Sexual risk behaviors and predictors of inconsistent condom use among crack cocaine users in the French overseas territories in the Americas

Abstract

The French overseas territories in the Americas are the French territories most affected by HIV. Crack cocaine users are particularly impacted. The objective of this study was to describe sexual risk behaviors and inconsistent condom use among crack cocaine users. A total of 640 crack cocaine users were interviewed. The sampling method was non-probabilistic. An anonymous standardized questionnaire of 110 questions was used. Over a third of persons did not use condoms systematically with casual sex partners (36.2%) or commercial sex partners (32.3%), and 64% did not use condoms systematically with regular sex partners. In a context of multiple sexual partnerships, frequent forced sex, and insufficient HIV testing, the lack of systematic condom use may favor the spread of HIV. Different predictive factors of inconsistent condom use were identified, depending on the type of partner, with a log binomial regression. However, among the predictor variables identified, two predictive factors frequently occur among the different type of sex partners: high perceived risk of HIV and self-perceived capacity to persuade a sex partner to use condom. These results pointed towards interventions improving access to rights and raising perceived self-efficacy.

Keywords

Crack cocaine users sexual risk behaviors French overseas territories HIV condom use

Introduction

French Guiana, Martinique and Guadeloupe are the French territories most affected by HIV, with an annual incidence of 59/100,000 person-years in 2009–2010 versus 17/100,000 in mainland France.¹ The epidemic probably predominates among vulnerable groups, notably crack cocaine users. The French territories in the Americas represent a hub of narcotics’ trafficking towards Europe. Consumption of crack cocaine is a widespread problem.^2–5 The links between crack cocaine and HIV have been reported by several authors.^6–8 These observations are consistent with the findings in the French territories where a study in 2011 at the addictology clinic in Cayenne showed an HIV prevalence of 15% among crack cocaine users.⁹ Another study in French Guiana showed that 16.8% of HIV-positive patients had previously tried crack cocaine.¹⁰ The link between crack and HIV is mostly explained by high risk sexual behaviors among crack cocaine users.^6,11 Researchers often report low levels of condom use among crack cocaine users.^7,12–14 A great variety of predictors of inconsistent condom use has been described in the literature, notably partner characteristics, school level, multiple sexual partnerships, financial dependence, intimate partner violence, partner refusal to use condom, living in a couple, desire to have a child, low empowerment, etc.^13,15,16

In 2012, a cross-sectional study was conducted in Guadeloupe, Martinique and French Guiana in order to describe the knowledge, attitudes, behaviors, and practices of crack cocaine users regarding HIV and other sexually transmitted infections. The present study aimed to describe sexual risk behaviors and inconsistent condom use among crack cocaine users.

Materials and methods

Study population

The study population consisted of persons aged 18 or older having used crack cocaine (‘crack’) in the past three months. The data collection took place between March 2012 and September 2012. The sample size was 640 persons interviewed in French Guiana, Guadeloupe and Martinique. These French territories are located in the West Indies for Guadeloupe (403,314 residents in 2012 according to the Institut National de la Statistique et des Etudes Economiques) and Martinique (388,364 residents in 2012) and in South America for French Guiana (239,648 residents in 2012).

Study questionnaire

An anonymous structured questionnaire of 110 questions was used. Because of the large number of persons who do not speak French, the questionnaire was translated into five languages (French, Spanish, English, Portuguese and Haitian Creole) and the investigators spoke at least one of these languages. It notably explored sociodemographic characteristics, sexual history, condom use, sexual risk behaviors and crack use. This questionnaire was based on a behavioral surveillance survey questionnaire, which was adapted to the context with non-government associations (NGO) working in the field.¹⁷ In order to avoid selection biases,¹⁷ HIV status was not part of the questions asked.

Study conduct

Study participants were interrogated face to face by trained interviewers. Individual interviews were conducted confidentially. The surveyors were all close to the field, often working with NGOs.

The study was conducted in partnership with the regional coordination for the fight against HIV and different NGOs working in the field of addiction or HIV. The study was part of the project ‘Mise en place d’un observatoire intercaribéen du VIH’ supported by the University Hospital of Pointe-à-Pitre. The project was funded by the European INTERREG Caribbean IV programme, by Cayenne Hospital, the French NGO AIDES and the University Hospital of Pointe-à-Pitre.

Sampling method

The sampling scheme combined convenience sampling at NGO premises receiving crack cocaine users in French Guiana, Martinique and Guadeloupe (75% of inclusions), and snowball sampling outside these structures in order to reach crack cocaine users who do not frequent any structures (25% of inclusions). The target population size was estimated by the NGOs working with crack cocaine users at 4800, the sample size objective was 600, which was expected to measure a proportion of 50% with a 4% precision and 95% confidence interval. Crack cocaine users were interviewed in a confidential location within the recruitment centres.

At the end of the interview, documentation on HIV prevention, NGO contact information, male condoms, lubricant and a backpack were given to the participants in order to compensate for the time spent answering the questionnaire.

Data analysis

Data analysis consisted of a descriptive analysis of the variables: mean and standard deviation for normally distributed variables, median for the non-normal distributed variables, and frequencies and percentages for qualitative variables. Normality was verified using the Shapiro–Wilk test. Bivariate and multivariate regressions were performed and prevalence ratios were used to identify predictive factors for inconsistent condom use with regular, casual and commercial partners. Regular partners were defined as partners with whom the interviewed person had sexual intercourse with regularly during a six-month period, commercial partners were defined as partners with whom the person had sexual intercourse in exchange for money, services, goods or drugs, and casual partners were those who were not regular and not commercial. Variables with a p < 0.20 in the bivariate analyses were included in stepwise models in order to select variables for the final model. We wanted to estimate the prevalence ratios, and as odds ratios are distorted estimates of the relative risks when exposure prevalence is high, we calculated the prevalence ratios directly with a log binomial model.¹⁸

Variance inflation factors were used to test for colinearity. Data were analyzed using STATA 13® (College Station, TX, USA).

Ethical and regulatory issues

The project was approved by the French regulatory authorities: the Comité d’Evaluation Ethique de l’INSERM (CEEI, avis no.12-056), the Comité Consultatif sur le Traitement de l’Information en matière de Recherche dans le domaine de la Santé (CCTIRS, n°12-084) at the French Ministry of Research and Higher Education, and the Commission Nationale Informatique et Liberté (CNIL, DR-2012-263).

Results

Sociodemographic characteristics

Overall, 640 crack cocaine users answered the questionnaire. Among them, 42% (n = 268) were interviewed in Guadeloupe, 41% (n = 265) in Martinique and 17% (n = 107) in French Guiana. The sample mostly included men (79%). The average age was 37.6 years (±9.5 years). Over three-quarters of the persons surveyed (77.3%) did not live as a couple, 37% had no one to rely on in case of difficulty, and 60% felt lonely, or very lonely.

The other nationalities mostly included Surinamese, Guyanese, Dominican, Brazilian, and Haitian nationals. Over one-third (37%) of respondents had no one to rely on in case of difficulty and 60% felt lonely, or very lonely. Over a third of them (37%) sometimes had difficulties in maintaining their personal hygiene (taking a shower, washing their clothes, etc.).

Table 1 shows the main sociodemographic characteristics of the sample.

Table 1.

Socio-demographic characteristics of the surveyed crack cocaine users (n = 640).

Characteristics	n	%
Gender (n = 639)
Male	505	79.03
Female	134	20.97
Age (years) (n = 640)
18–24	63	9.84
25–34	163	25.47
35–44	258	40.31
45–54	129	20.16
>55	27	4.22
	Mean: 37.6 (±9.5)
	Min: 18	Max: 66
Nationality (n = 640)
French	531	82.97
Other	109	17.03
Possession of a residence permit (n = 109)
No	59	56.19
Yes	46	43.81
Valid health insurance (n = 621)
Yes	414	66.67
No	207	33.33
Lives as a couple (n = 615)
Yes	120	19.51
No	495	80.49
Employment (n = 623)
No	321	51.52
Yes, with contract	105	16.85
Yes, without contract	197	31.63
Type of housing (n = 637)
Personal accommodation	382	59.97
Specialized structures	51	8.01
Precarious housing	204	32.02
Age at end of school (years) (n = 608)
	Mean: 17.11 (±2.86)
	Min: 8	Max: 29
16 or younger^a	281	46.22
Older than 16	327	53.78

16 is the age until which school is mandatory.

Sexuality and sexual risks

Of the 640 surveyed persons, 638 had previously had sex, 526 in the past 12 months. Nearly three-quarters of persons (74%) declared they were heterosexual, 4% declared they were homosexual, 12% declared they were bisexual and 10% did not wish to respond to that question. There were 232 persons (36.2%) who reported at least one commercial sex partner, 297 (46.4%) who reported at least one regular sex partner, and 246 persons (38.4%) who had a casual sex partner in the past 12 months. The persons surveyed reported a median number of 3.5 partners in the past 12 months (minimum number of partners: 0; maximum number of partners: 600). Nearly two-thirds (65.7%) declared having more than one sex partner during the past 12 months.

Seventy-one percent of crack cocaine users surveyed declared having used crack at least once during or before sexual intercourse in the past month. Nearly 19% had been forced to have sex, and this proportion was 53% among women and 10% among men (p < 0.001, Pearson’s Chi square test).

Nearly 10% of persons declared having had genital sores in the past 12 months (5.5% among men and 23.1% among women, p < 0.001, Pearson’s Chi square test), and among women 28% reported unusual vaginal discharge in the past 12 months. Over a quarter of persons (27%) had never performed an HIV test during their life.

We asked crack cocaine users to rate their own risk of contracting HIV, of whom nearly two-thirds (62%) of persons estimated they had a low risk of getting HIV because they had few partners or always used a condom.

Condom use

Inconsistent condom use was frequent among crack cocaine users for all types of partners: 64% did not systematically use a condom with regular partners, 36.2% did not with casual partners and 32.3% did not systematically use a condom with a commercial partner. Table 2 presents the frequency of use of condoms with the different types of sex partners.

Table 2.

Condoms use in crack cocaine users with regular, casual and commercial partners.

	Regular partner		Casual partner		Commercial partner
	n (%)	95.0%CI	n (%)	95.0%CI	n (%)	95.0%CI
Frequency of condom use in the past 12 months	N = 297^a		N = 246^a		N = 232^a
Every time	99 (33.3)	28.0–39.0	155 (63.0)	57.0–69.0	151 (65.1)	59.0–71.0
Never/sometimes	190 (64.0)	58.0–69.0	89 (36.2)	30.0–42.0	75 (32.3)	26.0–39.0
Condom use at last sexual intercourse	145 (48.8)	43.0–55.0	199 (80.9)	75.0–86.0	190 (81.9)	76.0–87.0

Differences between N and the sum of the two categories are due to non-responses.

The reasons given for not using condoms varied according to the type of partner. For regular partners, half of the participants mentioned trust in their partner (51.4%). For commercial sex partners, 22% of responders said the partner did not want to use any. Finally, for casual sex partners, other reasons are given such as use of crack or alcohol (24%).

Over 10% of the surveyed population did not think they could propose, or did not know if they would be able to propose condom use with a sex partner. A third (33.9%) of the sample did not think they could insist and persuade their sex partner to use a condom if he/she initially refused.

Table 3 presents the bivariate analysis used to determine which variables were selected in the multivariate models.

Table 3.

Bivariate analysis regarding inconsistent condom use.

	With regular partner		With casual partner		With commercial partner
	Crude prevalence ratios (95% CI)	p	Crude prevalence ratios (95% CI)	p	Crude prevalence ratios (95% CI)	p
Gender
Male	0.961 (0.697–1.325)	0.808	0.819 (0.446–1.505)	0.520	1.561 (0.981–2.485)	0.060
Female	1		1		1
Age	1.002 (0.986–1.018)	0.800	1.003 (0.981–1.025)	0.779	0.985 (0.961–1.009)	0.215
Perceived risk of HIV
High	1.592 (1.108–2.288)	0.012	2.756 (1.776–4.278)	0.000	3.430 (2.177–5.404)	0.000
Low/do not know/no answer	1		1		1
Lives in a couple
Yes	1.414 (1.058–1.889)	0.019	0.831 (0.469–1.472)	0.526	0.388 (0.168–0.897)	0.027
No	1		1		1
Multiple sexual partnerships
Yes	0.679 (0.506–0.910)	0.010	0.889 (0.556–1.421)	0.623	1.040 (0.590–1.833)	0.893
No	1		1		1
Comfortable asking a partner to use condom
Yes	0.725 (0.412–1.275)	0.264	0.464 (0.251–0.854)	0.014	0.593 (0.326–1.078)	0.087
No/do not know	1		1		1
Comfortable persuading a partner to use condom
Yes	0.704 (0.511–0.969)	0.031	0.375 (0.245–0.576)	0.000	0.305 (0.190–0.489)	0.000
No/do not know	1		1		1
Lives in a substandard housing
Yes	1.009 (0.722–1.409)	0.960	1.760 (1.158–2.674)	0.008	1.974 (1.250–3.119)	0.004
No	1		1
Has been forced to have sexual intercourse by a partner
Yes	0.943 (0.656–1.356)	0.752	0.552 (0.286–1.066)	0.077	1.477 (0.918–2.375)	0.108
No	1		1		1
Age at which left school
16 or younger^a	0.861 (0.645–1.149)	0.310	0.929 (0.601–1.434)	0.739	0.620 (0.388–0.992)	0.046
Older than 16	1		1		1
Has difficulties maintaining personal hygiene
Yes	1.264 (0.936–1.708)	0.126	1.854 (1.215–2.829)	0.004	1.804 (1.136–2.866)	0.012
No	1		1		1
Valid health insurance
Yes	1.024 (0.739–1.419)	0.887	1.165 (0.676–2.009)	0.582	0.592 (0.374–0.937)	0.025
No	1		1		1
Has ever had an HIV test
Yes	0.931 (0.656–1.320)	0.688	0.724 (0.464–1.129)	0.155	0.845 (0.517–1.383)	0.504
No	1		1		1
Has someone to rely on
Yes	1.115 (0.786–1.581)	0.541	0.636 (0.417–0.970)	0.035	0.570 (0.357–0.909)	0.018
No	1		1		1
Feels …
Alone/very alone	1.212 (0.907–1.619)	0.194	0.745 (0.469–1.183)	0.213	0.814 (0.496–1.336)	0.416
Supported/very supported	1		1		1

Bold values are significant < 0.20. ^a16 years is the age until which school is mandatory.

Table 4 shows the factors associated with inconsistent condom use with different types of partners during the last 12 months. Inconsistent condom use was correlated, whatever the type of partner, with thinking oneself to be at risk for HIV. Inconsistent condom use with regular partner was positively correlated with living as a couple and was negatively correlated with multiple sexual partnerships. Inconsistent condom use with a casual partner was associated with living in substandard housing, not being comfortable about persuading a partner to use a condom and having difficulties maintaining personal hygiene. Inconsistent condom use with a commercial partner was associated with leaving school before 16 and not being comfortable about persuading a partner to use a condom.

Table 4.

Predictive factors of inconsistent condom use during the past 12 months with regular, casual and commercial partners.

	Inconsistent condom use during the past 12 months/total (%)	Crude prevalence ratios (95% CI)	Prevalence ratios (95% CI)	p
With regular partners^a
Perceived risk of HIV
High	36/37 (97.3)	1.592 (1.108–2.288)	1.972 (1.325–2.934)	0.001
Low/do not know/no answer	154/252 (61.1)	1	1
Lives in a couple
Yes	83/103 (80.6)	1.414 (1.058–1.889)	1.498 (1.088–2.061)	0.013
No	102/179 (57.0)	1	1
Multiple sexual partnerships
Yes	111/194 (57.2)	0.679 (0.506–0.910)	0.669 (0.481–0.932)	0.017
No	75/89 (84.3)	1	1
With casual partners^b
Perceived risk of HIV
High	30/38 (78.9)	2.756 (1.776–4.278)	2.095 (1.275–3.442)	0.003
Low/do not know/no answer	59/206 (28.6)	1	1
Comfortable persuading a partner to use condom
Yes	43/176 (24.4)	0.375 (0.245–0.576)	0.529 (0.327–0.854)	0.009
No/do not know	41/63 (65.1)	1	1
Lives in a substandard housing
Yes	42/83 (50.6)	1.760 (1.158–2.674)	1.683 (1.068–2.650)	0.025
No	46/160 (28.7)	1	1
Had been forced to have sexual intercourse by a partner
Yes	10/44 (22.7)	0.552 (0.286–1.066)	0.574 (0.292–1.126)	0.107
No	79/192 (41.1)	1	1
With commercial partners^c
Perceived risk of HIV
High	34/44 (77.3)	3.430 (2.177–5.404)	2.301 (1.322–4.007)	0.003
Low/do not know/no answer	41/182 (22.5)	1	1
Comfortable persuading a partner to use condom
Yes	27/147 (18.4)	0.305 (0.190–0.489)	0.448 (0.249–0.807)	0.008
No/do not know	47/78 (60.2)	1	1
Age at which left school
16 or younger^d	46/109 (42.2)	0.620 (0.388–0.992)	0.695 (0.403–1.200)	0.192
Older than 16	28/107 (26.2)	1	1

Note: The fitness of the models was measured with the ratio deviance/degrees of freedom. Models fit well with a ratio below 1. Variance inflation factors ranged from 1.08 to 1.97 far below the threshold of 4.

Ratio D/df = 0.505 and 64 missing values.

Ratio D/df = 0.597 and 33 missing values.

Ratio D/df = 0.568 and 44 missing values.

16 is the age until which school is mandatory.

Discussion

Although great attention was given to minimizing the risks of non representativeness, the sampling method used was non-probabilistic and combined two recruitment techniques; thus, given the risks of selection biases, the present results cannot be extrapolated to all crack cocaine users in the area. Moreover, the study was designed as a knowledge, attitudes, beliefs and practices (KABP) survey. Responses were based on declarative data, which are potentially impacted by response and memory biases.^17,19,20 The study was cross sectional, which implies some limitations. This type of study cannot prove cause and effect relationships, and the timing of this snapshot is not guaranteed to be representative.²¹

Despite these limitations, the present results are among the first to focus on HIV among crack cocaine users in the French overseas territories in the Americas. Although NGOs working in the field of addiction work with crack cocaine users, their actions are not centered on HIV. Similarly, HIV prevention and testing NGOs are not necessarily centered on crack users. The present results reemphasize the high risk of this population in the surveyed territories and the importance of specific prevention. It also suggests some interventional strategies that could guide prevention efforts.

Cumulated vulnerabilities

The persons surveyed were often in very precarious situations. Generally, crack consumption is very stigmatized and persons become quickly isolated from society.^22,23 This is also the case in the studied French overseas territories,^3,4 where half of them did not work, and a third lived in precarious conditions (squat, street, spontaneous habitat, etc.). Few crack cocaine users lived in a couple, many had no one to rely on in case of difficulties, and often suffered from loneliness. Stigmatization induces barriers for access to rights and health.^22,24 In the present study, one-third of the surveyed persons had no valid health insurance at the time of the survey.

Irregular use of condoms and cumulated risks

The use of condoms was not consistent among those surveyed, and they did not use condoms systematically with their regular partner, as observed in the general population.²⁵ However, over a third of persons did not use condoms systematically with casual sex partners or commercial sex partners. In the context of multiple sexual partnerships, frequent forced sex, and insufficient HIV testing, the lack of systematic condom use may favor the spread of HIV.

The predictive factors of non-systematic condom use were studied for each type of sex partner with different risk factors. Thus with regular partners, living as a couple was linked to inconsistent use of condoms for multiple reasons (difficulty in negotiating condom use, fear of breaking trust with regular partners and bringing discord in the couple, desire to have a child, decrease of sexual satisfaction with condom use, crack and/or alcohol use, etc.).^16,26,27 However, having multiple sex partners was associated with systematic condom use with the regular partner, suggesting that persons were aware of the risks associated with multiple sex partners.

The factors associated with non-systematic condom use with casual sex partners were more linked to social precariousness. Hence, living in precarious housing or having difficulties in maintaining personal hygiene were associated with non-systematic condom use. The association between socioeconomic level and consistent condom use was already described in the literature.²⁸

Non-systematic condom use with commercial sex partners was negatively correlated with leaving school before 16, an unexpected and unexplained finding given that the reverse is usually described.²⁹ Further investigations would be required to explain this finding.

Risk perception and perceived self-efficacy

Despite these differences, two predictive factors of non-systematic condom use were found for almost all types of sex partners. Although several studies clearly show that a high perception of one’s risk of getting HIV is associated with change towards less risky sexual behaviors,^30,31 the present study found that perception of one’s high risk of getting HIV was linked to non-systematic condom use. Persons thus presumably inferred their high risk because they did not use condoms, but did not change their high risk behavior. Social, cultural or pharmacological factors may have sustained the risky behavior.^32,33 There seemed to be a gap between one’s perception of sexual risk and one’s preventive attitudes, notably condom use.

Self-perceived capacity to persuade a sex partner to use a condom was negatively associated with non-systematic condom use with casual and commercial sex partners. Thus, it seems that when persons felt capable of negotiating condom use with the partner, they used condoms more systematically. Self-perceived capacity is often cited as an important element to adapt health behaviors.^27,34,35 It is a component of various behavioral theories, such as the Health Belief Model, Social Cognitive Theory, or the stages of change model.^36–41 The perceived self-efficacy may be affected in situations of stigma, addiction, and violent relations.^22,42,43 Over a third of surveyed persons thus did not feel capable of insisting that a sex partner use a condom. Reinforcing this perception of self-efficacy could thus be an important intervention target, notably regarding negotiation skills and condom use.

Prevention strategies

These results may thus help optimize HIV prevention among crack cocaine users. Behavioral change may be achieved by acting on different levels^44,45 notably at the individual and institutional levels to improve access to rights and access to care.

At the individual level, risk reduction is a priority.⁴⁶ Improving self-perceived capacity and reducing stigma may help reduce risky sex behaviors among drug users.^34,47

At the institutional level, the objective should be to achieve risk reduction by reducing obstacles and improving social and physical environments to facilitate access to rights for crack cocaine users.^46,48

Conclusion

This is the first HIV KABP study of crack cocaine users, one of the most at risk populations for HIV. The study shows frequent multiple partnerships, commercial sex, and insufficient condom use and HIV testing. Although inconsistent condom use rates are slightly lower than those found in the literature (about 60% reported by Sherman¹³ and 56% reported by Booth⁷), they are in the same order of magnitude and remain a significant problem in this population. Although sexual behaviors and risk-taking differ significantly from one type of partner to another, few studies take this into consideration. Studies differentiating types of partners are needed in this population. The results thus emphasize the importance of implementing specific HIV prevention interventions in this population. Among the predictor variables identified, those linked to self-efficacy pointed towards interventions improving access to rights and raising perceived self-efficacy.

Footnotes

Acknowledgements

We would like to thank Barbara Gontier, Stéphane Barboza, Katia Delagneau, Jérôme Evanno and Arlette Suzanne, field coordinators, Marcus Day and the Caribbean Drug Abuse Research Institute in Saint-Lucia, the project team and all the participants of the study.

Declaration of conflicting interests

The authors declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.

Funding

The authors disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This present study was supported by the European INTERREG Caribbean IV Fund, by the Hospital of Cayenne, by the NGO AIDES and by the University Hospital of Pointe-à-Pitre.

References

Le Vu

Le Strat

Barin

et al.

Population-based HIV-1 incidence in France, 2003–08: a modelling analysis. Lancet Infect Dis 2010; 10: 682–687.

Charles-Nicolas

Lacoste

Ballon

. Le point sur l’addiction à la cocaïne et au crack. Annales Médico-psychologiques, revue psychiatrique 2009; 167: 504–507.

OSM/CIRDD. Enquête CAME, Crack à la Martinique: Etat des lieux. In: Centre d'information et de ressources sur les drogues et les dépendances/Observatoire de la santé de la Martinique, 2008.

ORSAG. Profils et trajectoires des usagers de crack en Guadeloupe. In: Observatoire Régional de la Santé de Guadeloupe, 2008.

OFDT. Phénomènes émergents liés aux drogues en 2004, tendances récentes sur le site de la Guyane. In: Observatoire français des drogues et des toxicomanies, 2005.

Leigh

Stall

. Substance use and risky sexual behavior for exposure to HIV: issues in methodology, interpretation, and prevention. Am Psychol 1993; 48: 1035–1035.

Booth

Watters

Chitwood

. HIV risk-related sex behaviors among injection drug users, crack smokers, and injection drug users who smoke crack. Am J Public Health 1993; 83: 1144–1148.

Edlin

Irwin

Faruque

et al.

Intersecting epidemics – crack cocaine use and HIV infection among inner-city young adults. New Engl J Med 1994; 331: 1422–1427.

Guillaume Y. Prévalence de la séropositivité au VIH chez les usagers de crack du CSAPA de Cayenne, 2012.

10.

Nacher

Vantilcke

Parriault

et al.

What is driving the HIV epidemic in French Guiana?

Int J STD AIDS 2010; 21: 359–361.

11.

Azevedo RCSd Botega

Guimarães

LAM

. Crack users, sexual behavior and risk of HIV infection. Revista Brasileira de Psiquiatria 2007; 29: 26–30.

12.

Pallonen

Timpson

Williams

et al.

Stages of consistent condom use, partner intimacy, condom use attitude, and self-efficacy in African-American crack cocaine users. Arch Sex Behav 2009; 38: 149–158.

13.

Sherman

Latkin

. Intimate relationship characteristics associated with condom use among drug users and their sex partners: a multilevel analysis. Drug Alcohol Depend 2001; 64: 97–104.

14.

Atkinson

Williams

Timpson

et al.

Multiple sexual partnerships in a sample of African-American crack smokers. AIDS Behav 2010; 14: 48–58.

15.

Falck

Wang

Carlson

et al.

Factors influencing condom use among heterosexual users of injection drugs and crack cocaine. Sex Transmit Dis 1997; 24: 204–210.

16.

Chakrapani

Newman

Shunmugam

et al.

Prevalence and contexts of inconsistent condom use among heterosexual men and women living with HIV in India: implications for prevention. AIDS Patient Care STDs 2010; 24: 49–58.

17.

Amon J, Brown T, Hogle J, et al. Behavioral surveillance surveys BSS. Guidelines for repeated behavioral surveys in populations at risk of HIV. 2000.

18.

Barros

Hirakata

. Alternatives for logistic regression in cross-sectional studies: an empirical comparison of models that directly estimate the prevalence ratio. BMC Med Res Methodol 2003; 3: 1–1.

19.

Cobat

Halfen

Grémy

. Déterminants de l’utilisation du préservatif et multipartenariat hétérosexuel aux Antilles et en Guyane françaises. Revue d'épidémiologie et de Santé Publique 2008; 56: 143–157.

20.

Magnani

Sabin

Saidel

et al.

Review of sampling hard-to-reach and hidden populations for HIV surveillance. AIDS 2005; 19: S67–S72.

21.

Levin

. Study design III: cross-sectional studies. Evidence-Base Dentist 2006; 7: 24–25.

22.

Ahern

Stuber

Galea

. Stigma, discrimination and the health of illicit drug users. Drug Alcohol Depend 2007; 88: 188–196.

23.

Furst

Johnson

Dunlap

et al.

the stigmatized image of the''crack head'': a sociocultural exploration of a barrier to cocaine smoking among a cohort of youth in New York City. Deviant Behavior 1999; 20: 153–181.

24.

Luoma

Twohig

Waltz

et al.

An investigation of stigma in individuals receiving treatment for substance abuse. Addict Behav 2007; 32: 1331–1346.

25.

Halfen S and Lydié N. Les habitants des Antilles et de la Guyane face au VIH/Sida et à d'autres risques sexuels. In: ORS Île de France, INPES, ANRS, 2014.

26.

Macaluso

Demand

Artz

et al.

Partner type and condom use. AIDS 2000; 14: 537–546.

27.

Nehl

Elifson

DePadilla

et al.

Sex partner type, drug use and condom use self-efficacy among African Americans from disadvantaged neighborhoods: are associations with consistent condom use moderated by gender?

J Sex Res 2015; 53: 1–11.

28.

Caballero Hoyos

Villaseñor Sierra

. Socioeconomic strata as a predictor factor for consistent condom use among adolescents. Revista de Saúde Pública 2001; 35: 531–538.

29.

Lagarde

Caraël

Glynn

et al.

Educational level is associated with condom use within non-spousal partnerships in four cities of sub-Saharan Africa. AIDS 2001; 15: 1399–1408.

30.

Maharaj

Cleland

. Risk perception and condom use among married or cohabiting couples in KwaZulu-Natal, South Africa. Int Fam Plan Perspect 2005; 31: 24–29.

31.

Carter

Harry

Jeune

et al.

HIV risk perception, risk behavior, and seroprevalence among female commercial sex workers in Georgetown, Guyana. Revista Panamericana de Salud Pública 1997; 2: 238–246.

32.

Adedimeji

Omololu

Odutolu

. HIV risk perception and constraints to protective behaviour among young slum dwellers in Ibadan, Nigeria. J Health Popul Nutr 2007; 25: 146–146.

33.

Akwara

Madise

Hinde

. Perception of risk of HIV/AIDS and sexual behaviour in Kenya. J Biosocial Sci 2003; 35: 385–411.

34.

Strecher

DeVellis

Becker

et al.

The role of self-efficacy in achieving health behavior change. Health Educ Behav 1986; 13: 73–92.

35.

Montoya

. Attitudes, norms, self-efficacy, and state of change among out-of-treatment female crack cocaine users: a pilot study. AIDS Educ Prevent 1997; 9: 424–441.

36.

Glanz

Rimer

Viswanath

. Health behavior and health education: theory, research, and practice, San Francisco: John Wiley & Sons, 2008.

37.

Rosenstock

Strecher

Becker

. Social learning theory and the health belief model. Health Educ Behav 1988; 15: 175–183.

38.

Bandura A. Self-efﬁcacy: the exercise of control. New York, NY: Freeman, 1997.

39.

Bandura A. Going global with social cognitive theory: from prospect to paydirt. In: Donaldson SI, Berger DE and Pezdek K (eds) The rise of applied psychology: new frontiers and rewarding careers. Mahwah, NJ: Lawrence Erlbaum, 2006, pp.53–70.

40.

Csillik A. Polémique actuelle autour du modèle transthéorique: ce modèle mérite-t-il d’être encore utilisé? Annales Médico-psychologiques, revue psychiatrique 2009; 167: 355–360.

41.

DiClemente RJ and Peterson JL. Preventing AIDS: theories and methods of behavioral interventions. Springer Science & Business Media, 1994.

42.

Pettifor

Measham

Rees

et al.

Sexual power and HIV risk, South Africa. Emerg Infect Dis 2004; 10: 1996–2004.

43.

Schomerus

Corrigan

Klauer

et al.

Self-stigma in alcohol dependence: consequences for drinking-refusal self-efficacy. Drug Alcohol Depend 2011; 114: 12–17.

44.

Vermund

Hayes

. Combination prevention: new hope for stopping the epidemic. Current HIV/AIDS Reports 2013; 10: 169–186.

45.

Padian

McCoy

Karim

SSA

et al.

HIV prevention transformed: the new prevention research agenda. Lancet 2011; 378: 269–278.

46.

Degenhardt

Mathers

Vickerman

et al.

Prevention of HIV infection for people who inject drugs: why individual, structural, and combination approaches are needed. Lancet 2010; 376: 285–301.

47.

Kang

S-Y

Deren

Andia

et al.

Effects of changes in perceived self-efficacy on HIV risk behaviors over time. Addict Behav 2004; 29: 567–574.

48.

El-Bassel

Wechsberg

Shaw

. Dual HIV risk and vulnerabilities among women who use or inject drugs: no single prevention strategy is the answer. Curr Opin HIV AIDS 2012; 7: 326–331.