Table 1.

Pre and post intervention values for specific model parameters

Intervention (all public sector unless otherwise specified)	Model parameters influenced	Pre-intervention value	Notes and reference for pre intervention	Post-intervention value	Notes and Reference for post intervention
Community Education	Patient delay (probability of seeking care with a 1 year delay)	41 · 79 days = 0 · 11 probability of a 1 year delay	[14-17]	21 days = 0 · 06 probability of a 1 year delay	Assume 50% reduction in delay days
DOTS expansion for diagnosis	Incorrect diagnostic test ordered by heath professional	0 · 603	[18]	0 · 351	[19]
	Diagnostic delay (probability of incurring a 1 year delay)	29 · 49 days = 0 · 081 probability of a 1 year delay	[14-17]	1 · 83 days = 0 · 005 probability of a 1 year delay	Used pre-intervention data and ratio of delay days "pre" and "post" intervention from [20] to obtain post-intervention estimate of delay days
	Loss to follow up during diagnostic work-up	0 · 254	[21-24] (Assume that loss to follow up is the same for regardless of provider)	0 · 140	Used pre-intervention data and ratio of outcomes "pre" and "post" intervention from [19] to obtain post-intervention estimated of loss to follow up
DOTS Expansion for Treatment	Incorrect treatment	0 · 791	[25]	0 · 129	[25] Scenario assumed that incorrect treatment was given regardless of DST availability
Non specific DOTS Expansion (NTP Strengthening)	Initial access- inaccessible provider (ie · probability that patient seeks care with alternative provider that is inaccessible to interventions)	0 · 055	[14,26-31]	0 · 025	Intervention assumed to have same impact as in private sector
Private Sector interventions	Incorrect diagnostic test ordered by private provider	0 · 622	[18]	0 · 362	[19]
	Diagnostic delay (private sector only)	0 · 11	[14-17] (# days pre-intervention)	0 · 007	Used pre-intervention data and ratio of delay days "pre" and "post" intervention from [20] to obtain post-intervention estimate of delay days and then used ratio of outcomes in public vs private sector from [14] [27,32,33] to extrapolate estimate for public system to private system
	Loss to follow up during diagnosis (private sector only)	0 · 254	[21-24]	0 · 140	Assumed to be same as in public sector (a 45% reduction). Used pre-intervention data and ratio of outcomes "pre" and "post" intervention from [19] to obtain post-intervention estimated of drop out
	Incorrect treatment by private provider	0 · 771	[34]	0 · 126	Used pre-intervention data and ratio of outcomes "pre" and "post" intervention from [25] to obtain post-intervention estimate of incorrect treatment
HIV/ ART therapy programmes	TB Death rate in HIV/TB co-infected	0 · 12	[35]	0 · 10	[35] [36,37] (see table S5 in Supplement appendix for more detail)
	TB Relapse rate HIV/TB co-infected	0 · 16	[36]	0 · 01	[35] [36,37] (see table S5 in Supplement appendix for more detail)
	TB Reactivation rate HIV/TB co-infected	0 · 0340	[38-40]	0 · 02	[41]
MDR-TB related interventions	DST performed	0 · 2	Assumption	0 · 5	Assumption
	MDR- loss to follow up rate in HIV negative cases	0 · 22	[36]	0 · 11	Assumption- reduce rate to 50%