Referee 2's Review:

**********
N.B. These are the comments made by the referee when reviewing an earlier version of this paper. Prior to publication, the manuscript has been revised in light of these comments and to address other editorial requirements.
**********

This is an extremely, interesting, important and novel paper.

Currently it requires further discussion and expansion in order to clarify the work and to fully explain the significance of the findings.

1) When prevalence & incidence are mentioned it should always be mentioned as to whether the authors are describing infection or disease.

2) Are the authors modeling length of treatment? If so this should be explained.

3) An explanation of each of the results for the different strategies should be presented – for example, why is DOTS on it’s own fairly ineffective? Is it because of the length of treatment? Or for some other reason?

4) The work by Blower and her research group should be more adequately referenced – as her models were the original TB models on which all the subsequent models by Chris Dye and Megan Murray were based. Blower et al. ‘s papers in Science (1996), Journal of Molecular Medicine, Nature Medicine (2004), and (1998) Theoretical Population Biology paper, should be referenced in the appropriate places in the text. For uncertainty and sensitivity methodology the authors should cite the Blower and Dowlatabadi paper, and the Sanchez and Blower paper. When the authors mention drug resistant TB they should cite Blower et al. ‘s papers in Science (1996), Journal of Molecular Medicine paper, and, her 2004 Nature Medicine paper. All of these papers are available on Sally Blower’s website.

5) In the verbal description of the model all of the transitions between boxes/states should be mentioned, currently only some of them are mentioned in the text.

6) The flow diagram of the model needs to be redrawn; currently it is very difficult to read.

7) The authors should describe (in words & in terms of their equations) how they implemented each of their control strategies.

8) The authors should provide details of how they calibrated their model.

9) The authors should clarify that they conducted several analyses (rather than several scenarios) using uncertainty and sensitivity analyses.

10) Median values should be included in the Tables and mentioned in the text (as well as the mean values).

11) The authors should include their Supporting Information in the main text.

12) The authors should briefly mention in the discussion what is likely to happen if the prevalence and incidence of TB in the jail is increasing rather than stable.

13) In Table 1 the magnitude of the k parameter needs to be explained.

14) Table 1 should include units for all parameters.

15) With regards to the model shouldn’t the T states be linked to the D states? Arent the T states TB cases and D the detected treated cases? Why are the E and L states linked to the D states? Arent the E and L states non-infectious non-disease states? Please clarify.

16) On the figures and throughout the text 9.9 should be replaced by 10.

17) When the pdfs are shown there is an error as the reference source is not found – this should be corrected.

In summary, this paper only needs clarification and expansion, but overall it is an excellent paper.