persons in the U.S. live in or pass through Texas. Foreign-born children are disproportionately affected by tuberculosis (TB) and account for two-thirds of U.S. childhood TB cases. Children are at greater risk for progression from TB infection to disease and experience greater morbidity and mortality from TB disease. This makes screening for and treatment of TB infection in children from high-prevalence areas an important public health intervention. Since 2007, children 2-14 years old emigrating from high-prevalence countries (TB incidence >20 cases /100,000 persons) have been tested for TB infection. Children ≥15 years old are additionally screened with a chest radiograph. The Centers for Disease Control and Prevention (CDC) recommends treatment of children with TB infection, as treatment reduces the risk of life threatening disease and prevents future transmission. There are few studies describing the epidemiology of TB in internationally displaced children relocating to the U.S.; there have been no studies centered on Texas. We describe the secular trends and comparative epidemiology of positive TSTs and IGRAs in children of different immigration statuses cared for through the Houston-area public health program. Methods This was a retrospective cross-sectional study of children <18 years-old evaluated by the Harris County Public Health Refugee Health Screening Program between January 1st, 2010 and December 31st, 2015 with the following immigration statuses: refugee, asylee, parolee, special immigrant visa holder, or victim of human trafficking. We analyzed factors associated with TB test positivity, infection and disease for children with these immigration statuses. Data are from the U.S. Committee for Refugees and Immigrants (USCRI), the Harris County Public Health Refugee Health Screening Program, and the Texas Children's Hospital TB clinic. Chi-square test or Fisher’s exact test were used for dichotomous variables, one-way ANOVAs for univariate analyses, and Wilcoxon rank sum or Kruskal-Wallis for continuous variables. Multivariate logistic regression was performed to further analyze factors associated with TB test positivity. To assess secular trends in usage and positivity, monthly totals were analyzed using linear regression and the Wilcoxon Sign Rank test. A p-value <0.05 was considered significant. Children < 5 years were typically tested using tuberculin skin test (TSTs) and older children typically using interferon-gamma release assays (IGRAs). The primary outcome was a positive test of TB infection (TST and/or IGRA). Children who were TST+/IGRA- with no known contacts, a normal chest radiograph and no signs or symptoms consistent with TB disease were typically considered uninfected. However, there were some children who met these criteria who were considered to have TB infection and were treated as such –this typically occurred earlier in the study period, and was most commonly due to young age (< 2 years old) or some variability in provider practice. Results The program evaluated 5,990 children, of whom 5870 (98%) were tested, predominantly (64%) with an IGRA alone. During the study period, IGRA use increased (p<0.001), though percentages of positive test results (IGRA or TST) did not decline significantly (p=0.10). Overall, 364 (6.2%) children had at least one positive test of infection: 143/1,842 (7.8%) tested with TST alone, 129/3,730 (63.6%) tested with IGRA alone, and 92/298 (30.9%) had at least one positive test result for those tested with both TST and IGRA. Among the 364 children with any positive test of infection, 4 (1.1%) were diagnosed with TB disease, 325 (89.3%) were diagnosed with TB infection, and 35 (9.6%) were considered uninfected. Three factors were significantly associated with a positive TST or IGRA result: region of origin, younger age group, and HIV infection. All children were more likely to have a positive TST compared to IGRA (OR 2.92, 95% CI: 2.37-3.59). Discordant test results were common (20%) and most often were TST+/IGRA- (95.0%). 35/57 (61.4%) of children who were TST+/IGRA- were considered uninfected and did not receive therapy for TB infection; none developed TB disease. The 22/57 (38.6%) TST+/IGRA- children who were treated for TB infection, were treated as such typically due to young age (< 2 years old) or variability in provider practice. Conclusions Positive TST results were twice as common as positive IGRA results and discordant TST/IGRA results were common. Positive TST results in BCG-immunized children frequently represent cross-reactivity and false positivity. Use of IGRAs as opposed to TSTs in BCG-immunized children would reduce false positive tests and allow for TB infection therapy to be targeted to those who would most benefit. These findings support 2018 changes in U.S. immigration guidelines that mandate IGRA use for recently immigrated children above 2 years of age.