Challenges in Identifying and Estimating Undiagnosed HIV Infection

Keywords:

• HIV
• surveillance
• testing
• undiagnosed

In the USA, approximately 18% of the estimated 1.1 million individuals with HIV infection are unaware of their status, although the proportion has decreased somewhat in recent years [101]. HIV testing strategies for reaching the undiagnosed are showing promise, yet a key challenge is that accurate estimates of the proportion of infected individuals remaining undiagnosed are hard to obtain: multiple methods have been used, yet no consensus exists as to the best approach [1]. In order to reduce the proportion of undiagnosed infections, we need to both apply effective testing strategies and identify the best methods for estimation.

Reaching and testing individuals unaware of their infection are critical to the success of HIV prevention in the USA. An estimated 49% of transmissions are from individuals living with HIV unaware of their infection [2]. Individuals with an undiagnosed HIV infection are more likely to transmit the virus to others, since they are more likely to be in the acute infection stage, with a higher infectiousness, and are less likely to take precautions than those who are aware of their status [3,4]. Identifying HIV infection at an early stage among all individuals – not just those with acute HIV infection – is critical for timely linkage to care, initiation of combination antiretroviral therapy when indicated and promotion of adherence to these regimens to reduce morbidity and transmission [5,6].

Multiple HIV testing strategies are designed to reduce the number of undiagnosed infections. These include test and treat, routine testing and increased frequency of testing. Early diagnosis and prompt treatment are the basis for a ‘test-and-treat‘ approach to reduce HIV transmission. Test-and-treat programs are characterized by aggressive methods to test a broad segment of a population in order to identify the undiagnosed and re-engage HIV-infected individuals in care. Early treatment improves overall health outcomes [7]. In addition, the reduced viral loads in individuals living with HIV who are on effective treatment greatly reduce the probability of transmission of the virus to others [8–10]. Individuals in care can also benefit from counseling for risk behaviors [102]. Community- and prison-based test-and-treat strategies are currently being evaluated for feasibility in the USA.

Routine testing is another strategy for reaching the undiagnosed, and is particularly effective in areas with high prevalence for multiple reasons. People who may not seek out testing often accept it when offered. Providers may not assess HIV risk among all patients; routine testing takes away the need for risk assessment. Routine testing also reduces the stigma associated with HIV testing; when all patients are offered testing, no individual or group is singled out. Routine HIV testing is cost-effective in populations with undiagnosed HIV prevalence over 0.1% [11]. Recently, the United States Preventive Services Task Force (USPSTF) issued a recommendation that “clinicians screen adolescents and adults ages 15–65 years for HIV infection” [12]. This was a grade A recommendation. Grade A and B recommendations from the USPSTF are the basis for certain expansions of no-cost-sharing coverage for clinical preventive services under the Patient Protection and Affordable Care Act of 2010. This means that, for many insured individuals, routine HIV testing may be covered by health insurance plans or policies. CDC has implemented routine HIV testing through the Expanded HIV Testing Initiative in 30 public health jurisdictions; in the first 3 years of this program, 2.8 million tests were conducted and more than 18,000 individuals were newly diagnosed with HIV infection [13].

In populations with high HIV incidence, frequent testing may be needed to stay abreast of undiagnosed infections. Men who have sex with men (MSM) account for the largest proportion (63%) of new HIV infections in the USA [103]. CDC guidelines identify MSM who should be tested more frequently according to their risk behaviors [11]. An analysis of a community-based sample of MSM in cities with high HIV and AIDS prevalence found that risk behaviors did not differentiate those newly diagnosed, and that frequent testing, perhaps as often as every 3–6 months, might be warranted among sexually active MSM to better identify undiagnosed infection [14]. CDC is currently funding two large-scale testing programs for MSM: one focusing on young MSM of color and the other focusing on African–American and Hispanic/Latino MSM of any age; these programs will provide model practices for reaching the undiagnosed in these subgroups.

Improved sensitivity of HIV diagnostic tests also factor into strategies for identifying those with undiagnosed infections. The fourth-generation combined antigen/antibody test identifies infections early [4]. Tests that identify infection early allow for referral to early and appropriate care to improve health outcomes, and can be used to stop chains of transmission.

Estimating the prevalence and characteristics of people with undiagnosed HIV infection at the national and local levels is critical to plan both routine and targeted testing programs; application of testing strategies should be informed by the best available data to characterize undiagnosed infections. Different methods of measuring undiagnosed HIV infection result in different estimates of the proportion undiagnosed; as noted above, there is no consensus on which method is best, and use of different methods may depend on available data [1]. Estimates of undiagnosed HIV infection in the USA have been derived from two surveillance systems: the National HIV Surveillance System (NHSS), which uses HIV case reports and modeling methods for estimation; and the National HIV Behavioral Surveillance System (NHBS), which estimates undiagnosed infections from prevalence surveys.

Using NHSS data on HIV and AIDS diagnoses and deaths among individuals with HIV, the prevalence of diagnosed and undiagnosed HIV infections is estimated with a back-calculation method [15]. First, data are statistically adjusted to mitigate the effects of delays in reporting new cases and deaths, incomplete reporting of diagnosed cases and cases reported without sufficient risk factor information to be classified into an HIV transmission category. Second, on the basis of the estimated annual number of HIV diagnoses and the severity of disease at diagnosis (i.e., whether the infection was classified as stage 3 [AIDS] in the same calendar year the HIV diagnosis was made), the extended back-calculation model is fitted to estimate the cumulative number of individuals who had been infected with HIV each year. The estimated overall HIV prevalence is calculated by subtracting the estimated cumulative number of deaths that have occurred among those infected from the estimated cumulative number of HIV infections. The prevalence of undiagnosed HIV infection is calculated by subtracting the estimated number of diagnosed HIV infections in living individuals from the estimated overall HIV prevalence.

This method of estimation using NHSS data serves as the metric for estimating the proportion of individuals with an undiagnosed HIV infection as an indicator for the National HIV/AIDS Strategy (NHAS). The NHAS goal is to increase the proportion of individuals who know their status to 90% by 2015 [104]. Current estimates based on the most recently published surveillance data for 2009 are that 18.1% remain undiagnosed [101].

Another method used to describe undiagnosed infections is based on self-reported HIV status and subsequent HIV testing (i.e., prevalence surveys). In the USA, this latter method is available from NHBS.

NHBS operates in 20 cities in the USA, with high rates of HIV and AIDS diagnoses. It is used to monitor prevalence and trends in risk behaviors, testing and use of prevention services among populations at high risk for acquiring HIV. Surveys and tests are offered without regard to self-reported HIV status. The percentage of individuals unaware of their HIV infection is defined as those newly diagnosed through NHBS who reported their HIV status as negative, indeterminate or unknown, or who reported never having an HIV test [14]. NHBS has produced estimates of undiagnosed infections of approximately 45% for MSM, injection drug users and heterosexuals [16].

It is important to distinguish between NHSS and NHBS measures of undiagnosed infection and use them correctly (i.e., not using NHBS data to monitor progress towards NHAS goals, which are based on NHSS methods). Each method has strengths and limitations: for example, the NHBS method is subject to biases inherent in self-reported data and is designed to provide information on high-risk populations; the NHSS models, while appropriate to generalize to the US population, are based on a number of assumptions. Understanding the unique contributions of each method and how they might be used together (e.g., to describe undiagnosed infection among MSM) will help to create a better description of undiagnosed HIV infection in the USA.

Reducing the proportion of individuals with an undiagnosed HIV infection is critical for improving health outcomes and reducing HIV transmission, addressing two of the goals of the NHAS. Testing programs are key to reducing the proportion of HIV-infected individuals who are undiagnosed. Addressing measurement issues in estimating the proportion who are undiagnosed is important to efficiently target resources and accurately assess progress. Continued development of methods for estimating this proportion and reaching those individuals with HIV testing and linkage to care are necessary for successful reductions in HIV infections in the USA.

Disclaimer

The findings and conclusions in this manuscript are those of the authors and do not necessarily represent the views of CDC.

Financial & competing interests disclosure

The authors have no relevant affiliations or financial involvement with any organization or entity with a financial interest in or financial conflict with the subject matter or materials discussed in the manuscript. This includes employment, consultancies, honoraria, stock ownership or options, expert testimony, grants or patents received or pending, or royalties.

No writing assistance was utilized in the production of this manuscript.