Table 1.

Definitions of key terms used in this paper

Transmission dynamics model: A model in which the force of infection changes over time due to direct and indirect effects from changes in the proportion of individuals living with transmissible HIV (i.e. virally unsuppressed) [26].

Basic reproduction number, 𝓡0: The average number of secondary transmissions from a person living with HIV in an otherwise completely susceptible population. If 𝓡0 >1, HIV has the potential to spread in the population, whilst if 𝓡0 <1, sustained HIV transmission is unlikely. Conceptually, it depends on the contact rate (c), the duration of time virally unsuppressed (D), and the transmission probability per contact (β). Other factors also affect 𝓡0, including population heterogeneity (vulnerability and exposure to HIV may vary across and within populations), and mixing patterns (how contact between groups varies, i.e. who mixes with whom) [27–30].

Force of infection, λ: The per capita incidence rate at which people susceptible in the population acquire infection [26]. It depends on the contact rate (c) (which can be conceptualised as accounting for mixing patterns by relevant population subgroups), the probability of transmission per effective contact (β), and the prevalence (I/N) of virally unsuppressed infection (I) among partners (N).

Structural determinants: The fundamental, foundational, underlying social, economic, political, cultural, organisational, and environmental determinants that affect HIV risks by shaping exposure patterns to risk and prevention factors (mediators) further downstream on the causal pathways [1].

Distal structural determinants: Macrolevel, aggregate structural determinants that affect whole populations, communities, or groups of individuals (e.g. key populations) [4, 31, 32]. They affect exposure to individual-level proximate structural determinants. Examples include laws and policies such as those governing sex work, sex between men, and drug use, but also alcohol and tobacco advertising, systemic and institutionalised racism, and inequitable norms surrounding gender, sexual identity, and substance use.

Proximate structural determinants: Structural factors experienced at an individual-level [4, 31, 32]. They are closer to and have more immediate effects on HIV risks. Examples include incarceration, stigma, discrimination, violence, housing instability, access, and availability of drugs.

Structural interventions: Interventions that promote the availability, accessibility, or acceptability of specific resources needed to prevent poor health outcomes or that reduce vulnerability to them [33]. They seek to mitigate the negative effects of structural determinants or prevent exposure to them (e.g. drug law reform that institutes drug treatment instead of incarceration). Structural interventions encompass both societal enablers and development synergies [34]. Societal enablers are social programmes, policies, and interventions that aim to remove barriers to accessing necessary health services. Examples include decriminalisation (e.g. of sex work, sex between men, and drug use/possession), community mobilisation, stigma reduction, and other specific interventions including the Avahan intimate partner violence intervention in India or the integration of self-help groups to empower FSW within the national sex worker programme in Zimbabwe [35, 36]. Development synergies are investments in other sectors that can have positive effects on HIV outcomes (e.g. HIV incidence, treatment use, mortality). Examples include investments in education, employment practices, gender equality, legal reform, as well as specific economic empowerment interventions, such as cash transfer interventions for women.

Exposure history: The specified duration of exposure as well as the time-varying intensities of exposure within different exposure periods (e.g. current, recent (< 6 months), and non-recent (≥ 6 months) exposures) [37]. Duration and time periods of exposure are usually based on the recall periods of the survey instruments that measure exposures, and intensities are based on the findings of analyses that assess the effects of exposure on causal pathways within those time periods.

Causal pathways: The chain of variables that causally link exposure to structural determinants and structural interventions to individual-level HIV risks.

Direct pathways: Causal pathways not involving mediators. These may represent that the mediators on the causal pathways are unmeasured and therefore unobserved.

Indirect pathways: Causal pathways that involve mediators.

Mediators: Intermediate variables on the causal pathways that link exposure to structural determinants and interventions to HIV risks. They are typically assumed or established as the main mechanisms through which exposure to structural determinants affects HIV vulnerabilities. Examples for structural determinants may include the number of sexual or injecting partners, the frequency of sex or sharing injection equipment, inconsistent condom use, and access to and uptake of HIV prevention and treatment, and others. Mediators may be observed or unobserved. The term ‘mediator’ to describe a variable is context specific. A variable that is a mediator on one causal pathway could be considered an independent exposure variable on another (e.g. pre-exposure prophylaxis (PrEP) use could be a mediator in analyses estimating the effect of exposure to HIV education on individual HIV acquisition risk and an exposure variable in analyses estimating the impact of PrEP use on HIV acquisition).

HIV outcomes: The last step in the causal pathways. These include HIV acquisition and onward HIV transmission, as well as individual-level HIV health outcomes such as HIV-related morbidity and mortality (e.g. disability-adjusted life years).