Using a HIV registry to develop accurate estimates for the HIV care cascade - the Singapore experience

Abstract

Introduction: Achieving UNAIDS 90-90-90 targets is a crucial step towards ending the AIDS epidemic. Many countries have published estimates of care cascades, although often with methodological limitations. We describe an approach that used the national HIV registry as a starting-point to determine the HIV care cascade and resulting UNAIDS 90-90-90 estimates for Singapore in 2014. Methods: HIV is a legally notifiable disease in Singapore. The anonymized HIV registry data provided for a back-calculation model from the European Centre for Disease Prevention and Control to obtain 2014 estimates for the total number of persons living with HIV (PLHIV), and the count in the registry for proportions diagnosed with HIV and linked to care. Using additional data collected for a simple random sample from the registry, outcomes in 2015 and 2016 were ascertained retrospectively to derive proportions for those retained in care, on antiretroviral therapy, and achieved viral suppression. Findings were extrapolated to derive national estimates and UN90-90-90 estimates. Bootstrapped samples from the model and sample were used to derive 95% confidence intervals. Results: An estimated 6900 (95% CI 6650, 7050) persons were living with HIV and AIDS in 2014. Of these, 4948 were diagnosed with HIV, and 4820 had been linked to care. The random sample of 500 persons was further analysed, and of these, 87.2% were retained in care, 84.6% on antiretroviral therapy, and 79.6% had suppressed viral loads. The proportions of HIV-infected individuals on antiretroviral therapy and achieving viral suppression were 60.7% (95% CI 58.4, 63.6) and 57.1% (95% CI 55.0, 60.5) respectively. The corresponding UNAIDS 90-90-90 estimates were 71.7% (95% CI 70.0, 74.2) of all persons diagnosed; 84.6% (95% CI 81.6, 87.4) of diagnosed persons being on antiretroviral therapy, and 94.1% (95% CI 91.6, 96.2) of persons on therapy having achieved viral suppression. Conclusions: A national HIV registry, alongside back-calculation and additional data from a sample, can be used to estimate attainment of UNAIDS 90-90-90 targets and identify system gaps. The registry had advantages of providing a well-established, comprehensive capture of diagnosed persons and easily accessible data. The same approach can be used elsewhere if similar data are available. © 2019 The Authors. Journal of the International AIDS Society published by John Wiley & Sons Ltd on behalf of the International AIDS Society.