The AIDS Healthcare Foundation (AHF), a Los Angeles-based AIDS service and activist organization with a global reach, has called for more transparency and accountability in the way UNAIDS estimates the number of people living with HIV/AIDS who are on life-saving antiretroviral treatment. In an article published by in The Lancet on 1 September and reproduced on the AHF website, the foundation states that it believes the announcement by UNAIDS on 31 May 2016 that two million more individuals worldwide are on AIDS treatment now than in the year before, bringing their estimate of those on treatment to 17 million people, may be exaggerated. AHF pointed to a study published in The Lancet led by an independent research body, the GBD group (named after the title of the study, “Global Burden of Disease Study 2015”). The GBD study produced estimates of treatment coverage rates that are different than those released by UNAIDS. The study, which involved numerous collaborators, used the Guidelines for Accurate and Transparent Health Estimates Reporting (GATHER) which were developed by the World Health Organization. AHF said that UNAIDS does not strictly adhere to GATHER. According to AHF’s Associate Director of Global Policy, Denys Nazarov, some of the estimates derived by the GBD study differ by a large margin from those prepared by UNAIDS. “For example, in 2014 UNAIDS estimated a much faster rate of decline in annual new infections than GBD. Globally, the GBD 2015 report estimates about 2.5 million new infections in 2014, whereas UNAIDS estimates about 2 million for the same time period.” Mr. Nazarov further states: “A more dramatic case was Kenya, where results from GBD 2015 show an increase in annual new infections from 60,000 in 2005, to 146,700 in 2014, whereas UNAIDS shows a decrease from 73,000 to 56,000 during the same period. Since we have HIV services in Kenya, who should we believe? I think the responsibility for full transparency needs to come from UNAIDS.” AHF said that it is crucial to have as accurate a number of those on treatment as possible in order to best deploy the resources needed to end the epidemic by 2030 as part of the 90-90-90 targets (90% of people with HIV have been diagnosed; 90% of those diagnosed receive ART; and 90% of those on ART achieve viral load suppression by year 2020). UNAIDS and GBD estimates have increasingly converged at the global level. Nevertheless, estimates differ substantially in several countries, particularly in middle-income and high-income countries, where the GBD estimates are based on data from vital registration systems and UNAIDS estimates are based on prevalence in high-risk groups and calculations of this prevalence as a fraction of the population in these groups. Jorge Saavedra, AHF’s Global Public Health Ambassador and former Head of the National AIDS Program of Mexico, said there are unfortunately some methods utilized that tend to overestimate the number of people on ART, such as only counting the volumes of procurement without taking into account if the drugs really reached the mouth of the patient. “On the other hand,” Saavedra added, “when big countries like Nigeria, India, or Russia, among others, do not release their full data, UNAIDS should embrace its leadership role and fully release all the assumptions used in order to estimate those numbers.” According to AHF President Michael Weinstein, “UNAIDS needs to be accountable to the entire international community and not just to governments and donors. Right now, it is not clear that a site-by-site audit inside a country would produce the same numbers as what is being reported by UNAIDS. Some countries do not provide the numbers, or do not have the numbers, therefore UNAIDS estimates are based on multiple layers of assumptions. They end up with numbers that are sometimes difficult to believe, but that definitely portray an optimistic view of the progress in the war on AIDS.” The accuracy of reported numbers is of importance to the Global Fund as it relies significantly on UNAIDS data. Not having access to accurate numbers can lead to programmatic challenges, inaccurate reporting of results, and doubts amongst stakeholders as to effectiveness of service delivery. The GBD studies and UNAIDS are the only two sources that provide comparable evaluations of levels and trends of the HIV/AIDS epidemic at both the global and country level. UNAIDS has provided global estimates on HIV/AIDS since 1997, and has developed two epidemiological programs to estimate incidence, prevalence, and mortality: (a) the Estimation and Projection Package (EPP); and (b) Spectrum. The article on the GBD study in The Lancet states that fostering a culture of greater transparency and accountability for HIV prevention and treatment programs will benefit everyone concerned with tackling HIV more effectively in the future. According to the WHO website, a GATHER checklist was developed by the WHO and researchers, including the Institute for Health Metrics and Evaluation (IHME) at the University of Washington in Seattle, and was published in The Lancet and PLOS Medicine. The checklist contains 18 best practices that sets the standard for disclosing how health estimates are developed. GATHER includes requirements for disclosing which data are used to calculate estimates, and for making them available to others. It also includes a requirement to disclose how the computer code used to crunch the numbers can be accessed, making it possible for others to reproduce estimates, making them more robust. Reaction from UNAIDS In response to a request from Aidspan to comment on the issues raised by AHF, UNAIDS provided a copy of a letter which Peter Ghys, UNAIDS’ Director of Strategic Information and Evaluation sent to Weinstein on 14 September. In the letter, Ghys said, “We are confident about the accuracy of UNAIDS figures.” Ghys added, “In only a few countries are there opportunities to directly validate national data on ART between different data sources. We note that the percentage of people receiving antiretroviral treatment for Kenya is similar when basing it on counts of people receiving ART in facilities divided by the estimated number of people living with HIV, or [when] basing it on a direct estimate of the ART coverage derived from 2012-2013 Kenya AIDS Indicator Survey.” Ghys said that the number of people on ART in 2015 in India and Nigeria is publicly available at aidsinfo.unaids.org. He added that the number of people on ART in 2015 for Russia (and China), although submitted late, is available on the same site. Ghys said that UNAIDS makes guidance available to countries to help them generate and validate the number of people on ART; and that UNAIDS leads a global validation process that compares reported data to other sources of information. “While this will not eliminate double-counting in all countries,” Ghys said, “it will catch large discrepancies.” UNAIDS also referred Aidspan to a page on its website which explains how UNAIDS compiles HIV data and estimates. The 14 September letter from Peter Ghys to Michael Weinstein is on file with the author. This is Larson Moth’s first article for GFO. Larson was recently appointed Aidspan’s Chief Editor. See also separate article in this issue about the Global  Burden of Disease Study 2015.