Tuberculosis prevalence survey reveals the high proportion of asymptomatic disease, and points to high-burden households as a site for potential intervention in peri-urban slums of Callao, Peru

      Background: To improve the reach of tuberculosis elimination, we must understand the population that does not seek medical contact. Prevalence surveys are widely used to measure undiagnosed TB disease, but rarely characterize the clinical and social risk factors and barriers to care that obstruct diagnosis, notification, and treatment.
      Methods & Materials: A cross-sectional prevalence survey of pulmonary TB disease was undertaken with sputum culture on Middlebrook 7H11 thin-layer agar in a cohort of 1668 households known to have been affected by TB in the past 10 years. Symptoms, co-morbidities, and health-seeking behaviour were assessed for all household members, and treatment records from community health posts were used to identify all courses of antitubercular therapy prescribed to household members.
      Results: Sputum was obtained from 6380 residents over 8 years old (1744 former patients; 4636 other residents). 101 participants were taking antitubercular therapy and no longer sputum-culture-positive. 51 culture-positive cases of prevalent TB disease were detected, of whom only 13 met the national clinical criteria for sputum testing (>2 weeks of productive or bloody cough). 11 had been tested for TB, 10 diagnosed, and 4 had been initiated on treatment at the time of the survey.
      Multiple treatment episodes within the household (whether due to repeated infection, re-initiation of uncompleted therapy, or change of treatment scheme) were associated with increased odds of prevalent disease in the survey. Participants who had undergone two or more treatment episodes had 8.6 times the odds of prevalent TB of those with none (95%CI:3.9-18.7;P < 0.001). Members of households with more that two treatment episodes had 3.2 times the odds of TB of those without—2.9 after adjustment for age, sex, poverty, HIV status, and drug and alcohol misuse (95%CI:1.6-5.3;P < 0.001).
      Conclusion: Accelerating currently recommended case management strategies could achieve a four-fold increase in treatment coverage for this population, but 75% of potentially infectious cases would still remain undetected. New strategies are needed to identify and engage populations at risk of undetected, minimally symptomatic disease. Meanwhile, follow-up of households affected by TB to provide preventative therapy for latent TB and active surveillance for new illness episodes will aid control and elimination of TB.