aNational Institute of Public Health, Cuernavaca, Mexico
bCenter for Economics Research and Teaching, Mexico City, Mexico
cUniversity of California, Berkeley, California, USA.
Received 22 August, 2006
Accepted 26 October, 2006
We appreciate the careful review of our paper  by Dr Volkow and her co-authors. We agree with some points they make and would like to provide some additional clarifications regarding others.
The period evaluated in our study, 1997–2001 is not, according to Dr Volkow, adequate to evaluate the rapid scaling up of treatment in Mexico, because the real scale-up access occurred only after 2000.
We not only agree with Dr Volkow, but highlight this fact in the paper. For example, in the introduction section: ‘There have been major efforts to improve access since 2001, thus the data presented here do not necessarily reflect the current standard of care for people living with HIV/AIDS in Mexico.’
Even although the period analysed in our study is before the official commitment to achieve universal access, the data reveal an impressive expansion of antiretroviral treatment during the period of observation, as reflected in the number of patients receiving any antiretroviral drugs and the number receiving triple therapy. We thus believe that this early period also merits analysis.
Dr Volkow argues that in the period evaluated, access to antiretroviral drugs was very limited, and in that context changes in treatment regimens were driven by accessibility more than prescription decisions.
We agree that interruptions in drug supply are a potentially important factor in explaining the pattern of care we observed. It was never our intention to imply any definite causality between the possible explanatory factors and the observed outcomes. We do not know and cannot know the reasons behind the observed heterogeneity in the antiretroviral treatment received by the patients in our study. Besides deficiencies in physician training, we discussed additional possibilities: ‘There are a number of other factors which can be related to suboptimal standard of care in Mexico. Interruption in drug supply is one crucial issue which could explain some of the results encountered in this study. Patients having to buy their medications and laboratory tests could be a significant factor in Mexico's experience before 2002.’
However, the main conclusion of our paper is that scaling up of treatment in any country must consider all the potential factors that could lead to results similar to those observed, in order to avoid them. Ensuring an uninterrupted drug supply and appropriate training and monitoring of prescription practices are surely among the most important determinants of success to be considered by any country.
Dr Volkow points out that there are probably errors of misclassification of non-recommended regimes, because we used 2000 guidelines to classify treatment combinations prescribed between 1997 and 2001.
This is a fair criticism, and we agree that it would have improved the paper had we used the earlier guidelines for the initial years. That said, we do not believe that the change would significantly affect the conclusions. The total number of observations in our sample is 8363, of which 73.7% are documented in the years 2000 and 2001. The potential misclassification thus only affects a quarter of the sample. We would also argue that pre-2000 regimens that are consistent with post-2000 guidelines also be considered appropriate, suggesting that the only relevant misclassifications are those regimens pre-2000 that are consistent with the pre-2000 guidelines and not with the post-2000 guidelines.
It was not clear for Dr Volkow how the levels of adherence were calculated in the ‘observed scenario’. In addition, she argues that the approach to estimating adherence is questionable and invalidates any conclusion, given that the source of data was a retrospective review of clinical charts, which most likely will under-register patient adherence.
A truly accurate retrospective measurement of adherence levels is impossible in our setting. We explicitly acknowledge this limitation in the paper and address it in several ways. The first is by constructing several different adherence scenarios. The base scenario counted as adherent months only ‘chart confirmed months of medications supplied to patients…’; the other two are more optimistic, including one with perfect adherence. Our principal conclusions are related to prescription practices, i.e. the number of changes of antiretroviral regimes, the number of non-recommended regimens, and the inefficiency of such practices. We also argue in the paper that a serious adherence problem probably exists. The fact that there was so little documentation in the records of efforts to promote, evaluate and document adherence suggests serious problems with quality of care, even if it does underestimate true adherence in the base scenario.
Dr Volkow suggests that our study did not consider differences in physicians' experience and training in our analysis.
We are not completely sure what she means by the comment. If she means to suggest that physicians with better training and more experience provided better patient care then we would agree, at least intuitively, and this is reflected in our study. We did not document physician characteristics and thus we can only hypothesize, as she does, that physician training and experience is probably an important determinant of quality care.
Most importantly, Dr Volkow and her colleagues are most concerned that our paper might be used as an excuse for policy makers to back off from commitments to expand and maintain coverage. Given other publications by our group [2,3] and work to promote universal access globally to care, it is clear that this was not our intent. The authors include an apparent quote from our paper in their letter ‘it is better not to treat, than to treat badly’, which we cannot find in our text. What we do say is ‘Such high levels of non-adherence suggest that a significant proportion of patients were receiving no benefit from their treatment and may have been better off with no ART until the system is able to ensure better adherence.’ Our desire is not to give anyone an excuse for backing away from their commitments, but rather to point out that the expansion of coverage that only tracks the number of patients receiving therapy rather than the number of patients receiving care of adequate quality is not in anyone's interest, not patients', not their families', not the health system's and not society's, a lesson learned decades ago with tuberculosis treatment.
1. Bautista S, Mane A, Bertozzi SM. Economic impact of antiretroviral therapy prescription decisions in the context of rapid scaling-up of access to treatment: lessons from Mexico. AIDS 2006; 20:101–109.
2. Stover J, Bertozzi S, Gutierrez JP, Walker N, Stanecki KA, Greener R, et al
. The global impact and net costs of scaling-up prevention programs for HIV/AIDS in low- and middle-income countries through. Science 2006; 311:1474–1476.
3. Gutiérrez JP, Johns B, Adam T, Bertozzi SM, Edejer TT, Greener R, et al
. Achieving the WHO/UNAIDS antiretroviral treatment 3 by 5 goal: what will it cost? Lancet 2004; 364:63–64.
© 2007 Lippincott Williams & Wilkins, Inc.