HIV testing is the cornerstone of HIV prevention and care and an essential step in the treatment of people living with HIV. In 2006, the Centers for Disease Control and Prevention (CDC) issued HIV testing guidelines that updated portions of the 2001 testing guidelines. The 2006 guidelines provide specific guidance for health care settings that promote routine opt-out HIV testing for all people between 13 and 64 years of age in populations with a prevalence of undiagnosed HIV of >0.1%.1,2 These guidelines are supported by evidence that timely awareness of serostatus and access to antiretroviral therapy (ART) can increase the life expectancy of persons with HIV by decades and may reduce HIV transmission due to reductions in risk behavior and treatment-based reductions in infectiousness.3,4 Transmission rates for persons unaware of their infection are estimated to be 3 times as great as those for persons aware of their infection.5 Additionally, for people without HIV, especially those in HIV-discordant partnerships, diagnosis and disclosure of a partner's status allow individuals and couples to make informed HIV prevention choices.6
Although there has been a modest increase in HIV testing since publication of the 2006 guidelines, the majority of Americans still have not been tested for HIV. Among those tested, approximately one-third of new HIV diagnoses are late diagnoses made simultaneously with or within 12 months of an AIDS diagnosis.7 In 2007, CDC announced the Expanded HIV Testing Initiative, which increased health department funding for HIV testing, early diagnosis, and linkage to care and prevention services and focused on jurisdictions with a relatively high proportion of AIDS diagnoses among blacks (funding opportunity announcement, FOA CDC-PS07-0768). More than a 3-year period (October 2007 to September 2010), CDC invested $102.3 million in 25 jurisdictions (23 states and large cities in the first year and 2 additional states in the second and third years) to conduct testing and related services. Health departments were required to focus 80% of their activities on promoting opt-out HIV screening in high morbidity clinical settings, such as emergency departments, sexually transmitted disease clinics, and community health centers. Up to 20% of resources could be used to test high-risk populations in nonclinical settings.8 Grantees also were required to implement sustainable HIV screening practices using either rapid or conventional testing.8 The primary objective of the Initiative was to increase HIV testing opportunities for populations disproportionately affected by HIV, primarily blacks, and to increase the proportion of HIV-infected persons who were aware of their infection and linked to appropriate medical care and prevention services. In 2008, blacks accounted for 51.2% of new HIV diagnoses, and the HIV diagnosis rate for blacks was 9 times as high as that of whites.9 Additionally, blacks and Hispanics and men who have sex with men (MSM) and injection drug users of all races and ethnicities are overrepresented among HIV-infected persons unaware of their status.10
Cost-effectiveness analyses support routine and expanded HIV testing.11–13 However, economic analyses are needed to determine the value of the benefits associated with investments in large-scale testing or the financial return on investment (ROI). Such analyses are helpful to government agencies that must optimize outcomes for public health given a set budget and may assist with resource allocation during times of economic constraint.14,15 In this article, we report the results of a ROI analysis that compares expenditures of the Expanded HIV Testing Initiative to benefits in terms of HIV transmissions averted and associated medical care costs.
We conducted a ROI analysis of the Initiative using expenditure and outcome data reported over the 3 years of the program. Using a published mathematical model of HIV transmission, we estimated the number of HIV transmissions averted by the Initiative based on the number of persons tested, newly diagnosed with HIV infection, and linked to care.16 We estimated the costs associated with these averted HIV infections and the ROI in the program. Benefits of HIV transmissions averted were valued using published estimated lifetime HIV treatment costs discounted to the time of infection.17 We adjusted all costs and expenditures to 2009 dollars and discounted future costs and effects at a 3% annual rate. We conducted our analysis in Microsoft Excel.
We calculated ROI for programmatic expenditures on HIV testing and related services by CDC and partners, such as state and local governments, and the larger health system. Return on CDC/partner investment was calculated as program benefits (the averted medical costs associated with HIV transmissions prevented) divided by expenditures by CDC and partners. Return on the health system investment was calculated as program benefits divided by all program expenditures, regardless of source, and including the medical costs of treating newly diagnosed index patients through the period of analysis. Policy interest focused on whether program benefits (treatment costs averted) exceeded the investment in the Initiative (an ROI value > $1, or a positive return on investment).
Program Expenditures and Treatment Costs
For the CDC/partner investment, we defined the program investment as the actual expenditures for the 3 fiscal years (FY) of the program, FY 2007–FY 2009, as reported by the grantees to CDC. We also calculated additional non-CDC expenditures for the Initiative based on a survey of 7 of the 25 funded jurisdictions regarding outside funding used to implement the Initiative. The average proportion of CDC expenditures for the Initiative to total expenditures for testing under the Initiative was 81% in these jurisdictions. Thus, we increased CDC expenditures accordingly to estimate expenditures by CDC and partners. For the health system investment, we also included costs incurred by the health system for HIV treatment provided to persons newly identified as HIV infected through the Initiative for the early awareness period, the estimated period of time before these persons would have been otherwise diagnosed with HIV. These costs were derived from the lifetime HIV treatment cost estimate.
Benefits of the testing program were defined as the value of HIV transmissions averted from persons newly identified with HIV through the Initiative. We used lifetime HIV treatment costs of $367,134 (adjusted to 2009 dollars) to value each transmission averted.17 We did not include benefits to the index client from earlier diagnosis. Additionally, it should be noted that regardless of testing interval, the benefits of averted medical costs due to HIV transmission accrue over a lifetime.
HIV Transmission Model
The mathematical model of HIV transmission incorporated the effects of awareness of HIV status on risk behavior and the use of ART.5,18 We modeled transmissions averted as the difference between the transmissions likely to occur by those who were unaware of their infection and the transmissions likely to occur by those who became aware of their infection earlier through the Initiative. We calculated transmissions averted from the time individuals with HIV learned of their infection through the Initiative until the estimated time on average those persons would have learned of their infection without the screening program, which we refer to as the alternative testing interval. We chose a 3-year testing interval for our base case, which is the mid-point between the 5-year interval commonly used in economic evaluations of HIV screening12,13 and the 1-year testing recommendations for high-risk groups.2 In sensitivity analyses, we assessed the effect of 1-year, 2-year, 4-year, and 5-year alternate testing intervals.
We used program outcome data from funded jurisdictions for FY 2007 through FY 2009 on the number of persons tested and the number of newly infected persons identified, informed of their results, and linked to care. Under the Expanded HIV Testing Initiative, a total of 2,786,739 people were tested for HIV; 18,432 (0.7%) were newly diagnosed with HIV; and 15,737 (91%) of these persons were notified of their test results (Table 1). We adjusted these data for the expected number of infected persons who would have been identified through background testing or testing that would have occurred in the absence of the Initiative. The background annual testing level (17%) was based on a CDC HIV prevention resource allocation model19 (A. Lasry, PhD, CDC unpublished data).
We then applied estimated transmission probabilities for aware versus unaware HIV-infected persons to calculate the total number of transmissions averted attributable to the Initiative. To estimate sexual HIV transmissions averted, we applied annual transmission rates for HIV-infected persons who were unaware of their HIV status,20 those who were aware of their status and on ART, and those aware of their status and not on ART20 (Table 1). The reduced likelihood of sexual transmission of HIV after a new diagnosis was due to a combination of the index person's engagement in fewer risky behaviors and in reduced transmission due to viral load suppression after treatment. We used the same method to estimate HIV transmissions averted due to injection drug use (IDU) as we did for sexual transmission. However, we adjusted IDU transmission rates only by awareness of infection and not by treatment status. IDU transmission rates, which were based on injection drug use behaviors alone (exclusive of sexual transmission by injection drug users), were drawn from the literature.13,21 The estimated proportions of HIV transmissions from sexual activity and injection drug use were derived from 2006 national HIV incidence surveillance data.22
In our model, 75% of newly identified HIV-infected persons were linked to care-based outcome data from the Initiative. We assumed that 80% of persons diagnosed were eligible to start ART at diagnosis based on program data on the number of new HIV diagnoses by setting, and data from the literature on the distribution of CD4 counts at diagnosis by testing setting. Thirty percent of persons tested through the Initiative were tested in emergency departments, 20% in sexually transmitted disease clinics, and 23% in community health centers or community-based organizations.8Based on data from White et al,23 we estimated that 50% of patients in emergency departments were tested at CD4 counts of 350 or fewer cells per microliter, 20% were tested at CD4 counts between 350 and 500 cells per microliter, and 30% were tested at CD4 counts greater than 500 cells per microliter. For sexually transmitted disease clinics and community health centers, we estimated that 35% of patients were tested at CD4 counts of 350 or fewer cells per microliter, 25% were tested at CD4 counts between 350 and 500 cells per microliter, and 40% were tested at CD4 counts greater than 500 cells per microliter.24,25 We then assumed that all patients with a CD4 count of 350 or fewer cells per microliter started ART at diagnosis, as did 90% of patients diagnosed with a CD4 count between 350 and 500 cells per microliter and 50% of patients diagnosed with a CD4 count greater than 500 cells per microliter. These assumptions, which assume relative adherence to current guidelines regarding initiation of ART, resulted in the estimate that 80% of all patients were eligible to start ART at diagnosis.26 We assumed newly diagnosed patients who did not start ART remained untreated for the duration of the period of the analysis.
The HIV transmission model is summarized in equation (1):
TS = sexual transmissions averted per person with HIV diagnosed and notified.
TIDU = IDU transmissions averted per person with HIV diagnosed and notified.
PS = proportion of transmissions due to sexual activity in the U.S. in 2006 (0.865).
PIDU = proportion of transmissions due to injecting drug use in the U.S. in 2006 (0.134).
N = number of persons newly diagnosed with HIV and notified of their infection adjusted for background testing (Table 1).
Sexual transmissions averted,
IDU transmissions averted,
TnoINITIATIVE = number of transmissions without the Initiative.
TINITIATIVE = number of transmissions with the Initiative.
D = duration individuals would be undiagnosed in the absence of the Initiative (3 years).
PL = proportion of newly diagnosed persons linked to care from the Initiative (0.75).
μSUA = sexual annual HIV transmission rates for unaware persons (0.1117).
μSAART = sexual annual HIV transmission rates for aware persons on ART (0.0097).
μSANART = Sexual annual HIV transmission rates for aware persons not on ART (0.0484).
μIUA = IDU annual HIV transmission rates for unaware persons (0.165).
μIA = IDU annual HIV transmission rates for aware persons (0.126).
Sensitivity and Threshold Analysis
We conducted additional sensitivity analyses to determine how robust our findings are to alternative testing intervals of 1–5 years and smaller differences in transmissions averted by adjusting the base case values downward 25%, 50%, and 75%. We also conducted a threshold analysis to determine the lowest prevalence of undiagnosed HIV infection at which ROI values would be $1.0 or positive, for the base case, 3-year, alternative testing interval.
Under base case assumptions that infected persons were diagnosed 3 years earlier because of the Initiative, we estimated that 3381 HIV infections were averted from persons newly diagnosed (Table 2).
Return on the Health System Investment
The health system investment included an estimated total expenditure of $599,096,000, which consisted of investments in the Initiative from CDC of $102,335,000; additional funding from sources, such as state and local government, of $24,049,000; and estimated medical costs for newly diagnosed persons of $472,712,000. Medical care costs from HIV infections averted were estimated to be $1,169,887,000 resulting in net benefits (total benefit minus total cost) of $570,791,000 and a return on investment of $1.95 for every dollar invested in the Initiative (Table 2).
The ROI ranged from $1.46 to $2.01 for the 1-year to 5-year alternative testing intervals. ROI values remained above $1, a positive return on investment, with a prevalence of undiagnosed HIV infection as low as 0.12% (data not shown) and with a 25% reduction in transmissions averted. ROI approached $1 ($0.98) at a 50% reduction in transmissions averted from base case values (Fig. 1).
Return on CDC's Investment
The total investment in the program from CDC and partners for the 3 years was $126,384,000. Medical care costs from HIV infections averted were estimated to be $1,169,887,000 resulting in net benefits of $ 1,043,503,000 or a return on investment of $9.26 for every dollar invested in the Initiative (Table 2). When considering only CDC's investment in the Initiative, the ROI was $11.43.
The number of HIV infections averted, net benefits, and ROI values increased with the length of the alternative testing interval or the time the infected persons would have been diagnosed on average without the Initiative. The return on CDC and partners' investment ranged from $3.27 to $14.54 for the 1-year to 5-year alternative testing intervals (Table 2). ROI values remained above $1 with the prevalence of undiagnosed HIV infection as low as 0.07% and with reductions in transmissions by 75% (data not shown).
Our findings showed considerable gains from investments in CDC's Expanded HIV Testing Initiative when quantified in dollar terms. The return on the health system investment, which takes into account the medical care and treatment costs for those diagnosed with HIV, was $1.95 saved for every dollar invested. Our ROI value, which reflects the value of investment in prevention, compared favorably with ROI values calculated for US investments in treatment of prevalent heath conditions such as heart attack (ROI, $1.10), stroke (ROI, $1.49), and type 2 diabetes (ROI, $1.55).14 Additionally, positive return on investment is a high threshold for health care interventions, as the majority of medical interventions that improve health require an investment greater than the value of the benefits (ROI < $1).27 Our sensitivity analysis indicated that net benefits may accrue to testing programs with much lower undiagnosed seroprevalence (0.12%) than the 0.7% observed in this program.
Our analysis is conservative for several reasons. Because we assumed all persons diagnosed under the Initiative would have been diagnosed at a later date, we did not assign a benefit in terms of extended survival and quality of life from earlier treatment to the index case. There is evidence from the cost-effectiveness analysis literature that HIV diagnosis at slightly earlier stages of disease can be cost-saving when the benefits of averted HIV transmissions are included.28 Cost-effectiveness analyses, however, measure actual costs, whereas ROI is based on expenditures. If we included benefits for the index case, if the Initiative resulted in an earlier diagnosis than we assumed, or if persons were diagnosed who would never have been diagnosed otherwise, our ROI estimates would be greater. Additionally, our analysis focused only on persons with newly diagnosed HIV infection. Other persons diagnosed through the Initiative were previously known by the health department to be infected. Some of these individuals may not have received their previous HIV test results although others may have been linked or relinked to care as a result of the program. Because data on relinkage were incomplete, we were not able to assess this potential benefit.
ROI changed with alternative testing intervals. If the average retesting interval was longer, the benefits of reduced transmission would have been greater and many clients would have accrued additional benefits from early initiation of ART. Although a 1-year testing interval is recommended for high-risk populations, it is not yet current practice on a population level and represents benefits that populations with higher frequencies of testing (eg, MSM) may experience. In addition, for the health system investment, our findings were sensitive to a reduction of 50% or more in HIV transmissions averted. However, even with a 50% reduction, the Initiative produced health gains valued at $0.98 for every dollar invested.
Our analysis was subject to a number of limitations. The lifetime medical costs we used to value the benefits of HIV transmissions averted assume individuals will remain in treatment and receive optimal care. Thus, the benefits could be overestimated in populations in which HIV-infected persons are not retained into care. We limited our calculations of HIV transmissions averted to a first generation of transmissions, which underestimated the total number of averted HIV transmissions attributable to the Initiative. However, we also counted HIV infections averted as permanently averted rather than delayed. We did not include acute phase detection of HIV infection because most screening tests detect HIV infection after the acute phase.29 Recently, approved fourth generation HIV enzyme immunoassays can detect infections as early as the acute phase.30 Detection of acute infections could have additional transmission benefits. In our model, the same transmission rates were used for MSM and heterosexual populations, so our estimates should be considered conservative with regard to MSM populations. Caution should be used when applying these findings more broadly than CDC's Expanded HIV Testing Initiative, as key parameters such as testing and diagnostic rates and expenditures were specific to the Initiative. There were also some aspects of the program that we did not explicitly value in our analysis such as the benefits of prevention services, including partner services. Additionally, we included expenditures for the first year that had lower testing rates due to start-up activities, which may have yielded more conservative ROI values for the project period. With continued testing, however, the program could gradually experience a decrease in HIV positivity over time, which would decrease ROI values.
We did not value downstream HIV treatment costs that may be incurred as a result of earlier diagnosis outside of the early awareness period conferred by the Initiative. A recent study has assessed the budget impact, or stream of financial costs, of expanded HIV screening in the United States on government programs over a 5-year period.31 These authors found that expanded testing would increase total HIV testing and treatment costs; however, they did not assign financial benefits to preventing HIV transmission to partners.31
Despite these limitations, the strength of our study is that we used actual expenditure and outcome data from a large-scale program and assessed expenditures by agencies and organizations other than CDC. We thereby were able to reduce the uncertainty associated with relying on data from other studies.
We find that CDC's Expanded HIV Testing Initiative yielded positive returns on investment over a broad range of assumptions, supporting large-scale HIV testing programs as beneficial from public health and economic standpoints. Return on investment analysis, along with cost-effectiveness analysis, can assist policy makers in optimizing public health given a set budget.
1. Centers for Disease Control and Prevention (CDC). Revised guidelines for HIV counseling, testing, and referral. MMWR Recomm Rep. 2001;50:1–58.
2. Centers for Disease Control and Prevention (CDC). Revised recommendations for HIV testing of adults, adolescents, and pregnant women in health-care settings. MMWR Recomm Rep. 2006;55:1–17.
3. Lohse N, Hansen AB, Pedersen G, et al.. Survival of persons with and without HIV infection in Denmark, 1995–2005. Ann Intern Med. 2007;146:87–95.
4. Donnell D, Baeten JM, Kiarie J, et al.. Heterosexual HIV-1 transmission after initiation of antiretroviral therapy: a prospective cohort analysis. Lancet. 2009;375:2092–2098.
5. Marks G, Crepaz N, Janssen RS. Estimating sexual transmission of HIV from persons aware and unaware that they are infected with the virus in the USA. AIDS. 2006;20:1447–1450.
6. Gorbach PM, Weiss RE, Jeffries R, et al.. Behaviors of recently HIV-infected men who have sex with men in the year post-diagnosis: effects of drug use and partner types. J Acquir Immune Defic Syndr. 2011;56:176–182.
7. CDC. Late HIV Testing—34 States, 1996—2005. MMWR Morb Mortal Wkly Rep. 2009;58:661–665.
8. CDC. Results of the Expanded HIV Testing Initiative—25 Jurisdictions, United States, 2007–2010. MMWR Morb Mortal Wkly Rep. 2011;60:806–810.
9. CDC. Vital signs: HIV testing and diagnosis among adults—United States, 2001–2009. MMWR Morb Mortal Wkly Rep. 2010;59:1550–1555.
10. Campsmith M, Rhodes PH, Hall HI, et al.. Undiagnosed HIV prevalence among adults and adolescents in the United States at the end of 2006. J Acquir Immune Defic Syndr. 2010;53:619–624.
11. Long EF, Brandeau ML, Owens DK. The cost-effectiveness and population outcomes of expanded HIV screening and antiretroviral treatment in the United States. Ann Intern Med. 2010;153:778–789.
12. Paltiel AD, Weinstein MC, Kimmel AD, et al.. Expanded screening for HIV in the United States–an analysis of cost-effectiveness. N Engl J Med. 2005;352:586–595.
13. Sanders GD, Bayoumi AM, Sundaram V, et al.. Cost-effectiveness of screening for HIV in the era of highly active antiretroviral therapy. N Engl J Med. 2005;352:570–585.
14. Luce BR, Mauskopf J, Sloan FA, et al.. The return on investment in health care: from 1980 to 2000. Value Health. 2006;9:146–156.
15. Grosse SD, Sotnikov SV, Leatherman S, et al.. The business case for preconception care: methods and issues. Matern Child Health J. 2006;10:(5 suppl):S93–S99.
16. Hutchinson AB, Patel P, Sansom SL, et al.. Cost-effectiveness of pooled nucleic acid amplification testing for acute HIV infection after third-generation HIV antibody screening and rapid testing in the United States: a comparison of three public health settings. PLoS Med. 2010;7:e1000342.
17. Schackman BR, Gebo KA, Walensky RP, et al.. The lifetime cost of current human immunodeficiency virus care in the United States. Med Care. 2006;44:990–997.
18. Marks G, Crepaz N, Senterfitt JW, et al.. Meta-analysis of high-risk sexual behavior in persons aware and unaware they are infected with HIV in the United States: implications for HIV prevention programs. J Acquir Immune Defic Syndr. 2005;39:446–453.
19. Lasry A, Sansom SL, Hicks KA, et al.. A model for allocating CDC's HIV prevention resources in the United States. Health Care Manag Sci. 2011;14:115–124.
20. Prabhu VS, Hutchinson AB, Farnham PG, et al.. Sexually acquired HIV infections in the United States due to acute-phase HIV transmission: an update. AIDS. 2009;23:1792–1794.
21. Zaric GS, Barnett PG, Brandeau ML. HIV transmission and the cost-effectiveness of methadone maintenance. Am J Public Health. 2000;90:1100–1111.
22. Hall HI, Song R, Rhodes P, et al.. Estimation of HIV incidence in the United States. JAMA. 2008;300:520–529.
23. White DE, Scribner AN, Schulden JD. Results of a rapid HIV screening and diagnostic testing program in an urban emergency. Ann Emerg Med. 2009;54:56–64.
24. Golden MR, Stekler J, Wood RW. Trends in the frequency of HIV testing and CD4 count at diagnosis among persons tested through a public health program in Seattle, WA USA, 1995–2008; Presented at: 16th Conference on Retroviruses and Opportunistic Infections; February 8–11, 2009; Montreal, Canada.
25. Lyons MS, Lindsell CJ, Hawkins DA, et al.. Contributions to early HIV diagnosis among patients linked to care vary by testing venue. BMC Public Health. 2008;8:220.
26. Panel on Antiretroviral guidelines for Adults and Adolescents. Guidelines for the Use of Antiretroviral Agents in HIV-1-Infected Adults and Adolescents. Washington, DC: Department of Health and Human Services; 2011:1–174. Available at: http://www.aidsinfo.nih.gov/ContentFiles/AdultandAdolescentGL.pdf
. Accessed April 2011.
27. Trogdon J, Finkelstein EA, Reyes M, et al.. A return-on-investment simulation model of workplace obesity interventions. J Occup Environ Med. 2009;51:751–758.
28. Prabhu VS, Farnham PG, Hutchinson AB, et al.. Cost-effectiveness of HIV screening in STD clinics, emergency departments, and inpatient units: a model-based analysis. PLoS One. 2011;6:e19936.
29. Owen SM, Yang C, Spira T, et al.. Alternative algorithms for human immunodeficiency virus infection diagnosis using tests that are licensed in the United States. J Clin Microbiol. 2008;46:1588–1595.
30. Patel P, Mackellar D, Simmons P, et al.. Detecting acute human immunodeficiency virus infection using 3 different screening immunoassays and nucleic acid amplification testing for human immunodeficiency virus RNA, 2006–2008. Arch Intern Med. 2010;170:66–74.
31. Martin EG, Paltiel AD, Walensky RP, et al.. Expanded HIV screening in the United States: what will it cost government discretionary and entitlement programs? A budget impact analysis. Value Health. 2010;13:893–902.