Figure 1.
Decision analysis model comparing Xpert MTB/RIF and smear diagnostic strategies guiding respiratory isolation decisions.
Definition of abbreviations: DS-TB = drug-sensitive TB. MDR-TB = multi-drug resistant TB. PTB = pulmonary TB. TB = tuberculosis. Legend: The large square represents a decision node, each circle a probabilistic node; and each triangle a terminal node. The two insets display the detailed distal branches leading to the terminal nodes for the two strategies being compared in the master tree. Applying the inputs in Tables 1 and 2, we used this decision analysis to generate the outcome data on economic and utilization costs presented in Tables 3 and 4. Note that for the Xpert strategy, the diagnostic performance of Xpert was estimated using the smear-positive sensitivity and specificity data for pluri-bacillary sputa and using the smear-negative sensitivity and specificity data for pauci-bacillary sputa. [19].
Table 1.
Input variables for epidemiologic and diagnostic parameters used in the base-case and sensitivity analyses.
Table 2.
Input variables for cost parameters used in the base-case and sensitivity analyses.
Table 3.
Average utilization and economic costs per patient.
Table 4.
Total annual utilization and economic costs for all patients.
Figure 2.
One-way sensitivity analyses comparing Xpert MTB/RIF and smear diagnostic strategies. Definition of abbreviations: AFB+, sputum acid-fast bacli smear-positive; AFB−, sputum acid-fast bacilli smear-negative.
Legend: †Ranges for sensitivity analyses are shown in parentheses. The vertical line at $2278 corresponds to the incremental savings in total costs per patient as shown in Table 3, dark grey bars show the estimated incremental savings of Xpert versus sputum smear when using the lowest value of the parameter range shown, and the light grey bars show the corresponding estimate when using the highest parameter value.