www.aids2014.org

Innovative peer-to-peer educational intervention to reduce HIV and other blood-borne infection risks in

difficult–to-reach people who inject drugs: results from the ANRS AERLI study

Patrizia Carrieri, Jean-Marie Le Gall, Marie Debrus, Baptiste Demoulin, Caroline Lions, Aurélie Haas, Marion Mora, Perrine Roux, Marie Suzan-Monti,

Bruno Spire

www.aids2014.org

• French harm reduction policies have contributed to vastly reduce HIV incidence among people who inject drugs (PWID)– ~1% of new HIV diagnoses in 2012

• However, complications related to injecting practices remain– High prevalence of Hepatitis C Virus (HepC) infections

• Needle/syringe and paraphernalia sharing– Other intravenous drug use related complications

• Local lesions : venous damage, abscesses, cellulitis, other skin infections

• Cardiovascular and pulmonary complications

Background (1)

www.aids2014.org

• Existence of persisting at-risk behaviors suggests that current Harm Reduction Programs are insufficient for PWID– Opiate substitution treatments– Needle/syringe programs– Support services

• Need to experiment with, evaluate and implement alternative and innovative strategies that effectively improve the health of this population

Background (2)

www.aids2014.org

• The objective was to assess the effectiveness and feasibility of a community-based training and education intervention on reducing injection-associated risks• Viral infections (HIV, HepC)• Other injection-related damages

• AIDES and Médecins du Monde staff members or volunteers were involved in study design, training of peers, performing the intervention, collecting and analyzing the data

Objective

www.aids2014.org

• National multi-site community-based study comparing the evolution of injection- related risks over one year in low-threshold services, hold by AIDES and MDM, between:– An intervention group, comprising people

benefitting from educational sessions in 8 participating services

– A control group, comprising people recruited in 9 services, working in accordance with current guidelines

Study design (1)

www.aids2014.org

Inclusion M6 M12

Study design (2)

Phone interviews with an independent trained interviewer not involved in the ES

Educational sessions (ES) with PWID = educational tailored intervention by NGO staff member/volunteer after observation of participants’ self-injection practices

Intervention group

Control group

At least 1 ES over 5 months

www.aids2014.org

• Eligibility criteria – ≥18 years old; injected drug at least once during the previous

week; willing to be contacted for a phone interview; able to provide written informed consent

• Data collection– Phone interviews at M0, M6 and M12 collecting :

• socio-demographic characteristics• behaviors related to HepC transmission risk (BBV-TRAQ)• experience with drug use; history of incarceration• access to care; HIV, Hep C and B testing and diagnosis• motivational factors for willing to participate in the intervention

– Data from the intervention sessions were collected by community-based association members

Methods (1)

www.aids2014.org

• Analyses– Outcomes

• Having at least one at-risk practice for HepC transmission in the previous month

• Having at least one local complication at the injection site– 2-step Heckman model to avoid possible selection bias

arising from the clustering of intervention (n=113) and control groups (n=127)

• 1st step: probit model to identify baseline factors associated with exposure to the ES to compute the inverse Mills ratio (IMR)

• 2nd step: probit mixed model to identify factors associated with each outcome after introducing the IMR to control for the clustering bias

Methods (2)

www.aids2014.org

Participants recruitedn=144

Intervention group M0n=113

Intervention group M6 n=75

Intervention group M12 n=44

Participants recruitedn=127

Control group M0 n=127

Control group M6 n=92

Control group M12 n=71

Study sample diagram

Participants who did not receive ES

N=31

Intervention sites n=8 Control sites n=9

288 ES

www.aids2014.org

Results (1) – Participants’ characteristics

† Chi-squared test or Wilcoxon test§ in years ⱡ AUDIT score ≥3 for women; ≥4 for men*during the previous 4 weeks1more than 1 HCV at-risk practice during the previous month2more than 1 problem at the injection site

Baseline characteristics (n (%) or median [IQR]), ANRS-AERLI study (n=240) Control group n=127 Intervention group n=113 p-value†

Gender

Male 99 (78) 88 (78) 0.99Age§ 31 [26-37] 30 [25-37] 0.38High School Certificate 23 (18) 36 (32) 0.01Precarious housing 17 (13) 32 (28) 0.01Age at first drug injection§ 19 [17-23] 19 [17-23] 0.73Harmful alcohol consumption ⱡ 69 (54) 63 (56) 0.77Heroin use* 32 (25) 46 (41) 0.01Cocaine/crack use* 50 (39) 55 (49) 0.19Morphine sulphate use* 36 (72) 60 (53) <0.001Buprenorphine use* 67 (53) 30 (27) <0.001Frequent daily injection 58 (46) 61 (54) 0.20HCV screening 108 (85) 84 (74) 0.04HCV risk practices1 34 (27) 49 (44) 0.01Problems at the injection site2 71 (56) 75 (66) 0.10HCV seropositivity 29 (23) 37 (33) 0.05HIV seropositivity 5 (4) 2 (2) 0.35

www.aids2014.org

Results (2) – Evolution of the outcomes

• Percentage of participants who reported at least one complication at the injection site

• Percentage of participants who reported at least one HCV at-risk practice

Follow-up

*

*

Follow-up

•p<0.05 : significant difference between M0 and M12 in the intervention group

•p<0.05 : significant difference between M0 and M6 in the intervention group

www.aids2014.org

Results (3) - Impact of education intervention on HCV risk practices

Multivariate model (n = 238, visits = 497) No of visits (%) or

median [IQR] multivariate analysis

coef [95%CI] p-

value Gender

Female Male

117 (22) 405 (78)

1

-0.56 [-1.04; -0.08]

0.022 Age§ 31 [26-38] -0.05 [-0.08; -0.02] 0.001 Harmful alcohol consumption ⱡ

No Yes

243 (47) 276 (53)

1 0.54 [1.15; 0.93]

0.007 Cocaine/crack use

No Yes

325 (62) 197 (38)

1

0.43 [0.03; 0.84]

0.036 Polydrug use*

No Yes

413 (79) 109 (21)

1

0.74 [0.26; 1.22]

0.003 Intervention

No Yes

290 (56) 232 (44)

1

1.96 [0.86; 3.07]

0.001 Follow-up

Baseline (M0) M6 M12

240 (46) 167 (32) 115 (22)

1

-0.18 [-0.69; 0.32] 0.23 [-0.28; 0.74]

0.480 0.382

Interaction Interv X Baseline Interv X M6 Interv X M12

1

-0.76 [-1.50; -0.05] -0.70 [-1.55; 0.15]

0.037 0.106

IMR 0.40 [-0.04; 0.84] 0.076

www.aids2014.org

Results (4) - Impact of education intervention on local complications at the injection site

Multivariate model (n=240, visits=520) No of visits (%) or

median [IQR] multivariate analysis

coef [95%CI] p-

value Age§ 31 [26-38] -0.04 [-0.06; -0.01] 0.010 Morphine sulfate use

No Yes

325 (62) 197 (38)

1

0.41 [-0.02; 0.83]

0.061 Intervention

No Yes

290 (56) 232 (44)

1

-0.81 [-1.80; 0.19]

0.113 Follow-up

Baseline (M0) M6 M12

240 (46) 167 (32) 115 (22)

1

0.23 [-0.19; 0.64] 0.14 [-0.31; 0.59]

0.292 0.529

Interaction Interv X Baseline Interv X M6 Interv X M12

1

0.13 [-0.53; 0.78] -1.01 [-1.77; -0.24]

0.708 0.010

IMR 0.78 [0.13; 1.43] 0.018

www.aids2014.org

• Limitations– Difficult to have a comparable control group because

of clustering– High diversity of PWID population– Lost to follow-up– Self-reports

• Positive impact of this innovative education intervention on injecting practices in terms of - A decrease in HCV at-risk practices- A reduction in local complications at the injection site

Discussion

www.aids2014.org

• Innovative aspects of this intervention- Peer-based intervention- Short intervention- Low-cost intervention

• Intervention that may be used in several contexts– Needle/syringe programs, safer injecting facilities,

outreach intervention in settings where prevalence of PWID is high

Conclusion

www.aids2014.org

• People who agreed to participate in the study• The ANRS-AERLI study group : M. Beaumont ; G.

Boyault ; P.M. Carrrieri ; M. Debrus ; A. Haas ;J.M. Legall; G. Maradan ; M. Mora ; M. Préau ; P. Roux : B. Spire ; M. Suzan-Monti

• All participating centers and their staff: M. Debrus (Paris); G. Boyault (Nevers); G. Penavayre (Pau); C. Labbé (Lille); C. Urdiales (Nîmes /Alès); J. Murat (Toulouse); C. Saramago (Grenoble); F. Tempez (Rennes); N. Perrin (Clermont Ferrand); G. Dubosc (Avignon); N. Rodier (Limoges); M. Louanges (La Roche sur Yon); F. Rigaud (Béziers); M. Alvès Da Costa (Nancy) 

• Funding

Acknowledgements

Thank you for your attention