Background/Purpose: Tumor necrosis factor inhibitors (TNFi) revolutionized treatment of various conditions, however they drastically increase the risk of latent tuberculosis (LTBI) reactivation. Many national medical bodies recommend screening for LTBI prior to initiating TNFi, and this has been incorporated as a Merit-Based Incentive Payment System (MIPS) measure. We determined screening rates for LTBI across the United States, prior to initiating a TNFi.

Methods: We retrospectively analyzed patients in Truven MarketScan from 2011-2015. This dataset contains deidentified inpatient and outpatient claims records on over 100 million patients. We included patients over18 years with at least 1 filled prescription for TNFi. To ensure these were new TNFi starts we excluded patients without a 6-month washout period (i.e. during which time they could not receive biologic DMARDs). Continuous enrollment in the database was required during the washout period and 3 months after TNFi initiation. Our primary outcome was the proportion of patients screened for TB during the 6-month washout period, either by interferon gamma release assays (IGRA) or tuberculin skin testing (TST). Sensitivity analysis was performed to extend the eligible screening period to 12 months pre-drug. Descriptive statistics were represented as means and medians for continuous variables and as percentages for categorical variables.

Results: We identified 76,128 patients starting a TNFi. The mean age was 44.7 years, the cohort was 61% female. Adalimumab and Etanercept were the most common TNFi. 50.9% of patients had a rheumatologic diagnosis, 22.4% gastrointestinal, 17.6% dermatologic, and 0.8% ophthalmic. Most patients received specialty care, and a rheumatologist was involved in 40.9% of cases. 40,282 (52.9%) were screened for TB in the 6-month washout. By extending the pre-drug washout to 12 months, the proportion of unscreened patients improved mildly to 59.3%. 48.5% were screened by IGRA, 27% by TST, and 24.5% unknown. Steroid and DMARD use, male sex, urban residence, low Charlson comorbidity score and specialty care were associated with increased TB screening rates. Patients cared for by a rheumatologist or dermatologist were more commonly screened than those seeing gastroenterologists. Care by an ophthalmologist was not associated with improved screening.

Conclusion: In the United States, screening for latent TB prior to initiating TNFi therapy was poor, such that only 52.9% received appropriate pre-drug screening. Our study population of over 75,000 patients starting a new TNFi represents nationwide, real world data across various specialties. As clinicians, these results suggest we need to improve compliance with guidelines and quality measures. Care by rheumatologists and dermatologists was associated with improved screening compared to that by gastroenterologists and ophthalmologists.