Setting: The burden of tuberculosis (TB) disease among household contacts of multidrug-resistant TB (MDR-TB) patients is poorly understood and might represent a target for transmission-interrupting interventions.
Design: This retrospective cohort study, conducted in Lima, Peru, from June to September 2008, estimated the incidence of TB disease among household contacts of MDR-TB patients in 358 households.
Results: Of 2112 household contacts in 80 households (22% of households), 108 (5%) developed TB disease during the study, giving an incidence rate of 2360 per 100 000 contact follow-up years for each of the first 3 years after exposure. Drug susceptibility tests (DST) were available for 50 diseased contacts, of whom 36 (80%) had MDR-TB. Forty-two pairs of index-contact DSTs were available, among which the contact had an identical or less resistant phenotype than the index case in 27 pairs. Multivariate Cox regression demonstrated that male contacts (hazard ratio [HR] 2.8, P < 0.05), with previous TB disease (HR 20.7, P < 0.001) and with associated (non-human immunodeficiency virus) comorbidities (HR 11.2, P < 0.001) were more likely to develop TB.
Conclusion: The high percentage of diseased household contacts highlights an opportunity for household-level interventions to prevent transmission, whether or not these cases were all attributable to the index case.