| Prevalence among simulated cohort |
|
|
|
|
| Proportion of patients known to have tuberculosis who require drug resistance testing ("MDR-suspect") |
0.0245 |
0.000 |
0.0254 |
Sachdeva et al. 2014 [7] |
| Proportion of presumptive TB patients previously treated for TB |
0.168 |
0.019 |
0.3522 |
Sachdeva et al. 2014 [7]; average and range observed across sites |
| Prevalence of true PTB among new presumptive TB patients |
0.15 |
0.08 |
0.23 |
Assumption |
| Prevalence of true PTB among previously treated presumptive TB patients |
0.266 |
0.183 |
0.3982 |
Assumption |
| Prevalence of RIF resistance among new patients |
0.022 |
0.019 |
0.0260 |
WHO report 2014 [1] |
| Prevalence of RIF resistance among previously treated patients and DR-suspects |
0.150 |
0.110 |
0.1900 |
WHO report 2014 [1] |
| Diagnostic test accuracy |
|
|
|
|
| Sensitivity of Xpert MTB/RIF for diagnosing TB |
0.89 |
0.85 |
0.92 |
Steingart et al. 2014 [4] |
| Specificity of Xpert MTB/RIF for diagnosing TB |
0.99 |
0.98 |
0.99 |
Steingart et al. 2014 [4] |
| Sensitivity of Xpert MTB/RIF for Rif detection |
0.95 |
0.90 |
0.97 |
Steingart et al. 2014 [4] |
| Specificity of Xpert MTB/RIF for Rif detection |
0.98 |
0.97 |
0.99 |
Steingart et al. 2014 [4] |
| Sensitivity of sputum smear microscopy (ZN) |
0.62 |
0.44 |
0.79 |
Steingart et al. 2006 [12] |
| Specificity of sputum smear microscopy (ZN) |
0.98 |
0.96 |
1.00 |
Steingart et al. 2006 [12] |
| Sensitivity of LPA in detecting rifampicin resistance |
0.99 |
0.96 |
1.00 |
Bwanga et al. 2009 [26] |
| Specificity of LPA in detecting rifampicin resistance |
0.99 |
0.98 |
1.00 |
Bwanga et al. 2009 [26] |
| Sensitivity of clinical diagnosis in case found negative sputum smear microscopy |
0.160 |
0.10 |
0.22 |
Vassall 2011[5] |
| Alternative source for Sensitivity of clinical diagnosis in case found negative sputum smear microscopy |
0.610 |
|
|
Walusimbi 2013 [24] |
| Specificity of clinical diagnosis in case found negative sputum smear microscopy |
0.942 |
0.93 |
0.97 |
Vassall 2011[5] |
| Alternative source for specificity of clinical diagnosis in case found negative sputum smear microscopy |
0.690 |
|
|
Walusimbi 2013 [23] |
| Proportion of sputum smear negative patients getting CXR |
0.037 |
0 |
1 |
Vassall 2011[5] |
| Ratio of proportion of Xpert-negative patients who get CXR, compared to proportion of smear-negative |
0.5 |
0 |
1 |
Assumption |
| Proportion of Xpert positive patients who are smear negative |
0.36 |
0.355 |
0.368 |
Sachdeva et al. 2014 [7] |
| Diagnostic test unit costs (in US$ 2013) |
|
|
|
|
| Cost of sputum smear microscopy (per patient tested) |
0.83 |
0.60 |
1.10 |
Microcosting study. Rupert et al. 2017 [15] |
| Cost of Xpert testing, per patient tested |
13.17 |
12.06 |
14.72 |
Microcosting study. Rupert et al. 2017 [15] |
| Cost of liquid culture, per patient tested |
13.42 |
10.51 |
16.21 |
Microcosting study. Rupert et al. 2017 [15] |
| Cost of line probe assay for DST, per patient tested (1) |
21.34 |
19.50 |
23.07 |
Microcosting study. Rupert et al. 2017 [15]; adjusted for 2.2% error rate |
| Cost of antibiotic trial in clinical diagnosis |
3.86 |
3.86 |
3.86 |
Vassall 2011[5] |
| Cost of CXR |
4.0 |
2.98 |
5.00 |
WHO planning and budgeting tool [17] |
| Cost of clinical diagnosis = cost of antibiotic trial in all patients + cost of CXR multiplied by the proportion of patients getting CXR |
|
|
|
|
| Treatment costs |
|
|
|
|
| Full first-line regimen (2RH/4EHRZ) |
$148 |
$93 |
$188 |
S1 |
| Full first-line retreatment regimen |
$189 |
$185 |
$305 |
S1 |
| Second-line standard regimen, average for 24 months |
$5,812 |
$4,204 |
$7,421 |
S1; Fitzpatrick 2012 [21] |
| Number of weeks on first-line treatment until confirmatory results of LPA are returned in new Xpert RIF positive patients |
2.5 |
|
|
Assumption |
| Loss to follow-up before treatment initiation (3) |
|
|
|
|
| Proportion of sputum smear positive patients not starting treatment in the baseline |
0.092 |
0 |
0.20 |
Sachdeva et al. 2014 [7] |
| Proportion of bacteriologically positive TB patients not starting treatment in the intervention phase |
0.099 |
0 |
0.20 |
Sachdeva et al. 2014 [7] |
| Proportion clinically diagnosed TB patients not starting treatment |
0 |
|
|
Assumption |
| Additional initial default in DR-TB patients |
0.10 |
0 |
0.2 |
Sachdeva et al. 2014 [7] |
