The Johns Hopkins University Bloomberg School of Public Health, Department of International Health (at time of publication)

"Unfortunately, in the field of behavioural HIV prevention, interventions are frequently implemented without clearly conceptualized operational models and the field lacks an accepted framework for how to classify intervention strategies."

This 32-page document is premised on the claim that there is little consistency in how interventions designed to address HIV and AIDS are described and defined. Based on a review of the scientific literature (and drawing on data collected from the Synthesizing Intervention Effectiveness project, which is being conducted jointly between the Johns Hopkins University School of Public Health and the World Health Organization (WHO)'s Department of HIV/AIDS), the author and his colleagues first examine how the field of HIV behavioural prevention science has been described and categorised historically. The initiatives explored include the following features: voluntary counselling and testing (VCT), condom social marketing, partner notification, family planning for HIV-positive women, mass media, messages focused on abstinence and/or abstinence-only, needle and syringe exchange, psychosocial support, and medical treatment's impact on HIV risk behaviour.

Citing specific examples of interventions reflecting the above-cited approaches, the author, Michael Sweat, finds that most interventions were combinations of multiple activities; for instance, several distinct and competing frameworks for defining interventions were used, including by the following: type of activity, mode of delivery, "target" population, setting, commodity provided, outcome sought, and/or theoretical orientation. Despite this, according to the author, investigators tended to identify the intervention by a single, often prominent, component. In addition, "frequently, detailed descriptions of interventions were missing completely and only a word or brief phrase was used to describe the activities conducted". Sweat argues that "the current ad hoc manner of conveying the content of interventions results in difficulty in establishing quality standards."

The solution proposed here is an ontology, which is a specific system of classification based on an explicit logic and purposefully designed with objective criteria in mind. Sweat and his colleagues here provide criteria for an ontology of HIV interventions, which is shaped by the following:

• "There is a need for concise definitions for commonly implemented intervention approaches, based on a consistent definitional logic.
• To enhance specificity, interventions should be foremost defined by activities or services, and commodities provided.
• When relevant, the secondary focus of intervention definition should be based on message content.
• Outcomes along the causal chain of events should not be the primary basis of the definition of an intervention.
• Defining interventions solely by the intervention's theoretical basis leads to a lack of clarity and should be avoided.
• A list of intervention characteristics is needed, which should be described in written reports to complement the use of shorthand phrases for interventions.
• Carefully designed consensus-building is needed to wed theory to practice....This is a continuing process, subject to updating as approaches in the field evolve and populations change."

The following steps are recommended for establishing clear definitions for HIV and AIDS interventions:

1. Identify brief (shorthand) definitions describing activities or services and commodities provided in the intervention. When relevant, also refer to key message content included with the intervention.
2. Through consensus meetings with key stakeholders and experts, provide a rich description of other requisite components of the intervention, including message content, the method of delivery, how to tailor the intervention to audiences and settings, the intervention's desired outcomes, and its theoretical basis.

In summary, Sweat and his colleagues propose that an intervention should be defined primarily by the activity or service and commodity provided. However, in applying interventions to specific settings they should be tailored to the realities of the epidemic scenario and target population, as recommended by the Joint United Nations Programme on HIV/AIDS (UNAIDS)' Practical Guidelines for Intensifying HIV Prevention [PDF] [1]. Key recommendations of the guidelines include knowing an epidemic in terms of its key drivers, vulnerable populations most in need, and epidemiological scenarios, as well as focusing on the source of new infections. The authors "propose that the preliminary list of interventions described herein be expanded and subjected to further review and critique by prevention scientists and practitioners. Once consensus is established and definitions are agreed upon for major interventions necessary for intensifying HIV prevention efforts, minimum quality standards should be developed for each intervention component and estimated costs for implementation established. Finally, further analysis of the optimal manner of bundling these interventions into logical combinations should be determined. With the establishment of clearly defined interventions with associated quality standards, estimated costs and optimal bundling strategies, it will be possible to provide improved recommendations to programme planners and policy-makers striving to intensify HIV prevention programmes. Consensus in these areas would enhance our ability to plan, allocate requisite resources - including human resources - advocate prevention, and monitor and evaluate the success of prevention programmes."

Various tables are included in the report, such as one listing many of the major HIV prevention intervention activities that are currently being implemented. This particular list defines and classifies intervention by the activity, service, or commodity, with additional information provided to allow tailoring of the intervention as needed. The interventions are also grouped into broad categories, based largely on each intervention's intended purpose(s). The table also describes frequently seen examples of the message content, delivery mode, "target" population, setting, intervention outcome, and theory associated with each intervention. The interventions are also grouped in the table according to the following categories: i) interventions that affect knowledge, attitudes, and beliefs and influence psychological and social correlates of risk; ii) harm reduction interventions that lower the risk of a behaviour, but do not eliminate the behaviour; iii) biological/biomedical interventions that strive to reduce HIV infection and transmission risk; iv) mitigation of barriers to prevention and negative social outcomes of HIV infection; and v) mitigation of biological outcomes of HIV infection. The paper also includes a summary of several hybrid interventions that bundle discrete intervention approaches. For example, condom social marketing typically includes commodity procurement and logistics management, mass media, and small media marketing tailored to specific audiences being addressed, and monitoring and evaluation providing feedback on programme effectiveness. "The logic for bundling these intervention components is based on the need not only to provide the commodity of condoms efficiently in the market place, but also to generate demand for condom sales and address social and cultural barriers to uptake and use of condoms." Several case studies of interventions being implemented, and the logical basis for their bundling, are featured in the concluding segments of the report.

UNAIDS website [8].