UC San Diego

Military personnel are at a higher risk of tuberculosis (TB) infection through military deployments, assignments to high incidence countries, and close contact with refugees and prisoners of war. Service members are routinely tested; those with active TB or latent tuberculosis infection (LTBI) are treated. The purpose of this dissertation was to (1) propose a method for LTBI surveillance among United States military personnel using pharmacy and medical records; (2) determine LTBI conversion rates and associated risk factors in Navy health care workers (HCWs) using this method; and (3) evaluate adverse events associated with LTBI treatment in this population. These studies used data from the Military Health System Data Repository, a collection of databases on health care services (inpatient, outpatient, and prescription information) provided to military personnel. Using both pharmacy and medical records, among 3,089,436 military personnel, 133,365 service members were identified with LTBI; 38.9% (n = 51,943) had LTBI indicated in both sources within 180 days of each other. Moderate concordance (kappa = 0.55) was observed between both sources for LTBI. The LTBI incidence among 18,975 Navy HCWs over the study period was1.7% (95% CI 1.5-1.9), with an incidence density of 4.6 per 1,000 person-years (95% CI 4.12-5.13). Using survival analysis, being non- white and foreign born were LTBI risk factors for HCWs, while overseas duty stations and Iraq or Afghanistan deployments were inversely associated with LTBI diagnosis. Among 74,537 service members prescribed LTBI treatment, 96% (71,427) were dispensed isoniazid and 4% (3,110) rifampin. Treatment-associated adverse events were reported in 7.0% of service members. During the study period, the number of service members starting treatment decreased over time, but the incidence of adverse events did not change. Isoniazid was a risk factor for any adverse event compared to rifampin using logistic regression. Prescribing vitamin B6 was associated with lower odds of adverse events. For service members older than 25 years, the odds of having any adverse event increased with age. This proposed method for TB surveillance can be beneficial for military LTBI providers and could be used in the absence of a global LTBI reporting system to monitor incidence and treatment- associated adverse events