The Case for Routine, Voluntary HIV Testing : The Case for Routine, Voluntary HIV Testing Rochelle P. Walensky, MD, MPH
Assistant Professor of Medicine
Harvard Medical School
Massachusetts General Hospital
Brigham and Women’s Hospital
Motivation : Motivation Unidentified infection:
1,000,000 people in the US living with HIV
300,000 undiagnosed
Poor follow-up:
25% of those testing HIV+ fail to return for results
Inadequate linkage to care:
only 2/3 of HIV-infected persons receive appropriate care
Too little, too late:
40% learn there are HIV-infected in the year prior to AIDS
5% learn it within the month prior to death.
CDC Guidelines : CDC Guidelines 2001 CDC guidelines recommend HIV counseling, testing and referral (CTR):
Routine, voluntary HIV CTR for all patients in hospitals with >1% HIV prevalence
Routine CTR in settings serving populations at increased HIV risk
Targeted HIV CTR in settings of <1% prevalence and low risk
The Problem. . . : The Problem. . . No definition of “HIV risk”
No guidance on targeting strategy
No analysis to support the 1% threshold
Minimal mention of scarce resources
The Bottom Line... : The Bottom Line... Expanded HIV CTR services are feasible and can have high yield.
HIV CTR can be justified at lower prevalences
Routine HIV CTR is a highly cost-effective use of HIV care dollars in the United States.
Failure to identify and link HIV infection has led to hundreds of thousands of years of life lost
The Inpatient Experience : The Inpatient Experience RP Walensky et al, Arch Int Med, 2002
The Inpatient Testing Experience Boston Medical Center 4/99-6/00 : The Inpatient Testing Experience Boston Medical Center 4/99-6/00 Patients admitted to the medical service were offered HIV counseling and testing
MA DPH funded on-site counselors
473 (6.4%) of the 7,356 medical admissions were voluntarily tested for HIV
The CTR program was compared to a period of historical control (1/98-3/99)
RP Walensky et al, Arch Int Med, 2002
Results Inpatient Testing : Results Inpatient Testing RP Walensky et al, Arch Int Med, 2002
Results: Monthly Positive Tests : Results: Monthly Positive Tests
The Atlanta Experience : The Atlanta Experience Del Rio, MMWR, 2001
Atlanta: Urgent Care Center 3/00-9/00 : Atlanta: Urgent Care Center 3/00-9/00 All urgent care patients 18-65 y/o were offered voluntary HIV CTR over a 24-week period.
Compared to historical control over same period one year prior.
Del Rio, MMWR, 2001
Atlanta Urgent Care : Atlanta Urgent Care
The Outpatient Experience : The Outpatient Experience Walensky et al, MMWR, 2004
Walensky et al, AJPH, 2005
Think HIV Objectives : Think HIV Objectives 1) Establish “Think HIV” in 4 Massachusetts urgent care centers
2) Identify and refer to care patients with undiagnosed HIV infection
3) Determine the seroprevalence of undiagnosed infection Walensky et al., MMWR, 2004
Walensky et al., AJPH, 2005
Results : Results January - September 2002
Think HIV offered >7,000 patients HIV testing
2,444 (37%) accepted testing
33,608 HIV tests statewide in same time period
HIV Prevalence ________________________________________________________ : HIV Prevalence ________________________________________________________ ______________________________________ p=0.016
HIV Prevalence: Previous HIV Testing : HIV Prevalence: Previous HIV Testing Think HIV
Time of previous test
<1 year prior 2.2%
>1 year prior 1.9%
Never tested before 1.9%
Referral to Care : Referral to Care 42/48 (88%) infected patients returned for test results
All 42 who returned for results linked to care
MA Dept of Public Health: Program Costs : MA Dept of Public Health: Program Costs
Cost for 9 months = $232,900
Cost per case identified = $4,850
Slide20 : “The price tag probably makes the program too expensive for most states. . . I don't think it will work in the UCC at a suburban mall”
Is routine HIV screening cost-effective? If so, at what HIV prevalence?
Cost-effectiveness of Routine HIV Testing : Cost-effectiveness of Routine HIV Testing
Objectives : Objectives To evaluate the clinical impact and cost-effectiveness of the current CDC HIV CTR guidelines in the outpatient/inpatient setting.
Methods : Methods Computer simulation model of HIV infection
Model includes HIV screening and symptom detection
Data from MACS, clinical trials, ACSUS
Prophylaxis for PCP, toxo, MAC, CMV, fungal infections
Compares clinical outcomes, costs, cost-effectiveness
Slide24 :
Methods: The Model
HIV diagnosed or undiagnosed
HIV RNA and CD4 cell count
Opportunistic infections and prophylaxis
Antiretroviral therapy
Acute Clinical
Event Death Chronic HIV
Infection Primary HIV
Infection
The CEPAC Model : The CEPAC Model Bacterial
pneumonia CMV Death QALYs: 8.47
Total costs: $142,400
Methods Overview : Methods Overview Screening Module
New HIV screening program Detection via background HIV Screening HIV Therapy
ART and OI prophylaxis Undiagnosed HIV-infected patient
Detection via development of an OI
Cost-effectiveness of Routine HIV Testing : Cost-effectiveness of Routine HIV Testing Outpatients AD Paltiel et al, NEJM, 2005
Three Target Populations : Three Target Populations
Undiagnosed Monthly
HIV HIV Prevalence Incidence
(%) (%)
High-Risk 3.0 0.1
CDC Threshold 1.0 0.01
US Overall 0.1 0.0012
Results Hi-Risk Population Cost-effectiveness High Risk : Results Hi-Risk Population Cost-effectiveness High Risk Population HIV Cost Cost-effectiveness
(QALMS) (QALMS) ($) ($/QALY)
Std Practice 250.89 219.84 $32,700 ---
Single EIA 251.26 220.74 $33,800 $36,000
EIA q 5 yrs 252.11 222.78 $37,300 $50,000
EIA q 3 yrs 252.40 223.46 $38,900 $63,000
Annual EIA 252.75 224.29 $41,700 $100,000
Paltiel et al, NEJM, 2005
Infections Averted High Risk Population - Drug User : Infections Averted High Risk Population - Drug User
Infections
Averted
Single EIA 300
EIA q 5 yrs 2,700
EIA q 3 yrs 3,600
Annual EIA 5,100 Paltiel et al, NEJM, 2005 Anticipate 44,000-60,000 new infections/100,000
Results Outpatient Cost-effectiveness : Results Outpatient Cost-effectiveness In a high risk population, HIV testing every five years had a cost-effectiveness ratio of $50,000/QALY gained
At the CDC threshold, HIV testing every five years had a cost-effectiveness ratio of $71,000/QALY gained
Even in the “US Overall Population” a one-time HIV test may be cost-effective: $113,000/QALY gained
Slide32 : “Failure to implement widespread
routine screening for HIV infection
represents a critical disservice
to patients who are currently infected,
those at risk for infection,
and the future health of the nation.” Bozzette, NEJM 2005
Cost-effectiveness of Routine HIV Testing : Cost-effectiveness of Routine HIV Testing Inpatients RP Walensky et al, AJM, 2005
Results Inpatient Cost-effectiveness : Results Inpatient Cost-effectiveness
Prevalence Population HIV+ Cost Cost-effectiveness
(QALMS) (QALMS) ($) ($/QALY)
1.0%
No Testing 204.10 72.30 $1,200 ---
Testing 204.20 81.77 $1,500 $38,600
0.1%
No Testing 205.30 72.30 $120 ---
Testing 205.31 81.77 $160 $50,000 RP Walensky et al, Am J Med, 2005
Slide35 : Unidentified
10% Background
Testing
37% Opportunistic
Infection
53% Unidentified
7% Background
Testing
25% Opportunistic
Infection
36% New Screening Program
32% Without HIV CTR Program With HIV CTR Program Results
Mechanisms of HIV Detection
Results: Testing Costs : Results: Testing Costs
Results: Testing Costs : Results: Testing Costs
The HIV Testing Pathway : HIV testing is a pathway of sequential processes:
Failure in any one process results in overall failure
The HIV Testing Pathway
HIV Testing Pathway : HIV Testing Pathway p(offer/accept) p(return/link) X Index of Participation (IOP)
Base case: Index of Participation (IOP) : Base case: Index of Participation (IOP) Offer/Accept
Return/Link to care
IOP
37%
88%
88%
37%
33% 33% Walensky et al, Med Dec Making, 2005
Results: Index of Participation : Results: Index of Participation 0.04, $43,400/QALY 0.33,
$38,600/QALY
Results: Index of Participation HIV Prevalence 1.0% : Results: Index of Participation HIV Prevalence 1.0% Base Case Walensky et al, MDM 2005
Index of Participation (IOP) : We examined three alternative ways to achieve an index of participation of 0.16:
All allow identical number of people through the testing program, at what cost?
Index of Participation (IOP)
Results: Index of Participation HIV Prevalence 0.1% : Results: Index of Participation HIV Prevalence 0.1% CE Ratio ($/QALY) IOP = p(offer/accept) x p(return/link)
Offer/accept ≥ Return/link
Offer/accept < Return/link
HIV Prevalence 0.1% p(offer/accept) = p(return/link) = 0.4 : HIV Prevalence 0.1% p(offer/accept) = p(return/link) = 0.4
Slide46 : Fungal proph. Fluconazole $123,700 Freedberg JAMA 1998
CMV proph. Valganciclovir $893,600 Paltiel Clin Inf Dis 2001
C-E Ratio
Intervention Agent ($/QALY)* Reference
PCP/Toxo proph. TMP-SMX $2,800 Freedberg JAMA 1998
ART AZT/3TC/EFV $11,700 Freedberg NEJM 2001
GART 2nd line --- $20,200 Weinstein Ann Int Med 2001
Inpt HIV screening --- $38,600 Current Analysis
MAC proph. Azithromycin $43,300 Freedberg JAMA 1998
HIV screening q5y --- $50,000 Current Analysis
high risk patients Cost-effectiveness Ratios for HIV Care *All costs adjusted to 2001 US dollars
Cost-effectiveness Ratios for Other Screening Programs : Cost-effectiveness Ratios for Other Screening Programs C-E ratio
Screening Program ($/QALY)* Reference
HIV screening inpatients $38,600 Current Analysis
HIV screening every 5 years
high risk patients $50,000 Current Analysis
Breast cancer screening Salzmann
Annual mammogram, 50–69 y/o $57,500 Ann Intern Med 1997
Colon cancer
FOBT + SIG q5y, adults 50–85 y/o $57,700 Frazier JAMA 2000
Diabetes Mellitus, Type 2
fasting plasma glucose, adults >25 y/o $70,000 CDC JAMA 1998 *all costs adjusted to 2001 US dollars
Conclusions : Conclusions Routine HIV testing programs in the inpatient and outpatient setting are feasible and can have a high yield of HIV case identification (2.0-6.8%).
C-E models show that screening every 5 years in high risk populations and even one-time HIV screening in the general US population is cost-effective.
C-E models show that inpatient HIV screening is highly cost-effective at an undiagnosed HIV prevalence of 1.0% (likely 0.1%).
Conclusions : Conclusions Investments in linkage to HIV care once patients are identified through screening programs should be paramount.
Identification and treatment of HIV infection can lead to per person survival benefits of over 13 years.
Expansion of routine HIV CTR programs nationally should be a public health priority.
Acknowledgements : Acknowledgements Massachusetts Department of Public Health
George E. Barton
Laureen Malatesta, PA
Jean F. McGuire, PhD
Catherine A. O’Connor, CNS
Site physicians, administrators, counselors, and patients
Massachusetts General Hospital
Kenneth A. Freedberg, MD, MSc
Boston University School of Public Health
Elena Losina, PhD
Boston University Medical Center
Paul Skolnik, MD
Jon Hall
CEPAC Investigators : CEPAC Investigators Harvard Medical School
Wendy Aaronson, MPH
Nomita Divi, MSc
Kenneth Freedberg, MD, MSc
April Kimmel, MSc
Elena Losina, PhD
Zhigang Lu, MD
Lauren Mercincavage
Sara Sadownik
Paul Sax, MD
Heather Smith
Rochelle Walensky, MD, MPH
Lindsey Wolf
Hong Zhang, SM
Hui Zheng, PhD Harvard SPH
Sue Goldie, MD, MPH
George Seage, DSc, MPH
Milton Weinstein, PhD
Cornell
Bruce Schackman, PhD, MBA
Yale
A. David Paltiel, PhD
Lille, France
Yazdan Yazdanpanah, MD, PhD
Acknowledgements : Acknowledgements National Institute of Mental Health
National Institute of Allergy & Infectious Diseases
National Institute on Drug Abuse
Centers for Disease Control and Prevention
Massachusetts Department of Public Health