Health Education Research, Vol. 18, No. 2, 145-155,
April 2003
© 2003 Oxford University Press
Beliefs as predictors of condom use by injecting drug users in treatment
Hamilton Health Sciences, PO Box 2000, Hamilton, Ontario L8N 3Z5, Canada and 1 Department of Psychology, University of Western Australia, Nedlands, West Australian 6009, Australia. E-mail: houldingc{at}hhsc.ca
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
This study was conducted to clarify (1) the extent to which health beliefs selected from Protection Motivation Theory can combine to correctly classify 72 injecting drug users (IDUs) as condom users or non-users and (2) which of the beliefs (‘vulnerability to a regular partner’, ‘vulnerability to a casual partner’, ‘self-efficacy’, ‘response efficacy’, ‘response costs’ and ‘social norms’) were most influential in this distinction. Results of a logistic regression indicated that these beliefs were significant predictors of condom use. Overall, 83.3% of participants were correctly classified according to condom use, with condom ‘non-users’ being more accurately predicted (94.0%) than ‘users’ (59.1%). ‘Vulnerability to a regular partner’ and ‘social norms’ were significant multivariate and univariate predictors of condom use, and ‘response costs’ were significant univariate predictors. IDUs were confident of their ability to use condoms, considered themselves highly vulnerable to HIV infection from casual partners and were confident in the efficacy of condoms to protect them from AIDS. However, the majority of IDUs were not condom users, particularly with ‘regular’ partners. Findings suggest that HIV prevention programmes should target beliefs regarding risks from known partners, perceived norms and negative consequences of condom use in order to increase condom use by IDUs in treatment.
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Injecting drug users (IDUs) continue to engage in needle use practices that place them at risk for HIV infection (Crofts et al., 1999
). There is therefore increased risk of HIV infection from unprotected sex with them. Sexual contact with IDUs has been identified as the primary source of HIV infection for non-drug using heterosexuals (Rhodes and Quirk, 1995). Whilst risky needle use by IDUs has decreased (Loxley et al., 1992), their sexual behavior has proved consistently more difficult to modify [e.g. (Loxley and Hawks, 1994)]. IDUs of particular interest are those who are sexually active and whose sexual behavior places them or their partners at risk, but who do not belong to traditionally high sexual risk groups (such as homosexual men or sex workers). Not identifying with one of these high-risk groups may be part of the reason that, while many in these identified groups have begun to change the safety of their sexual practices, IDUs as a whole have not (Ekstrand and Coates, 1990; Loxley and Hawks, 1994). Studies indicate that condom use is low (Loxley, 2000) and partner change rates relatively high in IDUs (Rhodes and Quirk, 1995). There is a need to improve understanding of factors underlying condom use in IDUs, to better inform future intervention efforts (Chapman, 1997).
Health-related beliefs have been considered when attempting to discern important influences on condom use and have accounted for significant amounts of variance in AIDS-relevant sexual behavior in other populations [e.g. 49% in gay men (Van der Velde and Van der Pligt, 1991)]. They are particularly important since they are altered more easily than other factors also associated with condom use (e.g. demographics) and interventions manipulating particular beliefs have lead to changes in AIDS-relevant behaviors [e.g. (Raghubir and Menon, 1998)].
Protection Motivation Theory (PMT) (Rogers, 1983) is one of a number of models of health-protective behaviors which has been applied to understanding condom use [e.g. by undergraduate students (Dinoff and Kowalski, 1999)]. PMT identifies ‘vulnerability’, ‘self-efficacy’, ‘response efficacy’, ‘response costs’ and ‘social norms’ (as a reward of maladaptive behavior) as key in eliciting ‘protection motivation’, which is measured by ‘intentions to engage in an adaptive behavior’ [in this case, condom use] (Rogers, 1983). Certain modifications of PMT in application to understanding condom use may be warranted, however. It appears that actual condom use is likely to be a better dependent variable than intentions to use condoms, since beliefs predicting these two can differ [e.g. Bryan et al., 1997; Godin et al., 1997)]. In addition, perceived ‘severity’ (of the consequences of developing AIDS) is also identified by PMT as an important predictor of health behavior, but is likely to be uniformly high, and therefore not useful when attempting to distinguish condom users and non-users [e.g. (Aspinwall et al., 1991)].
To our knowledge, PMT has never been applied to predicting IDUs’ condom use and the studies where sexual behavior/intentions of IDUs have been associated with various beliefs broadly similar to those identified within PMT have certain methodological constraints [e.g. (Kenen and Armstrong, 1992; Kowalewski et al., 1994; Loxley and Hawks, 1994; Bowen and Trotter, 1995; Corby et al., 1996; Lauby et al., 1998; Loxley, 2000)]. Beliefs were not always related to condom use (as opposed to intentions), they were not always quantified making it difficult to systematically compare the relative impact of particular beliefs [e.g. (Kenen and Armstrong, 1992)], were measured using only one item, compromising the reliability of the measure [e.g. (Abdul-Quader et al., 1990)] or when quantified were not always considered in the context of other beliefs (i.e. using a multivariate model) which compromises interpretation of results [cf. (Tabachnick and Fidell, 1996)]. Results from participants from a variety of risk categories were sometimes collapsed [e.g. (Kasprzyk et al., 1998)], and operationalization of beliefs differed widely and were not always consistent with that of PMT or any identified theory.
Nonetheless, certain tentative observations can be drawn from results. ‘Social norms’ and ‘self-efficacy’ were most consistently associated with condom use [e.g. (Lauby et al., 1998; Loxley, 2000)], having modest-moderate correlations with IDUs’ intentions to use [e.g. ‘social factors’ r = 0.20-0.33 (Kowalewski et al., 1994); ‘perceived behavioral control’ ß = 0.28-0.50 (Corby et al., 1996)]. ‘Vulnerability’ was a somewhat inconsistent predictor [e.g. (Kowalewski et al., 1994)], perhaps as a result of the way in which the variable has been operationalized. Rogers (Rogers, 1983) indicates that ‘vulnerability’ should be specifically tied to the health behavior and research additionally suggests that it should also be tied to a specific partner (‘I might get AIDS if I don’t wear condoms with my regular/casual partner’), rather than an abstract probability [e.g. ‘[what is your] perceived risk of getting AIDS?’ (Bowen and Trotter, 1995)] [cf. (Bosga et al., 1995; Reisen and Poppin, 1999)]. Thus, there is a need to (1) clarify the extent to which PMT health beliefs can correctly classify IDU condom users and non-users, and (2) clarify which of the health beliefs examined are most influential in this distinction using methodology which addresses some of the difficulties from previous research. The present study aims to address these needs.
Hypotheses
Hypothesis 1. The PMT beliefs will combine to distinguish condom users and non-users.
Hypothesis 2. ‘Self-efficacy’ and ‘social norms’ will be the strongest individual predictors of IDU condom use status.
Hypothesis 3. Condom use with regular partners will be predicted by perceptions of vulnerability to HIV infection from regular partners. Condom use with casual partners will be predicted by perceptions of vulnerability to HIV infection from casual partners.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Sample
A total of 117 IDUs were interviewed for the study. Participants were eligible to be included in the analysis if they were heterosexual, non-sex-workers, were not HIV-positive, did not have a sexual partner who was HIV-positive, had not been in a monogamous relationship for more than 5 years and had had vaginal/penile sex within the 6 months prior to the study. These criterion were applied to obtain a group of sexually active and at-risk individuals, with broadly similar levels of risk of HIV infection. Of the 117 IDUs, 45 participants were excluded for the following reasons: one subject was HIV-positive and homosexual, four were bisexual, six were sex workers, 22 had been in monogamous relationships for more than 5 years and 12 had not had sex in the previous 6 months. Thus, 72 participants were included within the study, representing 61.5% of the original sample.
The sample comprised 69.4% (50) male and 30.6% (22) female participants, with a mean age of 29.9 years (SD = 5.7) (Table I). Participants were recruited from addiction treatment programmes in Perth, Western Australia in November/December 1993. These included a methadone maintenance programme (83.3%, n = 60), an in-patient detoxification center (6.9%, n = 5) and a residential rehabilitation unit (9.7%, n = 7). Thus, the sample included only IDUs in treatment, with most participants being opiate users who were in daily contact with services over an extended period (i.e. the methadone maintenance programme) and who had received AIDS-relevant information and HIV testing.
|
Participants reported that they had had sex with exclusively regular partners (58.3%, n = 42), both regular and casual partners (19.4%, n = 14), and only casual partners (22.2%, n = 16) in the month in which they had last had sex.
Procedure
Participants were approached at the treatment centers. After being assured of the confidentiality and anonymity of their responses, the interviewer read the questions to subjects (due to reported high levels of functional illiteracy in this population) (Turner et al., 1989) and they completed the questionnaire. Ethical clearance was obtained from the relevant body at the University of Western Australia before data collection commenced.
Measures
The questionnaire requested information including demographics (e.g. age, sex), belief indices, condom use and other sexual behavior, and the selection criteria (e.g. HIV status).
Belief indices
Measures of beliefs related to HIV/AIDS and condoms were rated using either four or five statements on a seven-point Likert-type scale with end points ‘strongly agree’ and ‘strongly disagree’. Some statements were worded ‘negatively’ in order to avoid artificially high consistency within responses. Statements were adapted from scales utilized in previous research on health behaviors (Table II).
|
- Vulnerability—regular partner and vulnerability—casual partner. ‘Vulnerability’ refers to perceptions of risk of being infected with HIV if the recommended behavior is not adopted (Rogers, 1983). As mentioned, ‘vulnerability’ has been found to differ according to the type of partner (e.g. a regular partner or a casual one) and was therefore assessed separately for two partner types (regular ‘vulner[regular]’ and casual ‘vulner[casual]’) using four statements for each type [adapted from (Rippetoe and Rogers, 1987)]. ‘Regular partner’ was defined as a person with whom a ‘steady’ sexual relationship had been formed. ‘Casual partner’ was defined as a partner with whom sex was irregular. The statements were concerned with perceptions of risk of HIV infection from unprotected sex.
- Self-efficacy. Within PMT, ‘self-efficacy’ refers to beliefs about ability and effort required to carry out the recommended health behavior (Rogers, 1983; Rogers and Prentice-Dunn, 1997). Beliefs of perceived ‘self-efficacy’ were rated on five statements, based on a measure used by Rosenthal et al. (Rosenthal et al., 1989). Statements were concerned with subjects’ perceptions of their ability to use condoms for AIDS protection.
- Response efficacy. ‘Response efficacy’ is related to the belief that a recommended action is able to avert an undesirable threat (Rogers, 1983). Response efficacy was assessed using four statements adapted from Rippetoe and Rogers (Rippetoe and Rogers, 1987). The statements were concerned with respondents’ perception of the ability of condoms to protect them from HIV infection.
- Response costs. ‘Response costs’ of a health behavior are related to the consequences of such action which decrease the probability that the behavior will be consistently adopted (Rogers, 1983). Although this definition could encompass all other measures suggested by PMT, Rogers (Rogers, 1983) provides examples that help clarify the nature of this variable (e.g. monetary cost, inconvenience). ‘Response costs’ were assessed using four statements adapted from Jemmott and Jemmott (Jemmott and Jemmott, 1989) and Kenen and Armstrong (Kenen and Armstrong, 1992). Statements were concerned with participants’ perceptions of the negative consequences of condom use.
- Social norms. ‘Social norms’ were concerned with participants’ perception of the behavior and attitudes of others in regard to condoms use. Four statements were used which were adapted from Van der Velde and Van der Pligt (Van der Velde and Van der Pligt, 1991).
Sexual behavior
Information about sexual behavior was assessed using (1) five questions related to selection criteria (e.g. ‘Have you had sex with anybody else [apart from your regular partner] in the last 5 years?’) and (2) five questions based on the sexual behavior subscale of the HIV Risk Behavior Scale, including questions related to condom use with regular and casual partners (HRBS) (Darke et al., 1991). On six-point scales, subjects rated how many times they had used condoms when having sex with (1) regular partner(s) (‘no regular partner’, ‘everytime’–‘never’) and (2) casual partner(s) (‘no casual partner’, ‘everytime’–’never’). The scale was adapted for the present study to measure behavior in the last month in which the subject had had sex. (Only results from these questions related to condom use will be presented in the present study.)
Scoring
Belief indices
Scores on each of the six belief indices were calculated by summing pertinent ratings (with scores reversed for negatively stated questions). For all indices, a high score was consistent with AIDS risk-reduction behavior.
Condom use
‘Condom use’ classification was ascertained using responses to questions regarding use of condoms with regular and casual partners. Because responses to both questions were bi-modally distributed (Figure 1), participants were classified as either ‘users’ or ‘non-users’ of condoms. ‘Condom users’ were defined as those who reported using condoms with their partners at least ‘sometimes’. Thus, ‘non-users’ were those who indicated that they never or rarely used condoms when having sex with their reported sexual partner. The 14 participants who reported having sex with both regular and casual partners were classified according to their behavior with their regular partner, since this represented the sexual contact with most repeated exposures, which in turn represents a greater risk for HIV infection (Hulley and Hearst, 1989). (Only five of the 14 with both types of partner had different behavior with the two partner types—they were all condom users with casual but not regular partners. A logistic regression was run with these five subjects excluded and results were very similar to those in the present study.)
|
Data analysis
To ascertain whether ‘condom use’ status could be predicted by combination of the belief indices, a multiple logistic regression was conducted.
To enhance the interpretation of this multivariate model, univariate correlations between belief indices and condom use were calculated using Pearson’s r correlations (Tabachnick and Fidell, 1996).
To ascertain the ability of perceptions of ‘vulnerability to casual/regular partners’ to predict condom use with each of these partner types, univariate correlations were calculated using Pearson’s r correlations.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Descriptive statistics
Table III
displays means, SDs and internal reliability (Cronbach’s 
) for each belief index. Scores on vulner[casual], self-efficacy and response efficacy were skewed, with mean above mid-range. Most people perceived themselves to be at high risk of being infected with HIV from a casual partner, were highly confident in their ability to use condoms and were highly confident in the ability of condoms to protect from HIV infection. Further, individuals perceived themselves to be at greater risk from unprotected sex with casual partners than regular. All belief indices had acceptable internal reliability scores, except ‘response efficacy’.
|
Participant behavior
A minority of the sample were classified as ‘condom users’ (30.6%, n = 22), with most being ‘non-users’ (69.5%, n = 50). Table IV
illustrates differences in condom use according to partner type. Of the 56 participants reporting sex with regular partners, 21.4% were condom users and 78.6% were condom non-users. In contrast, of the 30 participants who had sex with casual partners, 56.7% of participants were condom users and 43.3% were non-users (Table IV).
|
Evaluation of the model
A logistic regression was conducted in order to evaluate the model, which demonstrated that the model could predict whether participants were users or non-users of condoms (Model
2(6) = 21.7, P = 0.0014). Overall, 83.3% of the cases were correctly classified as either ‘users’ or ‘non-users’ (Table V). In addition, the model was able to predict condom non-users more accurately than users (94.0 versus 59.1%). This should be considered in light of the fact that 69.4% of subjects could be correctly classified by chance (since 69.4% of subjects were non-users). Thus, the model increased the probability of correct classification from chance by a similar proportion for each outcome (69.4–94.0% for non-users and 30.6–59.1% for users).
|
Perceptions of vulnerability to infection from a regular partner (‘vulner[regular]’) and perceptions of normative support for condom use (‘social norms’) were the only variables which contributed significantly to the prediction of condom use when all other variables were in the equation (B = -0.1, Wald = 5.5, P = 0.02 and B = -0.2, Wald = 4.1, P = 0.04 for ‘vulner[regular]’ and ‘social norms’, respectively; Table VI
). People were more likely to be condom users if they thought they were at risk from unprotected sex with regular partners or those who perceived that normative support existed for condom use.
|
Univariate correlations
A matrix of the correlations between belief indices and condom use was calculated (Table VII
). Moderate significant correlations were found between condom use and ‘vulner[regular]’, ‘social norms’ and ‘response costs’ (r = -0.4, P = 0.001, r = -0.3, P = 0.007 and r = -0.4, P = 0.001, respectively). Those who perceived themselves to be vulnerable to infection from regular partners perceived support from others for condom use and perceived low costs associated with condom use were more likely to use them. There were a number of inter-correlations between beliefs.
|
Perceptions of vulnerability and partner type
For the 56 IDUs who had had sex with regular partners, condom use with these regular partners was significantly correlated with ‘vulner[regular]’ (r = -0.4, P = 0.002). Those who perceived themselves to be vulnerable to infection from regular partners were more likely to be condom users. There were no significant univariate predictors of condom use for the 30 IDUs who had had sex with casual partners.
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
This study indicated that:
- In general, IDUs did not use condoms consistently, particularly with regular partners.
- Beliefs combined to correctly classify IDUs as condom users or non-users.
- IDUs who did not perceive themselves to be at risk from regular partners, who did not perceive others to support condom use and who perceived higher ‘costs’ of using condoms were less likely to use them.
- Condom use with regular partners was predicted by perceptions of vulnerability to them. Condom use with casual partners was not significantly related to perceptions of vulnerability to infection from them.
- IDUs (1) considered themselves to be highly vulnerable to HIV infection from unprotected sex with casual partners, (2) were confident in their ability to use condoms and (3) were confident in the ability of condoms to protect them from HIV.
Beliefs as predictors
The present study confirmed that IDUs’ beliefs regarding HIV and condom use can be used to correctly classify a large proportion of the sample as either ‘users’ or ‘non-users’ of condoms, with condom non-users being more accurately predicted than users. Thus, beliefs are important in understanding IDUs’ condom use and can inform interventions to increase safer sex in this population. Results also confirmed concerns regarding condom use and sexual behavior by IDUs, whose behavior placed them and their partners at continued risk of HIV infection. Reasons why this might be the case can be considered by examining individual beliefs significant in predicting condom use.
Consistent with results from a qualitative study of a similar population (Loxley 2000), the study found that perceived ‘vulnerability to infection from a regular partner’ and ‘social norms’ were the only significant predictors of condom use status. The moderately strong univariate relationship between the two beliefs and condom use was comparable with previous research [e.g. (Kasprzyk et al., 1998)]. The ability of ‘vulner[regular]’ to predict condom use may be because it refers to risk from specific and familiar (i.e. regular) partners rather than the abstract, global risk suggested by the more anonymous term ‘casual partner(s)’ [cf. (Reisen and Poppin, 1999)]. IDUs (and others) may be inclined to use partner selection, rather than condom use, as their primary protection from HIV infection and rely on their ability to ‘tell’ if someone is high risk by looking at them (Moore and Rosenthal, 1993; Loxley 2000). Further, several participants mentioned results from HIV screening tests. Negative results from this screening appeared to provide a sense of security—that since infection has not occurred thus far, both partners were ‘safe’ and there was no perceived need for behavior change. This is problematic, since a negative result only reflects sero-status 3 months prior to the last test and is only as ‘safe’ as the protective behavior (sexual and needle use) of both sexual partners since then (Tessier et al. 2001). These beliefs should be borne in mind when counseling IDUs regarding HIV behaviors. There may also be a tendency to focus on relative risks from needle use. IDUs might correctly assess that unsafe needle use is more risky than unprotected sex—although a reduced risk does not mean a negligible one (Loxley and Davidson, 1991). Further research could examine needle use and sexual behavior of IDUs and their partners to assess perceived and actual levels of risk, and what influences these.
The findings that (1) condoms are used more often with casual partners than regular, (2) there is generally high perceived risk of infection from casual partners, but that (3) the perception of risk from casual partners does not differentiate users and non-users is interesting. It may be that a heightened sense of risk is understood between casual partners and therefore it is more acceptable to ask an acquaintance to use condoms than a known partner who ‘should’ be trustworthy. Thus, ‘social norms’ and ‘response costs’ may be the more important influence on condom use in casual partnerships. This could be clarified by future research examining predictors of exclusively casual partners.
Our results suggest that although ‘response costs’ reflect behavior, they are superseded by the individuals’ perception of their vulnerability and whether others are using condoms [cf. (Lauby et al., 1998)]. It seems that interventions shifting perceptions of risk from known partners and of ‘social norms’ may in turn affect perceptions of the ‘costs’ of using condoms.
The generally high rating of ‘self-efficacy’ may explain its unexpected non-significance as a predictor of condom use. Education campaigns may have successfully improved both confidence in condom use and perceived effectiveness of condoms in protecting from HIV infection. The low internal reliability of ‘response efficacy’ makes interpretation of results tentative, however.
The moderately strong inter-correlations between beliefs suggests that there may be conceptual overlap between indices, and/or that individuals are consistent in the degree to which they have ‘taken on board’ messages related to condom use and HIV infection. In addition, it may mean that intervention efforts influencing one belief may lead to changes in a number of others.
Together, it appears that IDUs in treatment are confident that they are able to use condoms, and believe that they are effective at preventing HIV infection, but (given that they are not at risk from their partners and that no-one else is using them) many do not do so. It appears we need to personalize the risk of HIV infection for individuals from known partners, make condom use the assumed norm within these relationships and to a lesser degree address the perceived costs associated with condom use [cf. (Loxley and Davidson, 1998)]. In other populations, interventions have successfully increased people’s perceptions of their vulnerability to HIV infection and increased their intentions to use condoms [e.g. (Raghubir and Menon, 1998)], and peer educators may be particularly effective vehicles of change (Loxley 2000).
Theoretical and methodological constraints
PMT and other cognitive models of health behavior have been criticized for ignoring the process of weighing up which is said to take place when deciding whether to use condoms, and assuming that individuals will use them if they understand themselves to be at risk (Loxley and Davidson, 1998). Rogers and Prentice-Dunn (Rogers and Prentice-Dunn, 1997), however, clarify that PMT does not assume rationality in the decision maker, but that decision making is affected by cognitive and motivational biases. Thus, although IDUs may acknowledge that they are somewhat at risk from unprotected sex with a given partner, they will balance this against factors such as the embarrassment of suggesting or insisting on condom use.
Beliefs within the study were correlated with behavior which had occurred up to 6 months prior to participants being interviewed, meaning that beliefs may not have been the same as those held at the time they had had sex. Further, when examining correlational relationships between variables (as in the present study) it is not possible to establish the direction of causality or whether causal mechanisms exist at all. Therefore, it would be preferable to use a longitudinal method to measure beliefs close in time and prior to sexual encounters [cf. (Reisen and Poppen, 1999)].
One of the aims of the present study was to use belief indices with multiple items to improve confidence in the reliability of responses. The only moderate internal reliability of belief indices (especially response efficacy) may have weakened observed relationships between them and condom use. This underlines concerns regarding the reliability of findings from previous research, which rely only on one item to measure particular beliefs, and perhaps suggests that the sample and/or population are less consistent reporters than ‘normal’ adult populations.
The size of the sample was small and drawn only from clinical populations, which limits the reliability and generalizability of the results and precluded further analysis regarding beliefs predictive of behavior for other subgroups (such as gender and partner type or treatment programme) which may be influential [e.g. (Kazprzyk et al., 1998)]. Being in treatment predicted ‘very risky sexual behavior’ in women (but not men) in one previous study (Loxley et al., 1998), which may mean the women in the present study population are more risky than female IDUs not in contact with treatment. Future research should use larger samples and could examine beliefs particular to drug use type or programmes, and should seek to include IDUs not in contact with treatment programmes.
These results suggest that interventions to increase condom use should increase the perception of vulnerability to infection from known sexual partners and make condom use the assumed ‘norm’. Future research should make use of a longitudinal design examining predictors specific to subgroups within the given population and making use of more reliable measures of beliefs.
|
Acknowledgments |
|---|
Data was collected as part of the requirements for a Masters in Psychology (clinical) at the University of Western Australia by the first author and was co-supervised by the second. The authors would like to acknowledge the contributions of all participants in the study, Wendy Loxley for co-supervising the thesis and the cooperation of the three treatment sites in gathering data. The authors would also like to thank Peter Braunberger for helpful comments on earlier versions of this paper.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
Abdul-Quadar, A. S., Tross. S., Friedman, S. R., Kouzi, A. C. and Des Jarlais, D. C. (1990) Street-recruited intravenous drug users and sexual risk reduction in New York City. AIDS, 4, 1075–1079.[ISI][Medline]
Aspinwall, L. G., Kemeny, M. E., Taylor, S. E., Schneider, S. G. and Dudley, J. P. (1991) Psychological predictors of gay men’s AIDS risk reduction behaviour. Health Psychology, 10, 432–466.[CrossRef][ISI][Medline]
Bosga, M. B., de Wit, J. B., de Vroome, E. M., Houweling, H., Schop, W. and Sanfort, T. G. (1995) Differences in perception of risk for HIV infection with steady and non-steady partner among homosexual men. AIDS Education and Prevention, 7, 103–115.[ISI][Medline]
Bowen, A. M. and Trotter, R., II (1995) HIV risk in intravenous drug users and crack cocaine smokers: predicting stage of change for condom use. Journal of Consulting and Clinical Psychology, 63, 238–248.[CrossRef][ISI][Medline]
Bryan, A. S., Aiken, L. S. and West, S. G. (1997) Young women’s condom use: the influence of acceptance of sexuality, control over the sexual encounter, and perceived susceptiblity to common STDs. Health Psychology, 16, 468–479.[CrossRef][ISI][Medline]
Chapman, L. S. (1997) Reactions to the article. The Art of Health Promotion, 1, 7–8.
Corby, N. H., Jamner, M. S. and Wolitski, R. J (1996) Using the Theory of Planned Behavior to predict intention to use condoms among male and female injecting drug users. Journal of Applied Social Psychology, 26, 52–75.[CrossRef]
Crofts, N., Aitken, C. And Kaldor, J. (1999) The force of numbers: why hepatitis C is spreading among Australian injecting drug users while HIV is not. Medical Journal of Australia, 170, 220–221.[ISI][Medline]
Darke, S., Ward, J., Hall, W., Heather, N. and Wodak, A. (1991) The reliability and validity of a scale to measure HIV risk-taking behaviour among intravenous drug users. AIDS, 5, 181–185.[ISI][Medline]
Dinoff, B. L. and Kowalski, R. M. (1999) Reducing AIDS risk behavior: the combined efficacy of Protection Motivation Theory and the Elaboration Likelihood Model. Journal of Social and Clinical Psychology, 18, 223–239.
Ekstrand and Coates (1990) Maintenance of safer sexual behaviors and predictors of risky sex: the San Francisco Men’s Health Study. American Journal of Public Health, 80, 973–977.
Godin, G., Fortin, C., Michaud, F., Bradet, R. and Kok, G. (1997) Use of condoms: intention and behaviour of adolescents living in juvenile rehabilitation centre. Health Education Research, 12, 289–300.
Hulley, S. B. and Hearst, N. (1989) The worldwide epidemiology and prevention of AIDS. In Mays, V. M., Albee, G. W. and Schneider, S. F. (eds), Primary Prevention of AIDS. Sage, Newbury Park, CA, pp. 47–71.
Jemmott, L. S. and Jemmott, J. B., III (1989) Applying the Theory of Reasoned Action to AIDS risk reduction behavior: condom use among black women. Nursing Research, 40, 228–234.
Kasprzyk, D., Montano, D. E. and Fishbein, M. (1998) Application of an integrated behavioural model to predict condom use: a prospective study among high risk HIV risk groups. Journal of Applied Social Psychology, 28, 1557–1583.[CrossRef]
Kenen, R. H. and Armstrong, K. (1992) The why, when and whether of condom use among female and male drug users. Journal of Community Health, 17, 303–317.[CrossRef][Medline]
Kowalewski, M. R., Longshore, D. and Anglin, M. D. (1994) The AIDS Risk Reduction Model: examining intentions to use condoms among injection drug users. Journal of Applied Social Psychology, 24, 2002–2027.[CrossRef]
Lauby, J. L., Semaan, S., Cohen, L., Gielen, A., Pulley, L. V., Walls, C. and O’Campo, P. (1998) Self-efficacy, decisional balance and stages of change for condom use among women at risk for HIV infection. Health Education Research, 13, 343–356.
Loxley, W. (2000) Double risk: young injectors and sexual relationships. Sexual and Relationship Therapy, 15, 297–310.[CrossRef]
Loxley, W. and Davidson, R. (1991) Why do injecting drug users take risks: barriers to safer injecting and sexual behaviour in Perth. National AIDS Bulletin, 9, 32–38.
Loxley, W. and Davidson, R. (1998) How rational is needle sharing to young injecting drug users. Addiction Research, 6, 499–515.
Loxley, W. and Hawks, D. V. (1994) AIDS and injecting drug use: very risky behaviour in a Perth sample of injecting drug users. Drug and Alcohol Review, 13, 21–30.
Loxley, W., McDonald, C. and Marsh, A. (1992) The Australian National AIDS and Injecting Drug Users Study, Perth, 1990. Technical Report, National Centre for Research into the Prevention of Drug Abuse, Perth, WA.
Loxley, W., Bevan, J. and Carruthers, S. (1998) Sex, gender, drugs and risk: the Australian study of HIV and injecting drug use. International Journal of Drug Policy, 9, 255–262.[CrossRef]
Moore, S. and Rosenthal, D. (1993) Sexuality in Adolescence. Routledge, London.
Raghubir, P. and Menon, G. (1998) AIDS and me, never the twain shall meet: the effects of information accessibility on judgements of risk and advertising effectiveness. Journal of Consumer Research, 25, 52–63.[CrossRef]
Reisen, C. A. and Poppin, P. J. (1999) Partner-specific risk perception: a new conceptualization of perceived vulnerability to STDs. Journal of Applied Social Psychology, 29, 667–684.[CrossRef]
Rhodes, R. and Quirk, A. (1995) Where is the sex in harm reduction? International Journal of Drug Policy, 6, 78–82.
Rippetoe, P. A. and Rogers, R. W. (1987) Effects of components of Protection Motivation Theory on adaptive and maladaptive coping with a health threat. Journal of Personality and Social Psychology, 52, 596–604.[CrossRef][ISI][Medline]
Rogers, R. W. (1983) Cognitive and physiological processes in fear appeals and attitude change: a revised theory of protection motivation. In Cacioppo, J. T. and Petty, R. E. (eds), Social Psychophysiology: A Sourcebook. Guilford Press, New York, NY, pp. 153–176.
Rogers, R. W. and Prentice-Dunn, S. (1997) Protection Motivation Theory. In Gochman, D. S. (ed.), Handbook of Health Behavior Research I: Personal and Social Dimensions. Plenum Press, New York, pp. 113–132.
Rosenthal, D. A., Moore, S. and Flynn, I. (1989) Adolescent self efficacy, self esteem and sexual risk-taking. Journal of Community and Applied Social Psychology, 1, 77–88.
Tabachnick, B. G. and Fidell, L. S. (1996) Using Multivariate Statistics. Harper Collins, New York.
Tessier, D., Dion, H., Grossman, D. W. and Rachlis, A. (2001) HIV Care: A Primer and Resource Guide for Family Physicians. College of Family Physicians of Canada, Mississauga, Ontario.
Turner, C. F., Miller, H. G. and Moses, L. E. (1989) AIDS Sexual Behavior and Intravenous Drug Use. National Academy Press, Washington, DC.
Van der Velde, F. W. and Van der Pligt, J. (1991) AIDS-related health behavior: coping, protection motivation and previous behavior. Journal of Behavioral Medicine, 14, 429–451.[CrossRef][ISI][Medline]
Received on May 1, 2001; accepted on December 13, 2001
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  B. Stanton, M. Cole, J. Galbraith, X. Li, S. Pendleton, L. Cottrel, S. Marshall, Y. Wu, and L. Kaljee Randomized Trial of a Parent Intervention: Parents Can Make a Difference in Long-term Adolescent Risk Behaviors, Perceptions, and Knowledge Arch Pediatr Adolesc Med, October 1, 2004; 158(10): 947 - 955. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||