People living with HIV who know their HIV status

There are two recommended methods for estimating the proportion of people living with HIV who know their HIV status. The method used depends on the availability of data in the country.

Direct estimates from HIV case surveillance systems:

• For the numerator: in countries with well-functioning HIV case surveillance systems, the number of people living with HIV who know their status is the same as the number of people diagnosed with HIV and reported to the surveillance system who are still alive.
• For the denominator: estimation models such as Spectrum are the recommended source for the number of people living with HIV. If a model other than Spectrum is used, documentation of the estimation method and uncertainty bounds should be provided.
• On case surveillance methods: an HIV case surveillance system is considered to be well functioning if reporting from all facilities providing confirmatory HIV testing, care and treatment services has been in place for at least five years, and if people who have died, been lost to follow-up or emigrated are removed from the numerator. Only confirmed diagnoses of HIV should be counted, although countries should adjust for reporting delays by including an estimate of the number of people diagnosed but not yet reported during the latest calendar year (if necessary). Mechanisms should be in place to de-duplicate people diagnosed and reported multiple times or from multiple facilities.

Modelled estimates:

• For the numerator: the approach to modelling the estimate of the number of people who know their HIV-positive status among people living with HIV depends on the availability of data in the country:
• For countries with robust case surveillance and vital registration systems, the number of people who know their HIV-positive status can be derived using the CSAVR incidence estimation option in Spectrum or the European Centre for Disease Control HIV Modelling Tool (https://ecdc.europa.eu/en/publications-data/hiv-modelling-tool) importable to Spectrum. Alternatively, estimates from other country-specific models based on case surveillance and clinical data may be input to the national Spectrum epidemic file if the methods have been peer-reviewed and published.
• For countries with one or more national household population serosurveys that reported the number of respondents living with HIV who know their HIV-positive status or have ever been tested for HIV, UNAIDS recommends modelling knowledge of HIV-positive status using the Shiny90 model, integrated in Spectrum.
• Knowledge of HIV-positive status based on only self-reported data or historical household population survey data about testing history but without HIV serology should not be used.
• For the denominator: estimation models such as Spectrum are the preferred source for the number of people living with HIV. If a non-Spectrum model is used, UNAIDS will work with the country to emulate the external estimate of people living with HIV in Spectrum, for validation and for inclusion in regional and global estimates.
• On estimating the number of children who know their status in countries with modelled estimates based on household survey data: household surveys are often restricted to respondents of reproductive age, but Spectrum offers a standardized method to estimate knowledge of HIV status among children aged 0–14 years, considering national annual data on children on antiretroviral therapy and treatment interruptions, alongside estimated mortality among children on antiretroviral therapy and ageing-out from the child cohort.