A. David Paltiel, PhD; Rochelle P. Walensky, MD, MPH; Bruce R. Schackman, PhD; George R. Seage III, ScD, MPH; Lauren M. Mercincavage, AB; Milton C. Weinstein, PhD; Kenneth A. Freedberg, MD, MSc
Acknowledgments: The authors thank Douglas K. Owens, MD, and several anonymous reviewers for their comments on various drafts of the manuscript. They also thank their colleagues on the Cost-Effectiveness of Preventing AIDS Complications (CEPAC) project team for their valuable guidance: April Kimmel, MSc; Elena Losina, PhD; Alethea McCormick, ScD; Paul Sax, MD; Heather E. Hsu; and Hong Zhang, SM.
Grant Support: By the National Institute of Mental Health (R01MH65869), the National Institute of Allergy and Infectious Diseases (K23AI01794, K24AI062476, R01AI42006, P30AI42851), the National Institute on Drug Abuse (R01DA015612, K01DA0717179), the Doris Duke Charitable Foundation (Clinical Scientist Development Award), and the Centers for Disease Control and Prevention (S1396-20/21).
Potential Financial Conflicts of Interest: None disclosed.
Requests for Single Reprints: A. David Paltiel, PhD, Department of Epidemiology and Public Health, Yale School of Medicine, 60 College Street, New Haven, CT 06520-8034; e-mail, email@example.com.
Current Author Addresses: Dr. Paltiel: Department of Epidemiology and Public Health, Yale School of Medicine, 60 College Street, New Haven, CT 06520-8034.
Drs. Walensky and Freedberg and Ms. Mercincavage: Division of General Medicine, Massachusetts General Hospital, 50 Staniford Street, 9th Floor, Boston, MA 02114.
Dr. Schackman: Department of Public Health, Weill Medical College of Cornell University, 411 East 69th Street, New York, NY 10021.
Drs. Seage and Weinstein: Harvard School of Public Health, 677 Huntington Avenue, Boston, MA 02115.
Author Contributions: Conception and design: A.D. Paltiel, R.P. Walensky, G.R. Seage III, M.C. Weinstein.
Analysis and interpretation of the data: A.D. Paltiel, R.P. Walensky, B.R. Schackman, G.R. Seage III, L.M. Mercincavage, M.C. Weinstein, K.A. Freedberg.
Drafting of the article: A.D. Paltiel, G.R. Seage III, M.C. Weinstein, K.A. Freedberg, R.P. Walensky.
Critical revision of the article for important intellectual content: A.D. Paltiel, R.P. Walensky, B.R. Schackman, G.R. Seage III, L.M. Mercincavage, M.C. Weinstein, K.A. Freedberg.
Final approval of the article: A.D. Paltiel, R.P. Walensky, B.R. Schackman, G.R. Seage III, L.M. Mercincavage, M.C. Weinstein, K.A. Freedberg.
Statistical expertise: G.R. Seage III, M.C. Weinstein.
Obtaining of funding: A.D. Paltiel, K.A. Freedberg.
Administrative, technical, or logistic support: L.M. Mercincavage.
Collection and assembly of data: L.M. Mercincavage.
An extensive literature supports expanded HIV screening in the United States. However, the question of whom to test and how frequently remains controversial.
To inform the design of HIV screening programs by identifying combinations of screening frequency and HIV prevalence and incidence at which screening is cost-effective.
Cost-effectiveness analysis linking simulation models of HIV screening to published reports of HIV transmission risk, with and without antiretroviral therapy.
Published randomized trials, observational cohorts, national cost and service utilization surveys, the Red Book, and previous modeling results.
U.S. communities with low to moderate HIV prevalence (0.05% to 1.0%) and annual incidence (0.0084% to 0.12%).
One-time and increasingly frequent voluntary HIV screening of all adults using a same-day rapid test.
HIV infections detected, secondary transmissions averted, quality-adjusted survival, lifetime medical costs, and societal cost-effectiveness, reported in discounted 2004 dollars per quality-adjusted life-year (QALY) gained.
Under moderately favorable assumptions regarding the effect of HIV patient care on secondary transmission, routine HIV screening in a population with HIV prevalence of 1.0% and annual incidence of 0.12% had incremental cost-effectiveness ratios of $30 800/QALY (one-time screening), $32 300/QALY (screening every 5 years), and $55 500/QALY (screening every 3 years). In settings with HIV prevalence of 0.10% and annual incidence of 0.014%, one-time screening produced cost-effectiveness ratios of $60 700/QALY.
The cost-effectiveness of screening policies varied within a narrow range as assumptions about the effect of screening on secondary transmission varied from favorable to unfavorable. Assuming moderately favorable effects of antiretroviral therapy on transmission, cost-effectiveness ratios remained below $50 000/QALY in settings with HIV prevalence as low as 0.20% for routine HIV screening on a one-time basis and at prevalences as low as 0.45% and annual incidences as low as 0.0075% for screening every 5 years.
This analysis does not address the difficulty of determining the prevalence and incidence of undetected HIV infection in a given patient population.
Routine, rapid HIV testing is recommended for all adults except in settings where there is evidence that the prevalence of undiagnosed HIV infection is below 0.2%.
Two unsolved problems in HIV screening policy are the maximum cost-effective screening frequency and the minimum HIV prevalence for cost-effective screening.
The authors used a decision model to estimate the cost-effectiveness of same-day rapid test HIV screening, considering outcomes experienced by the infected person and his or her sexual contacts. One-time screening was cost-effective when the prevalence of HIV was as low as 0.20%. Repeated screening every 5 years was cost-effective with an annual incidence of 0.0075% and an HIV prevalence of 0.45%.
The authors did not count HIV transmission from infected contacts.
Screening for HIV is cost-effective when HIV prevalence is similar to that of average-risk populations.
Table 1. Summary of Key Model Input Parameters and Sources for Efficacy of Antiretroviral Therapy and Rapid Test Protocol*
Table 2. Summary of Key Model Input Parameters and Sources for Target Population Characteristics and Effect of Patient Care on HIV Transmission*
Table 3. Mechanisms of Detection through Alternative Screening Practices, Baseline Population Scenario*
Table 4. Survival, Cost, and Cost-Effectiveness Results for HIV Screening Strategies in the Baseline Population*
Recommended strategy regions: $50 000 per quality-adjusted life-year threshold.
The figure recommends an HIV screening policy as a function of both the HIV prevalence in the target population (vertical axis) and the impact of HIV patient care on secondary transmission, ΔR0 (horizontal axis). ΔR0 can be interpreted as the lifetime number of secondary HIV infections averted when an HIV-infected person in a susceptible population is identified, counseled, and linked to treatment via HIV screening. Each prevalence value is associated with a specific incidence assumption (see Methods section for details). The figure recommends HIV screening policies, assuming that society is prepared to pay up to $50 000 per additional quality-adjusted life-year of health for its citizens. The dotted lines represent the 3 transmission impact scenarios described in Table 2: “favorable impact,” “no effect of screening and treatment on transmission impact,” and “adverse impact.” The curves denote the circumstances under which a given HIV screening strategy is preferred. For example, assuming no impact on secondary transmission, a one-time screening is recommended for prevalences greater than 0.28% (solid circle). Assuming a favorable transmission impact, the one-time screening threshold falls to 0.20% (solid square); with an adverse transmission impact, it increases to 0.40% (solid triangle). The threshold population for screening every 5 years (assuming favorable transmission impact) is HIV prevalence of 0.45% and annual incidence of 0.0075% (solid diamond).
One-time screening versus no specific screening program: sensitivity to cost-effectiveness threshold.
The figure identifies the evolution of the boundary between current practice (that is, no specific screening program) and one-time HIV screening as a function of 3 factors: 1) the prevalence of HIV in the target population (vertical axis); 2) the impact of care on secondary transmission, ΔR0 (horizontal axis); and 3) the value that society is prepared to pay to purchase an additional quality-adjusted life-year (QALY) of health for its citizens (as measured by the threshold cost-effectiveness ratio). Each prevalence value is associated with a specific incidence assumption (see Methods section for details). The figure reports results for threshold cost-effectiveness ratios ranging from $25 000 to $100 000 per QALY. The dotted lines represent the 3 transmission impact scenarios described in Table 2: “favorable impact,” “no effect of screening and treatment on transmission,” and “adverse impact.” The curves represent the borders of regions over which a given HIV screening strategy is preferred. For example, assuming that society is willing to pay up to $50 000/QALY and an adverse transmission impact, one-time screening is recommended for prevalences above 0.40% (solid circle); if society is willing to pay even more (up to $75 000/QALY), one-time screening is recommended for prevalences above 0.15% (solid square). Assuming no effect of screening and treatment on transmission and a societal willingness to pay $75 000 per additional QALY, one-time screening is recommended for prevalences above 0.10% (solid triangle). At a societal willingness to pay of $100 000/QALY, one-time screening is preferred under almost all plausible scenarios.
Study flow diagram.
Appendix Table 1. Incremental Effects of Model Updates: Individual-Patient–Level Analysis*
Appendix Table 2. Incremental Effects of Model Updates: Population-Level Analysis*
Appendix Table 3. Effects of Data Updates*
Appendix Table 4. Effects of Biennial HIV Screening*
Vasiliy V Vlassov
Russian branch of the Nordic Cochrane Centre
December 6, 2006
New Old screening
We believe that thorough analysis must go before the decisions, but in case when analysis support the decision it give a relief. This modeling study rise two questions "“ both addressed in the discussion "“ how reliable are estimates of the prevalence and was the old approach to screening for HIV right.
Russia still enjoys a rather low prevalence of HIV, but Russian HIV epidemic is the fastest all over the world, -- it is a common and widespread belief: that there are 940 000 [560 000 - 1 600 000] persons living with HIV/AIDS according to UNAIDS. Although the registered (one could say: evidence based) number of HIV+ in Russia is 347 222 . Two natural questions stem from huge difference between the data and estimate(s): (1) where the estimate(s) came from, and (2) what are their prospects, for instance, reconsiderations and reasons for up/downgrades. We suppose that the basis for neglecting data is low reputation of Russia or charm of a genre of Russian thriller. These estimates supported from within Russia itself ironically by the major official authority in this area Dr. V. Pokrovski, head of the national HIV/AIDS center, a special service to control and combat the epidemic. Pokrovski many times declared that true number of HIV+ persons in Russia is 2-5 higher than registered one. Although neither Pokrovski himself nor his aides in Russia and supporters abroad ever provided rational reasons for these estimates.
Recently published UNAIDS technique to estimate unknown HIV+ population from known populations of drug users, sex workers, and gays seems too approximate to improve the data in Russia. This estimate relies on the other estimates (1) of prevalence in risk groups, and (2) of their sizes. Although the populations's size of sex workers, gay/lesbian people, intravenous drug users is hardly known in Russia and elsewhere.
Meanwhile, Russia inherited from the USSR the system of extensive testing of citizens without a barrier of consent: blood donors, pregnant women, all inpatients etc. It is reasonable to suppose that no other big country does such testing. More likely Russian HIV prevalence data are more reliable than in other countries. Although the projects aimed on estimating the completeness of the Russian registration failed in fundraising, the support to common sense is appearing from unexpected sides. The efficacy of Russian case reporting system was de-facto recognized by the U.S. Centers for Disease Control and Prevention recommendations calling for routine HIV testing without specific consent in all doctors' offices, clinics, and hospitals, unless patients explicitly refuse or "opt out.", what is mimicking the Russian style. The WHO recently is also supporting the Russian style system by its recommendation of provider-initiated testing.
Until recently Russian government paid little attention to AIDS, and it seems like an adequate behavior first years "“ keeping in mind that incidence was low for the long time. The government relied on erected at the end of Soviet era surveillance system; keeps it and averted to destroy it in spite of appeals of international advisers who claim that system is ineffective and violates human rights. Two events moved HIV/AIDS upward the agenda list: (1) rising prevalence of HIV+, and (2) introduction of antiretrovirals. The latter made the system of registration sensible, and gave good reason for the optimism of its designers and builders, who are yet alive and deserve great esteem and respect.
B. Denisov, Senior Researcher, Lab of Population Economics and Demography, Mosow University (firstname.lastname@example.org)"¦ V. Vlassov, Director, Russian Branch of the Nordic Cochrane Centre (email@example.com)
(1) Paltiel AD, Walensky RP, Schackman BR, Seage GR, III, Mercincavage LM, Weinstein MC et al. Expanded HIV Screening in the United States: Effect on Clinical Outcomes, HIV Transmission, and Costs. Ann Intern Med 2006; 145(11):797-806.
(2) UNAIDS. Russian Federation: Indicators, Estimates and Country Assessment). Accessed Dec 6, 2006. Available from: URL:http://www.unaids.org/en/Regions_Countries/Countries/russian_federation.asp
(3) Russian Federal AIDS Centre. Officially Registered HIV Cases in the Russian Federation 1 January 1987 through 30 June 2006. Accessed Dec 6, 2006. 2006 Available from: URL:http://www.afew.org/english/statistics/HIVdata-RF.htm
(4) Walker N, Grassly NC, Garnett GP, Stanecki KA, Ghys PD. Estimating the global burden of HIV/AIDS: what do we really know about the HIV pandemic? Lancet 2004; 363(9427):2180-2185.
(5) The Russian HIV/AIDS Case Reporting System. European Population Conference, 21-24 June 2006, Liverpool, UK. Accessed Dec 6, 2006.
(6) CDC. Advancing HIV Prevention. New Strategies for a Changing Epidemic. Accessed Dec 6 2006. 2006 Available from: URL:http://www.cdc.gov/hiv/topics/prev_prog/AHP/default.htm
(7) WHO. WHO and UNAIDS Secretariat Statement on HIV testing and counseling, Aug 14, 2006. Accessed Dec 06 2006. Available from: URL:http://www.who.int/hiv/toronto2006/WHO- UNAIDSstatement_TC_081406_dh.pdf
Hartmut B. Krentz
Southern Alberta Cohort/University of Calgary
December 15, 2006
Impact on the HIV Care Budget of Expanded HIV Screening
To the Editor: The cost effectiveness of expanding rapid routine HIV screening to all adults, as promoted in the recent CDC recommendations (1), has been clearly described by Paltiel et al (2). We wished to model the predicted cost impact on the HIV care budget from transferring such recommendations into policy. Expanded screening should capture undiagnosed HIV patients primarily in populations not currently targeted. HIV care costs are heavily influenced by the stage of HIV disease (3). We used the 13 % of patients in our regional population diagnosed with HIV by existing healthy population based screening (insurance, blood donation, immigration and pregnancy), as a guide for measuring the anticipated acuity of new patients and then stratified their costs by CD4 count using our costing data (4). Of all healthy HIV infected patients found on population based screening 6% had initial CD4 counts of <75, 21% between 75"“200, 54% between 201-500, and 20% >500/mm3. Mean per patient per year costs in US$ for each CD4 strata was $29,460, $15,528, $13,020 and $12,756 respectively for fiscal year 2004/05. Our model estimates that if 100% of the predicted 25 % undiagnosed HIV patients in our population were identified by screening and initiated care, the cost to the HIV care budget would increase by 27.4%. More realistically at 75%, 50% or 25% detection levels, cost increases would be 21%, 13.4% and 7% respectively. These estimates may vary somewhat depending on the cost of care within a region, and both the true number and also the health of the undiagnosed HIV population. The direct cost, however, of providing medical care to those newly diagnosed by a screening program needs to be included in any budget prediction if the cost effectiveness for screening described by Paltiel et al (2) is to be achieved.
1 Branson BM, Handsfield HH, Lampe MA, Janssen RS, Taylor AW, Lyss, SB, et al Revised recommendations for HIV testing of adults, adolescents, and pregnant women in health-care settings. MMWR Recomm. Rep. 2006; 55: 1-17
2 Paltiel AD, Walensky RP, Schackman BR, Seage GR, Mercincavage LM, Weinstein MC, Freedberg KA Expanded HIV Screening in the United States: Effect on Clinical Outcomes, HIV Transmission, and Costs. Ann Intern Med 2006; 145: 797-806.
3 Chen RY, Accortt NA, Westfall AO, Mugavero MJ, Raper JL, Cloud GA et al Distribution of Health Care Expenditures for HIV-Infected Patients. CID 2006:42:1003-10.
4 Krentz HB, Auld C, Gill MJ. The changing direct costs of medical care for patients with HIV/AIDS, 1995-2001. CMAJ 2003; 169(2):106-110
David H Lander
Virginia College of Osteopathic Medicine
December 27, 2006
Re: Impact on the HIV Care Budget of Expanded HIV Screening
To the Editor:
Paltiel et al made another useful contribution to the debate regarding HIV screening1. I find it hard to imagine that we will not in the future expand the scope of HIV testing, in a variety of settings, and I support the general concept.
The advantages of rapid tests ("higher levels of test acceptance, follow-up, and linkage to care") seem appealing, but the authors caution that "rapid testing may exacerbate the distress associated with false-positive results." How often would false positives occur?
Using the high-end of 1% HIV prevalence used in their analysis ("that reflects the pre-September 2006 guidelines for HIV screening"), the "positive predictive value" (PPV) of the rapid test used in their model would be only 28.7%. In other words, if about 1% of the patients in a given primary care setting had HIV, then a positive rapid test would be correct about one out of three times, and incorrect the other 2 times. With a lower prevalence of 0.1% (the value they cite as an estimate of the US general population prevalence), the PPV would be 3.8%, producing about 25 false-positive results for each true-positive.
How would a primary care clinic or practice deal with this scenario? How does one get consent from a patient, obtain a positive test result, and then probably have to explain to the patient about the low (or possibly tiny) chance that the test is actually correct?
Perhaps the patient would be told out-front: "If your test is negative we will believe it, and we are done; but if it is positive, then that result is too inaccurate to accept, so we will send off the specimen for further testing." One wonders if the predictably large number of false-positives and the ensuing quantity of "distress" would lead the practitioners to question the practicality of such rapid testing, and lead to screening with old fashioned non-rapid tests.
At the risk of seeming obsessive, I considered the following unusual scenario that would (none-the-less) be bound (if only rarely) to occur if mass screening of low-prevalence patients was performed with such rapid tests. First, the patient is told the test is positive in the office or clinic, and is understandably greatly distressed, but then greatly relieved to hear it is probably not correct. Days later, the confirmatory Western blot result comes back positive, to his horror. But when the work-up is pursued further with a RNA PCR measurement, it turns out that the Western blot was a false- positive (as was found in 20 of 421 positive Western blot samples obtained in a study of blood donor screening2). Surprise, you really don't have HIV!
I raise these concerns about false-positive results not to discourage wider screening, but to plead for care in this potentially tricky endeavor.
David Lander MD FACP FACEP Associate Professor, Virginia College of Osteopathic Medicine Blacksburg, Virginia
1. Paltiel AD, Walensky RP, et al. Expanded HIV screening in the United States: effect on clinical outcomes, HIV transmission, and costs. Ann Intern Med. 2006; 145:797-806
2. Kleinman S, Busch MP, et al. False-positive HIV-1 test results in a low-risk screening setting of voluntary blood donation. JAMA 1998 280:1080-1085.
A. David Paltiel
Yale University School of Medicine
January 18, 2007
Impact of Expanded HIV Screening
We share Dr. Lander's concern regarding false positive results with rapid HIV tests, especially in populations of low prevalence, and agree that guidelines for communicating findings to patients will be useful. However, we believe that Dr. Lander's presentation of the issue is overstated. First, we deliberately accentuated the false-positive problem by adopting a conservative specificity assumption (97.5%). Today's rapid HIV tests have higher reported specificities (99.3% to 99.6%) and, therefore, more favorable predictive values. Second, current approaches to screening for other chronic diseases (mammography for breast cancer, for example) suggest that diagnostic tests with high false- positive rates can be appropriately managed in the clinical setting. Practitioners can explain that while a negative result is a reliable indicator of the absence of HIV infection (setting aside the 3-month pre- seroconversion "window" period), an initial positive result is not conclusive for HIV, but highlights the need for more specific tests.
Rapid HIV tests have similar sensitivity and specificity to standard antibody tests. They provide results within 20 minutes, eliminating the high rate of failure to return for results (25% in persons testing HIV- positive and 33% in persons testing HIV-negative at publicly funded U.S. clinics ). However, unlike standard antibody tests, positive results obtained via rapid testing are reported to the patient before they can be confirmed by repeat tests and Western Blots. The tradeoff is clear: wait one or two weeks, knowing that up to a third of cases will be lost to follow-up; or report preliminary results to patients and link them to care, knowing that this may cause short-term distress in a small percentage of those tested. We find that the benefits of rapid testing more than offset the downside risks, even when we assume a low-specificity test and assign large economic and quality-of-life costs to false-positive findings.
We agree with Drs. Krentz and Gill that "cost-effective" does not mean "cheap" and that planners must account for the direct costs of providing medical care to newly diagnosed cases. We included these costs in our analysis. We, too, found that the economic impact of expanded HIV screening lies less in the cost of the test than in the downstream treatment costs triggered when a new case is diagnosed. This highlights the need for a coordinated, comprehensive commitment of resources, at both the state and federal levels, to finance the impact of expanded HIV screening on publicly funded HIV programs in the U.S.
 Walensky RP and Paltiel AD. Rapid HIV testing at home: Does it solve a problem or create one? Annals of Internal Medicine (2006) 145:459 -462.
 U.S. Preventive Services Task Force. Recommendations and Rationale: Screening for Breast Cancer. http://www.ahrq.gov/clinic/3rduspstf/breastcancer/brcanrr.htm Accessed: January 16, 2007.
 Update: HIV counseling and testing using rapid tests "” United States, 1995. MMWR Morb Mortal Wkly Rep 1998;47: 211-5.
Paltiel AD, Walensky RP, Schackman BR, et al. Expanded HIV Screening in the United States: Effect on Clinical Outcomes, HIV Transmission, and Costs. Ann Intern Med. 2006;145:797–806. doi: https://doi.org/10.7326/0003-4819-145-11-200612050-00004
Download citation file:
Published: Ann Intern Med. 2006;145(11):797-806.
HIV, Infectious Disease.
Results provided by:
Copyright © 2020 American College of Physicians. All Rights Reserved.
Print ISSN: 0003-4819 | Online ISSN: 1539-3704
Conditions of Use