1 R E V I E W Drug and Alcohol Review (September 2010), 29, DOI: /j x Perceptions of discriminatory treatment by staff as predictors of drug treatment completion: Utility of a mixed methods approachdar_ LOREN BRENER 1, WILLIAM VON HIPPEL 2, COURTNEY VON HIPPEL 2, ILYSE RESNICK 1 & CARLA TRELOAR 1 1 National Centre in HIV Social Research, University of New South Wales, Sydney, Australia, and 2 School of Psychology, University of Queensland, St Lucia, Australia Abstract Introduction and Aims. Staff interactions with their clients are an important factor in the quality of care that is provided to people in drug treatment.yet there is very little research that addresses staff attitudes or clients perceptions of discrimination and prejudice by staff with regard to treatment outcomes.this research aimed to assess whether perceptions of discrimination by staff predict drug treatment completion. Design and Methods. The study used a mixed methods approach. Ninety-two clients in residential rehabilitation facilities in Sydney were administered a series of quantitative measures assessing drug history, severity of drug use, treatment history, perceptions of staff discrimination and treatment motivation. Clients were followed up regularly until an outcome (dropout or completion) was obtained for the full sample. Results. Perceptions of discrimination were a significant predictor of treatment completion, with greater perceived discrimination associated with increased dropout. Qualitative interviews with 13 clients and eight health-care workers from these treatment services were then conducted to gain insight into how perceived discrimination may impact on treatment experiences. Clients and staff discussed how they would address the issue of perceived discrimination during the current treatment experience. Discussion and Conclusions. Adopting a mixed methods approach facilitated exploration of the impact of perceived discrimination on treatment from both clients and health-care workers perspectives.this methodology may also enhance interpretation and utilisation of these findings in drug treatment. [Brener L, von Hippel W, von Hippel C, Resnick I, Treloar C. Perceptions of discriminatory treatment by staff as predictors of drug treatment completion: Utility of a mixed methods approach. Drug Alcohol Rev 2010;29; ] Key words: drug treatment, perceived discrimination, attitudes, treatment outcomes, mixed methods. Drug treatment retention and completion has consistently been associated with positive outcomes, including improved mental and physical health, positive changes in drug use, decreases in criminal activity and increased employment [1,2]. Understanding the factors that may contribute to treatment completion is particularly important in light of the high dropout rates among clients in drug treatment [3,4]. In a 2004 national survey of residential rehabilitation programs in England, the average completion rate was 48%, while in a California-based study the completion rate was 55% [5,6]. Dropout rates in inpatient drug treatment programs can range from 19.7% to 63.22% . A range of client-centred variables have been associated with treatment dropout, including being male, younger age, minority group status, greater cognitive dysfunction, treatment readiness or motivation [8 11], less severe substance use [12 14] and antisocial or borderline personality disorders . In contrast, little emphasis has been placed on program and staff-related factors that may be involved in treatment dropout despite their potential significance in contributing to client retention in drug treatment [9,10,16]. Loren Brener PhD, Research Fellow,William von Hippel PhD, Professor, Courtney von Hippel PhD, Lecturer, Ilyse Resnick BSc, PhD Candidate, Carla Treloar PhD, Associate Professor. Correspondence to Dr Loren Brener, National Centre in HIV Social Research, University of New South Wales, Sydney 2052, Australia. Tel: ; Fax: ; Received 22 September 2009; accepted for publication 17 January 2010.
2 492 L. Brener et al. Staff attitudes can influence the quality of care provided to people in drug treatment and may therefore play a role in treatment retention [17 19]. Previous research regarding staff-related variables in drug treatment has focused on staff skills as well as similarities in staff and client beliefs about treatment processes and outcomes [20 22]. For example, concordance between clients and staff views about treatment implementation and outcome predicts treatment engagement by clients [23,24]. Similarly, therapeutic alliance during drug treatment has been found to predict retention [21,24]. Perceiving staff as non-threatening and supportive appears to be important for ongoing engagement of clients in treatment  and continuing after-care . Health-care workers report that injecting drug users are often their most difficult clients, as they expect them to be more dangerous, less cooperative, more aggressive, less truthful, less likely to complete treatment and more demanding than other clients [26,27]. It is thus unsurprising that health workers show evidence of negative unconscious attitudes toward their drug-injecting clients even when they report no conscious negative attitudes towards them . Indeed, these negative unconscious attitudes predict job turnover among nurses working in drug treatment , and may lead to nonverbal negativity that is detectable by drug-injecting clients . Perceived prejudice and discrimination in the health sector can impact on receipt of care and health-seeking behaviour. People who fear discrimination may be less likely to disclose their health condition if it is associated with stigma and hence may be less likely to receive appropriate care for that condition [31,32]. Fear of discrimination by health staff is also a deterrent from adhering to health regiments [33,34]. In an effort to address the impact of perceived discrimination on treatment, the research team adopted a mixed methods approach, using both qualitative and quantitative methodologies. This process of methodological triangulation increases the validity and interpretative potential of the research, using independent but complementary research strategies [35,36]. Sequential triangulation was adopted whereby the results of the quantitative study were used to plan and inform the qualitative phase of study . Hence the aim of the quantitative arm of the study was to examine whether perceptions of discrimination by staff predicted treatment completion for heroin-using clients in residential rehabilitation programs. Findings from this part of the study were then extended to in-depth qualitative interviews that explored the impact of perceived discrimination on drug treatment for clients and staff. Quantitative research Method Participants. Participants were 92 heroin users recruited from six residential rehabilitation treatment facilities in and around Sydney.The majority of participants who were informed of the study by staff agreed to participate. Exceptions to this were clients who were too ill to participate. A researcher provided assistance for participants to complete the measures on a laptop computer at the treatment facility in a private space. Participants were reimbursed $25 for their time. Recruitment. Residential rehabilitation services were targeted as the treatment setting for assessment because their residential nature allows clients to be tracked so a treatment outcome (completion or dropout) can be ascertained. Service directors were informed of the study, and the participation of their services requested. Clients who met the criteria for inclusion in the study were then informed by staff and invited to participate. Only clients who used heroin as their primary drug were included, as heroin users face substantial stereotyping from the media and from other drug users . The consent process included a request for permission for researchers to contact the service during the clients scheduled treatment period to establish treatment outcome. All participating clients agreed to this. The study received ethics approval from the University of New South Wales Human Research Ethics Committee (UNSW HREC). Procedure. The survey included measures of age of first heroin use, duration of treatment (6, 8 or 12 months), frequency of drug use in the month prior to treatment (four-item scale, including items about heroin use, polydrug use with lower scores indicating more frequent use, alpha = 0.70), the Severity of Dependence Scale  (higher scores indicate more severe use, alpha = 0.81), client perceptions of discrimination by staff (five-item scale, including items, such as, Staff at my current drug treatment program do not treat me with respect, and, Staff discriminate against me because I am a heroin user, with higher scores indicating greater discrimination, alpha = 0.68) and a seven-item treatment motivation scale  (higher scores indicate greater treatment motivation, alpha = 0.74). Results Sample characteristics, including gender, age, level of education, main source of income and age at first heroin use, are reported in Table 1. To test the primary
3 Discriminatory treatment and drug treatment completion 493 Demographic variable Table 1. Sample characteristics Mean age 35 (SD = 7.13), range Gender Male 75 Female 17 Education level Primary school only 7 Up to an including year Up to and including year Diploma or trade certificate 9 Attended or completed university 6 Income source Full-time work 8 Part-time/casual work 5 The dole/other temporary benefit 58 Disability Pension 16 Other 5 Treatment completion Yes 35 No 57 Age of first heroin use M = 19 (SD = 4.78), range Perceived discrimination M = 1.68 (SD = 0.68), range 1 4 Severity of dependence M = 3.03 (SD = 0.70), range Treatment motivation M = 4.26 (SD = 0.50), range Frequency of drug use M = 5.94 (SD = 1.30), range 2 7 Length in treatment M = 2.04 (SD = 1.00), range 1 3 hypothesis that clients perceptions of discrimination by staff predict their treatment completion, a correlational analysis was conducted. As can be seen in Table 2, frequency of drug use, treatment motivation and perceived discrimination all predicted final completion, as did duration of the treatment program. To assess whether these variables predicted treatment completion over and above treatment program duration, a series of logistic regressions were estimated. These analyses revealed that perceived discrimination remained a significant predictor of treatment completion (Wald = 4.15, P < 0.05), as did treatment motivation (Wald = 3.91, P < 0.05), but frequency of drug use did not (Wald = 2.11, P > 0.10), when controlling for length of the treatment program. Qualitative research The findings of the survey indicate the importance of perceived discrimination in determining whether a client completes treatment. While this finding contributes to our understanding of treatment completion, it cannot tell us why clients may choose to leave treatment if they feel that staff treat them badly, nor how they develop these perceptions. It is the latter information that would be important to inform interventions designed to retain people in treatment. To further explore the findings of the survey data, the impact of perceived discrimination on the treatment encounter was examined in qualitative interviews with 13 clients and eight health workers from the same treatment facilities. The aim of this phase of the research was: (i) to develop a greater understanding of client perceptions of staff discrimination toward drug users and how these impact their treatment experiences; and (ii) to assess how staff could interpret these findings and take account of perceived discrimination within the treatment encounter. This phase of the study also provided the opportunity to provide feedback to clients and staff regarding the findings of the quantitative study and to get their perspective on ways to understand and manage perceived discrimination during treatment. Method Participants Thirteen clients participated in this arm of the study. Ten were male and three were female with an age range from 19 to 53 years. All except one (the 19-year-old) had participated in at least one other treatment program before coming to the current residential rehabilitation. Eight frontline drug treatment workers participated: five were male and three were female with experience in drug and alcohol treatment ranging from 1 to 20 years. Clients were reimbursed $25 for their time and staff were given a $20 gift voucher. Procedure The recruitment procedure was similar to the quantitative study. Participants were recruited from the same six treatment facilities in the quantitative phase; however, they had not undertaken the quantitative survey. Staff at the treatment facility approached potential participants (e.g. previous heroin users) and invited them to participate. Most agreed to participate. In terms of staff, the director of each facility approached staff and asked them to participate. Clients and staff were chosen from all of the six original survey recruitment sites. Interviews were conducted at the treatment facility by one of two trained researchers and lasted approximately an hour. Interviews were open-ended and guided by focal areas developed from the findings of the quantitative study. The focus of the client interviews
4 494 L. Brener et al. Table 2. Correlations with treatment completion Severity of dependence Perceived discrimination Treatment motivation Frequency of drug use Treatment completion Length in treatment Age Gender Perceived discrimination Treatment motivation *** Frequency of drug use Treatment completion * 0.26* 0.29** Length in treatment * 0.47*** 0.28** Age * Gender ** ** Education *P < 0.05, **P < 0.01, ***P < Correlations involving completion, gender and education are Spearman s rho. All other correlations are Pearson s r. was around issues related to their understanding of prejudice and discrimination, whether they had experienced discrimination when accessing treatment and their experiences with staff in the current treatment facility. Both clients and health workers were then told the main finding of the quantitative data, that perception of staff discrimination impacts on client treatment outcomes, and then asked to think about how this could be addressed in the current treatment. Additional approval for this phase of the study was received from the UNSW HREC. Data were recorded and transcribed verbatim. All identifying information was removed from the transcripts and participants given pseudonyms. Thematic analysis was used to identify key themes in the transcripts of clients and staff . These themes around the focal area of perceived prejudice and discrimination are reported below. This represents an exploratory methodology to feed back the quantitative findings to both sets of participants and to obtain their opinions on how best to interpret and utilise these findings in the context of treatment. Results All clients reported that they had experienced discrimination in the past either in drug treatment or at other health facilities. Clients tended to acknowledge that these experiences were also shaped by their perceptions of society s view of drug users. I think it s just my own, I guess, guilt and shame..., and that it s society who view you as pieces of shit-... Like kind of projected, and its like people leaving fits lying around and giving heroin users a bad name, methadone clinics, you know, where people are fighting out the front. And that s kind of why society looks down on us. (Samantha, client, 32 years) Clients noted that these perceptions were also linked to issues around low self-esteem and this could lead them to be very sensitive to the way others treated them.... discrimination and self esteem are really closely tied...anyway,...i feel like I m letting myself down with my drug using. I ve got track marks onmyarms...then I feel that I m not worthy to fit into maybe the upper class of society. So feeling that I will be discriminated against for being put in a stereotypical sort of setting as a drug user, to be somebody that steals, somebody that robs (Tim, client, 32 years) It is these experiences of discrimination that may create the expectation for clients in drug treatment that they will experience negative treatment and discrimination in their current and future health-care encounters. Additionally the reported feelings of self-blame and low self-worth illustrate the heightened sensitivity that people who use drugs may have towards others who they think may judge them [42,43]. Hence, they may be more sensitive to staff attitudes based on past experience of discrimination, their feelings of shame associated with drug and use, and perceptions that they both deserve and will be exposed to discrimination and negative treatment. Indeed, clients did associate feelings of perceived discrimination based on society s perceptions of drug users with decisions not to access or remain engaged with health care and/or drug treatment. that is just the thing of a heroin addict, that you have, I guess, the society view that we are...costing the tax payers money and that kind of shit. Not worth trying to get help,...so I guess people would feed into that negative kind of view and maybe that would stop them from reaching out I guess. (Samantha, client, 32 years).
5 Discriminatory treatment and drug treatment completion 495 Staff seemed to understand that their clients may have some negative expectations based on prior treatment experiences involving both real and perceived discrimination, and that there was a need to address this issues within the current encounter. They also expressed the difficulties involved in working with clients expectations of how they are treated by staff....wehavehadclients who come in here and expect to be treated a certain way because of their using and addiction. And they do come with guard up and ready for a battle. It can be really difficult to get them to take that guard down and to know they are here for support and service. (Cathy, worked in drug and alcohol sector for 1 year) Our clients feel the weight of that judgment when they come in here and its discernable in our treatment in our program. It s something that we name, something we acknowledge and work with. (Jill, worked in drug and alcohol sector 20 years) Staff were very open to and interested in the idea that their clients may perceive their behaviour as discriminatory and provided suggestions for working with clients to improve treatment completion, including allowing clients to voice their concerns in a safe space and dealing with any history of past discrimination, real or perceived. Staff also noted the importance of reflecting on the way they interacted with clients to ensure that their behaviour (e.g. enforcing rules of the treatment facility) was not misinterpreted or negatively construed by clients. Clients, on the other hand, identified that staff who themselves had been drug users would be critical to resolving issues around client perceptions of how they are treated by staff and around discrimination. Clients reported that this experience made staff credible and that these staff were more able to genuinely interpret client issues, particularly real or perceived discrimination. Discussion Most research on treatment outcomes for people who use illicit substances focuses on variables associated with the client, such as motivational status, frequency of drug use and psychological health. This research illustrates the importance of the interaction between staff and client variables in determining outcomes, specifically client perceptions of discrimination by staff. The data show that perceived discrimination is a significant predictor of treatment completion; clients who felt staff did not treat them well were less likely to complete treatment. This finding points to the significance of exploring other non-client-related variables in understanding treatment outcomes. People who inject drugs may interact with others in particular ways because they anticipate that they will be discriminated against. Studies on perceived stigma have found that expectations can influence what is attended to in a social situation and the inferences that are drawn about others . Similarly, fear of rejection and discrimination can create strained and uncomfortable interaction with those who could potentially stigmatise, such as treatment staff . Research has shown that the more the stigmatising condition is viewed as pivotal in shaping social encounters, the greater the expectation that interactions with others will be influenced by this stigmatised condition . As the qualitative data show, drugusing clients are already sensitised to discrimination and may expect to be treated negatively by treatment staff. These feelings could also be based on past negative experiences in the health sector [47 49]. Hence the findings of this study point to the need for drug treatment staff to be aware of the potential health consequences of perceived discrimination. As the qualitative data suggest, both staff and clients need to consider any past history of discrimination and negative health-care experiences that could shape clients perceptions of staff within the current treatment encounter. Additionally staff who have a history of past drug use are perceived by clients to be more sensitive to their experiences especially in relation to concerns about prejudice and discrimination. Suggestions for methodological directions in treatment outcome research It is important to highlight the limitations of the current research. The quantitative sample was small and results cannot be generalised beyond residential rehabilitation facilities. Additionally, this study was exploratory and relied on the self-report of clients to assess their perceptions of discrimination by staff. Future research should include corresponding ratings from staff regarding their own attitudes toward clients so that behaviours and attitudes can be compared and issues around discrimination can be assessed as real or perceived. It would also be helpful to have implicit measures of attitudes toward clients . These measures might be predictive of clients perceptions if self-report measures of prejudice from staff prove to be too fraught with social desirability to be of use . Despite these concerns, this study has sought to understand issues impacting on treatment outcomes beyond those usually considered in the literature. This is of particular importance as completion rates in drug treatment are low. In this study only 38% of clients
6 496 L. Brener et al. completed treatment. Client-centred variables have typically been the focus of prior research, presumably because these were hypothesised to be useful in producing better outcomes or tailoring programs to target specific subgroups. There has also been some previous work on staff-level variables, such as skills and beliefs. However, literature on the interaction of client and staff experiences is limited. As has been argued previously in relation to health care, focusing only or separately on the individual client or health-care worker obscures social, organisational and structural factors that may influence outcomes . The position we have taken in this research acknowledges the role of perceptions in treatment for people engaged in a stigmatised and illegal activity . To examine how people who use drugs can participate effectively in health care, drug treatment and in research itself requires a shift in conceptual and methodological approach. What we have explored are social aspects of the interaction between clients and staff. By including both groups as participants we can juxtapose their views and by feeding back results to elicit additional responses, we have attempted to further explore these views with a qualitative approach. In so doing we are not proposing ways that client perceptions can be modified. Rather, we are examining how the structure of the service (including its staffing profile) can be enhanced to maximise its readiness and fitness for duty (, p. 248). Methodologically, we have addressed this through a mixed-method design where the quantitative data were a precursor to further exploration through a qualitative approach. We see this type of conceptualisation and methodological approach as potentially helpful in enabling research to increase the quality of care afforded to people who use illicit substances and, in so doing, help ensure that they feel supported in and able to complete drug treatment programs. Acknowledgements This work was supported by an Early Career Grant from the Faculty of Arts and Social Sciences, UNSW. The National Centre in HIV Social Research is supported by a grant from the Australian Government Department of Health and Ageing. References  Moos RH, Moos BS. Long-term influence of duration and intensity of treatment on previously untreated individuals with alcohol use disorders. Addiction 2003;98:  Etheridge RM, Craddock SG, Hubbard RL, Rounds- Bryant JL. The relationship of counselling and self-help participation to patient outcomes in DATOS. Drug Alcohol Depend 1999;57:  Rabinowitz J, Marjefsky S. Predictors of being expelled from and dropping out of alcohol treatment. Psychiat Serv 1998;49:  Stark MJ. Dropping out of substance abuse treatment: a clinically oriented review. Clin Psychol Rev 1992;12:  Meier P. A national survey of retention in residential rehabilitation services. National Treatment Agency for Treatment Substance Misuse. London: National Treatment Agency,  Moos R, King M. Participation in community residential treatment and substance abuse patients outcomes at discharge. J Subst Abuse Treat 1997;14:  Harris R, Linn M, Pratt T. A comparison of dropouts and disciplinary discharges from a therapeutic community. Int J Addict 1990;15:  Anderson S, Berg J. The use of a sense of coherence test to predict drop-out and mortality after residential treatment of substance abuse. Addict Res Theory 2001;9:  Ball S, Carroll KM, Canning-Ball M, Rounsaville BJ. Reasons for dropout drug abuse treatment: symptoms, personality, and motivation. Addict Behav 2006;31:  Claus R, Kindleberger L. Engaging substance abusers after centralized assessment: predictors of treatment entry and dropout. J Psychoactive Drugs 2002;34:  Maglione M, Chao B, Aglin M. Correlates of outpatient drug treatment drop-out among methamphetamine users. J Psychoactive Drugs 2000;32:  Evans E, Li L, Hser Y. Client and program factors associated with dropout from court mandated drug treatment. Eval Program Plann 2009;32:  Joe G, Simpson D, Broome K. Retention and patient engagement models for different treatment modalities in DATOS. Drug Alcohol Depend 1999;57:  Ryan R, Plant R, O Malley S. Initial motivations for alcohol treatment: relations with patient characteristics, treatment involvement, and dropout. Addict Behav 1995;20:  Daly JE, Pelowski S. Predictors of dropout among men who batter: a review of studies and implications for research and practice. Violence Vict 2000;15:  Grosenick JK, Hatmaker CM. Perceptions of the importance of physical setting in substance abuse treatment. J Subst Abuse Treat 2000;18:  Caplehorn J, Hartel D, Irwig L. Measuring and comparing the attitudes and beliefs of staff working in NewYork methadone maintenance clinics. Subst Use Misuse 1997;32:  Humphreys K, Noke J, Moos RH. Recovering substance abuse staff members beliefs about addiction. J Subst Abuse Treat 1996;13:75 8.  Reid G, Crofts N, Hocking J. Needs analysis for primary health care among the street drug using community in Footscray. Melbourne: The Centre for Harm Reduction, Macfarlane Burnet Centre for medical Research,  Gaume J, Gmel G, Faouzi M, Daeppen J. Counselor skill influences outcome of brief motivational interventions. J Subst Abuse Treat 2009;37:  Meier PS, Donmall MC, McElduff P, Barrowclough C, Heller RF. The role of early therapeutic alliance in predicting drug treatment dropout. Drug Alcohol Depend 2006;83:  Melnick G, Wexler HK, Chaple M, Banks S. The contribution of consensus within staff and client groups as well as concordance between staff and clients to treatment engagement. J Subst Abuse Treat 2006;31:
7 Discriminatory treatment and drug treatment completion 497  Melnick G, Wexler HK, Cleland CM. Client consensus on beliefs about abstinence: effects on substance abuse treatment outcomes. Drug Alcohol Depend 2008;93:30 7.  Petry RK, Bickel WK. Therapeutic alliance and psychiatric severity as predictors of completion of treatment for opioid dependence. Psychiatr Serv 1999;50:  Harris AHS, McKellar JD, Moos RH, Schaefer JA, Cronkite RC. Predictors of engagement in continuing care following residential substance use disorder treatment. Drug Alcohol Depend 2006;84:  Brener L, von Hippel W, Kippax S, Preacher K. The role of physician and nurse attitudes in the health care of injecting drug users. Subst Use Misuse (in press).  McLaughlin D, McKenna H, Leslie J. The perceptions and aspirations illicit drug holders hold toward health care staff and the care they receive. J Psychiatr Ment Health Nurs 2000;7:  Brener L, von Hippel W, Kippax S. Health care workers and discrimination towards injecting drug users with hepatitis C: does greater contact lead to less prejudice. Int J Drug Policy 2007;18:  von Hippel W, Brener L, von Hippel C. Implicit prejudice toward injecting drug users predicts job turnover intentions among drug and alcohol nurses. Psychol Sci 2008;19:7 11.  Greenwald AG, Poehlman TA, Uhlmann E, Banaji MR. Understanding and using the Implicit Association Test: III. Meta-analysis of predictive validity. J Perso Soc Psychol 2009;97:  Link BG, Phelan JC. Stigma and its public health implications. Lancet 2006;367:  Paterson BL, Backmund M, Hirsch G, Yim C. The depiction of stigmatization in research about hepatitis C. Int J Drug Policy 2007;18:  Bogart LM, Bird ST, Walt LC, Delahanty DL, Figler JL. Association of stereotypes about physicians to health care satisfaction, help-seeking behaviour, and adherence to treatment. Soc Sci Med 2004;58:  Facione N, Miaskowski C, Dodd M, Paul S. The selfreported likelihood of patient delay in breast cancer: new thoughts for early detection. Prevent Med 2000;34:  Denzin N. The research act: a theoretical introduction to sociological methods. Chicago: Aldine,  Kimchi J, Polivka B, Stevenson J. Triangulation in action. Admin Sci Quart 1991;24:  Morse J. Principles of mixed methods and multimethod research design. In: Tashakkori A, Teddlie C, eds. Handbook of mixed methods in social and behavioral research. London: Sage, 2003:  Elliott A, Chapman S. Heroin hell of their own making : construction of heroin users in the Australian press Drug Alcohol Rev 2000;19:  Gossop M, Darke S, Griffiths P, et al. The Severity of Dependence Scale (SDS): psychometric properties of the SDS in English and Australian samples of heroin, cocaine and amphetamine users. Addiction 1995;90:  Joe G, Simpson D, Broome K. Effects of readiness for drug abuse treatment on client retention and assessment of process. Addiction 1998;93:  Braun V, Clarke V. Using thematic analysis in psychology. Qual Res Psychol 2006;3:  Buchanan J, Young L. The war on drugs a war on drug users? Drugs Educ Prev Policy 2000;7:  Lawless S, Kippax S, Crawford J. Dirty, diseased and undeserving: the positioning of HIV positive women. Soc Sci Med 1996;43:  Strenta AC, Kleck RE. Physical disability and the perception of social interaction: it s not what you look at but how you look at it. Perso Soc Psychol Bull 1984;10:  Link BG, Struening EL, Neese-Todd S, Asmussen S, Phelan JC. The consequences of stigma for the self-esteem of people with mental illnesses. Psychiatr Serv 2001; 52:  Strenta AC, Kleck RA. Physical disability and the attribution dilemma: perceiving the causes of social behavior. J Soc Clin Psychol 1985;3:  Day C, Ross J, Dolan K. Hepatitis C-related discrimination among heroin users in Sydney: drug user or hepatitis C discrimination? Drug Alcohol Rev 2003;22:  Hopwood M, Treloar C, Bryant J. Hepatitis C and injecting-related discrimination in New South Wales, Australia. Drugs Educ Prevent Policy 2006;13:  Treloar C, Hopwood M. Infection control in the context of hepatitis C disclosure: implications for education of healthcare professionals. Educ Health 2004;17:  Wolfe D. Paradoxes in antiretrovial treatment for injecting drug users: access, adherence and structural barriers in Asia and the former Soviet Union. Int J Drug Policy 2007;18: