The Impact of HIV/AIDS on Medical Schemes in South Africa Roseanne da Silva, University of the Witwatersrand Lara Wayburne, Deloitte IAAHS 2007 16 May 2007 CTICC www.iaahs2007.com
Agenda HIV modelling Medical scheme population Prescribed Minimum Benefits Models Family unit algorithm Sample population Results Discussion
The four stages of HIV The progression of the infection is measured in terms of the CD4 count, and the viral load (the amount of virus in the blood). The CD4 count generally determines the rate of deterioration of the patients resistance Clinical Indicator Average Duration Symptom Stage1 CD4>500 4 to 6 years Asymptomatic Stage2 CD4 between 350 and 500 2 to 3 years Some opportunistic infections Stage3 CD4 between 200 and 350 2 to 3 years Opportunistic infections Stage4 CD4<200 6 months to 1 year AIDS
Modelling HIV/AIDS UNAIDS ASSA2003 Suite: Census 2001 for national and provincial populations and prior censuses for fertility rates. Antenatal prevalence data for 2003 Mortality data for 2002/3 Data from the 2002 Nelson Mandela/Human Sciences Research Council household survey The Reproductive Health Research Unit survey of sexual behaviour and prevalence of youth in South Africa.
ASSA 2003 The ASSA2003 version includes five interventions: improved treatment for sexually transmitted infections (STIs); information and education campaigns (IEC); voluntary counselling and testing (VCT); mother-to-child transmission prevention (MTCTP); anti-retroviral treatment (ART).
35 30 25 20 15 10 5 0 ANC Prevalence HIV Prevalence trends among antenatal clinic attendees in South Africa, 1990-2005 1990 1991 1992 1993 1994 1995 1996 1997 1998 1999 2000 2001 2002 2003 2004 2005 Year Actual ASSA2003 Prevalence (%)
Some Results ANC data The prevalence measured in 2005 was 30.2% (95% confidence interval 29.1% to 31.2%) Population extrapolation estimate of 5.54m HSRC household survey 10.8% or 4.8m (4.2-5.3) UNAIDS 5.5m 18.8% of adults
UNAIDS analysis of sub-saharan epidemic People living with HIV New infections AIDS deaths Women living with HIV Children living with HIV People in need of ART Proportion of Global Epidemic in sub- Saharan Africa 64% 66% 71% 75% 90% 72%
Population Forecasts (ASSA2003) Millions 60 Total population 58 AIDS No AIDS 56 54 52 50 48 46 44 2000 2002 2004 2006 2008 2010 2012 2014 2016 2018 2020 Year
Update from the ASSA2003 Model for 2006 AIDS deaths per day New HIV infections per day People living with HIV AIDS deaths New HIV infections Adults on ART Children on ART 947 1 443 5 372 476 345 640 526 771 154 832 20 050 Source: AIDS Analysis Africa Online
Population prevalence by category 2005 2010 2015 2020 Total Population 11.0% 11.8% 12.0% 11.9% Adult Population 18.8% 19.7% 19.6% 19.3% Labour force 21.5% 22.2% 21.4% 20.2%
Medical Scheme Population Select population Approx 16% of South African population Twin peaks age profile Socio-economic status Racial profile
Evolution of the Prescribed Minimum Benefits Legislated set of minimum benefits that each registered medical scheme is compelled to offer as part of each benefit option Stipulated in Annexure A of the Regulations to the Medical Schemes Act (Act 131 of 1998) January 2000: VCT, Co-trimoxazole, screening and preventative therapy for TB, STI diagnosis and treatment, palliative care, treatment of opportunistic infections, MTCTP, PEP following occupation exposure or sexual assault January 2005: medical management and medication, including the provision of anti-retroviral therapy to the extent provided for in the national guidelines Link between public sector and private sector provision
Age profile of the Medical Scheme Population (Sources: ASSA 2003 Lite Model, REF grid) REF industry age profile and estimated HIV prevalence REF industry age pro file per 1000 beneficiaries HIV prevalence by age 100 2.5% per 1,000 beneficiaries 80 60 40 20 2.0% 1.5% 1.0% 0.5% HIV prevalence 0 Under 1 5-9 15-19 25-29 35-39 45-49 55-59 65-69 75-79 85+ Age band 0.0%
Racial split of the Medical Scheme Population (Source: McLeod et al, 2004) 100% 90% 80% 70% 60% 50% 40% 30% 20% 10% 0% White Other Coloured African/Black Indian/Asian Medical Schemes Potential SHI Public Sector Total Population
Racial adjustments Adjustment to national racial profile to reflect medical scheme profile reduction in HIV prevalence rates Estimated national HIV prevalence, adjusted for racial composition 20% 18% 16% 14% 12% 10% 8% 6% 4% 2% 0% 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 HIV prevalence Unadjusted, to tal po pulatio n Unadjusted, adults (20-64) Adjusted, total population A djusted, adults (20-64)
Modelling ASSA2003 Select Model Model staging system: HIV stages 1-4, ART stages 1-4 (Source: Rosenberg et al, 2000) Construction of HIV prevalence scenarios: various multiplicative adjustments to HIV prevalence rates for new entrants and HIV incidence rates Multiplicative adjustments To HIV prevalence rates for new entrants To HIV incidence rates Skill level 1 2 3 4 5 1 2 3 4 5 Scenario 1 100% 100% 100% 100% 100% 100% 100% 100% 100% 100% Scenario 2 124% 90% 30% 18% 18% 124% 90% 30% 18% 18% Scenario 3 100% 100% 100% 100% 100% 50% 50% 50% 50% 50% Scenario 4 100% 80% 20% 10% 10% 100% 80% 20% 10% 10%
Family Unit Model and Assumptions Referred to studies on HIV sero-discordant couples Principal member (P) with spouse (S): Assume each HIV+ P has a corresponding HIV- S Transmission dynamics: Male female: 44% probability of male infecting female, females twice as likely to contract HIV from their HIV+ male partner (Sources: Quinn et al, 2000, Hugonnet et al, 2002, Carpenter, 1999) Female male: 22% probability of female infecting male (Source: Quinn et al, 2000) Ignore infection of S other than from P Applied over 4 year periods starting from [ ] P without S: infection independent of family structure
Family Unit Model and Assumptions New born child dependants (C) infection through MTCT: HIV+ P or S, 75% know HIV status: 2% probability of MTCT (assuming ART administered during pregnancy, labour and delivery, and elective caesarian section if high viral load, ART administered to child.) (Source: CDC, 2006) HIV+ P or S, 25% not know HIV status: 25% probability of MTCT (Source: CDC, 2006) Cover age of 14: infection independent of family Modelled using National model Adult dependants (A): infection independent of family Modelled separately using Select model Aggregate HIV prevalence rates for each beneficiary group
Sample Population Demographic data obtained from 3 open medical schemes in 2004 2.202 million lives, 31.3% of beneficiaries Data at beneficiary level: member number, date of birth, gender, benefit option, province Stable population assumed Calibration
14% 12% 10% 8% 6% 4% 2% 0% Age Distribution of Sample Population 0-4 5-9 10-14 15-19 20-24 25-29 30-34 35-39 40-44 45-59 50-54 55-59 60-64 65-69 70-74 75+ Scheme 1 Scheme 2 Scheme 3 Aggregate
Socio-Economic Classification Grade Description Patterson Peromnes TASK Castellion bands grades grades grades 1 Employees living in single sex hostels and construction camps 2 Semi-skilled employees A, B 12 to 19 1 to 8 1 to 7 3 Skilled employees C 8 to 11 9 to 13 8, 9 4 Middle management D 5 to 7 14 to 18 10 to 12 5 Senior management E, F 1++ to 4 19 to 28 13, 14 Benefit options are proxies for job categories Benefit option movement ignored
12% 10% 8% 6% 4% 2% 0% Results HIV prevalence by Scenario Scenario 1 Scenario 2 Scenario 3 Scenario 4 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009
Results HIV Prevalence Scenario 4: 20%, 80% ART Uptake 4.5% 4.0% 3.5% 3.0% 2.5% 2.0% HIV stage 1 HIV stage 2 HIV stage 3 HIV stage 4 ART stage 1 ART stage 2 ART stage 3 ART stage 4 1.5% 1.0% 0.5% 0.0% 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009
Results HIV Prevalence Scenario 4: 60%, 80% ART Uptake 8.0% 7.0% 6.0% 5.0% 4.0% 3.0% HIV stage 1 HIV stage 2 HIV stage 3 HIV stage 4 ART stage 1 ART stage 2 ART stage 3 ART stage 4 2.0% 1.0% 0.0% 1985 1987 1989 1991 1993 1995 1997 1999 2001 2003 2005 2007 2009
Comparison of Modelling Results Johnson and Dorrington (2003, appendix), modelled 3 scenarios: A: no change in profile by age, gender, skill level or race B: change in skill profile C: change in race profile Peak of 7,5% expected in 2008 (A) B, C not differ 9% 8% 7% 6% 5% 4% 3% 2% 1% substantially from A 0% Scenario A Scenario C Scenario B 1995 1997 1999 2001 2003 2005 2007 2009 2011 2013 2015
Financial Impact Cost structure based on prices in private health care sector in 2005 Costed elements: ART Other meds (immune boosters pre-art and with ART) Pathology testing MTCTP per birth Outpatient costs (doctor consultations) Hospitalisation (in-patient and out-patient) Costs applied to each HIV and ART stage MTCTP costs calculated on a per birth basis
Financial Assumptions Use of brand drugs and generics Inflation assumptions Salvage regimen (ART stage 4) Managed cost scenario only
Treatment Protocols Utilisation per infected person per annum ART Immune modulators Pathology testing Doctor consultations Hospitalisation HIV stage 1 HIV stage 2 0 0 Monthly Monthly Biannually Biannually Biannually Biannually Cost per HIVpositive or AIDSsick beneficiary per annum HIV stage 3 0 Monthly Biannually Biannually Cost highest in: HIV stage 4 0 Monthly Biannually Biannually -ART stage 4 -Children AIDS ART stage 1 Monthly Monthly Biannually Biannually -HIV stage 4 -ART stage 3 ART stage 2 Monthly Monthly Biannually Biannually ART stage 3 Monthly Monthly Quarterly Quarterly ART stage 4 Monthly Monthly Quarterly Quarterly Children pre-aids 0 Monthly Biannually Biannually Children AIDS Monthly Monthly Quarterly Quarterly
Estimated Cost per HIV and ART stage per annum Chidlren A IDS Children pre-a IDS ART stage 4 ART stage 3 ART stage 2 ART stage 1 HIV stage 4 HIV stage 3 HIV stage 2 HIV stage 1 R 0 R 10,000 R 20,000 R 30,000 R 40,000 R 50,000 R 60,000
Numbers infected: Scenario 2 & Scenario 4 160 000 140 000 120 000 Numbers infected 100 000 80 000 60 000 40 000 20 000 - Scenario 2: 20%, 80% Scenario 2: 60%, 80% Scenario 4: 20%, 80% Scenario 4: 60%, 80% 2005 2006 2007 2008 2009 2010
Results: Impact of ART initiation assumption: cost pbpm Cost per beneficiary per month R 30 R 25 R 20 R 15 R 10 R 5 Scenario 2: 20%, 80% Scenario 2: 60%, 80% Scenario 4: 20%, 80% Scenario 4: 60%, 80% R 0 2005 2006 2007 2008 2009 2010
Cost components 2005: Scenario 4-60%, 80% Relative costs fairly similar across scenarios 20% assumption leads to higher proportion of costs for MTCT and children Children 1% MTCTP 14% Adults 85%
Cost of HIV as % of gross and average contributions: Scenario 4 2.5% % of gross or average contributions 2.0% 1.5% 1.0% 0.5% Gross 20%, 80% Gross 60%, 80% Average 20%, 80% Average 60%, 80% 0.0% 2005 2006 2007 2008 2009 2010 Scenario 1: 4% - 9% Scenario 2: 1.5% - 3% Scenario 3: 3% - 6.5%
Discussion Further research Change in medical scheme population Take-up and compliance rates Unmanaged costs REF implications Low Income Medical Scheme (LIMS) Need for risk-pooling of SMMEs