HIV RISK BEHAVIOR AND BELIEFS OF HIV-SEROPOSITIVE DRUG USERS
C. Hartgers, P. Krijnen, J.A.R. van den Hoek, R.A. Coutinho, J. van der Pligt
Journal of Drug issues 1992

INTRODUCTION

Sharing of needles and syringes is the main route of transmission of human immunodeficiency virus (HIV) among drug injectors [Friedland et al. 1985], while heterosexual transmission is the main route from injectors to non-injectors [Des Jarlais et al. 1985, Moss 1987].

Behavioral change is dependent on knowledge, availability of means, and motivation to change. Studies among Amsterdam drug users (DUs) (Korf et al 1990, Hartgers et al 1991) suggest that Amsterdam DUs are well aware of HIV transmission paths. Furthermore, Amsterdam has a large needle and syringe exchange program [Hartgers 1989] and condoms are easily available.

In relation to injecting drug use, many European and U.S. efforts to prevent new HIV infections focus on sexual and injecting risk behavior of persons who are not infected (both injectors and sexual partners of injectors). Behavioral changes among these HIVnegatives are - given that knowledge and means are present - dependent on a desire for self-protection against HIV infection. However, in order to prevent new HIV infections, it seems equally important to focus prevention programs at the - smaller - group of HIV-seropositive (HIV+) DUs. One reason for safe behavior among HIV+ DUs might be the belief that AIDS can be postponed by avoiding re- or other infections. However, to our knowledge, no evidence for this belief has been reported. Next to this desire for self protection, safe behavior of HIV-positive drug users depends on the wish to protect their injecting and/or sexual partners from acquiring HIV, which could be termed a wish for "other-protection".

Provided that knowledge and means are present, models of health behavior tend to focus on motivation for self-protection. However, some of the concepts employed in these models may be relevant for other protective behavior, for example the perceived self-efficacy with regard to a certain behavior (i.e., the conviction that one is able to perform a certain behavior consistently) [Bandura 1977], the perceived efficacy of the advocated ("safe") response in reducing HIV risk (response efficacy) [Rogers 1983], the attitude towards the behavior (Fishbein and Ajzen, 1975) and the intention to perform safe behavior [Fishbein and Middlestadt 1989] The present paper presents the results of a study among HIV+ DUs who are aware of their serostatus. Central to the aims of the study were the extent to which these drug users put others at risk of HIV infection, their perceptions of the risk of transmitting HIV, their intentions with regard to future behavior, and the characteristics of HIV+ DUs who put others at risk.

METHOD

We studied HIV+ DUs who participated between June and November 1990 in a longitudinal HIV study among DUs in Amsterdam. This ongoing cohort-study [Van den Hoek et al. 1988; Van den Hoek et al. 1989a; Van Haastrecht et al. 1991] includes voluntary and confidential HIV-testing and counselling, combined with a standardized interview conducted by trained professionals. Participants can either participate once or take part in the follow-up study (in which study-visits are scheduled every 4 months). Enzyme-linked immunosorbent assays (ELISA'S) are used for HIV-testing. Confirmation of a positive specimen is performed by competitive ELISA's and/or by immunoblotting, as previously described [Van den Hoek et al. 1988].

All HIV+ DUs who were familiar with their serostatus and who came for a follow-up visit in the above mentioned period were given an additional questionnaire (see below) besides the standardized interview on medical history, demographic factors, drug use and sexual history. This resulted in a sample of 122 HIV+ DUs, who came for a 2nd to 16th visit in the cohort study, and who reported 'current' behavior; i.e. behavior in the period since the preceding interview (median 4.2 months, range 3.3-12.6).

Univariate statistics include the X 2 test of independence, Fisher's exact test and the Mann-Whitney test (M-W) and Kruskal-Wallis test (K-W) for two or more independent samples (employing scales as described in table 1 and 2).

Predictors of unsafe vaginal sex were estimated by stepwise logistic regression modeling.

RESULTS
 

Demography; heterosexual and injecting risk behavior
The sample consists of 61% male and 66% Dutch DUs. Their mean age in years is 32.6 (s.d. 5.5) and they have lived a mean number of 16.7 years in Amsterdam (s.d. 11.8). Hav ing permanent housing is reported by 93%; currently working as a prostitute is reported by 15% (only females, hereafter indicated as female prostitutes). Almost all (98%) report a history of injecting. They have used heroin regularly for a mean number of 10.0 years (s.d. 4.7); current daily methadone use is reported by 79%.

Ninety eight (80%) DUs in the sample are (at the time of their visit) classified in CDC-I or II (acute or asymptomatic HIV infection), 18 (15%) in CDC-III (generalized lymphadenopathy) and 6 (6%) in CDC-IV (3 with symptomatic HIV disease, 3 with AIDS).

Sixty two of the 122 HIV+ DUs (51%) are heterosexually active, i.e. they report having had vaginal sex with a steady sexual partner (n=43), with a casual partner (n.=16) and/or with clients (n=18) in the previous months. Heterosexual activity is not significantly different between DUs in the different CDCsubgroups. Six DUs (all among the heterosexually active, is 10% of this group) report having had a sexually transmitted disease (STD) in the previous four months. Unsafe sex (defined as not (always) using condoms during vaginal sex in the preceding months) is reported by 29 HIV+ DUs (24%, 47% of the sexually active). When eight, who report only unsafe vaginal sex with an HIV+ steady partner, are excluded, this results in 21 HIV+ DUs who could have transmitted HIV via their sexual practices (hereafter called sex risk group or SR-group: 34% of sexually active HIV+ DUs). The SR-group consists of 14 females and 7 males. Thirteen HIV+ DUs (8 females, 5 males) report unsafe sex with a private steady partner, 4 (2 females, 2 males) with a private casual partner, and 4 HIV+ DUs (all female prostitutes) with a client. The 13 HIV+ DUs who had unsafe sex with a steady partner report that 8 (62%) of these steady partners have never injected. Four (19%) of the 21 SR-DUs report having had an STD in the previous months, as compared to 2 (5%) among the other 41 sexually active HIV+ DUs (Fisher exact test, p=0.09). Ninety five HIV+ DUs (78%) are current injectors. Four (=4%) of these report current needle sharing, defined as letting a needle and syringe, which they had already used themselves, be used by somebody else. These 4 persons do not report unsafe vaginal sex.

Risk estimates, beliefs, attitude and behavioral intentions
With regard to the risk estimates, 57 HIV+ DUs (47%) think they run a chance to get syphilis in the future (score a2), while 54 (44%) and 25 (21%; 26% of current injectors) think they might infect someone else with HIV in the future (score >2), respectively through sex or through needle sharing (see Table 1 for frequencies). All in all, 60 HIV+ DUs (49%) think they might infect someone with HIV in the future, either through sex, through needle sharing or through both.

The SR-group (n=21), when compared with the 41 heterosexually active and the 60 not heterosexually active subjects, think more often they might infect someone with HIV through sex in the future (76%, 44% and 33% with score >2 respectively, K-W p=0.0005).

With regard to response efficacy, 58 persons (48%) are uncertain (score c3) about the efficacy of using condoms (the advocated response to avert HIV risk) in preventing HIV transmission (Table 1). The SR-group, when compared with the other heterosexually active and the heterosexually non-active subjects, are least uncertain about condom efficacy (33%, 46% and 53% with score <3 respectively; K-W, p=0.02). A negative attitude towards condoms (score> 4) is reported by 38 persons (31%) (Table 1). The SR-group, when compared with the other two groups, does not have a significantly more negative attitude (38%, 20% and 37% with score >4 respectively; K-W, p=0.08). When only the first two groups are compared (the SR-group and the other heterosexually active), there is also no significant difference (M-W, p=0.48). Table 2 shows that 16 (94%) of the 17 female prostitutes in the sample who answered these questions have a strong or very strong intention to always use condoms with clients, but a high self-efficacy score (>4) with regard to condom use with clients is obtained by only 7 (41%). SR-prostitutes have a less strong intention to use condoms with clients than the other prostitutes (M-W, p=0.03), but there is no significant difference with regard to self-efficacy.

There were 56 HIV+ DUs with current private partners (i.e. who report a steady and/or casual heterosexual partner(s) while not being a female prostitute), of whom 42' answered the questions concerning intention, and self-efficacy with regard to condom use with future casual partners. Most (n=37, 88%) have a strong or very strong intention to use condoms with casual partners, but a high self-efficacy score was obtained for only 12 (29%). There were no significant differences between the SR-group and the other subjects with regard to intention and self-efficacy.

Finally, the subgroup of 95 current injectors were asked about their behavioral intentions about needle sharing: 87 (93%) had a strong or very strong intention not to share. It is worth noting that all four needle sharers reported a very strong intention not to share.

Correlates of heterosexual risk behavior
Since unsafe sexual behavior was more prevalent than unsafe injecting behavior, we only examined demographic and drug use correlates of sex risk (SR) behavior. Four demographic variables were univariately related to SR behavior (p<0.05): female gender, younger age, a non-Dutch nationality and being a female prostitute. No significant correlations with drug use variables (like current cocaine use, alcohol use or injecting) were found. Since all prostitutes in the present sample are female, a new variable (gender/prostitution) was made with three categories: males, female prostitutes and other females. In a logistic regression analysis, this variable and nationality were independent predictors of unsafe vaginal sex (with p<0.05). No interaction was found. The model could not be improved by entering response efficacy or attitude. With both nationality and gender/prostitution in the model, persons of non-Dutch nationality have an increased risk (OR=3.14, CI=1.07-9.24) when compared to Dutch DUs. Female prostitutes, when compared to females, have an increased risk of having unsafe vaginal sex (OR=3.49, CI=0.92-13.30. Males, when compared to females, have a decreased risk (OR=0.65, CI=0.18-2.31). However, the finding that female prostitutes are a high risk group does not mean that they have only unprotected contacts with clients. Four of the eight prostitutes in the SR-group had unsafe vaginal sex with clients, while the other four had safe sex with clients but unsafe sex with private partners.

' Initially, the questionnaire was unclear at this point. Interviewers assumed at first that HIV+ DUs with a steady partner did not have to answer these questions, hence the 14 (56 minus 42) missing values.

DISCUSSION

With regard to current behavior, three quarters of our sample of HIV+ DUs inject and half are heterosexually active. This confirms other findings that a substantial proportion of HIV-infected DUs are heterosexually active [Van den Hoek et al. 1990, 1992; Schrager et al. 1991].

Before discussing our findings, we would like to reiterate that the present sample consists exclusively of HIV+ DUs aware of their serostatus, and that the dependent variable is unsafe vaginal sex (i.e. the risk of HIVtransmission could not be excluded). Most research of heterosexual behavior among DUs concerns DUs with unknown serostatus and concerns condom use in general, irrespective of the likelihood of transmission. A further complication is that some studies are restricted to 100% condom use during heterosexual contacts as outcome variable, while others employ any condom use as outcome variable. We have tried to take these differences into account in comparing the results. Because the data are self-reported, our findings may be biased due to memory loss or a tendency to give socially desirable answers. Another limitation of our study is that the sample consists of DUs volunteering for an HIV test combined with an epidemiological study of HIV. Therefore, one should be careful in generalizing the present findings to the larger population of HIV+ DUs.

The vast majority of the present sample of HIV+ DUs have 'good' intentions with regard to their behavior. Nevertheless, one fifth knowingly put others at risk of HIV infection in the previous four months, the majority through sexual contact. The rate of condom usage among these HIV+ DUs is much higher than reported in surveys of DUs with unknown serostatus [Donoghoe et al. 1989, Paulussen et al. 1990, Magura et al. 1990, Klee et al. 1990, Saxon et al. 1991, Lewis & Watters, 1991]. This is in line with findings that HIV+ IVDUs are more likely to use condoms than HIV-negative IVDUs [McCoy 1990, Van den Hoek et al. 1990, 1992].

The proportion of DUs in the present sample who disclosed their HIV-positive status to their sexual and/or injecting partner is unknown. To our knowledge, there are no studies among DUs adressing this issue. In a study of sexually active HIV+ homosexual men, Marks et al. [1991] found a high prevalence of non-disclosure of HIV infection in combination with unsafe sex.

Although needle sharing is only reported by a minority of the present sample, this behavior is very effective in transmitting HIV. Unsafe vaginal sex was reported both with injecting and non-injecting partners. Among current IDUs, heterosexual transmission seems to play a minor role, as compared to transmission by injecting risk behavior [Des Jarlais et al 1987, Chaisson 1987, Schoenbaum 1989, Battjes 1990, Nelson 1991, Van Ameyden et al. 1992]. However, this does not imply that the risk of heterosexual transmission among current IDUs is zero. The considerable risk of heterosexual transmission from injectors to non-injectors, especially through male-to-female transmission, is well documented [France et al. 1988, Fordyce et al. 1991, Padian et al. 1991]. Therefore, the prevalence of unsafe vaginal sex in the present sample gives cause to concern, especially since almost half of the sample acknowledge that they could possibly transmit HIV through unsafe sex in the future. Such an acknowledgement comes close to being a behavioral expectation, and it has been argued that a behavioral expectation is a better predictor of behavior than intention [Warshaw & Davis 1985, Sheppard et al. 1988].

Approximately one third of these HIV-positive DUs report a clearly negative attitude toward condoms. This compares favorably to IVDUs with unknown serostatus - with a low level of condom use - of whom 77% expressed a strong dislike for condoms [Jones & Vlahov 1989]. Condom use among IVDUs with unknown serostatus has been found to be related to attitude toward condoms [Paulussen et al. 1990] and to a greater personal acceptance of condoms (Magura et al. 1990]. In the present sample, no relation was found between unsafe vaginal sex and a negative attitude toward condoms.

The perceived self-efficacy with regard to condom use is low among the DUs in the present sample, which means that most of these HIV+ DUs do not feel confident about their ability to use condoms consistently, either with clients or with casual partners. Furthermore, response efficacy associated with condom use is low: half of these HIV+ DUs are not certain that condoms are an effective means to prevent HIV infection. Respondents most often referred to rupture or slipping off of condoms during intercourse.

Contrary to our expectations, the HIV+ DUs in the SR-group did not have lower levels of self- and response efficacy than the other HIV+ DUs. This appears to be similar to findings by Magura et al. (1990) and Huang et al. (1989) but differs from findings by Abdul Quader et al. (1990): the first study fails to find an association between condom use and perceived condom efficacy against AIDS, the second finds that health beliefs play a relatively minor role in predicting condom use among IVDUs as compared to situational factors, while the third finds health beliefs associated with self-reported sexual risk reduction. All three studies concern DUs with unknown serostatus.

In the present study, neither social cognitive characteristics nor drug use characteristics are associated with unsafe vaginal sex; the strongest indicators are 'being a female prostitute' and 'having a non-Dutch nationality'. The first risk indicator is in line with Van den Hoek et al. (1989b), who found that sexually transmitted diseases are highly prevalent among addicted female prostitutes in Amsterdam, despite a (self-reported) high frequency of condom use. Both risk correlates indicate a marginal social position, which supports the view that social and cultural factors often interfere with adoption of preventive health behaviors (Levine and Sorenson, 1984, Mondanaro 1987, Mays and Cochran 1988, Worth 1989, Korf et al. 1990, Singer et al. 1990, Sibthorpe et al. 1991]. Although Schilling et al. (1991) find modest positive outcomes of an intervention aiming to improve condom use skills among 91 female methadone patients with unknown serostatus, the present findings suggest that the value of using persuasive methods and of skills training - to ameliorate attitude, response and self-efficacy - is yet uncertain among these Amsterdam HIV+ DUs. More detailed research seems necessary to investigate the underlying social and cultural processes that generate the observed relations and to develop an effective prevention approach. For example, Stone et al (1989) recommend to tailor information to the group, to aim interventions at couples (IVDUs and partners) and at their peer-groups, and to stress both self-protective and altruistic (i.e., other-protective) themes.

In general, it seems potentially worthwhile to focus prevention efforts at injecting and sexual HIV risk behavior of HIV+ drug users, especially since more and more DUs learn their serostatus after HIV-testing. Since the potential behavioral outcome of not using condoms is quite different for HIV-positives, HIV-negatives and persons unaware of their serostatus, the determinants of not using condoms are likely to differ between these groups (see Fishbein & Middlestadt 1989). However, in many of the above cited studies of condom use by IVDUs, no information is provided on serostatus nor on knowledge of serostatus, which makes it difficult to understand and interpret the results. For these reasons, it seems important to take serostatus and knowledge of serostatus into account in studying condom use among IVDUs and their partners and in developing effective intervention strategies with regard to condom use.

ACKNOWLEDGEMENTS

This study was supported by the Netherlands Foundation for Preventive Medicine (grant no. 28-1258). The authors thank B. Frblich, B. Scheeringa-Troost and R. Lopez Dias for interviewing and collecting blood samples; Dr. J. Goudsmit and M. Bakker for carrying out the laboratory tests; H.J.A. van Haastrecht for data-management and M. ter Pelle for preparing the manuscript.

REFERENCES

Abdul-Quader, A.S., S. Tross, S.R. Friedman, A.C. Kouzi and D.C. des Jarlais 1990 Street-recruited intravenous drug users and sexual risk reduction in New York City. AIDS 4:1075-1079.

Ameijden, E.J.C. van, J.A.R. van den Hoek., H.J.A. van Haastrecht and R.A. Coutinho 1992 The harm reduction approach and risk factors for HIV seroconversion   in injecting drug users, Amsterdam. American Journal of Epidemiology, in press.

Battjes, R.J., R.W. Pickens, Z. Amsel and L.S. Brown 1990 Heterosexual Transmission of Human immunodeficiency Virus among Intravenous Drug Users. Journal of Infectious Diseases 162: 1007-1011.

Bandura, A. 1977 Self-efficacy: towards a unifying theory of behavior change. Psychological Review 84:191-215.

Bandura, A. 1987 Perceived self-efficacy in the exercise of control over AIDS infection. In Primary Pre vention of AIDS. Psychological Approaches, Mays, V.M., Albee, G.W. and Schneider, S.F. (eds.). Newbury Park:128-141, Sage.

Chaisson, R.E., A.R. Moss, R. Onishi, D. Osmond and J.R. Carlson 1987 Human Immunodeficiency Virus Infection in Heterosexual Intravenous Drug Users in San Francisco. American Journal of Public Health 77:169-172.

Des Jarlais, D.C., S.R. Friedman and W. Hopkins 1985 Risk Reduction for the Acquired Immunodeficiency Syndrome Among Intravenous Drug Users. Annals of Internal Medicine 103:755-759.

Des Jarlais, D.C., E. Wish, S.R. Friedman, R. Stoneburner, S.R. Yancovitz, D. Mildvan, W. EI-Sadr, E. Brady and M. Cuadrado 1987 Intravenous drug use and the heterosexual transmission of the human immunodeficiency virus, Current trends in New York City. New York State Journal of Medicine:283286.

Donoghoe, M.C., G.V. Stimson and K.A. Dolan 1989 Sexual Behaviour of Injecting Drug Users and Associated Risks of HIV Infection for Non-injecting Sexual Partners. AIDS care 1:5158.

Fishbein, M. and I. Ajzen 1975 Belief, attitude, intention and behavior. an introduction to theory and research. Reading, Mass.: Addison-Wesley.

Fishbein, M. and S.E. Middlestadt 1989 Using the Theory of Reasoned Action as a Framework for Understanding and Changing AIDS-Related Behaviors. In Primary Pre vention of AIDS. Psychological Approaches, Mays, V.M., Albee, G.W., Schneider, S.F. (eds.) Newbury Park:93-110, Sage.

Fordyce, E.J., S. Blum, A. Balanon and R.L. Stoneburner 1991 A method for estimating HIV transmission rates among female sex partners of male intravenous drug users. American Journal of Epidemiology 133:590-598.

France, A.J., C.A. Skidmore, J.R. Robertson, R.P. Brettle, J.J.K. Roberts, S.M. Burns, C.A. Foster, J.M. Inglis, W.B.F. Galloway, S.J. Davidson. 1988 Heterosexual spread of human immunodeficiency virus in Edinburgh. British Medical Journal 296:526-529.

Friedland, G.H., C. Harris, C. Butkus-Small, D. Shine, B. Moll, W. Darrow, and R.S. Klein 1985 Intravenous Drug Abusers and the Acquired Immunodeficiency Syndrome (AIDS). Archives of Internal Medicine 145:1413-1417.

Haastrecht, H.J.A. van, J.A.R. van den Hoek, C. Bardoux, A. Leentvaar-Kuypers and R.A. Coutinho 1991 The course of the HIV epidemic among Intravenous Drug Users in Amsterdam, The Netherlands. American Journal of Public Health 81:59-62.

Hoek, J.A.R. van den, R.A. Coutinho, H.J.A. van Haastrecht, A.W. van Zadelhoff and 1. Goudsmit 1988 Prevalence and risk factors of HIV infections among drug users and drug-using prostitutes in Amsterdam. AIDS 1:55-60.

Hoek, J.A.R. van den, H.J.A. van Haastrecht and R.A. Coutinho 1989a Risk reduction among intravenous drug users in Amsterdam under the influence of AIDS. American Journal of Public Health 79:1355-1357.

Hoek, J.A.R. van den, H.J.A. van Haastrecht, B. Scheeringa-Troost, J. Goudsmit and R.A. Coutinho 1989b HIV infection and STD in drug addicted prostitutes in Amsterdam; potential for heterosexual HIV transmission. Genitourinary Medicine 65:146-150.

Hoek, J.A.R. van den, H.J.A. van Haastrecht and R.A. Coutinho 1990 Heterosexual behavior of intravenous drug users in Amsterdam; implications for the AIDS epidemic. AIDS 4:449-453.

Hoek, J.A.R. van den, H.J.A. van Haastrecht and R.A. Coutinho 1992 Little change in sexual behavior in injecting drug users in Amsterdam. Journal of Acquired Immune Deficiency Syndromes, in press.

Huang, K.H.C., J.K. Watters and P. Case 1989 Predicting compliance with HIV risk reduction behaviors among heterosexual intravenous drug users: relative contributions of health beliefs and situational factors. Presented to the Fifth International Conference on AIDS, Montreal, Canada June 49.

Jones, C. and D. Vlahov 1989 Why don't intravenous drug users use condoms? Journal of Acquired Immune Deficiency Syndromes 2:416-417.

Klee, H., J. Faugier, C. Hayes, T. Boulton and J. Morris 1990 Sexual partners of injecting drug users: the risk of HIV infection. British Journal of Addiction 85:413-418.

Korf, D.J., E.C. Buning, M. Elorche en B.M. Hes 1990 AIDS, drugs en belemmeringen bij veilig gedrag. Tijdschrift voor Alcohol en Drugs 16:201-209

Levine, S. and J.R. Sorenson 1984 Social and cultural factors in health promotion. In Behavioral Health: A hand book of Health Enhancement and Disease Prevention Health: Matazarro, J.D., Weiss, S.M., Herd, J.A., Miller, N.E. and S.M. Weiss (Eds.). New York: 222-229, Wiley.

Lewis, D.K. & J.K. Watters 1991 Sexual risk behavior among heterosexual intravenous drug users: ethnic and gender variations. AIDS 5:77-83.

Magura, S., J.L. Shapiro, Q. Siddigi and D.S. Lipton 1990 Variables influencing condom use among intravenous drug users. American Journal of Public Health 80:82-84.

Marks G., J.L. Richardson and N. Maldonado 1991 Self-disclosure of HIV infection to sexual partners. American Journal of Public Health 81:1321-1322.

Mays, V.M. and S.D. Cochran 1988 Issues in the perception of AIDS risk and risk reduction activities by black and hispanidlatina women. American Psychologist 43:949-957.

McCoy, H.V., C.Y. McKay, L. Hermanns and S. Lai 1990 Sexual behavior and the risk of HIV infection. American Behavioral Scientist 33:432-450.

Mondanaro, J. 1987 Strategies for AIDS prevention: motivating health behavior in drug dependent women. Journal of Psychoactive Drugs 19:143149.

Moss, A.R 1987 AIDS and intravenous drug use: the real heterosexual epidemic. British Medical Journal 294:389-390

Nelson, K.E., D. Vlahov, S. Cohn, M. Odunmbaku, A. Lindsay, J.C. Anthony and E.W. Hook III 1991 Sexually transmitted diseases in a population of intravenous drug users:association with seropositivity to the Human Immunodeficiency Virus (HIV). The Journal of Infectious Diseases 164:457-463.

Padian, N.S., S.C. Shiboski and N.P. Jewell 1991 Female-to-male transmission of Human Immunodeficiency Virus. Journal of the American Medical Association 266:1664-1667.

Paulussen, T.G.W.M., G.J. Kok, R.A. Knibbe and A. Cramer 1990 Determinanten van aan AIDS gerelateerde risicogedragingen van intraveneuze druggebruikers. Tijdschrift voor Sociale Gezondheidszorg 68:129-136.

Rogers, R.W. 1983 Cognitive and physiological processes in fear appeals and attitude change: a revised theory of protection motivation. In Social psycho- physiology, A sourcebook, J.T. Cacioppo and R.E. Petty (eds.) New York, London: 153-176, The Guilford Press.

Saxon, A.J., D.A. Calsyn, S. Whittaker & G. Freeman jr. 1991 Sexual behaviors of intravenous drug users in treatment. Journal of Acquired Immune Deficiency Syndromes 4:938-944.

Schilling, R.F., N. Ei-Bassel, S.P. Schinke, Gordon K. and S. Nichols 1991 Building skills of recovering women drug users to reduce heterosexual AIDS transmission. Public Health Reports 106:297303.

Schoenbaum, E.E., D. Hartel, P.A. Selwyn, R.S. Klein, K. Davenny, M. Rogers, C. Feiner and G. Friedland 1989 Risk factors for human immunodeficiency virus infection in intravenous drug users. New England Journal of Medicine 321:874879.

Schrager, L., G. Friedland, C. Feiner and P. Kahl 1991 Demographic characteristics, drug use, and sexual behaviour of IV drug users with AIDS in Bronx, New York. Public Health Reports 106:78-84.

Sheppard, B.H., J. Hartwick and P.R. Warshaw 1988 The theory of reasoned action: A metaanalysis of past research with recommendations for modifications and future research. Journal of Consumer Research 15:325-343.

Sibthorpe, B., D. Fleming, H. Tesselaar and J. Gould 1991 Needle use and sexual practices: differences in perception of personal risk of HIV among intravenous drug users. Journal of Drug Issues 21:699-712.

Singer, M., C. Flores, L. Davison, G. Burke, Z. Castillo, K. Scanlon and M. Rivera 1990 SIDA: The economic, Social, and Cultural Context of AIDS among Latinos. Medical Anthropology Quarterly (New series) 4:72114.

Stone, A.J., D. Morisky, R. Detels and H. Braxton 1989 Designing interventions to prevent HIV1 infection by promoting use of condoms and spermicides among intravenous drug abusers and their sexual partners. AIDS Education and Prevention 1:171-183.

Warshaw, P.R. and F.D. Davis 1985 Disentangling behavioral intentions and behavioral expectation. Journal of Experimental Social Psychology 21:213-228.

Worth, D. 1989 Sexual Decision-Making and AIDS: Why Condom Promotion among Vulnerable Women is Likely to fail. Studies in Family Planning 20: 297-307.