THE INCIDENCE OF SEXUALLY TRANSMITTED infections (STIs) continues to rise in North America with prevention strategies remaining a challenge.1–4 Over the last 2 decades, these strategies have been based on the assumption that increased perceptions of personal STI/HIV risk motivate individuals to adopt safer sexual behaviors, even though studies have not consistently supported this relationship.5–9 Behaviors promoted as effective in preventing STI/HIV transmission include sexual abstinence, consistent use of latex condoms, and choosing a sexual partner who is uninfected with an STI/HIV. Of particular concern is evidence suggesting that many individuals are selecting partners they believe to be “safe” or uninfected.10,11 These beliefs, however, are often based on intuitively “just knowing” whether a partner is safe, and not based on definitive evidence such as STI test results.12–17 Qualitative studies propose several theoretical explanations for the use of these beliefs.
Evidence suggests that people are using stereotypical beliefs based on visible and inferred personal characteristics (e.g., appearance, profession, education) and the type of relationship one has with that partner (e.g., a known partner, a trusted partner) to evaluate their partner's level of STI/HIV risk.9,10,18–21 For example, a person met in a bar may be perceived as being more likely to be infected with an STI/HIV than someone met at a fitness centre. If the impression of a sexual partner is not consistent with the image of someone infected, the possibility of that partner being infected may be discounted. Sexual partners who are seen in a positive light, particularly those who possess the qualities sought in a relationship partner, (e.g., intelligent, polite), are unlikely to be considered a source of risk for STI/HIV infection.9,12,20,21
Other qualitative research has identified the use of inaccurate heuristics (known partners and trusted partners are safe partners) as an explanation for some people's failure to use safer sexual behaviors. Familiarity of the partner is often used as an index of trust and safety. As an individual accumulates more information about a partner, feelings of familiarity increase resulting in less concern about the risk of STI transmission.17,22–25 Qualitative studies have consistently found that partners who are known, liked, trusted, and considered similar to oneself are more likely to be perceived as safe.12,16 The decision to not practice safer sex with partners may also be based on the individual's reluctance to link risk or disease with loving or caring. As individuals are drawn to their partners by feelings of attraction and affection, these feelings interfere with rational risk assessments and may result in the partner not being seen as a potential source of infection.21,23,26 This occurs even when the only information known about the partner is irrelevant to STI status.14 These findings are consistent with the early research on interpersonal attraction which identified familiarity, similarity to oneself, and trust as predictors of attraction and intimacy.27 In this study, we have attempted to integrate the stereotypical beliefs and heuristics that have been found through qualitative research to influence perceptions of partner safety.12,16,17,20,21 We have labeled assumptions that are based on partner attributes and relationship characteristics such as familiarity, trust, similarity, likeability, appearance, and seriousness of the relationship, as partner safety beliefs (PSBs).
Few quantitative studies have explored whether individuals are using PSBs in their assessment of sexual partner STI/HIV risk, and whether this assessment subsequently influences perceptions of personal STI/HIV risk. Therefore, the primary purpose of this research was to determine: (a) whether individuals use PSBs as a means for evaluating the STI/HIV risk of a sexual partner, and (b) the influence of demographic characteristics, other known STI risk factors (e.g., condom use and number of sexual partners), and PSBs, on an individual's perception of his or her own STI/HIV risk.
Materials and Methods
Study Design, Sample, and Recruitment
This study used a cross-sectional, correlational survey design to collect data on factors used by heterosexual men and women to determine the STI/HIV safety of a sexual partner. Participants were recruited from a free, public sexually transmitted disease (STD) clinic located in a large urban centre. Approximately 1000 new and returning clients visit the clinic each month seeking testing or treatment for STIs. To decrease the heterogeneity of the sample, participation was limited to STD clinic clients who: (a) were adults 19 years of age and over, (b) self-identified as heterosexuals, (c) reported having had penetrative penile vaginal or penile rectal sex only with partners of the opposite sex during the last 6 months, (d) believed they were HIV negative, (e) were attending the clinic for an initial assessment visit (e.g., were not yet diagnosed) and, (f) were able to understand spoken and written English. HIV-positive clients were excluded from this study because we hypothesized that their concerns and perceptions of risk would differ from those who believed themselves to be HIV negative.
Eligible clients completed a 26-item questionnaire in private during their clinic visit. Each completed questionnaire was placed in a sealed envelope and exchanged at reception for a $10 honorarium. Permission to conduct this study was obtained through the University of British Columbia's ethics review board. Data collection occurred between September 2004 and January 2005. A total of 430 clinic clients met the eligibility criteria of which 317 (74%) completed questionnaires. The most commonly cited reason for not participating among those eligible was time constraints. No demographic information was collected on the screening form, prohibiting comparison between participating and nonparticipating eligible subjects.
Partner Safety Beliefs Scale.
The use of partner attributes and relationship characteristics for evaluating the potential infectivity of a hypothetical sexual partner was assessed using 16 statements representing 7 theoretical categories: familiarity, trust, similarity, likeability, superficial traits (e.g., physically attractive, looks healthy/clean), assumed sexual history, and seriousness of the relationship (Table 1). We derived some statements from scales reported in previous STI/HIV research10,15 and developed others from qualitative studies.12,16,17,20 Participants were asked to rate their level of agreement with each statement using a 5-point Likert response scale ranging from 1 (strongly disagree) to 5 (strongly agree). The 16 responses were summed to obtain a total PSBS score with a range of 16 to 80. Higher scores represented a higher level of endorsement, indicating greater reliance on personal attributes and relationship characteristics as a determinant of STI/HIV safety. Exploratory factor analysis of the scale, using unrotated principal component extraction, supported a single factor structure accounting for 46% of the variance (Table 1 for factor loadings). Internal consistency was high with a Cronbach's α coefficient of 0.92.
Personal STI/HIV Risk.
Participants were asked to rate their chances of contracting an STI (including HIV) within the next year if they did not use a condom. Responses were measured using a 10 cm visual analogue scale (VAS) anchored at 0% (no chance) and 100% (very likely).
Condom Use in the Last 6 Months.
Participants were asked to estimate the percentage of time they had used condoms for STI/HIV prevention in the last 6 months (considering all partners). Percentage condom use was measured on a 10 cm VAS anchored at 0% (never) and 100% (always).
Current Sexual Partners.
Participants were asked to choose 1 of 6 categorical options that best described their number of current sexual partners. Responses were dichotomized as 1 partner and 2 or more partners.
Sexual Partners in the Last 6 Months.
Participants were also asked to estimate how many people they had sex with in the previous 6 months. Responses were dichotomized as 0 to 4 partners, and 5 or more partners, with the latter representing the highest 15% of responses.
Drug and Alcohol Use in the Last 6 Months.
Participants were asked to select 1 of 8 responses, ranging from never to every day, to describe how often they had been drunk on alcohol or high on drugs during the previous 6 months. Responses were dichotomized to create 2 new variables Frequent Drug Highs (yes/no) and Frequently Drunk (yes/no), with frequent being defined as more than once or twice a month.
Univariate and bivariate statistics (t-tests, χ2 tests, ANOVAs) were computed for the main study variables. Correlation coefficients were calculated to determine significant bivariate relationships before performing hierarchical linear regression analysis to examine the influence of potential explanatory variables on perceived personal STI/HIV risk. Explanatory variables were chosen according to the Hosmer-Lemeshow method of including only those with correlational P values less than 0.10. Independent variables were entered into the model in the following theoretical order: (a) gender, education, and income; (b) condom use in last 6 months, number of current sexual partners, and number of sexual partners in the last 6 months; and (c) PSBS scores. Thus, we were able to examine the influence of PSBSs after accounting for demographic characteristics and other known STI risk factors.
In general, participants tended to be male (62%), white (85%), single (77%), and educated with a university degree (Table 2). Respondents ranged in age from 18 to 66 years (M = 31.0; SD = 9.2). Less than half the sample (47%) reported having a previous STI; 32% and 42% reported high frequencies of drug highs and alcohol intoxication, respectively. Results for frequent drug highs and frequently drunk were above the national average, suggesting that the sample may not be representative of the general population.28 Females reported higher levels of perceived personal STI/HIV risk, whereas males reported higher numbers of sexual partners (currently and during the last 6 months), and were more frequently drunk or on a drug high. There were no significant differences by gender for the PSB items or total PSBS scores.
Endorsement of STI/HIV Safety Beliefs and PSBS Scores
Table 3 presents the frequencies of endorsement for the PSB statements. Four of the 16 statements were endorsed by more than 60% of the participants; we considered these to be “frequently endorsed.” They included 2 statements representing familiarity (I felt I knew the person; I knew about the person's lifestyle), 1 statement representing trust (I felt I could trust the person), and 1 representing sexual history (I felt I knew the person's sexual history). The mean score of the PSBS was 50.80 (SD 13.00) indicating that, on average, participants were relying on PSBs when evaluating the sexual safety of their partners (a neutral response would be represented by a score of 48).
Personal STI/HIV Risk and Condom Use
There was a large variance in estimates of personal STI/HIV risk (M = 49%, SD = 32%) and condom use (M = 56%, SD = 31%). Sixty percent of the participants indicated that they used condoms less than 75% of the time; yet only 38% rated themselves as having more than a 50% chance of contracting an STI/HIV within the next year if they did not use a condom.
Demographic, STI/HIV Risk Factors, and PSB Scores Predicting Personal Risk
Correlation coefficients for main study variables are reported in Table 4. A higher estimate of personal risk was associated with being female, more frequent condom use, lower PSBS scores, number of current sexual partners, and number of partners in the past 6 months. One-way analysis of variance indicated that personal risk was also associated with education F(2306) = 4.00 (P <0.05) and income F(2304) = 8.40 (P <0.01).
The multiple regression model included PSBS scores, and 6 demographic and other known STI risk factor variables: gender, education, income, general condom use, number of current sexual partners, and number of partners in the last 6 months. All variables except gender and condom use were found to be significant predictors of personal risk, with the final model accounting for 21.3% of the variance (Table 5). Results showed that higher perceived risk was associated with lower levels of education (i.e., no postsecondary or trade school) and an annual income less than $50,000. Results for numbers of partners were as expected—those with more partners estimated their personal risk as being higher. Higher PSBS scores were associated with lower perceived personal risk indicating that greater reliance on PSBs is associated with lower perceptions of risk.
The results of our study suggest that clients of this STD clinic are using partner attributes and relationship characteristics (familiarity, trust, and assumed knowledge of a partner's sexual history) as an index for evaluating sexual partner safety. These results are consistent with qualitative research findings indicating that when people feel they “just know” their partner, they judge the partner to be safe even in the absence of any STI/HIV testing.9,14,15
Previous research has shown that trustworthiness of a partner is one of the most influential factors for engaging in unprotected sex.16 Although most individuals are confident in their assessment of their partner's character, their assessments of their partner's STI risk factors have been shown to be inaccurate when compared to the partner's self-reported behaviors.29 Moreover, once trust has been “established,” individuals assume their partner is safe, even when there is evidence that the partner has engaged in risky behaviors.10,14,15,22,25,30,31 The influence of trust in the determination of partner safety is particularly problematic because risky sexual behavior within a relationship is seen as a symbol of trust.
Reasons for the use of these presumed indicators of partner safety are unclear. There is evidence that STI/HIV prevention campaigns have been successful in increasing public awareness about the epidemiologic predictors of STI/HIV (i.e., number of partners, previous STIs, condom use), but this has generally not translated into behavior change.32–34 Moreover, few researchers have considered whether individuals use these predictors when evaluating their own STI/HIV risk. An exception is a study of adolescent women that found participants were aware of what defines high risk sexual behavior, but those engaging in the behaviors rated their chances of contracting an STI/HIV as low because they did not consider their partner a risk.35 One proposed explanation cites the early HIV prevention recommendations to “know your partner well” and adopt safer sexual behaviors in circumstances involving sex with anonymous, casual, or high risk partners (e.g., sex workers).22,23 This message may have contributed to an increase in risky behavior if it was interpreted to mean that superficial knowledge about a sexual partner is sufficient for assessing his/her STI risk.
Our multivariate analyses showed that the number of sexual partners in the last 6 months and number of current sexual partners were predictive of personal risk, suggesting that participants were aware of these known risk factors. However, PSBs contributed to their estimates of personal STI/HIV risk even after accounting for these risk factors and other demographic characteristics. As PSBS scores increased, perceived personal STI/HIV risk decreased. By assessing the STI/HIV infectivity of a partner using inferred attributes and characteristics, some individuals may be choosing to believe that this is an effective strategy for avoiding exposure to STIs or HIV.
To be effective, strategies for selecting safe partners require accurate knowledge of both partners' status, and the willingness of both to honestly discuss their status. However, previous studies have shown that knowledge of a partner's sexual history is often gathered only indirectly through ambiguous communication,21 consistent with our findings about assumed knowledge of sexual history. Discussion of one's prior sexual activities is often considered taboo, especially during impression formation.18,22,36 Although people may vaguely ask about their partner's sexual history, very few ask directly about STIs or HIV status.37 Furthermore, those who have been tested for STIs are often reluctant to share their own results or disclose their sexual history when starting a new relationship38 for fear of appearing more of a risk to potential partners and creating a threat to the relationship. There is a large body of research that addresses communication between sexual partners, but more studies are needed to determine how interpersonal relationships influence perceptions of risk and sexual decision making.
This study was limited by its correlational design; a longitudinal study would provide stronger support for causal links between study variables. The reliance on self-report measures of sensitive information may be subject to social desirability response bias. In addition, recall biases may have contributed to inaccuracies in the data collection as participants were required to recall information over the previous 6 months rather over a shorter time period (e.g., 4 weeks) which is considered to be more reliable.39,40 Although the PSBS measured whether participants would consider a partner to be safe, it was not determined whether participants would engage in unprotected sex, given a safe partner assessment. While participants were asked to think of a hypothetical sexual partner when completing the questionnaire, it is not known whether participants were thinking of a specific partner, and if so, whether they had actual or assumed knowledge of that partner's sexual history and STI status. Future research in this area would be strengthened by collecting data on the type of STI/HIV information actually discussed between partners.
This study has several key implications for STI/HIV prevention education and further research. Prevention efforts must begin to acknowledge that feelings of familiarity and trust toward a sexual partner make it less likely that those partners will be viewed as a health threat, consequently reducing the impact of relevant clinical information related to safer sexual behaviors. Prevention strategies must be developed to counter people's tendencies to use assessments based on partner attributes and relationship characteristics in determining a sexual partner's safety. Campaigns should also publicize the possibility of asymptomatic infections and encourage a culture of frequent STI screening. In view of our findings that men perceived themselves to be at lower STI risk than females, despite having had more partners and higher frequency of drug use, efforts should be made to raise their level of awareness regarding their risk for STI, and to increase their frequency of STI screening. Most importantly, the context of a relationship must be acknowledged, with measures incorporated to make individuals aware of behavioral tendencies during the developmental stages of a relationship, particularly the tendency to rely on feelings of familiarity, trust, and assumed knowledge of a partner's sexual history, rather than testing. Developing interventions that target assumptions of safety and dispel incorrect beliefs about the selection of safe partners is needed to promote safer sexual behavior. The PSBS developed in this study could be a useful tool for testing the effectiveness of these new educational strategies.
1. Health Canada. 1998/1999 Canadian sexually transmitted diseases (STD) surveillance report. Volume: 26S6. Ottawa: Division of STD Prevention and Control, Bureau of HIV/AIDS, STD and TB, Centre for Infectious Disease Prevention; 2000.
2. Health Canada. HIV/AIDS Epi updates April 2003. Ottawa: Centre for Infectious Disease Prevention and Control; 2003.
3. Health Canada. Women's health surveillance report: A multi-dimensional look at the health of Canadian women. 2003.
4. Centres for Disease Control and Prevention (CDC). Tracking the hidden epidemics: Trends in STDs in the United States. 2000.
5. Albarracin D, Johnson BT, Fishbein M, et al. Theories of reasoned action and planned behavior as models of condom use: A meta-analysis. Psychol Bull 2001; 127:142–161.
6. Ellen JM, Adler N, Gurvey JE, et al. Adolescent condom use and perceptions of risk for sexually transmitted diseases: A prospective study. Sex Transm Dis 2002; 29:756–762.
7. Gerrard M, Gibbons FX, Bushman BJ. Relation between perceived vulnerability to HIV and precautionary sexual behavior. Psychol Bull 1996; 119:390–409.
8. Agocha VB, Cooper ML. Risk perceptions and safer-sex intentions: Does a partner's physical attractiveness undermine the use of risk-relevant information? Pers Soc Psychol Bull 1999; 25:746–759.
9. Fisher JD, Fisher WA. The information-motivation behavioral skills model of AIDS risk behaviour change: Empirical support and application. In: Oskamp S, Thompson SC, eds. Understanding and Preventing HIV Risk Behaviour: Safer Sex and Drug Use. Thousand Oaks: Sage, 1996:100–127.
10. Clark LF, Miller KS, Harrison JS, et al. The role of attraction in partner assessments and heterosexual risk for HIV. In: Oskamp S, Thompson SC, eds. Understanding and Preventing HIV Risk Behavior: Safer Sex and Drug Use. Thousand Oaks: Sage, 1996:80–89.
11. Hoffman V, Cohen D. A night with Venus: Partner assessments and high-risk sexual encounters. AIDS Care 1999; 11:555–66.
12. Hammer JC, Fisher JD, Fitzgerald P, et al. When two heads aren't better than one: AIDS risk behavior in college-age couples. J Appl Soc Psychol 1996; 26:375–397.
13. Keller ML. Why don't young adults protect themselves against sexual transmission of HIV? Possible answers to a complex question. AIDS Educ Prev 1993; 5:220–233.
14. Maticka-Tyndale E. Sexual scripts and AIDS prevention: Variations in adherence to safer-sex guidelines by heterosexual adolescents. J Sex Res 1991; 28:45–66.
15. Misovich SJ, Fisher JD, Fisher WA. The perceived AIDS-preventive utility of knowing one's partner well: A public health dictum and individuals' risky sexual behaviour. Can J Hum Sex 1996; 5:83–90.
16. Skidmore D, Hayter E. Risk and sex: Ego-centricity and sexual behaviour in young adults. Health Risk Soc 2000; 2:23–32.
17. Swann WBJ, Silvera DH, Proske CU. On knowing your partner: Dangerous illusions in the age of AIDS? Pers Relatsh 1995; 2:173–186.
18. Gold RS, Skinner MJ. Judging a book by its cover: Gay men's use of perceptible characteristics to infer antibody status. Int J STD AIDS 1996; 7:39–43.
19. Gold RS, Skinner MJ, Hinchy J. Gay men's stereotypes about who is HIV infected: A further study. Int J STD AIDS 1999; 10:600– 605.
20. Offir JT, Fisher JD, Williams SS, et al. Reasons for inconsistent AIDS-preventive behaviors among gay men. J Sex Res 1993; 30:62–69.
21. Williams SS, Kimble DL, Covell NH, et al. College students use implicit personality theory instead of safer sex. J Appl Soc Psychol 1992; 22:921–933.
22. Misovich SJ, Fisher JD, Fisher WA. Close relationships and elevated HIV risk behavior: Evidence and possible underlying psychological processes. Rev Gen Psychol 1997; 1:72–107.
23. Kelly JA, Kalichman SC. Increased attention to human sexuality can improve HIV-AIDS prevention efforts: Key research issues and directions. J Consult Clin Psychol 1995; 63:907–918.
24. Swann WBJ, Gill MJ. Confidence and accuracy in person perception: Do we know what we think we know about our relationship partners? J Pers Soc Psychol 1997; 73:747–757.
25. Fortenberry JD, Tu W, Harezlak J, et al. Condom use as a function of time in new and established adolescent sexual relationships. Am J Public Health 2002; 92:211–213.
26. Kirkman M, Rosenthal D, Smith AMA. Adolescent sex and the romantic narrative: Why some young heterosexuals use condoms to prevent pregnancy but not disease. Psychol Health Med 1998; 3:355–370.
27. Berscheid E, Reis HT. Attraction and close relationships. In: Gilbert DT, Fiske ST, Lindzey G, eds. The Handbook of Social Psychology. Vols. 1 and 2, 4th ed. New York: McGraw-Hill, 1998:193–281.
28. Federal, Provincial and Territorial Advisory Committee on Population Health. The statistical report on the health of Canadians. Ottawa: Health Canada, Statistics Canada, Institute for Health Information. Health Canada. Ottawa: Statistics Canada, Institute for Health Information; 1999.
29. Stoner BP, Whittington WL, Aral SO, et al. Avoiding risky sex partners: Perception of partners' risks v partners' self reported risks. Sex Transm Infect 2003; 79:197–201.
30. Gebhardt WA, Kuyper L, Greunsven G. Need for intimacy in relationships and motives for sex as determinants of adolescent condom use. J Adolesc Health 2003; 33:154–164.
31. Thorburn S, Harvey SM, Ryan EA. HIV prevention heuristics and condom use among African-Americans at risk for HIV. AIDS Care 2005; 17:335–344.
32. Fisher JD, Fisher WA. Changing AIDS-risk behavior. Psychol Bull 1992; 111:455–474.
33. Barden-O'Fallon JL, de Graft-Johnson J, Bisika T, et al. Factors associated with HIV/AIDS knowledge and risk perception in rural Malawi. AIDS Behav 2004; 8:131–140.
34. James S, Reddy SP, Taylor M, et al. Young people, HIV/AIDS/STIs and sexuality in South Africa: The gap between awareness and behaviour. Acta Paediatr 2004; 93:264–269.
35. Ethier KA, Kershaw T, Niccolai L, et al. Adolescent women underestimate their susceptibility to sexually transmitted infections. Sex Transm Infect 2003; 79:408–411.
36. Comer LK, Nemeroff CJ. Blurring emotional safety with physical safety in AIDS and STD risk estimations: The casual/regular partner distinction. J Appl Soc Psychol 2000; 30:2467–2490.
37. Crosby RA, Yarber WL, Meyerson B. Prevention strategies other than male condoms employed by low-income women to prevent HIV infection. Public Health Nurs 2000; 17:53–60.
38. Wolitski RJ, Branson BM. “Gray area behaviors” and partner selection strategies. In: O'Leary A, ed. Beyond Condoms: Alternative Approaches to HIV Prevention. New York: Kluwer Academic, 2002:173–198.
39. Poppen PJ, Reisen CA. Perception of risk and sexual self-protective behavior: A methodological critique. AIDS Educ Prev 1997; 9:373–390.
40. Weinstein ND, Nicolich M. Correct and incorrect interpretations of correlations between risk perceptions and risk behaviors. Health Psychol 1993; 12:235–245.