Funding Support n National Institute on Drug Abuse

Funding Support n National Institute on Drug Abuse, Grant No. 1 R 01 DA 14703; Anonymous donor in NYC – Research and evaluation n Ford Foundation (Beijing and Hanoi), Global Fund, National and Provincial budgets – Interventions Address for Correspondence: Theodore M. Hammett, Ph. D. Abt Associates Inc. 55 Wheeler Street Cambridge, MA 02138 USA ted_hammett@abtassoc. com World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Background n Cross-Border movement n Drug Trans-shipment Routes Heroin Use Heroin Injection HIV Transmission Increasing HIV Prevalence Cross-Border Transmission (Molecular epidemiology: Beyrer et al. , AIDS, 2000) n Small-scale movement across long, porous border n Trade, migratory employment, “floating population” n Intermarriage, ethnic minority groups on both sides n Emerging evidence of sexual transmission and epidemic generalization. World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Geographic Setting of Cross-Border Project China Ha Giang Guigang Area of Detail Hanoi Vietnam China Ning Ming City Key: Puzhai Shilang Aidian Thanh Tongmian Dong Dang Cao. Loc Town Loc Binh Hop Thanh Large Project Site Small Border Site Lang Son City Vietnam PDI Site World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Near the border at Aidian, China World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Border scene at Aidian, China World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Border traffic, Vietnam side World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Interventions n Peer-based education of IDUs, sex workers, and sexual partners of IDUs n Distribution of needles/syringes and condoms – directly and through redemption of pharmacy vouchers n Collection/safe disposal of used needles/syringes n Community education n Support for drug use cessation Full implementation of IDU interventions began in 2002; Women at Risk interventions began in 2006. § Will present mainly on results of IDU interventions. World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Pharmacy/clinic in Tongmian, China World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Needle/Syringe Distribution by a Village Doctor, Near Ning Ming City, China World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Discarded wrappers and needles/syringes, Lang Son City World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Basket for collection of used needles/syringes, Loc Binh, Vietnam World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Cross-Border collaboration § Joint meetings every 6 months-1 year to share experiences, strategies, and evaluation results – host country alternates. § Bonds formed between Vietnamese and Chinese partners. § Peer educators from Tan Thanh, Vietnam site cross border every week to work with Vietnamese sex workers in Puzhai (Pingxiang), China in collaboration with Chinese officials and agencies. § Study tours from China to Vietnam. § Policy advocacy in both countries. World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Evaluation n Methods n Pre- and multiple-post intervention follow-up — Cross-sectional surveys of IDUs —Baseline, 6 -, 12 -, 18 -, 24 -, and 36 -month completed in China and Vietnam — HIV testing (with counseling) — HIV incidence estimation (repeat participants/new injectors) n KAP surveys in community (HIV knowledge, knowledge of/attitudes toward project) — same intervals as IDU surveys n Process data — Peer educator logs (contacts, collection of used needles/syringes) — Pharmacy/clinic logs (vouchers redeemed) World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Process indicators § Interventions reach ~50 -87% of IDUs, depending on site and survey wave. § § ~20, 000 – 25, 000 new needles/syringes provided per month. § Police have not directly interfered with interventions, but periodic crackdowns on IDUs may reduce participation – monthly average number of needles/syringes provided has gone down in China sites. § Public health scale intervention: Providing ~80 new needles/syringes per IDU per year (compares favorably to NYC where prevalence/incidence substantially reduced). ~4, 000 – 6, 000 condoms distributed per month. ~15, 000 – 20, 000 used needles/syringes collected per month. IDUs in Vietnam generally prefer pharmacy vouchers (except in sites with few participating pharmacies), while in China they prefer direct distribution. World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Drug-Related Risk Behaviors, Past 6 Months: Lang Son Province, Vietnam World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Drug-Related Risk Behaviors, Past 6 Months: Ning Ming County, China World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Other risk factors n Contrary to popular mythology, many IDUs are sexually active. n More IDUs in Ning Ming report patronizing sex workers and having unprotected sex with casual partners. n Self-reported cross-border drug purchase and injection are surprisingly infrequent, but pose very high risk for IDUs crossing from China to Vietnam (data not shown). World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

HIV Prevalence by Site World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Comments on HIV Prevalence Patterns n HIV Prevalence among IDUs is stable or declining in Lang Son and Ning Ming over 36 months, unlike pattern in other nearby sites (Yunnan, China and Quang Ninh and Haiphong, Vietnam). n In Lang Son and Ning Ming, there is a gradient of HIV prevalence among the sites running in descending order from sites farthest from the border in Vietnam to the sites the farthest from the border in China (data not shown). n There is a similar gradient for self-reported length of injection history by distance from the border, i. e. , IDUs in Lang Son tend to have been injecting longer than those in Ning Ming (data not shown). n These gradients are consistent with the spread of heroin injection and HIV along drug transshipment routes, but more slowly than might have been expected. n Stability of gradient pattern and actual prevalence rates, together with declining incidence, suggests control of cross-border and intra-country transmission of HIV. World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Estimating Incidence in New Injectors § § New injectors: those reported having injected <3 years; § Assume all HIV seropositive new injectors became infected halfway between time of first injection and time of interview; § Incident cases of new infections equal number of HIV seropositive new injectors in each survey; § Time at risk equals sum of times from first injection to interview for all HIV seronegatives plus 1/2 time from first injection to interview for HIV seropositives. Assume all new injectors HIV seronegative when they start to inject; World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

New Injectors: Lang Son Province %HIV+ (Prev. ) Mean Yrs Injecting Est. HIV Incidence baseline 31 % 1. 57 yr. 20 %/yr. 6 mo. 27 % 1. 31 yr. 21 %/yr. 12 mo. 21 % 1. 29 yr. 17 %/yr. 18 mo. 19 % 1. 37 yr. 14 %/yr. 24 mo. 13 % 1. 41 yr. 9 %/yr. 36 mo. 6% 1. 29 yr. 4 %/yr. World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

New Injectors: Ning Ming County %HIV+ (Prev. ) Mean Yrs Injecting Est. HIV Incidence baseline 16 % 1. 50 yr. 11 %/yr. 6 mo. 22 % 1. 22 yr. 18 %/yr. 12 mo. 11 % 1. 25 yr. 9 %/yr. 18 mo. 9% 1. 30 yr. 7 %/yr. 24 mo. 11 % 1. 36 yr. 8 %/yr. 36 mo. 0% 1. 29 yr. 0 %/yr. World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

HIV Incidence Among New Injectors, by Site World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata:

Knowledge and Attitudes in the Community: Ning Ming County, China Baseline 6 Month 12 Month 18 Month 24 Month 36 Month (n=321) (n=343) (n=346) (n=312) (n=317) (n=298) Transmitted through needle sharing 93% 95% 89% 93% 84% 93% 0. 04 Not transmitted by touching or hugging an HIV+ person 75% 84% 86% 89% 90% 93% <0. 001 Not transmitted through eating with HIV+ person 55% 67% 71% 67% 76% 81% <0. 001 Will reduce discarded needles/syringes 57% 71% 58% 63% 74% 68% 0. 003 Will reduce spread of HIV 75% 79% 76% 70% 73% 76% 0. 238 Will not increase drug use 62% 61% 68% 63% 54% 0. 003 “Correct” Answers World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata: p-value

Knowledge and Attitudes in the Community: Lang Son Province, Vietnam Baseline 6 Month 12 Month 18 Month 24 Month 36 Month (n=250) (n=250) Transmitted through needle sharing 100% 97% 98% 100% 97% 0. 771 Not transmitted by touching or hugging an HIV+ person 90% 86% 87% 93% 94% 86% 0. 484 Not transmitted through eating with HIV+ person 79% 87% 85% 96% 89% 88% <0. 001 Will reduce discarded needles/syringes 97% 98% 99% 100% 96% 0. 71 Will reduce spread of HIV 97% 98% 99% 96% 0. 505 Will not increase drug use 71% 67% 69% 82% 84% 77% 0. 002 “Correct” Answers World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata: p-value

Conclusions n This is an innovative project, offering important opportunities to control a cross-border HIV epidemic and increase cross-border collaboration. n The interventions are reaching substantial proportions of IDUs and providing public health scale distribution of new needles/syringes. n Drug-related HIV risk behaviors are being reduced. n HIV prevalence among IDUs is stable or declining. n Cross-border prevalence differential has been maintained; HIV prevalence has not risen to levels found in other nearby places. n HIV transmission has not been halted, but incidence among new injectors is coming down. n Evaluation results have had positive effects on policy development, in terms of support for harm reduction strategies. World Bank Inter-Country Consultation, Kolkata: April 11, 2007 Kolkata: