Referee 1's Review (Rony Zachariah):

**********
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.
**********

WHAT HAPPENS TO PATIENTS ON ANTIRETROVIRAL THERAPY WHO "TRANSFER OUT" TO ANOTHER FACILITY?"

This is very well written, original (and easy to read) paper that contains excellent practical information related to data monitoring of antiretroviral treatment (ART) from a real world situation in a resource-limited, high prevalence country.

The paper is extremely relevant from a operational perspective as this is one of the first studies to my knowledge that looks at the issue of "transfer outs", others having concentrated mainly on losses to follow up and deaths . The messages are very clear and the recommendations on avoiding double counting is applicable to similar settings in sub-Saharan Africa.

The study investigators also managed to actually trace the great majority of patients within this operational setting and thus the results are very valid and relevant. The paper should be published and would benefit not only those who monitor ART but more widely as the principle is applicable to other chronic disease eg tuberculosis

I have but a few minor comments that the authors might wish to consider:

Minor specific comments

1. Introduction: Para 2, Line 3. ...Reports on treatment outcomes.....make much referral to the problem of loss to follow up. Please rephrase as follows "Reports on treatment outcomes......often raise the problem of loss to follow up..

2. Methods, Background: When a patient is transferred out to a new ART facility, he goes with his master card but he/she receives a new ART number at the new clinic where he/she arrives. Please indicate how the old number and information can be traced on hand....ie Is a new ART master card opened and the old one clipped on? is the new ART number written beside the old one using the old card ?

3. Table 1. Please indicate abbreviations in full below the Table eg WHO= World Health Organisation etc

Row 5 and 10. It would be better to give the P values for each of the indicator variables in these two rows and the value of the chi square for trend could appear at the base of the table.

4. Table 1 Row 9: Please indicate range for median time.

5. Figure 1. Risk table: Please rephrase the sentence "Duration of risk after ART" to "Number at risk after starting ART".