Abstract:BACKGROUND: After successful intensive interventions to rapidly increase HIV awareness, coverage of antiretroviral therapy (ART), and viral suppression, HIV programmes in eastern and southern Africa are considering scaling back of some interventions, such as widespread general population HIV testing. We aimed to model whether scaling back of general population HIV testing in South Africa could result in a resurgence of the HIV epidemic or substantial slowing of declines in HIV incidence, resulting in increased long-term ART.METHODS: In this modelling study, we used the Thembisa 4.5 model (a deterministic compartmental model of HIV transmission in South Africa) to project the South African HIV epidemic to 2100 assuming the continuation of 2022 epidemiological conditions and HIV programme implementation. We assessed how implementing reductions in general population HIV testing services in 2025 (while maintaining antenatal, symptom-based, and risk-based testing modalities and other HIV prevention services at 2022 levels) would affect HIV incidence and prevalence among people aged 15-49 years, the year in which incidence would reach one per 1000 people aged 15-49 years (the threshold for virtual elimination of HIV), and associated costs, as well as numbers of additional new HIV infections and AIDS-related deaths. We also modelled the effects of delaying reductions in general population testing services by 5-year increments. Additionally, we modelled the potential effects of reductions in general population testing services in combination with increases or decreases in ART interruption rates (ie, the annual rate at which people who are on ART discontinue ART) and condom usage in 2025-35.FINDINGS: If general population HIV testing services and the HIV risk environment of 2022 were maintained, we projected that HIV incidence would steadily decline from 4·95 (95% CI 4·40-5·34) per 1000 population in 2025 to 0·14 (0·05-0·31) per 1000 in 2100, and that the so-called virtual elimination threshold of less than one new infection per 1000 population per year would be reached in 2055 (95% CI 2051-2060). Scaling back of general population HIV testing services by 25%, 50%, or 75% in 2025 delayed time to reaching the virtual elimination threshold by 5, 13, or 35 years, respectively, whereas complete cessation of general population testing would result in the threshold not being attained by 2100. Although the incidence of HIV continued to fall when general HIV testing services were reduced, our modelling suggested that, with reductions of between 25% and 100%, between 396 000 (95% CI 299 000-474 000) and 2·50 million (1·97 million-2·98 million) additional HIV infections and between 115 000 (94 000-135 000) and 795 000 (670 000-926 000) additional AIDS-related deaths would occur between 2025 and 2075, depending on the extent of reduction in testing. Delaying reductions in general population HIV testing services for 5-25 years mitigated some of these effects. HIV testing accounted for only 5% of total programmatic costs at baseline; reducing testing moderately reduced short-term total annual costs, but increased annual costs after 25 years. Increases in ART interruption and reductions in condom usage were projected to slow the decline in incidence and increase the coverage of general HIV testing services required to control transmission but did not cause rapid resurgence in HIV infections.INTERPRETATION: Our modelling suggests that scaling back of general population HIV testing would not result in a resurgence of HIV infections, but would delay attainment of incidence-reduction targets and result in long-term increases in HIV infections, AIDS-related deaths, and costs (via increased need for ART provision). HIV programmes need to balance short-term potential resource savings with long-term epidemic control objectives.FUNDING: Bill & Melinda Gates Foundation.

Modelling the Epidemiological Impact of Scaling Up HIV Testing and Antiretroviral Treatment in China.

Cost-effectiveness of Different Intervention Strategies of HIV in Zhejiang, China.

Scaling Up Human Immunodeficiency Virus Screening and Antiretroviral Therapy Among Men Who Have Sex With Men to Achieve the 90-90-90 Targets in China

Tuberculosis in China: A longitudinal predictive model of the general population and recommendations for achieving WHO goals.

Optimizing health resource allocation for improving timely HIV diagnosis in China

Exploring the seasonality and optimal control strategy of HIV/AIDS epidemic in China: The impact of seasonal testing

[Impact of the 90-90-90 Goal and Pre-Exposure Prophylaxis on HIV Transmission and Elimination in Men Who Have Sex with Men in China: A Mathematical Modeling Study].

HIV/AIDS in China: the Numbers Problem

Projecting the Dynamic Trends of Human Immunodeficiency Virus/Acquired Immunodeficiency Syndrome: Modeling the Epidemic in Sichuan Province, China

Dynamic analysis and optimal control of HIV/AIDS model considering the first 95% target

Impact of Antiretroviral Therapy on the Spread of Human Immunodeficiency Virus in Chaoyang District, Beijing, China: Using the Asian Epidemic Model

Improvement in the 95-95-95 Targets Is Accompanied by a Reduction in Both the Human Immunodeficiency Virus Transmission Rate and Incidence in China

Clinical epidemiology of HIV/AIDS in China from 2004-2011.

[Expanding Test and Normalization for HIV/AIDS Prevention and Control].

Changing epidemiological patterns of HIV and AIDS in China in the post-SARS era identified by the nationwide surveillance system

Effect of Different Interventions for Latent Tuberculosis Infections in China: a Model-Based Study

Future HIV epidemic trajectories in South Africa and projected long-term consequences of reductions in general population HIV testing: a mathematical modelling study

Modeling and prediction of HIV in China: transmission rates structured by infection ages

The transmission of drug-resistant strains of HIV in heterosexual populations based on genetic sequences

Evolution Of Information-Driven Hiv/Aids Policies In China

Expanding human immunodeficiency virus testing: whole society participation and multilevel promotion