Authors, using data from a population-based cancer registry (PBCR) in Colombia, investigated determinants of breast cancer survival in local population of Manizales, a mid-size city. Although the base population is not very large, the key determinants have a strong effect on survival and therefore these results are important and underscore the importance of health inequities and their impact on outcomes. Essentially, quality of care one can access is driven by what one can afford. The analytical approach is sufficiently robust to support the inference and manuscript is reasonably well-written.
Key areas where additional analyses and improvements are needed are given below.
Screen-detected proportion (and by stage) in screening eligible needs to be described.
Screening history is an important variable that is missing, subgroup analyses for the screening eligible group with this variable also included in the multivariate model.
A table of stage of presentation, not all 4 groups but I & II vs III & IV (as columns) and the following variables as rows would be helpful to understand overall data.
Insurance regime, Socioeconomic level, Age, residence (+/-) and for screening eligible (screened vs not).
Analysing role of screening will elucidate the impact of lack of awareness (not necessarily quantitatively) since not all late diagnosis may be due to lack of access, some of it may be due to lack of awareness.

The sentence on page 12, line 48 is unclear – needs to be reworded.

Minor:
If Chi-square values are given, then degree-of-freedom needs to be stated as well.
Some spelling and grammar errors.
Introduction could be shortened to move some of the material to discussion section.

Rebuttal and revision are now satisfactory.