To Treat or Not to Treat Suspected TB
Although today’s readily available and rapid laboratory cultures improve diagnosis of Mycobacterium tuberculosis, the bacteria causing tuberculosis (TB), it still takes 2-3 weeks to learn the outcome, and researchers say this delay in initiating treatment increases deaths.
A study supported by the Agency for Health Care Policy and Research found that treatment of HIV-infected patients who have negative acid-fast bacteria (AFB) smears decreased deaths by an average of 2%.1 When community prevalence of multidrug-resistant TB (MDR-TB) exceeds 9.6%, starting drug-resistant therapy for AFB smear-positive patients (before results from culture and drug-resistant tests are available) minimizes costs and decreases risk of mortality at an additional cost of $8,000 per life saved. This figure is still lower than costs per life-year saved for health care interventions such as screening blood donors for HIV.
Treatment Can Reduce Risks
Investigators point out that the average patient’s risk of death and health care costs while waiting 2-3 weeks for results of laboratory tests can be reduced by treating all TB-suspect patients who have a positive AFB sputum smear or are infected with human immunodeficiency virus (HIV). They recommend treating HIV-positive patients even if the AFB smear is negative, because such patients are at high risk for TB.
According to study authors, the potential for death from drug toxicity in treating numerous people without TB is slightly outweighed by the death potential of not treating those with smear-negative disease while waiting for culture results.
Researchers conclude that early treatment with drug-resistant therapy for AFB smear-positive patients rather than the usual four-drug regimen recommended (ethambutol, isonicotinic acid, pyrazinamide, and rifampin) minimizes both risk of mortality and costs. A quinolone drug is added to the above regimen for MDR-TB.
1. Brewer S, Heymann J, Ettling M, et al. An effectiveness and cost analysis of presumptive treatment of Mycobacterium tuberculosis. Am J Infect Control 1998;26:232-238.