Introduction: The Centers for Disease Control and Prevention has recommended using a nucleic acid amplification test (NAAT) for diagnosing pulmonary tuberculosis (TB) but there is a lack of data on NAAT cost-effectiveness.
Methods: We conducted a prospective cohort study that included all patients with an AFB smear-positive respiratory specimen at Grady Memorial Hospital in Atlanta, GA, USA between January 2002 and June 2008. We determined the sensitivity, specificity, and positive and negative predictive value of a commercially available and FDA-approved NAAT (amplified MTD, Gen-Probe) compared to the gold standard of culture. A cost analysis was performed and included costs related to laboratory tests, hospital charges, anti-TB medications, and contact investigations. Average cost per patient was calculated under two conditions: (1) using a NAAT on all AFB smear-postive respiratory specimens and (2) not using a NAAT. One-way sensitivity analyses were conducted to determine sensitivity of cost difference to reasonable ranges of model inputs.
Results: During a 6 1/2 year study period, there were 1,009 patients with an AFB smear-positive respiratory specimen at our public urban hospital. We found the NAAT to be highly sensitive (99.6%) and specific (99.1%) on AFB smear-positive specimens compared to culture. Overall, the positive predictive value (PPV) of an AFB smear-positive respiratory specimen for culture-confirmed TB was 27%. The PPV of an AFB smear-positive respiratory specimen for culture-confirmed TB was significantly higher for HIV-uninfected persons compared to those who were HIV-seropositive (152/271 [56%] vs. 85/445 [19%]; RR = 2.94, 95% CI 2.36-3.65, p<0.001). The cost savings of using the NAAT was $2,003 per AFB smear-positive case.
Conclusions: Routine use of the NAAT on AFB smear-positive respiratory specimens was highly cost-saving in our setting at a U.S. urban public hospital with a high prevalence of TB and HIV because of the low PPV of an AFB smear for culture-confirmed TB.