Archives

  • 2018-07
  • 2019-04
  • 2019-05
  • 2019-06
  • 2019-07
  • 2019-08
  • 2019-09
  • 2019-10
  • 2019-11
  • 2019-12
  • 2020-01
  • 2020-02
  • 2020-03
  • 2020-04
  • 2020-05
  • 2020-06
  • 2020-07
  • 2020-08
  • 2020-09
  • 2020-10
  • 2020-11
  • 2020-12
  • 2021-01
  • 2021-02
  • 2021-03
  • 2021-04
  • 2021-05
  • 2021-06
  • 2021-07
  • 2021-08
  • 2021-09
  • 2021-10
  • 2021-11
  • 2021-12
  • 2022-01
  • 2022-02
  • 2022-03
  • 2022-04
  • 2022-05
  • 2022-06
  • 2022-07
  • 2022-08
  • 2022-09
  • 2022-10
  • 2022-11
  • 2022-12
  • 2023-01
  • 2023-02
  • 2023-03
  • 2023-04
  • 2023-05
  • 2023-06
  • 2023-07
  • 2023-08
  • 2023-09
  • 2023-10
  • 2023-11
  • 2023-12
  • 2024-01
  • 2024-02
  • 2024-03
  • The approach used of estimating effectiveness relative to

    2019-04-29

    The approach used, of estimating effectiveness relative to a matched counterfactual, can be regarded as an attempt to reach the ideal estimation of relative risk, namely the comparison of a population to itself with the exposure removed. However, it is not possible to know exactly 4EGI-1 what would have happened if Avahan had not intervened, and the absence of true empirical baseline data increases uncertainty. In some districts pre-existing interventions were present, which might have led to increased condom use in the absence of Avahan. Although our conservative counterfactual makes allowances for this, the absence of data from districts without any Avahan intervention makes the attribution of effectiveness to the programme with absolute certainty difficult. Additionally, limitations arise from the use of reconstructed condom trends based on self-reported condom use, and although we tried to allow for social desirability biases within the modelling, as well as cross-validating with non-survey methods, it is not possible to know if this issue has been fully accounted for, although the use of Bayesian 4EGI-1 testing provides further evidence that these trends are credible in these settings. Effectiveness estimates are dependent on the sizes of the high-risk populations, and although mapping studies were specifically done in Avahan districts, the accurate mapping of hidden populations is challenging. Migration from non-Avahan districts could reduce estimated effectiveness. Antiretroviral therapy could also be changing the epidemic, leading to higher HIV prevalence as survival improves. However, although long-term projections could be affected by increasing access to antiretroviral therapy, coverage remained low until after 2008, by which time IBBA surveys in most districts had been completed. Moreover, modelling work suggests that the increase in HIV infections averted by antiretroviral therapy on top of the effect of Avahan is small (unpublished). Finally, although the IBBA districts comprise almost a third of all Avahan districts, they were not chosen randomly, so might not be representative. However, these data limitations, such as the absence of baseline data, are neither intrinsic nor unique to our approach, and reflect the realities of programme implementation and real-life assessment. Our use of a simulated matched counterfactual for each district means that non-random district selection is less problematic than for approaches in which Avahan and non-Avahan districts are compared, such as in the study by Ng and colleagues, or community-based randomised controlled trials, for which it might not be possible to find comparable control districts, leading to imbalance. A further issue for these alternative assessment designs, which rely on non-Avahan control districts, is the scaling-up of targeted interventions by NACO in non-Avahan districts since 2007, meaning that such analyses are effectively comparing Avahan with NACO interventions. Finally, although a step-wedge design can be useful for assessing intermediate outcomes, the present combined approach might be more suitable for assessing HIV interventions in populations, since changes in HIV prevalence and incidence might not be measurable for a long time. In summary, using mathematical modelling to quantitatively synthesise HIV and STI prevalence data with key setting-specific behavioural indicators, we have shown strong and plausible evidence for a large intervention effect of Avahan, which increased over time, based on Habicht and colleagues\' scale for assessing the strength of evidence for effectiveness of public health interventions. This effect occurred through increased condom use, brought about by removing barriers to use via intervention components including distribution and social marketing of condoms, peer outreach, STI treatment, structural intervention, and community mobilisation. In an era focused on antiretroviral therapy as prevention, these results show that behaviour-focused, core-group-targeted HIV preventive interventions can be rapidly and successfully implemented at scale. The low coverage of such programmes in many regions of the world should be addressed, since with high coverage these programmes have the potential to substantially reduce concentrated HIV epidemics. Contributors Conflicts of interest
    Acknowledgments This research was funded by the Bill & Melinda Gates Foundation. The views expressed herein are those of the authors and do not necessarily represent the official policy or position of the Bill & Melinda Gates Foundation, Imperial College London (London, UK), or the London School of Hygiene & Tropical Medicine (London, UK). Data from the integrated behavioural and biological assessment (IBBA) surveys are accessible through an application process from the National AIDS Research Institute website. We are grateful to FHI-India and the National AIDS Research Institute (Pune, India) for the IBBA data collection and to Karnataka Health Promotion Trust for data collection and analysis. We thank Lalit Dandona and Rakhi Dandona, lead investigators of the Guntur population-based HIV study, for provision of data from their study. We also thank Eric Demers, who did analyses to estimate model parameters; Sharmistha Mishra, Kate Mitchell, and Nadine Schur for their help with the figures; and the London School of Hygiene & Tropical Medicine for use of their high-performance computing facility.