The Economics and Financing of Harm Reduction David

The Economics and Financing of Harm Reduction David Wilson and Nicole Fraser, Global HIV/AIDS Program, World Bank David Wilson, University of New South Wales, Australia Tuesday 10 June 2013 IHRA 2013

Overview q. Why worry? q. What works and what does it cost? q. What’s the coverage? q. How much is spent on harm reduction? q. How much is needed to scale-up harm reduction? q. What’s the cost-effectiveness/return on investment?

Why worry?

Prevalence of Injecting Drug Use Mathers et al, Lancet (2008)

Prevalence of HIV among PWID Mathers et al, Lancet (2008)

HIV prevalence among PWID in Eastern and Central Asia 80% 72. 00% 60% 42. 00% 37. 20% 40% 2. 14% 0. 00% 0. 40% 0. 60% 1. 40% 1. 50% 1. 60% Slo va ki Bu a lga ria Cr oa t Ro ia ma nia Be lar u Ge s or gia Lit hu an ia Cz ec h Ky rgy sta n La tvi Ka a za kh sta Az n er ba ija n Ar me nia Ta jik i Uz stan be kis tan M old ov a Ru ssi Uk a rai ne Es to nia 0% 5. 00% 8. 15% 9. 20% 17. 00% 15. 60% 14. 70% 13. 40% 13. 00% Source: Bradley Mathers, Lancet 2008

HIV infections in PWID as share of infections in Eastern Europe and Central Asia 77% 80% 70% 62% 60% 67% 66% 69% 57% 54% 49% 50% 40% 33% 28% 30% 38% 36% 20% 10% Ru ja n ai er b Az Ky rg y zs ta n n st a kh za Ka Uz be kis t an ia an Lit hu gia Ge or e in Uk ra n st a iki va Ta j M ol do a en i Ar m ia tv La us la r Be ss ia n Fe de ra t io n 0% Source: Own calculation based on data from Euro. HIV (2007)

HIV prevalence among sex workers in Central Asia

Surging HIV epidemic among PWID in Greece

HIV, HCV and TB • PWID have higher HCV and TB rates • 10 million PWID may have HCV - surpassing HIV infection • HIV+ PWID 2 to 6 -fold higher risk of TB infection • TB risk 23 -fold higher in prisons Global State of Harm Reduction, 2012

What harm reduction interventions work and what do they cost? q. Three proven priority interventions q. NSP q. OST q. ART q. WHO, UNODC and UNAIDS - three priority interventions plus HCT, condoms, IEC, STI, HCV and TB prevention/treatment

What we know about NSP No. of injecting episodes Needles and Syringes Programs No effect Safe injection No effect centers Injecting risk Sexual HIV Cost-effective behavior risk incidence Tilson H. et al. Institute of Medicine 2007 Palmateer N. et al. Addiction 2010 Kerr T. et al. 2007 Hedrich D. et al. 2010 Yes Palmateer N. et al. Addiction 2010 Tilson H. et al. Institute of Medicine 2007 Jones L. et al. 2010 Yes Andersen MA et al. 2010 Bayoumi AM, Zarig GS 2008 Source: L. Degenhardt Lancet July 2010

What we know about NSP q. HIV prevalence in 99 cities worldwide (Mac. Donald et al, 2003) 19% per year in cities with NSP 8% in cities without NSP q. International evidence shows NSP effective (Wodak, 2008)

What we know about OST (versus compulsory detention) No. of injecting episodes Injecting risk Sexual HIV Cost effective behavior risk incidence Yes OST Tilson H. et al. (2007) Growing L. et al. (2008) Faggiano F. et al. (2009) Mattick RP et al. (2009) Tilson H. et al. 2007 Growing L. et al. No effect (2008) Tilson H. et al. (2007) Sullivan LE. Et al. (2005) Tilson H. et al. (2007) No Detention WHO 2010 Pearshouse R. et al. 2010 Open Society Institute (2010) WHO 2010 Pearshouse R. et al. (2010) OSI(2010) Constella (2008) Source: L. Degenhardt Lancet July 2010

What we know about OST (versus compulsory detention) q. Compulsory detention common especially in Asia and Eastern Europe q. Detention costly q. Minimum cost $1, 000 annually in Asia – mainly security q. Average OST cost $585 annually q. Two evaluations in progress in Malaysia and Vietnam

Effectiveness of community OST versus compulsory detention q Preliminary data from Malaysia q 95% relapse after compulsory detention q 7% relapse in community OST

What we know about OST q All RCTs of OST positive (Mattick et al, 2003) q Large observational studies show OST decreases heroin use and criminal activity (Mattcick, 1998) q OST reduces injecting and increases safe injections (Cochrane Syst. Review; Gowing, 2008; Mattick, 2009) q Amsterdam cohort study (Van den Berg, 2007) showed OST and NSP reduced HIV incidence by 66% q Recent meta-analysis (Mcarthur BMJ 2012) shows OST reduces HIV incidence by 50%

What we know about ART in PWID

What we know about combined NSP+OST+ART Modelling evidence: NSP+OST+ART combination: 5 -year impact on HIV incidence Source: Degenhardt et al, 2010

What are the cost ranges? NSP 1 q NSP costs $23– 71 /yr , but higher if all costs included q NSP costs vary by region and delivery system (pharmacies, specialist programme sites, vending machines, vehicles or outreach) 2 1 UNAIDS 2007 resource estimations; Schwartlaender et al 2011. 2 UNSW estimates, based on 10 studies identified in the 6 regions

What are the cost ranges? OST q OST cost : Methadone 80 mg: $363 - 1, 057 / yr; Buprenorphine, low dose: $1, 236 – 3, 167 /yr 1 q Few OST cost studies but consistently far higher than NSP 2 1 UNAIDS 2007 resource estimations; Schwartlaender et al 2011. 2 UNSW estimates, based on 10 studies identified in the 6 regions

What are the cost ranges? ART 1 q. ART cost: UNAIDS estimate $176 q. Estimated costs by authors $1, 000 -2, 000 per HIV+ PWID 2 1 UNAIDS 2007 resource estimations; Schwartlaender et al 2011. 2 UNSW estimates, based on 10 studies identified in the 6 regions

What is the current coverage of NSP, OST and ART in PWID?

Harm reduction data challenges q. Limited population size estimates q. Inconsistent service quality data q. Surveys miss hidden populations q. ATS increasingly used and injected but missed in surveys q. Significant but undocumented scale-down of services Sources: UNGASS country progress reports 2012; Mathers et al. , 2010; Global State of Harm Reduction, 2012

NSP coverage q 86 countries and territories implement NSPs q 3 new NSPs since 2010 – South Africa, Tanzania, Laos-PDR q. High coverage limited to Western Europe, Australia and Bangladesh (>200 NS/PWID/year) The Global State of Harm Reduction, 2012

NSP available as per policy (Black: community and prison, red: community only) Global State of Harm Reduction, 2012

Gaps in NSP coverage q. NSP coverage < 20% in all regions - globally, <2 clean needles distributed /PWID /month q. Since 2010, NSP provision scaled back in several countries in Asia (Pakistan, Nepal and Cambodia) and Eurasia (Belarus, Hungary, Kazakhstan, Lithuania and Russia) q 72 countries with PWID without NSPs (1) Global State of Harm Reduction, 2012; (2) based on Mathers et al. , 2010

Over 14 million PWID (90%) may not access NSP Source: Authors’ literature and estimations, based on Mathers et al. , 2010

OST coverage q. OST in 77 countries worldwide q 7 new countries since 2010 (Cambodia, Bangladesh, Tajikistan, Kenya, Tanzania, Macau, Kosovo) q. Primarily methadone and buprenorphine but also other formulations - slow-release morphine, codeine, heroin-assisted treatment Global State of Harm Reduction, 2012

OST available as per policy (Black: community and prison, red: community only) Global State of Harm Reduction, 2012

Gaps in OST coverage q 6– 12% of PWID access OST q. Coverage limited in much of CIS and Asia q. OST unavailable in 81 countries with PWID q. ATS use increasing – and limited ATS harm response Global State of Harm Reduction, 2012

Almost 15 million PWID (92%) may not use OST Source: Authors’ literature and estimates, using Mathers et al. , 2010

ART coverage in HIV+ PWID q. Large regional discrepancies q. Uptake highest in Western Europe (89%) and Australasia (50%) q. Elsewhere ART coverage < 5% q. Largest gaps in Eastern Europe & Central Asia (1 million) and South, East & South-East Asia (700, 000) Source: Authors literature review and estimates, using Mathers et al. 2010

About 2. 5 million HIV+ PWID (85%) may not access ART Source: Authors’ literature and estimates, using Mathers et al. 2010

What is the global coverage of harm reduction services? An estimated 10% access NSP An estimated 8% access OST Few PWID access all three priority interventions About 14% of HIV+ PWID access ART Female PWID far lower access than males Source: Authors’ literature review and estimates, using Mathers et al. 2010

How much is spent on harm reduction? q. Estimated $160 million in LMIC in 2007 (3 cents per PWID per day): 90% from international donors q. Global Fund largest HR funder (estimated $430 million 2002 -2009) > 50% to Eastern Europe and Central Asia Sources: Stimson et al 2010 (three cents report), UNAIDS 2009; UNAIDS Progress report 2012; Global State of Harm Reduction, 2012; Bridge et al 2012

Global Fund PWID investments by region (US$) Asia 17% Thailand 15% Viet Nam 14% China 30% Ukraine 10% Russ Fed 8% Kazakhstan Latin America Middle East & North Africa 166, 700, 000 Sub-Saharan Africa 366, 100, 000 10, 200, 000 24, 000 900, 000 7, 800, 000 Western Europe Eastern Europe & Central Asia Sources: Bridge 2012, summarised in Global State of Harm Reduction, 2012

How much is needed to scale up priority harm reduction interventions? q Very preliminary resource estimates based on regional estimates of current NSP /OST /ART coverage, population sizes and unit costs q Mid and high target scenarios costed NSP coverage (%) Needles / PWID /year OST uptake (%) ART uptake of HIV+ PWID (%) Current estimated level 10 22 8 14 Scenarios: Mid target 20 100 20 25 High target 60 200 40 75

How much needed to scale up priority harm reduction interventions – preliminary estimates ECA SSE LAC MNA SSA WEST 11. 7% 19. 1 M 11. 45 M 11. 5% 26. 84 M 153. 60 M 2% 2% 17% 8. 33 M 26. 84 M 1. 35 M 4. 35 M <1% 5. 3 M 15. 98 M <1% 715 M 1. 47 B 5. 9% 360 M 872 M <1% 427. 63 M 857. 41 M 1% <1% 27. 8% 23. 17 M 344. 01 M ---47. 57 M 689. 75 M 954. 74 M 1. 1% 1. 16 B 3. 59 B 3. 6% 856 M 2. 88 B 1% <1% 690. 29 M 34. 09 M 518. 09 M 2. 13 B 102. 28 M 1. 58 B NSP Coverage 20% 80% OST Coverage 20% 40% ART Coverage 25% 75% 16. 63 M 238. 30 M 78. 5% -------

Summary: Estimated annual cost of scaleup of NSP, OST and ART for PWIDs Mid target 20% NSP coverage 20% OST coverage 25% ART coverage High target 60% NSP coverage 40% OST coverage 75% ART coverage South, East & South East Asia 527 M 1, 49 B Latin America & Caribbean 625 M 1, 47 B Middle East & North Africa 26 M 55 M W- Europe, N- America & Australasia 17 M 1, 19 B Eastern Europe & Central Asia 1. 04 B 2, 51 B Sub-Saharan Africa 414 M 901 M 2, 65 B 7, 62 B Total per year 1: Mathers et al, Lancet (2010) 2: Scale-up calculations by UNSW

Annual scale-up costs by region and intervention q. Costs dominated by Eastern Europe and Central Asia 1: Mathers et al, Lancet (2010) 2: Scale-up calculations by UNSW

Cost-effectiveness and relative return on investment ranges by region () number of studies in literature Western Europe, North America & Australasia CE 1: $402 -$34, 278 (9) ROI 2: $1. 1 -$5. 5 (3) Eastern Europe & Central Asia CE 1: $97 -$564 (3) ROI 2: $1. 4 (1) The Middle East & North Africa CE 1: $1, 456 -$2, 952 (1) Latin America & The Caribbean South, East & South East Asia CE 1: $71 -$2, 800 (7) ROI 2: $1. 2 -$8. 0 (4) Sub-Saharan Africa 1: Cost per HIV infection averted 2: Total future return per $1 invested (3% discount rate)

Harm reduction cost-effectiveness q. Harm reduction cost-effective in all regions, with costs per HIV infection averted from $100 -$1, 000 q. Harm reduction returns positive, with total future returns per $ from $1. 1 – $8. 0 (3% discount rate) Also q. Unit costs fall as interventions scaled-up q. Combined, integrated interventions reduce overheads q. Intervention synergies increase effectiveness

Australia’s example: Economic benefits of a supportive legal and policy environment q. Australia invested A$243 million in NSP q. Prevented estimated 32, 050 HIV infections and 96, 667 HCV cases q A$1. 28 billion saved in direct healthcare costs q. Including patient/client costs and productivity gains and losses, net present value of NSPs is $5. 85 billion ROI - A$27 per A$1 invested Source: Return on Investment 2, Department of Health and Ageing, Australian Government

CONCLUSION q. Inaction costly q. NOT the equivalent of nothing happening q. Hard to reverse epidemic once established q. Whereas harm reduction is q. Effective - in terms of HIV cases averted q. Cost-effective - in terms of healthy years gained and costs q. Social benefits exceed treatment costs q. And benefits the whole population q. Substance abuse treatment can benefit more non- drug users than drug users q. Global best buy