“There’s always going to be that political filtering”: the emergence of Second Generation Surveillance for HIV/AIDS, data from Uganda, and the relationship between evidence and global health policy
Richards, Douglas Alexander
Background: It is widely acknowledged that Uganda was the first country in sub-Saharan Africa to experience a significant decline in HIV seroprevalence in the 1990s. Framed as the initial ‘success story’ in the history of the global HIV/AIDS pandemic, the behavioural mechanisms and policies accounting for the Ugandan HIV decline have been extensively debated over the past 25 years. With reference to broader debates about the role of evidence in policy, this thesis aims to examine contested explanations for the decline in HIV prevalence in Uganda and the role of evidence in the development of global HIV prevention policy in the 1990s. The thesis examines diverse explanations for Uganda’s HIV decline and how these came to be framed in the context of the emergence of Second Generation Surveillance (SGS), a global HIV/AIDS surveillance framework introduced by UNAIDS/WHO in 2000. Official accounts describe SGS as having been developed on the basis of Ugandan behavioural evidence presented during a key meeting of HIV/AIDS policymakers which took place in Nairobi in 1997. This meeting provides a focal point for examining the role of evidence in global HIV prevention policy and the relationship between evidence and policy pertaining to low income countries in the 1990s. Methods: A review of UNAIDS/WHO documents and 29 in-depth interviews with HIV/AIDS experts from Uganda and international organisations were analysed. Results: UNAIDS documents present SGS as a technocratic, problem-solving response to limitations in established HIV surveillance approaches, developed at a UNAIDS-sponsored workshop in Nairobi, Kenya, in 1997. These official accounts present the emergence of SGS as evidence-based and reflecting a clear consensus that developed during the Nairobi workshop. While interview data suggest agreement around the need for improved HIV surveillance systems, they indicate a more complex picture in terms of the extent to which SGS was evidence-based and highlight contested interpretations of this evidence among HIV experts. Findings from interviews suggest that the introduction of SGS by UNAIDS/WHO may be understood as serving both technical and broader strategic purposes. As indicated in UNAIDS/WHO policy documentation, SGS was intended to improve older global HIV surveillance methodologies via the triangulation of multiple data sources. The introduction of SGS also appears to have served two broader purposes, functioning as something akin to a marketing tool to help promote the institutional identity of UNAIDS, while also signalling a shift towards a ‘multisectoral’ approach that aimed to unify epidemiological and social scientific disciplinary approaches. While interviewees’ accounts coincide in describing a decline in HIV prevalence during the 1990s, they present divergent interpretations of this evidence which became significant in the development of SGS. One interpretation focused on a reduction in multiple partnerships within the Ugandan population as the key change driving the decline in HIV prevalence, while a contrasting explanation focused on increased use of condoms as the primary cause of this decline. Interviewees’ accounts suggest a process of competition, whereby different actors sought to secure the primacy of their interpretation in institutional understandings of Uganda’s HIV decline and in the development of SGS. Claims of disciplinary bias and institutional marginalisation appear to have contributed to the subordination of explanations focused on a decline in multiple sexual partners, while the policy entrepreneurship of one key actor appears influential in explaining the ascendency of explanations focused on increased condom use. Despite these contestations around the evidence used to inform the development of SGS, UNAIDS documents and peer-reviewed publications from this period emphasise one interpretation (that of increased condom uptake) which thus appears as the official explanation for the success of HIV control in Uganda. The transition from the WHO’s Global Programme on AIDS (GPA) to UNAIDS, and the initiation of a multisectoral HIV prevention approach, appear as important contextual and institutional influences in the interpretation of evidence for Uganda’s HIV decline. The failure of the partnership reduction explanation to align with the evolving institutional and political orthodoxy, and the potential for this explanation to challenge UNAIDS’ new focus on multisectoral HIV prevention, may help to explain why it did not inform subsequent HIV/AIDS policy and does not appear in official accounts of SGS’s development. In contrast, explanations focused on increased condom use were consistent with UNAIDS’ HIV prevention policy agenda (including its emphasis on multisectoral approaches) and appeared to reinforce the organisation’s need for increased financial resources to mitigate HIV/AIDS via the distribution and promotion of condoms. Discussion: This study demonstrates that the development of SGS, and the politics of evidence supporting its introduction, are more complex than existing UNAIDS/WHO accounts describe. Official explanations of the development of SGS provide a simplistic account of how evidence informed policy in a linear and rational way. In contrast, findings from this thesis suggest that SGS served multiple policy functions (i.e. marketing, promotion of institutional credibility, and a demonstration of disciplinary integration) in the context of the recently-formed UNAIDS, and that the role and interpretation of evidence in this context were highly contested. Consistent with the work of Kingdon (1995) and more recently Stevens (2007), this study suggests that personal, political and institutional factors play important roles in shaping how evidence is presented and linked with policy. These findings suggest that more nuanced understandings of the relationship between evidence and policy are needed to explain HIV/AIDS policy development within both sub-Saharan African and at a global level.