13CEP BENEFIT-RISK ANALYSIS OF ISONIAZID (INH) FOR TREATMENT OF LATENT TUBERCULOSIS INFECTION (LTBI)

Sunday, October 18, 2009
Grand Ballroom, Salons 1 & 2 (Renaissance Hollywood Hotel)
Mohsen Sadatsafavi, MD, MHSc1, Carlo A. Marra, PharmD, PhD1, Larry D. Lynd, PhD1, Fawziah Marra, PharmD1, Jiamei Liu, MSc1, Onofre Mendoza, MD., PhD2, Michael Tan, MA3, R. Kevin Elwood, MD4 and J. Mark FitzGerald, MD, FRCPC3, (1)University of British Columbia, Vancouver, BC, Canada, (2)Queen's university, Vancouver, BC, Canada, (3)Vancouver Coastal Health Research Institute, Vancouver, BC, Canada, (4)BC Centre for Disease Control, Vancouver, BC, Canada

Purpose: In North America, contacts of active TB cases who have a positive PPD skin test (>5mm induration) are offered preventive therapy with INH. However, there are concerns about the adverse effects of INH versus the risk of developing active TB. The objective of this study was to undertake a quantitative benefit-risk analysis of treatment of LTBI, with INH, for different groups of contacts using an incremental net-benefit (INB) approach.

Method: We developed a Markov model to compare treatment of LTBI in contacts compared to no treatment over a five year time horizon.  Contacts were stratified on four variables: BCG status (positive vs. negative), ethnicity (Canadian-born non-aboriginal, foreign born, and Aboriginal Canadian), type of contact (household vs. casual) and age group (older vs. younger than 10y). We calculated quality-adjusted life years (QALYs) gained due to delayed or prevention of active TB vs. QALYs lost due to the adverse events to INH using societal utilities for each health state. Risk for development of TB, compliance, and prevalence of immunization with BCG were taken from the longitudinal database of TB contacts in British Columbia. Other model parameters were obtained from the literature.

Result: INH was net-beneficial in treatment of LTBI subgroups at different risk of developing active TB. It was most beneficial in household Aboriginal contacts (net QALY gain 0.177 for BCG+, 0.0983 for BCG-), household contacts<10yo (0.0872), and casual Aboriginal contacts and <10yo (0.0869). The subgroup that benefitted least from prophylactic INH therapy was Canadian-born casual contacts (0.0062 for BCG+, 0.0098 for BCG-).  The number needed to treat (NNT) to avoid one case of active TB, varied from 3.38 for pediatric Aboriginal close contacts and 996 for BCG+ Canadian born, casual contacts. The chance of INH therapy having a positive INB never dropped below 60% in the probabilistic sensitivity analysis.

Conclusion: Stochastic risk-benefit analysis provides a valuable tool for assessing the merit of healthcare interventions. From this perspective, INH prophylaxis is recommended for all contacts that are tested positive in screening, even when the risk for the development of TB was low.

Candidate for the Lee B. Lusted Student Prize Competition