The International HIV-Cognition Working Group has published an account of its own establishment and its updated position on how cognitive impairment in people with HIV should be named and classified. The paper, a single-author review by Sam Nightingale of the HIV Mental Health Research Unit at the University of Cape Town, appeared in Current HIV/AIDS Reports (volume 23, article 15) on 9 April 2026. It restates the six consensus recommendations the group first published in Nature Reviews Neurology on 13 June 2023, chief among them that HIV-associated brain injury, which the group divides into active and legacy forms, be conceptually separated from the other causes of brain injury that occur in people with HIV. It is a framework paper: no trial, no cohort, no new measurements.
The target is the 2007 research nosology for HIV-associated neurocognitive disorders, published in Neurology on 3 October 2007 by Antinori and colleagues and widely known as the Frascati criteria. Research using those criteria found that around 45 per cent of people with HIV met a HAND category; a 2020 systematic review and meta-analysis of 18 studies in Frontiers in Neurology put the pooled figure at 44.9 per cent. The working group's argument is that this figure is inflated by design. The 2023 statement noted that the criteria can falsely classify more than 20 per cent of cognitively healthy people as impaired, and the new review makes the point arithmetically: a measured low cognitive performance prevalence of 30 per cent reads very differently once the false classification rate is known to be 26 per cent. The group also argues that psychosocial confounds to cognitive testing are inadequately handled and that multifactorial brain injury is too readily attributed to HIV.
The review sets out what the group would put in place instead. A classification of cognitive impairment would require at least two of three elements: low cognitive performance, cognitive symptoms, and objective evidence of brain injury on neurological investigation. Cognitive symptoms would mean any change in cognition noticed by the person or by someone close to them, whether or not it affects daily functioning, which the HAND criteria required. Nightingale describes the approach as a pragmatic reframing intended to harmonise research terminology with clinical diagnosis, and notes that HAND was research criteria never intended for clinical use. The working group was initiated at Cape Town and settled at 20 members, nine of them (45 per cent) based in low-income and middle-income countries, including three representatives of the community of people living with HIV. Its approach was adopted by the European AIDS Clinical Society in its version 12.0 guidelines in October 2023.
The field has not settled. In correspondence published in Nature Reviews Neurology on 16 January 2024, Lucette Cysique, Bruce Brew and colleagues argued for a careful update of the HAND criteria rather than their replacement, warning that abandoning them would make new data incomparable with decades of existing research, and that the active and legacy distinction is not well defined and would be hard to implement where neuroimaging and other investigations are not available. The working group replied in the same issue that it does not seek to undermine the lived experience of people with HIV who have or are at risk of cognitive impairment, and that those individuals are better served by criteria that delineate mechanism and give clearer prognostic information. The 2026 review accepts that cognitive impairment remains an important complication of HIV, particularly as the population with HIV ages.