Table 1.
Decision analysis model parameters related to the natural history, epidemiology, and treatment of LTBI and TB
| Estimates | Reference | |
|---|---|---|
| Probabilities | Probability (%)* | |
| •Probability of diagnosing LTBI | 25 | Assumption |
| •Probability of completing LTBI treatment among those who start it | 70.8 | [69] |
| •Probability of reactivation of longstanding LTBI to TB disease (non-smokers) | 0.05/year | [70–72] |
| •Probability of rapid progression following TB infection (non-smokers)** | 4.5/year for 2 years | Dynamic model# |
| •Probability of diagnosing active disease | 90 | Assumption |
| •Probability of spontaneous resolution of untreated TB disease | 25 | [16] |
| •Probability of relapse following spontaneously resolved TB | 2.5 | [16, 73] |
| •Probability of dying from untreated TB | 19/year | [74, 75] |
| •Probability of relapse after treated TB disease in first year | 1.5 | [76] |
| •Protective immunity from previous TB disease | 55 | Dynamic model# |
| QALYs for LTBI and active TB | Value (range) | |
| •Utility score for individual with active disease (treated) | 0.85 (0.70–0.90) | [40] |
| •Utility score for individual with active disease (untreated) | 0.68 (0.65–0.72) | [39] |
| •Utility score for individual with latent infection, during treatment | 0.97 (0.95–1.00) | [40] |
| •Utility score for individual with latent infection (untreated) | 1 | Assumption† |
INH isoniazid, LTBI latent tuberculosis infection, QALYs quality-adjusted life-years
*Probabilities are one-time percentages, unless stated otherwise
**After 2 years, risk of developing active TB falls to 0.05% annually
†Persons with LTBI are asymptomatic
#Calibrated in dynamic model