We conducted a cross-sectional study of pregnant women initiating prenatal care at a university-affiliated public hospital between January 5, 2009 and March 15, 2010. Study staff approached women seeking prenatal care at the Women’s Health Center at San Francisco General Hospital. Women were eligible if they were at least 16 years of age; pregnancy was confirmed by urine, blood or ultrasound testing; and primary language was English or Spanish. In addition, only women with a negative HIV antibody test during the current pregnancy were eligible for study participation. Exclusion criteria included prior severe reaction to tuberculin skin testing and treatment for latent or active tuberculosis during the current pregnancy.
After obtaining written informed consent, study staff administered a structured questionnaire assessing demographic information, medical history, TB history and risk factors for tuberculosis. Following the questionnaire, participants underwent phlebotomy and then tuberculin skin testing at the University of California San Francisco Clinical and Translational Science Institute Clinical Research Center at San Francisco General Hospital.
Immediately following phlebotomy, trained nurses from the research center placed a tuberculin skin test on all study participants using a standardized protocol. The tuberculin skin test was performed on the volar surface of the forearm according to the Mantoux method using 5 TU of tuberculin (Tubersol).(22
) tuberculin skin test results were interpreted 48 to72 hours after placement and maximum area of induration was measured in millimeters by the research nurses using Sylvac calipers.(23
) Using the CDC tuberculin skin test interpretation criteria(24
), we defined a positive test if the induration measured 10mm or greater for participants considered recent immigrants (<5 years) from a high-prevalence country (defined by World Health Organization as prevalence greater than 20/100,000 persons),(25
) injection drug users, or residents and employees of high-risk congregate settings (homeless shelter, prison, hospitals, or drug rehabilitation unit). For all other study participants, induration of 15 mm or greater was considered positive.
We used the QuantiFERON TB Gold In-Tube (Cellestis ©, Australia) assay, which measures interferon gamma levels in response to presentation of synthetic peptides.(18
) According to the manufacturer’s specifications, blood was collected into each of the three blood collection tubes, which include a TB antigens (ESAT-6, CFP-10, and TB 7.7) tube, a mitogen (phytohemagglutinin) tube and a negative control tube. Within a maximum of 7 hours of collection, tubes were incubated at 37°C for 16 to 24 hours. Tubes were centrifuged and plasma was collected and frozen at −70°C. ELISA was performed in batches of 28 samples per plate(18
) and interpreted according to the manufacturer’s protocol. A positive interferon gamma release assay test was defined as greater than or equal to 0.35 IU/ml after accounting for nil control and mitogen control results. A negative interferon gamma release assay test was defined as less than 0.35 IU/ml, after accounting for nil control and mitogen control results. The uncertainty zone was defined as 0.2 IU/ml to 0.5 IU/ml.(15
) The uncertainty zone has been documented in other studies of interferon gamma release assays as the range of results around the manufacturer’s defined cut-off for which there is a higher likelihood of assay reversion or conversion on repeat testing.
In the enrollment questionnaire, we asked participants if they had received birth bacille Calmette-Guérin vaccination. However, twenty-five percent of patients did not know their birth bacille Calmette-Guérin vaccination status. Therefore, after enrollment was completed, we assigned birth and booster bacille Calmette-Guérin vaccination status to study participants according to the bacille Calmette-Guérin World Atlas.(26
) This website, affiliated with McGill University and the Public Health Agency of Canada, allows one to search for country-specific bacille Calmette-Guérin vaccination policies and practices, including whether birth and/or school-aged booster vaccinations are given, estimated coverage countrywide, and how long the practice has been in place. This information enabled assignment of bacille Calmette-Guérin vaccination based on reported country of birth, birth year and year of immigration to the US.
Sample size was calculated based on an estimated 30% prevalence of positive tuberculin skin test results among pregnant women engaged in prenatal care at San Francisco General Hospital.(27
) At an alpha of 0.05 with 90% power to detect kappa of 0.80, we needed 199 patients to participate in the study.
Correlation between tuberculin skin test and interferon gamma release assay was assessed using kappa statistics. Chi-square test was used to compare the frequencies of positive test results among different groups of study recruits. Multivariable logistic regression was used to identify variables independently associated with positive tuberculin skin test and positive interferon gamma release assay results. Multivariable logistic regression was further used to identify variables associated with discordant tuberculin skin test and interferon gamma release assay results. Additional regression analysis was performed to evaluate predictors of women with discordant tuberculin skin test positive and interferon gamma release assay negative results. The correlation between tuberculin skin test and interferon gamma release assay was further explored using assay cut-offs based on the uncertainty zone. All statistical tests were run using STATA (version 11, College Station, TX).
Prior to study initiation, the University of California San Francisco Committee on Human Research approved the protocol. All study participants provided written, informed consent.