Objective South African health care workers are at growing risk of tuberculosis (TB). This study sought occupational risk factors for latent TB infection (LTBI).

Methods A sample of public sector facility staff in Cape Town completed a questionnaire and underwent 3 tests for LTBI: (1) tuberculin skin test (TST) (skin induration > 10 mm) (2) QuantiFERON-TB Gold In-Tube (QFT-GIT) and (3) TSPOT. TB test. These were repeated one year later and annual rate of test conversion calculated. Occupational factors associated with baseline LTBI and conversion were sought, adjusting for age and gender and stratified by primary care vs TB hospitals.

Results 505 staff participated from 7 facilities. LTBI prevalence was high: TST 84%; QFT-GIT 65%; and TSPOT. TB 60%. Predictors of positive TST in primary care were employment duration >20 years [OR = 4.17 (95% CI 1.12–15.62); hospital staff with training on self-protection from TB infection were less likely to test positive [OR = 0.38 (0.16–0.91)]. Predictors of a positive QFT-GIT test in primary care were involvement in sputum collection [OR = 3.25 (1.28–8.09)] and employment >20 years [OR = 2.42 (1.09−5.38)], while again there was a protective training effect in hospital staff [OR = 0.41 (0.22−0.77)]. Predictors of a positive TSPOT. TB in primary care were providing home-based care to TB patients [OR = 4.14 (1.60 − 10.70)], and, paradoxically, working at a facility which advocated cough etiquette [OR = 2.06 (1.04 − 4.10)] or provided surgical masks to coughing patients [OR = 3.65 (1.16 − 11.51). The conversion rates were: TST 38% (95% CI 24–55) and QFT-GIT and TSPOT. TB both 22% (15–30). There were no consistent occupational predictors of conversion.

Conclusion LTBI prevalence and conversion are very high in this population, suggesting occupational risk. Occupational factors included duration and intensity of exposure (primary care, sputum collection, home visits), suggesting targets for infection control. However, more research is needed on occupational risk.