In 2002, the Government of Botswana rolled out a national programme for the treatment of AIDS with antiretroviral
therapy (ART). In 2005, the impact of this ART scale-up programme was assessed by the National AIDS Committee
of the Botswana Ministry of Health (MOH), together with the World Health Organization (WHO), the Joint United
Nations Programme on HIV/AIDS (UNAIDS), and the University of California, San Francisco’s (UCSF) Institute for
Global Health, using country-enhanced monitoring and evaluation methodology tailored specifically to the situation
in Botswana.
Using data from multiple existing sources in a process called “triangulation,” the researchers were able to develop
a model to determine the impact of ART in Botswana. Preliminary results indicate that during the 3 years since its
inception, the ART programme in Botswana has achieved reductions in mortality of adults aged 25–54 years.
Reduced mortality is associated with early initiation of district ART programmes and with the overall rate of ART
uptake in the district.a
Widespread application of ART has been shown to decrease mortality among individuals and constrained samples
of the population. These previously stated findings are the basis for current clinical guidelines (1). This study
demonstrates that in Botswana, where excess mortality attributable to AIDS has been estimated to be 83% of all
deaths (2), ART has the potential to cause a population-level change in mortality rates.
The benefits of a triangulation methodology as applied in Botswana are twofold. First, the use of pre-existing data
sources allows the study to be relatively rapidly executed and concluded. This is of particular importance for
studies with significant policy or programmatic impact. Second, the systematic collection and examination of data
from multiple sources reveals new questions to be studied, permits verification, and reduces the likelihood of data
and researcher bias. The limitations imposed by the quality of the existing data remain, but are mitigated by this
methodology.
The Botswana experience has also identified some of the prerequisites for the effective application of triangulation.
It is necessary to be flexible during the analysis, and to consider complementing triangulation studies with additional
qualitative and quantitative research, where existing data are not sufficient to answer some questions. Based upon
the application of triangulation in Botswana, the engagement of high-level policy-makers and administrators
throughout the early part of the triangulation process is critical to the success of data identification and collation,
and remains important through the analysis phase.b A 1-week training course for representatives from a range of
institutions was initiated to build capacity in Botswana for future application of triangulation methods.