Effectiveness of the U.S. National HIV Testing Day Campaigns in Promoting HIV Testing: Evidence from CDC-Funded HIV Testing Sites, 2010

Abstract

Objectives

We assessed if HIV testing and diagnoses increased during the week of National HIV Testing Day (NHTD) and if characteristics of people who were tested varied compared with control weeks.

Methods

We analyzed HIV testing data from the 2010 National HIV Prevention Program Monitoring and Evaluation system to compare NHTD week (June 24–30, 2010) with two control weeks (January 7–13, 2010, and August 12–18, 2010) for the number of HIV testing events and new HIV-positive diagnoses, by demographics and other HIV-related variables. Characteristics associated with testing during NHTD week compared with control weeks were identified using Chi-square analyses.

Results

In 2010, an average of 15,000 more testing events were conducted and 100 more new HIV-positive diagnoses were identified during NHTD week than during the control weeks (p<0.001). Compared with control weeks, people tested during NHTD week were significantly less likely to be aged 20–29 years and non-Hispanic white and significantly more likely to be (1) aged ≥50 years, (2) non-Hispanic black or African American, (3) men who have sex with men, (4) low-risk heterosexuals, (5) tested with a rapid HIV test, or (6) tested in a non-health-care setting.

Conclusion

In 2010, CDC-funded HIV testing events and new HIV-positive diagnoses increased during NHTD week compared with control weeks. HIV testing programs increased the use of rapid tests and returned a high percentage of test results. NHTD campaigns reached populations disproportionately affected by HIV and further expanded testing to people traditionally less likely to be tested. Incorporating strategies used during NHTD in programs conducted throughout the year may assist in increasing HIV testing and the number of HIV-positive diagnoses.

Human immunodeficiency virus (HIV) continues to be a major public health problem in the United States. In 2009, the Centers for Disease Control and Prevention (CDC) estimated that 1.2 million people were living with HIV in the U.S. However, 18% are unaware they are infected and 32% of people diagnosed with HIV were also diagnosed with acquired immunodeficiency syndrome (AIDS) within 12 months of their diagnosis, which indicates they may have been infected with HIV for years before being diagnosed.¹ HIV testing identifies infected people and is the entry point to a continuum of HIV medical care and prevention services that improve health outcomes, including survival, and reduce the risk of HIV transmission. CDC recommends routine HIV testing in health-care settings for all people aged 13–64 years, annual HIV testing for those at high risk of infection, and routine testing as a part of prenatal care for all pregnant women.²

CDC supports these HIV testing recommendations with a number of initiatives^3,4 designed to expand HIV testing to populations disproportionately affected by HIV, especially non-Hispanic black or African American people (hereafter, non-Hispanic black people), Hispanic or Latino people, and men who have sex with men (MSM).⁴ CDC-funded HIV testing events are an important part of national HIV prevention efforts.⁵ In 2010, 3.2 million CDC-funded HIV testing events were conducted, diagnosing 13,000 newly identified people living with HIV.⁶

June 27 of each year is designated as National HIV Testing Day (NHTD). NHTD was founded in 1995 by the National Association of People with AIDS (NAPWA), an advocacy group for people living with HIV/AIDS. The objectives of NHTD are to encourage people to (1) get tested for HIV; (2) become aware of their HIV status; (3) get linked to prevention, care, and treatment services; and (4) work to reduce the stigma associated with HIV.^5,7 These objectives are in line with the major goals of the National HIV/AIDS Strategy, released by the White House in 2010.⁸ Through 2012, NAPWA organized and promoted NHTD activities with the support of CDC and other public and private partners, such as state and local health departments and community-based organizations. While the theme of each year's NHTD remains the promotion of HIV testing, local organizations determine the specific activities they conduct during NHTD. Most NHTD events offer free HIV testing and include mass media and social marketing campaigns to improve knowledge about HIV testing, care, and prevention services; address barriers to getting tested; and encourage testing.⁷

Because NHTD is a nationally recognized day to encourage HIV testing, it is important to document the effect this type of public health campaign has on increasing HIV testing and new HIV-positive diagnoses. Mass media campaigns can promote behavioral change for a wide range of HIV/AIDS preventive actions, including testing. A Cochrane review of 14 mass media interventions for promoting HIV testing found that these interventions have immediate and overall effects on promoting HIV testing in the targeted populations.⁹ However, most of these mass media interventions were designed and tested for specific populations or in limited geographical areas. In comparison, NHTD campaign activities are not predetermined at a national level, vary from place to place, and may incorporate a broad range of mass media campaigns. Therefore, the effectiveness of NHTD campaigns on increasing HIV testing nationally is unclear.

Two previous reports have assessed HIV testing before and after NHTD.^5,10 In 2000, CDC compared the number of HIV tests and HIV-positive tests the week before and the week of NHTD from 1995 through 1998 among 43 CDC-funded health departments. CDC found that 4,266 more HIV tests were conducted and 57 more HIV-positive tests were identified during the week of NHTD compared with the week before NHTD.⁵ Similar results were found in a 2013 study by the HIV/AIDS and Hepatitis Program of the Florida State Bureau of Communicable Diseases. From 2003 through 2012, greater numbers of HIV tests and HIV-positive tests were documented, annually, during the week of NHTD compared with the week before NHTD.¹⁰

We sought to update and expand the scope of the previous analyses. We assessed the effect national public health campaigns, such as NHTD, have on increasing the number of CDC-funded HIV testing events and the number of new HIV-positive diagnoses in the U.S. by asking two questions: (1) During the week of NHTD, how many additional CDC-funded HIV testing events were conducted and new HIV-positive diagnoses identified compared with the control weeks? and (2) Do characteristics of people tested during the week of NHTD differ from people tested during the control weeks? Specifically, do populations (e.g., MSM, non-Hispanic black people, and Hispanic or Latino people) disproportionately affected by HIV experience higher testing during the week of NHTD than during other times of the year?

The previous studies documented the importance of NHTD campaigns in increasing the uptake of HIV testing and diagnosis of people living with HIV.^5,10 However, the CDC study was conducted more than a decade ago and did not examine the characteristics of people tested. While the Florida study was more recent, it reflected data from only one state and did not provide comparisons to examine the potential differences in the populations served the week before and the week of NHTD. In contrast with the previous studies, we selected control weeks throughout the year rather than the week before NHTD to reduce the risk of missing a true difference in HIV testing (Type II error) due to implementation of NHTD activities in the weeks leading up to NHTD.

METHODS

Data source

CDC-funded health departments collect and report to CDC standardized test-level National HIV Prevention Program Monitoring and Evaluation (NHM&E) data. We conducted a secondary data analysis on a subset of 2010 NHM&E HIV test-level data with a valid session date for the HIV testing event. We defined an HIV testing event as a sequence of one or more HIV tests conducted with the client to determine his or her HIV status. During one testing event, a client may be tested once (i.e., one rapid test or one conventional test) or multiple times (i.e., one rapid test followed by one conventional test). However, the testing event data reported do not include personal identifiers; therefore, it is impossible to link multiple testing events to a unique client. Of 59 CDC-funded health departments, 52 reported analyzable test-level data to CDC for 2010. Seven health departments (Alabama; Chicago, Illinois; Los Angeles, California; Massachusetts; Mississippi; North Carolina; and the U.S. Virgin Islands), which contribute 21% of all testing events reported to CDC, were excluded because they did not report test-level data to CDC.⁶

Outcomes and covariates

We assessed demographic, HIV transmission risk, and HIV test characteristics. Demographic variables included gender, age, and race/ethnicity. We calculated race/ethnicity by combining the race and ethnicity variables using the following categories: Hispanic or Latino, non-Hispanic white, non-Hispanic black, and other. HIV transmission risk categories were ordered hierarchically based on the presumed likelihood of HIV transmission of self-reported sexual or drug use behaviors and included: (1) MSM/injection drug users (IDUs), (2) MSM, (3) IDUs, (4) high-risk heterosexual contact, (5) low-risk heterosexual contact, (6) no acknowledged risk, and (7) other.¹¹ Test type was categorized as rapid, conventional, or other. Test site type was categorized as health care, non-health care, or other. Provision of the test result was categorized as yes or no (missing or invalid records were excluded).

We used the self-reported previous HIV test result, test type, and current HIV test result variables to assess HIV status. We categorized HIV status as negative, newly identified preliminary HIV-positive (i.e., a testing event with a positive test result from a rapid HIV test and no positive test result from a conventional HIV test, and the client does not self-report having previously tested HIV-positive), and newly identified confirmed HIV-positive (i.e., a testing event with a confirmed HIV-positive test result from a conventional HIV test and the client does not self-report having previously tested HIV-positive). Newly identified HIV-positive testing events represent the total number of newly identified preliminary HIV-positive testing events and newly identified confirmed HIV-positive testing events.

Analysis

We used SAS^® version 9.3 to conduct all analyses.¹² HIV testing events were selected and categorized based on the session date. NHTD week was defined as June 24–30, 2010; control week 1 was defined as January 7–13, 2010; and control week 2 was defined as August 12–18, 2010. The control weeks were selected using three criteria. First, no known HIV testing or awareness campaigns, including NHTD, were conducted nationally within four weeks of these time periods. Second, no major national holidays (i.e., Memorial Day, Labor Day, Thanksgiving, or Christmas) were within one week of these time periods. Third, the weeks represent different times of the calendar year to help control for the effects of seasonal factors that may affect HIV testing.

We compared HIV testing events by week according to demographic, HIV transmission risk, and HIV test characteristics and report the number of testing events and column percentages. We conducted Chi-square analyses to identify the changes in the number of testing events, number of newly identified HIV-positive testing events, and characteristics associated with testing during NHTD week compared with the control weeks. We conducted Chi-square analyses when comparing the two control weeks to determine if the control weeks varied significantly by the characteristics analyzed (data not shown). The control weeks were compared separately with NHTD week because of differences in testing by age and test type and for groups (e.g., non-Hispanic white people and high-risk heterosexuals) with large numbers of testing events. We applied the Bonferroni correction for multiple comparisons to produce adjusted p-values for the characteristics (n=36 tests) associated with testing during NHTD week compared with the control weeks to reduce the potential to detect a difference in HIV testing when no true difference occurred (Type I error). We calculated the absolute difference between the column percentages and reported only if the unadjusted p-values comparing NHTD week with both control weeks were <0.001 and the adjusted p-values were <0.01. These standards were selected to highlight the characteristics with the most consistently significant differences between NHTD week and control weeks and to minimize the risk for Type I error due to multiple comparisons.

RESULTS

In 2010, the 52 health departments conducted 161,844 CDC-funded testing events during the three weeks of interest. During NHTD week, 63,914 testing events were conducted compared with 48,748 testing events during control week 1 and 49,182 testing events during control week 2, for differences of 15,166 and 14,732 more testing events, respectively (p<0.0001). During NHTD week, 467 newly identified HIV-positive testing events were reported compared with 367 during control week 1 and 356 during control week 2, for a total of 100 and 111 more newly identified HIV-positive testing events during NHTD week compared with control weeks 1 (p<0.0005) and 2 (p<0.0001), respectively (Figure).

Figure.

Number of (A) HIV testing events^a and (B) newly identified HIV-positive testing events,^b by week:^c NHM&E, 52 health departments, U.S. and Puerto Rico, 2010

^aCompared with control weeks 1 and 2, 15,166 and 14,732 more testing events were conducted during NHTD week, respectively (p<0.0001).

^bCompared with control weeks 1 and 2, 100 and 111 more newly identified HIV-positive testing events were reported during NHTD week (p<0.0005 and p<0.0001), respectively.

^cNHTD week was defined as June 24–30, 2010; control week 1 was defined as January 7–13, 2010; and control week 2 was defined as August 12–18, 2010.

HIV = human immunodeficiency virus

NHM&E = National HIV Prevention Program Monitoring and Evaluation

NHTD = National HIV Testing Day

We found a significant relationship between test week and age group, race/ethnicity, and transmission risk (Table 1). Compared with control weeks 1 and 2, there was a significant decrease in the percentage of people tested during NHTD week who were aged 20–29 years (–4.5% and –1.8%, respectively; p<0.0001) and non-Hispanic white (–2.8% and –1.4%, respectively; p<0.0001) and a significant increase in the percentage of people who were aged ≥50 years (2.7% and 1.6%, respectively; p<0.0001) and non-Hispanic black (2.9% and 1.9%, respectively; p<0.0001). Compared with control weeks 1 and 2, there was a significant increase in the percentage of people tested during NHTD week who were MSM (0.8% and 1.0%, respectively; p<0.0001) and low-risk heterosexuals (2.0% and 1.8%, respectively; p<0.0001).

Table 1.

Number of HIV testing events by week and select characteristics: NHM&E, 52 health departments, U.S. and Puerto Rico, 2010

^aChi-square tests were conducted to compare the distribution of people tested by week for each group compared with all other people (i.e., the percentage of males tested during NHTD week was compared with the percentage of females, transgendered individuals, and people with missing gender, combined).

^bBonferroni adjusted p-values were produced to reduce the potential for Type I error due to multiple comparisons; adjusted p-values were not produced for the total.

^cTransmission risk categories include only males and females; transgendered individuals and people with missing gender are excluded.

HIV = human immunodeficiency virus

NHM&E = National HIV Prevention Program Monitoring and Evaluation

NHTD = National HIV Testing Day

MSM = men who have sex with men

IDU = injection drug user

NA = not applicable

We found a significant relationship between test week and test type and test setting (Table 2). Compared with control weeks 1 and 2, there was a significant increase in the percentage of people tested during NHTD week who were tested with a rapid HIV test (6.6% and 4.9%, respectively; p<0.0001) and tested in a non-health-care setting (13.1% and 10.1%, respectively; p<0.0001). Compared with control week 1, a higher percentage of people who were tested during NHTD week received their test results (3.0%, p<0.0001); no significant difference was found during control week 2. There was no significant difference in the percentage of testing events that were newly identified HIV-positive (preliminary and confirmed) during NHTD week compared with control weeks 1 (p=0.667) and 2 (p=0.8935).

Table 2.

Number of testing events by week and HIV test-related characteristics: NHM&E, 52 health departments, U.S. and Puerto Rico, 2010

^aChi-square tests were conducted to compare the distribution of people tested by week for each group compared with all other people (i.e., the percentage of males tested during NHTD week was compared with the percentage of females, transgendered individuals, and people with missing gender, combined).

^bBonferroni adjusted p-values were produced to reduce the potential for Type I error due to multiple comparisons; adjusted p-values were not produced for the total.

HIV = human immunodeficiency virus

NHM&E = National HIV Prevention Program Monitoring and Evaluation

NHTD = National HIV Testing Day

NA = not applicable

DISCUSSION

Our data show that there was an average of 15,000 more CDC-funded HIV testing events and 100 additional new HIV-positive diagnoses during NHTD week (63,914 testing events) compared with the control weeks (average of 48,965 testing events). Increasing the number of people who know their HIV status is the first step to HIV care and treatment. Importantly, no change in the newly identified HIV-positivity was found between NHTD week and the control weeks, indicating that testing activities during NHTD week diagnosed more people living with HIV and maintained equivalent testing efficiency compared with the control weeks. These findings are consistent with those of the two previous studies that assessed the effect of NHTD.^5,10 CDC reported that from 1995 to 1998, an average of 4,266 more HIV tests were conducted and an average of 57 more HIV-positive tests were reported during NHTD week (average of 42,795 tests) than during the week before NHTD (average of 38,529 tests).⁵ For the period 2003–2012, the Florida State Bureau of Communicable Diseases reported an average of 1,986 (range: 846–2,901) more tests and 18 (–20–49) more HIV-positive tests during the week of NHTD than the week before NHTD.¹⁰ The increase in the number of testing events we report (14,949 average increase) is substantially larger than the increase found during the 1995–1998 period (4,266 average increase).⁵ The first CDC report was conducted during the first years of NHTD. The larger increase in testing observed in 2010 may be due to increased implementation or funding of NHTD, increased ability to test in community settings due to improved testing technology (e.g., rapid testing), and, perhaps, increased normalization of HIV testing among the general population.

Since 1995, in observance of NHTD, NAPWA and thousands of organizations in the U.S., including public health departments, community groups, and other institutions, have conducted special programs and activities with an overarching focus on HIV testing.⁷ NHTD campaigns not only address access to testing but also seek to address other barriers to testing. Schwarcz et al. reported that fear was the most frequently cited barrier to testing, as was being unaware of improved HIV treatment and risk for HIV.¹³ Mass media campaigns, often included as part of the NHTD activities, can help address these barriers. Studies that evaluated the effects of mass media campaigns have shown -significant improvements in information-seeking for HIV testing among African American women,¹⁴ hotline calls about and receipt of HIV testing among adolescents,¹⁵ and recall of HIV testing-related campaign material and HIV testing at partner clinics among Latinos.¹⁶ However, while another study was able to test thousands of people and identify hundreds of people living with HIV, awareness of the campaign was limited.¹⁷

In 2010, NHTD expanded HIV testing to a broader group of populations than are tested by CDC-funded HIV testing programs at other times during the year. We found that older people, non-Hispanic black people, MSM, and people with low-risk heterosexual contact were significantly more likely to be tested during NHTD week than during the control weeks. In addition to increasing knowledge and awareness, a primary goal of NHTD is to reduce stigma associated with HIV. Reaching people in the general population, both at higher and lower risk for HIV infection, will help to make testing more common and acceptable, thereby potentially reducing HIV-associated stigma.

NHTD campaigns may have successfully increased testing partly because they are tailored to the population each organization serves and are locally relevant.⁷ Not surprisingly, this analysis showed increased testing among two priority populations during NHTD week. Non-Hispanic black people are disproportionately affected by HIV, and a larger percentage of non-Hispanic black vs. non-Hispanic white people are unaware of their infection.¹ In this analysis, approximately 8,000 more non-Hispanic black people (2%–3% increase) were tested during NHTD week than during either control week. MSM are also disproportionately affected by HIV, and this analysis indicates that 1,800 more MSM (1% increase) were tested during NHTD week than during either control week.

NHTD campaigns do not appear to expand testing to all groups at increased risk for HIV, such as Hispanic or Latino people. This analysis found stable testing among Hispanic or Latino people during NHTD week compared with control weeks. There may be several reasons why testing among Hispanic or Latino people did not increase during NHTD week. First, Hispanic or Latino people are a heterogeneous population, given that they differ according to cultural heritage, race, socioeconomic status, national origin, and self-identification.¹¹ NHTD campaigns may not address the diversities of the various Hispanic or Latino groups. Second, compared with non-Hispanic black or white people, Hispanic or Latino people are more likely to experience significant psychosocial barriers to HIV testing and other HIV prevention services.^18,19 For example, in a 2009 national survey, Hispanic or Latino people were more likely to report not feeling at risk, not understanding where to get tested, and worrying about confidentiality as reasons for not being tested.¹⁸ Third, Hispanic or Latino people are more likely to be tested in clinical settings;^11,20 however, testing events were more likely to be in non-health-care settings during NHTD week than during control weeks. While NHTD campaigns may not be able to address all of these barriers, expanding testing locations and increasing development of disease prevention messages that are culturally appropriate may help reduce barriers to testing and increase individual HIV risk awareness, subsequently boosting HIV testing among Hispanic or Latino people.²¹

Increased use of rapid test technology (5%–7% increase) during NHTD week may have contributed to the NHTD campaigns' ability to maintain program effectiveness in returning HIV test results to an equal or greater percentage of people, despite testing more people than during control weeks. Tests during NHTD week were more likely to be rapid tests and to be conducted in a non-health-care setting (10%–13% increase). A previous study found that a rapid testing distribution program enabled diagnosis of HIV in people who might not otherwise have been tested and enabled them to test more clients for HIV due to greater acceptance of testing and staff availability for testing more clients.²²

Limitations

Our findings were subject to at least four limitations. First, this analysis excluded data from seven high-burden jurisdictions (contributing 21% of all testing events reported to CDC) because test-level data were not reported to CDC in 2010. Second, this analysis was limited to CDC-funded testing events. Overall, it is estimated that approximately 15%–20% of tests conducted in the U.S. are publicly funded.²³ Although it is unclear what proportion of testing events during NHTD week are publicly funded, it is reasonable to assume that a substantial proportion of the HIV testing events conducted during NHTD week may not be CDC-funded and, thus, were not included in this analysis. Third, the primary outcomes of this study may not be comparable with the outcomes of the other studies that assessed NHTD because these outcomes may be affected by changes in the national implementation of NHTD over time and by differences in demographic characteristics and variability in NHTD implementation at the local level.^5,10 Lastly, testing during control weeks may be different than testing during NHTD week for factors that could not be accounted for, such as seasonality and other HIV-related awareness events. Previous studies used the week before NHTD as the control week; however, our data indicated that NHTD campaigns may have begun to ramp up before the week of NHTD. To address this potential spillover effect, we selected two control weeks instead of one and chose different times of the year during which no other nationally known HIV testing or awareness campaigns were conducted. Our results indicate that the two control weeks did not differ in the number of testing events conducted and varied minimally by the characteristics assessed.

CONCLUSIONS

In 2010, NHTD campaigns increased the number of people living with HIV who know their HIV status by expanding CDC-funded HIV testing to populations disproportionately affected by HIV and people traditionally not targeted for HIV testing. In addition, NHTD campaigns increased the use of rapid HIV tests and returned a high percentage of test results compared with control weeks. These findings -underscore the importance of continuing to use new testing technology (e.g., rapid testing) in community-level HIV prevention programs to reach people who may not otherwise test for HIV and to ensure the timely return of test results. Furthermore, the findings support the adoption of strategies used during NHTD by other campaigns conducted throughout the year to further increase HIV testing and knowledge of HIV infection.