Influence of anonymous HIV testing on national HIV surveillance in the Republic of Korea (2000 to 2015): a retrospective analysis

Background Owing to the continuous increase in the number of new human immunodeficiency virus (HIV) infection in Korea, public health centers (PHCs) have performed anonymous tests since 1989. No study has examined the patterns of anonymous HIV testing performed at PHCs and the characteristics of HIV infection detected in those tests. We aimed to assess the influence of anonymous HIV testing on Korea’s national HIV surveillance. Methods HIV screening test data from 253 PHCs over a 16-year period were classified into 13 groups based on reason for testing. For anonymous HIV test takers (Anonymous), the HIV positivity per 10,000 tests was calculated, as repetitions could not be distinguished. Those with suspected HIV infection voluntarily underwent HIV testing and revealed their identity (Suspected). HIV prevalence was calculated as the number of HIV-positive persons per 10,000 test takers. Analyses were performed using chi-square and Cochran-Armitage trend test with SAS 9.4. Results Approximately 400,000 HIV screening tests were performed at PHCs annually, which remained unchanged in the past 10 years. The proportion of anonymous testing increased from < 3.0% before 2014 to 4.8% in 2014 and 6.1% in 2015. While the number of HIV cases increased, the number of anonymous HIV-positive test results per 10,000 tests decreased from 68.8 in 2010 to 41.8 in 2015. The HIV prevalence among the suspected was approximately 20.0 per 10,000 test takers before 2014, which steeply increased to 71.6 in 2015. Those with suspected HIV were predominantly men, aged 20 years, foreigners, and metropolitan city dwellers in the last 6 years. The high prevalence of persons with suspected HIV resulted in a doubling of HIV prevalence at PHCs between 2014 and 2015. Conclusions Anonymous and Suspected, which were driven by similar motives, impacted each other. Increase in HIV prevalence among the suspected led to a higher HIV prevalence among all test takers in PHCs and higher proportions of HIV infection nationwide, which could be attributed to the increase in the number of anonymous tests performed in PHCs. HIV positivity among the anonymous and HIV prevalence among the suspected are key indexes of the national HIV surveillance in Korea.


Background
While new cases of human immunodeficiency virus (HIV) infection are declining globally, 36.7 million people were diagnosed with HIV, and around 1.8 million new cases are reported annually. Worldwide, 940,000 people died of acquired immune deficiency syndrome (AIDS) in 2017, indicating that the public health burden associated with HIV infection is extremely high [1]. In 2014, the Joint United Nations Program on HIV/AIDS set the 90-90-90 target, which should be achieved by 2020; an interim check on this goal in 2016 showed a value of 70-77-82. The first goal (90%) was related to the proportion of people with HIV who were aware of their HIV status; the first goal is linked to the second and third goals. Seven countries have currently achieved the first goal, including Malaysia and Thailand [2].
In Korea, since the first case of HIV infection was diagnosed in 1985, approximately 14,000 people have been diagnosed with HIV infections between 1985 and 2015, and the number of persons newly diagnosed with HIV infection continues to increase annually. The annual number of HIV cases across Korea has increased to 1000 cases since 2013. Over 99% of HIV infection cases in Korea are caused by sexual contact, with a male to female sex ratio of 9.7 to 1.0; of the total HIV-infected persons diagnosed in Korea, 10% are from foreign countries [3]. HIV infection is predominantly detected through HIV tests performed in hospitals, public health centers (PHCs), and blood banks [4]. In Korea, blood donors at blood banks and donors in hospitals undergo mandatory HIV testing prior to transfer of bodily fluids or body parts [5]. At risk groups for sexually transmitted infection (STI), including commercial sex workers, prisoners, and military enlistees undergoing medical exams, are routinely screened for HIV. Individuals comprising the general population voluntarily undergo HIV testing during routine health checkups. Approximately 10 million HIV screening tests are performed annually in Korea [6,7]. A total of 253 PHCs in Korea conduct approximately 400,000 HIV screening tests annually, accounting for 3-4% of the total proportion of HIV screening tests in Korea [8].
In a previous study on trends in HIV prevalence among overall visitors to PHCs from 2000 to 2009, we reported that the proportion of new cases of HIV infection detected at PHCs each year had increased to 18% of total new HIV infection cases [9]. Although the proportion of HIV screening tests performed at PHCs has recently decreased, the proportion of HIV cases detected at PHCs has increased to 20%. To identify the reasons for this increasing proportion of new HIV cases detected at PHCs, it is important to monitor the magnitude of the HIV testing and to examine the patterns of newly diagnosed cases of HIV infection at PHCs. The objective of this study was to determine the magnitude and changes in numbers of anonymous HIV test takers (the anonymous), which refers to individuals who voluntarily undergo HIV tests but do not reveal their identity, and analyze changes in anonymous HIV test patterns at PHCs. We then compared the trend of the anonymous with the trend of a similar group: those with suspected HIV infection who voluntarily underwent HIV testing and revealed their identity (the suspected).

Methods
HIV test data collection from the national HIV surveillance system at PHCs Since 1985, various HIV testing policies have been formulated for early HIV detection in Korea. Initially, government-led HIV testing policies mainly focused on mandatory screening for high risk HIV-infected groups. This was subsequently expanded to include mandatory testing in groups of workers sanitation related employment, such as food industry sanitation workers. In October 1989, an anonymous testing scheme was initiated at PHCs in Korea. However, since changes were made to the voluntary testing scheme as part of the HIV testing policy in 1998, changes have been observed in overall HIV testing patterns [10][11][12]. In March 2008, the Prevention Act of Acquired Immunological Deficiency Syndrome amended the legislation pertaining to an anonymous surveillance system for HIV/AIDS prevention, which was first enacted in 1987 [6]. Currently, the Korean government is formulating HIV-prevention policies to promote counseling and anonymous screening tests for early HIV detection.
PHCs are one of the main screening test sites for HIV in Korea, and refer HIV-positive samples, which were evaluated using enzyme-linked immunosorbent assay (ELISA), particle agglutination (PA), and rapid tests, to the local Institute of Health & Environment (IHE) centers for confirmation of HIV infection. These confirmatory tests were performed using antigen ELISA, antibody ELISA, PA, and Western blots. Confirmed results were reported to the division of HIV and tuberculosis (TB) control of the Korea Centers for Disease Control and Prevention (KCDC) and were recorded in the HIV database of the KCDC [13]. HIV test results were conveyed to the test taker directly at the PHC, by telephone, or through a web interface. Of note, the anonymous HIV test results were communicated to the test taker using a personal telephone access code.
PHCs manage the data related to HIV testing through an electronic medical records system, the Health Care Information System (HCIS). The HCIS program was initiated in 2000 and was expanded to all PHCs in 2005 [14]. We collected the HIV testing data from the HCIS of 253 PHCs between 2000 and 2015. The following data were collected: institutional code, specimen code, sex, year of birth, reason for HIV testing, HIV screening test results, differentiation code, and confirmatory HIV test code. If the result of a HIV screening test was positive, a confirmatory test code was assigned to the IHE center referral for confirmatory HIV testing. The confirmatory test code comprises the referral year, referral institution, and referral order. The differentiation code was a parameter used for the identification of a person's frequency of test taking within a year. In these codes, the use of personal information was avoided to preserve patient privacy. These data variables were accounted for in the previous manuscripts [9,14]. Data on HIV-positive cases were collected from the Division of HIV and TB control of the KCDC. These data included sex, year of birth, the reason for HIV testing, and confirmatory HIV test code. Data from HIV testing and data on the HIV-positive cases were matched using confirmatory HIV test codes. In the data recorded manually before the installation of the HCIS program, only the number of HIV tests without variables was obtained.

Participants
The reasons for HIV testing were classified into 13 groups according to occupation, health status, or motive for testing: health checkup, medical certificate, prenatal checkup, referral by a doctor, the suspected, tuberculosis patient, prisoner, partner of a person with HIV infection, commercial sex worker, bar employee, tea-room employee, massage parlor employee, and the anonymous [14]. The anonymous refers to individuals who voluntarily underwent HIV testing to evaluate their HIV status after risk events but did not reveal their identity. By contrast, the suspected refers to individuals with suspected HIV infection who voluntarily underwent testing only to detect HIV infection and revealed their identity. Data were collected on the anonymous tests performed from 2000 to 2015 and the 12 reasons for HIV testing. If the anonymous test takers were informed that their HIV test results were positive, their reason for HIV testing could be changed from "anonymous" to "suspected" to facilitate referral for public services at PHCs.

Statistical methods
The annual HIV prevalence among all HIV test takers except the anonymous was defined as the number of confirmed HIV cases per 10,000 HIV test takers, which were calculated using a differentiation code. The frequency of HIV testing for each test taker, that is, the number of repeat HIV testing per person, was measured every year. In case the test taker provided two or more reasons for undergoing HIV testing within 1 year, the person was classified into the reason for testing of the test that provided positive HIV results. If a test taker who cited two or more reasons had negative test results for all tests, the reason for testing was classified as follows: HIV infection suspected group (the suspected or referral by doctor), HIV test recommended group (patients with tuberculosis, prisoner), STI risk group (commercial sex worker, bar employee, tea-room employee, or massage parlor employee), or general group (health checkup, medical certificate, or prenatal checkup). To analyze the HIV prevalence, data on the anonymous were excluded, because the frequency of anonymous tests per person within 1 year could not be measured. Therefore, the annual HIV positivity of the anonymous was calculated and defined as the number of HIV-positive cases per 10,000 anonymous HIV tests [14].
The trends in HIV prevalence of HIV test takers and HIV positivity of the anonymous at PHCs were assessed through a series of cross-sectional annual analyses [9]. We analyzed HIV prevalence among the test takers by sex (male, female), age (< 20, 20-29, 30-39, 40-49, 50-59, ≤60 years), nationality (Korean, foreigner), and region (metropolitan cities, small cities, or rural areas). HIV positivity of the anonymous was analyzed only by region. Data with missing values (about 0.7%) were excluded from the analysis of HIV prevalence. To assess the difference in HIV prevalence by epidemiological variables, we conducted a multivariate logistic regression analysis (sex, age, nationality, and region). Sixteenyear trends in the HIV prevalence and HIV positivity of the anonymous were analyzed using a Cochran-Armitage trend test [15]. All statistical analyses were performed using SAS 9.4 and R. Ethical approval was obtained from the KCDC Institutional Review Board (approval no. 2016-07-06-PE-A).

Anonymous HIV testing status at PHCs
Approximately 400,000 HIV tests were conducted annually across 253 PHCs nationwide, and the HIV test scale remained unchanged during the last 10 years. Of the HIV tests performed at the PHCs, the annual proportion of HIV tests of the anonymous and the suspected was less than 10% until 2008; this value increased to 11-16% from 2009 (P < 0.001). The annual proportion of anonymous tests accounted for 1-3% of the total HIV tests performed at the PHCs; this value increased by 2-3 times in 2014 and 2015, to values of 4.8 and 6.1%, respectively. The annual proportion of HIV tests of the suspected increased from 5 to 13% from 2000 to 2013 and decreased to 7.8 and 8.1% in 2014 and 2015, respectively (Table 1).

Changes in HIV positivity among anonymous test takers from 2000 to 2015
The HIV positivity among the anonymous was higher than the prevalence among the suspected and total test takers from

Characteristics of HIV prevalence among persons with suspected HIV infection
To identify the characteristics of HIV prevalence among the suspected in 2014 and 2015, the relevant data were analyzed by sex, age, nationality, and region from 2010 to 2015. While the sex ratio of the suspected was generally balanced, the HIV prevalence within 6 years was higher in men than in women. HIV prevalence in men increased sharply in 2014 and 2015 (P < 0.001). From 2010 to 2015, the age distribution of the suspected was highest among those in their 20s and 30s, and the prevalence of this age group increased steeply across the six study years (P < 0.001). In 2014 and 2015, persons aged 20 years, 40 years, and over 80 years had higher HIV prevalence than other age groups. Among the suspected, 5000-6000 foreigners were tested in 2010 and 2011, but the prevalence of tested foreigners decreased in 2012; however, the HIV prevalence started increasing sharply in this population in 2013. Locally, the prevalence among the suspected from small cities or rural areas who underwent HIV testing was higher than those from

Discussion
Our results show that while the number of anonymous HIV tests increased in Korea, the HIV positivity among the anonymous decreased in 2014 and 2015. Additionally, while the number of the suspected who underwent HIV testing decreased, the HIV prevalence among the suspected increased in the last 2 years. In recent years, the total number of HIV tests performed in the anonymous and the suspected remained unchanged. However, the decision to undergo anonymous testing, which was driven by similar motives in both the anonymous and the suspected, impacted each other. The increase in HIV prevalence among the suspected led to the increase in the HIV prevalence among all test takers at PHCs, which are the primary testing sites of the national HIV surveillance system in Korea.
The number of individuals with newly diagnosed HIV infection has been continuously increasing in Korea, with more than 1000 HIV-infected people reported annually since 2013: 953 people in 2012, 1114 in 2013, 1191 in 2014, and 1152 in 2015. More than 25% of all newly diagnosed HIV infections in Korea were diagnosed through HIV screening tests at PHCs [3]. According to the results of this study, of the new HIV infection cases diagnosed at PHCs, more than 50% were observed in the suspected in 2008; this proportion increased to 71.3 and 77.9% in 2014 and 2015, respectively. Therefore, most of the HIV infections identified at the PHCs were diagnosed in the suspected. This was because the PHCs actively provided various opportunities for HIV testing, including free HIV testing and rapid diagnostic tests for early HIV detection.
The patterns of HIV testing and HIV detection continued to change over the time period of this study. The decrease in HIV positivity among the anonymous after 2009 was possibly due to the active HIV anonymous testing program. Anonymous tests have been conducted in hospitals in Korea since 2008, and the number of anonymous tests at PHCs increased after 2009. In addition, as a result of the expanded program for rapid diagnostic testing used to encourage people to undergo anonymous tests at PHCs, the number of anonymous HIV tests doubled in 2014 and 2015. In a pilot project on rapid diagnostic testing introduced at four PHCs in Seoul in 2014, the number of anonymous tests increased nine-fold and the number of confirmed HIV infection cases increased four-fold [16]. Since then, the number of PHCs that provided rapid diagnostic tests increased to 34; by 2015, the proportion of tests that were anonymous at PHCs in Seoul increased to 80% (a total of 22,114 tests). The proportions of HIV testing and HIV detections of the anonymous and the suspected in Seoul varied dramatically in 2014 and 2015. Due to activation of the anonymous     [17]. Therefore, for early HIV detection, the United Kingdom modified the environment or facilities of the examination clinics and improved accessibility to anonymous testing, as well as increasing education and promotional activities. These changes were attributed to a large increase in the number of anonymous HIV test takers who identified as male and homosexual. Consequently, new HIV infections among homosexuals decreased by 29% in London and by 11% outside London in 2015 [18]. In Korea, among men who had sex with men (MSM) and transgenders, HIV voluntary counseling and testing (VCT) to diagnose HIV infection early and to prevent HIV infection has been performed in Seoul and Pusan [19].
In Korea, foreigners accounted for about 10% of the total HIV-positive cases, and this proportion is increasing every year [3]. Our study showed that the prevalence of HIV among foreigners rapidly increased in 2014 and 2015 and was higher than the prevalence among Korean patients. The increase in number of foreigners diagnosed with HIV corresponds to an increase in multicultural families, foreign workers, and tourists. In Europe, a large number patients newly diagnosed with HIV infection were migrants from Africa and other countries [20].
This study has two limitations worth noting. First, as the HCIS was installed in all PHCs in 2005, there may be biases pertaining to the estimated proportions, HIV positivity among anonymous test takers, and the prevalence before 2005, as these were calculated based on data from the 72-210 PHCs with the HCIS. Second, repeat anonymous tests for anonymous test takers and HIV tests for those with suspected HIV were either performed in the same PHC or in other PHCs. It is possible that some test results of the anonymous and the suspected could be of duplicate cases that were included in our analysis.

Conclusion
This study demonstrated that changes in the national HIV testing strategy affected the status of HIV detection in Korea. In particular, the introduction of the rapid diagnostic tests at PHCs since 2014 attributed to the increase in the number of anonymous tests. People fear the stigma associated with HIV/AIDS, and they prefer to be tested for HIV infection anonymously, which provides quick results without revealing their identity.
As anonymous HIV testing has been expanded to medical institutes since 2008 and HIV self-testing has been performed since 2015, it is impossible to determine the status of anonymous HIV test across Korea. Therefore, HIV positivity among the anonymous and HIV prevalence among the suspected at PHCs are important indexes. Hence, these factors should be continuously monitored and improved in other sites. In the future, more than 90% of Koreans with HIV should be made aware of their HIV status through provision of easier access to anonymous testing. This will aid in the prevention of HIV infection spread in Korea and serve as a foundation to support the treatment of people with HIV infection who do not progress to AIDS through the administration of antiviral drugs and healthcare. Availability of data and materials Not applicable.
Ethics approval and consent to participate The Institutional Review Board of the Korea Centers for Disease Control and Prevention approved the conduct of this study (no. 2016-07-06-PE-A) and waived the requirement for a written consent.

Consent for publication
Not applicable.