AIDSQuest: The HIV/AIDS Survey Library
Topic H: Risk and Prevention Behaviors

Overview
Risk and prevention behaviors frequently form the core of surveys related to HIV/AIDS. Behaviors such as sex without condoms and maintaining multiple sex partners place people at risk of HIV, and promoting safer sexual behavior is a key aspect of most AIDS programs. Drug and alcohol use place people at risk either directly, through sharing needles, or indirectly, via inhibited judgment. Partner violence is also increasingly recognized as a factor that leads to increased risk of HIV. Risk of mother-to-child transmission is influenced by factors such as breast-feeding practices. Risk and prevention behaviors have been separated into five sections: sexual risk and prevention practices, drugs and alcohol, mother-to-child transmission, violence, and sex work and trafficking.

Measurement issues
It is important to include a time period when measuring behaviors that is short enough so that respondents remember experiences (i.e., to minimize recall bias). For example, it is common to ask about the number of sexual partners during the past year, while asking about the number of drug use or sexual experiences during the past month.

There is some substantial evidence that self-reported behaviors can accurately reflect actual behaviors. However, reasonable concerns about over- or under-reporting of various behaviors still remain. The use of other data collection techniques can provide additional support to reported behaviors on a survey. Data can be "triangulated,"—collected about the same information via multiple sources, such as observation of couple interactions, or clinic service statistics.

Definitions of sexual partners vary, and it is important to include a definition in the survey itself, so that all respondents refer to the same definition. Differences between "primary," "regular," or "marital" partners and "secondary," "casual" or "non-regular" partners are particularly important, as behaviors related to the two different types of partners are often very different. Therefore, including questions about both types of partners is common. Many surveys ask about a range of sexual behaviors, reflecting a continuum of HIV risk, including anal sex, vaginal sex, and oral sex. Many surveys have also begun to explore patterns of sexual networking rather than simply individual sexual behaviors. This may provide additional information about those at risk. It is important to explore differences in behaviors between youth and adults, as well as between men and women. There is a difference between questions that ask about sexual orientation and same-sex sexual behavior. The first addresses a self-identity and the other addresses behaviors only. Depending upon the goal of the survey, one or both pieces of information would be important.

Condom use is most commonly measured over a period of a few months, and can be measured as a proportion of use during acts of intercourse in the past three months or how often condom use takes place over the last three months (always, sometimes, never), with the first example being more specific and accurate. Questions about condom use during last sex are also common. It may be useful to ask both questions, to compare and validate responses.

Violence is most often separated into multiple categories, the most common being physical and sexual violence. Other types of violence include psychological/emotional. If you are studying violence, you should specify which type of violence you are referring to and include separate questions about the different types, so that estimates can be made of all types separately, as well as combinations of them. The topic of violence should be introduced with care, because of its sensitive nature. WHO and PATH have created a practical guide regarding how to conduct research on this topic, which is included here as an appendix.

As with other types of questions, it is important to keep comparability in mind when designing a survey. This is especially true if you are going to ask similar questions within the same survey, collect more than one round of data (e.g., pre-and post-tests) or if you will want to compare your results to other studies. For example, condom use questions can be asked in different ways:

"How often did you use a condom with your partner during the last 3 months?"

and

"Did you use a condom your partner during your last sexual experience?"

Responses to each would be different and could not be compared.

The same situation could happen even when using the same question but different answer choices: "How often did you use a condom with your partner during the last 3 months?" with the answer choices including: 10 times vs. every day, once a week, once a month vs. often, somewhat often, and not often. Responses could not easily be compared.

Examples of questions related to risk and prevention behaviors have been drawn from the questionnaires below:

1. Sexual risk and prevention practices

• Horizons—Targeted vs. General Population Interventions for STD Control

• Horizons—Female Condom User Study

• CDC—Behavioral Surveillance Workgroup Core Questions

• PSI—Condom Social Marketing Consumer Profile

• DHS—AIDS module

• FHI—Behavioral Surveillance Survey

• UNAIDS/MEASURE—Evaluation HIV/AIDS Prevention Indicator Survey

• WHO—WHO Multi-country Study on Women’s Health and Life Experiences

• Population Council—RAPIDS Baseline Survey: Questions for youth aged 15-18 

2. Drugs and alcohol

• FHI—Behavioral Surveillance Survey

• CDC—Youth Risk Behavior Survey

• CDC—Behavioral Surveillance Workgroup Core Questions

• UCSF—Project Access Qualitative Guide

3. Mother-to-child transmission

• Horizons—Testing Clinic- and Community-based Strategies for Reducing Mother-to-Child Transmission of HIV

4. Violence

• Horizons—HIV-associated Violence: Implications for HIV Counseling and Testing Programs

• MEASURE Evaluation—Perform System of Indicators Survey in Uttar Pradesh, India-Men's Questionnaire

• FHI—Behavioral Surveillance Surveys

• DHS—AIDS module

5. Sex work and trafficking

• Horizons—Determining Needs and Opportunities to Address Trafficking of Girls and Women

• Horizons—STI/HIV Intervention Study Among Miners and Sex Workers in Carletonville

• FHI—Behavioral Surveillance Surveys: Female Sex Worker Module

Print this page

@ E-mail this page

This page updated 21 August 2006

---