Introduction

After more than 30 years of the HIV pandemic and great advances in HIV knowledge and antiretroviral treatment, the HIV epidemic continues to increase all over the world, causing a high number of deaths, especially in developing countries [1]. In Spain, the number of people living with HIV rises every year, as it is estimated that up to 18% of infected people are unaware of their own status [2]. These numbers are similar in other European countries [3]. In Western Europe, up to 45–50% of new HIV diagnoses are made late (with a CD4 count < 350 cells/mcl) and nearly 30% already having advanced disease at the diagnosis (< 200 CD4 cell/mcl) [4,5]. It is necessary to reduce the prevalence of people living with HIV who are unaware of their status, in order to bring down the incidence of new HIV infections [6,7].

The data above show that the current diagnostic strategies are not good enough. Unfortunately, the proven tools for HIV prevention are few [8] and difficult to implement, so we must focus on early diagnosis and early antiretroviral therapy (ART). An early HIV diagnosis implies not only personal benefits but also includes the community, since late diagnosis favors HIV transmission [9] and increases the use of resources and healthcare costs [10,11]. Maximum coverage of a screening program is essential to assure HIV “universal testing and universal treatment.” Some studies show that HIV-infected individuals who were diagnosed late presented to a healthcare facility multiple times in the two years prior to diagnosis, demonstrating a high percentage of missed diagnostic opportunities [12,13].

HIV testing guidelines stipulate different approaches, which vary from universal screening for people from 13 to 64 in all healthcare settings, unless participants decline (‘opt-out’ screening), to a targeted strategy in which HIV testing is recommended for individuals presenting with one of the indicator conditions [14–16]. Universal ‘opt-out’ testing strategies have been proposed to reduce the incidence of late diagnosis [17]. Although this non-targeted HIV testing strategy was shown to be cost-effective in populations with an HIV prevalence of at least 0.1% [18], the overall budget may be unaffordable for some countries. Furthermore, some studies exploring the routine ‘opt-out’ HIV screening strategy in Europe observed a low percentage of hidden HIV infection, as it cannot always identify early HIV infections [19]. In the United States, striking efforts have been made to establish non-targeted HIV screening in clinical practice. These operational programs have only been successful in some institutions with dedicated resources [20]; however, poor efficiency is expected in low prevalence areas [21].

The main area of discrepancy in HIV screening guidelines is the cost-effectiveness of routine testing among a population that is not known to be at higher risk [14,15]. Spanish guidelines recommend HIV testing in a variety of situations, with most implying previous risk of HIV exposure or indicator conditions, mainly but not exclusively in primary care centers [22]. Many studies have described results of HIV testing strategies in different healthcare scenarios, but few have face-to-face contact for comparison [23–25]. HIV testing in hospital emergency departments has been extensively evaluated [21,23], while sparse information is available to primary care centers or sexually-transmitted disease clinics [24,25]. Primary care centers are the health resource most frequently used by Spanish population to carry out HIV screening [26]. It is estimated that 85% of the population attend primary care every year. In addition, in this study screening has been implemented in the emergency department, since it is a good setting to perform screening in a large number of the population, especially in high risk individuals who are less likely to have consistent access to primary care, such as undocumented immigrants, drug users, and people with unstable housing [27].

The aim of our study is to evaluate the impact of a structured HIV-testing intervention (DRIVE program) in two clinical settings: a hospital emergency department and a primary care center in the same healthcare area, compared to HIV testing performed in routine clinical practice.

Methods

Results

DRIVE program: Baseline characteristics, testing coverage, and new HIV diagnoses

A total of 5,329 participants were included in the DRIVE program, of whom 2,684 (50.4%) were women. The median age was 37 (IQR: 28–47) years old. Spaniards made up 74.9% of participants, and 69.3% of HIV tests were performed in the primary care center. Only 4 participants were excluded: 2 came out as false positives, while 2 indeterminate. No cases were excluded because of invalid questionnaires. Three other participants were excluded because a prior diagnosis was identified. Table 1 shows the baseline characteristics of participants and the comparison between the emergency department and the primary care center.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 1. Participant characteristics according to the healthcare setting in the DRIVE program.

https://doi.org/10.1371/journal.pone.0220375.t001

All participants filled in the RE&IC questionnaire, and more than half (51.2%) reported at least one positive answer. Both men and younger individuals were more likely to present a positive result, as were those who had previous contact with primary care, patients already tested for HIV in the past, and participants included from the emergency department (Table 2).

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 2. Results of the RE&IC questionnaire in the DRIVE program, according to respondent characteristics.

https://doi.org/10.1371/journal.pone.0220375.t002

Twenty-two HIV-positive people were identified, which corresponds to an overall new HIV diagnosis rate of 4.1‰, with a higher percentage in the emergency department than the primary care center (8.6‰ vs. 2.2‰; p = 0.001). All of them had at least one positive answer on the RE&IC questionnaire. The new HIV diagnosis rate was higher among men (6.8‰ vs. 1.5‰; p = 0.002), and foreigners (7.5‰ vs. 3‰; p = 0.027). Individuals from Eastern Europe and previously HIV-tested participants were more likely to be diagnosed with HIV, as summarized in Table 3.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 3. Differences between participants according to HIV test results.

https://doi.org/10.1371/journal.pone.0220375.t003

Routine clinical practice vs. the DRIVE program: Testing coverage and new HIV diagnoses

The estimated HIV testing coverage was significantly higher in the DRIVE program than in routine clinical practice (7.17% vs. 0.96%, p < 0.001); it was better in the primary care center than in the emergency department, given the two strategies. The rate of new HIV diagnoses was also better in the DRIVE program (29.6 vs. 3.1 per 100,000 patients attended, p < 0.001), as shown in Table 4.

Download:

• PPT
  PowerPoint slide
• PNG
  larger image
• TIFF
  original image

Table 4. Estimated HIV testing coverage and new HIV diagnoses rates, according to the observation period and the healthcare setting.

https://doi.org/10.1371/journal.pone.0220375.t004

Discussion

The DRIVE study is one of the largest HIV testing programs in Spain over the last several years. It highlights the benefits of performing structured HIV screening, in terms of improved testing coverage, number of HIV diagnoses, and late diagnosis rates.

Testing for HIV is currently one of the main components of the strategies to control the HIV epidemic. The European region of the World Health Organization (WHO) is close to achieving the UNAIDS 90-90-90 target [31], with the main difficulty in terms of reaching the first 90 [32]. Therefore, the development of HIV testing programs is essential to meet the proposed objectives. During the first years of the epidemic, HIV-risk screening instruments [33] and self-reported sexual behavior risk factors [34] were evaluated, mainly for public health purposes. In recent years, HIV infection risk assessment has been more extensively studied. A score based on epidemiological characteristics and risk behaviors was shown to accurately categorize patients into groups with increasing probabilities of HIV infection [35], and was demonstrated to be strongly associated with new HIV diagnoses compared to non-targeted screening in the emergency department [23]. In line with the European Centre for Disease Prevention and Control (ECDC) guidance [36], some studies show that indicator conditions guiding HIV testing is an effective strategy to address the HIV epidemic in Europe as well as reducing the number of undiagnosed HIV infections [29,30]. Our questionnaire is the first tool to date that assesses the risk of HIV infection by combining the risk of exposure and indicator conditions, thus improving its accuracy [28]. According to UNAIDS estimations, 1.5 to 2.3 million people are living with HIV in Latin America [37], most of them Spanish-speaking. Spanish is used by 370 million people across the globe, but until our study, there was not a validated questionnaire in Spanish to assess the risk of HIV infection [28].

The DRIVE program, through the use of the RE&IC questionnaire, has achieved a testing coverage which is more than five times higher than non-targeted screening. Men, people aged 30–50 years, and participants with previous contact with primary care or who were previously tested for HIV were more likely to present a positive risk questionnaire. This is consistent with the population groups in which the epidemic is currently concentrated in our setting [5], and indicates that this tool could easily discriminate the patients who are most at risk according to the current guidelines [36]. Testing coverage in the DRIVE program was much higher in the primary care center than the emergency department. Primary care centers are most frequently used by the Spanish population to carry out HIV screening [26], and numerous studies agree about the high prevalence of missed opportunities for HIV testing in this setting [12,38]; implementation of HIV testing programs, such as DRIVE, would have a tremendous impact on screening rates and the reduction of missed diagnostic opportunities. A potentially useful strategy to increase coverage in emergency room scenarios might be the automatic implementation of the questionnaire within the electronic medical screening system, as it has been done in previous studies showing good results [23].

The use of rapid tests and nurse participation has facilitated the development of the screening; as such, immediate results are obtained, counseling is provided, and patients are referred to the HIV Infectious Diseases Service in a single visit. The implementation of HIV early diagnosis programs through the use of rapid tests in primary care centers has been a pioneering intervention in Spain [39]. These strategies promote the screening of high percentages of the population at risk. Previous studies demonstrated a similar diagnostic capacity for rapid tests (if performed correctly) and conventional diagnosis by ELISA and Western Blot. Virtually all blood rapid diagnostic tests reach a sensitivity and specificity close to 100%, although erroneous results may still appear [40].

The DRIVE program achieved a rate of new diagnoses seven times higher than in routine clinical practice, with a percentage of new HIV diagnoses higher than expected, according to the prevalence of undiagnosed HIV infection reported in previous studies in our healthcare area [41]. New HIV diagnoses were more frequent among men, Eastern Europeans (in line with the current epidemiological data [1]), and people previously tested for HIV. The latter may indicate a greater self-awareness of the risk of HIV infection, or an increased detection of risk in this population by healthcare providers. The rate of new HIV diagnoses out of 1000 tests performed was higher in the emergency department, which may be related to the greater risk of HIV exposure and indicator conditions reported by participants in this setting. This data corroborates that emergency departments should not be excluded from HIV screening programs, as indicated in the CDC testing guidelines [36].

The higher HIV infection prevalence in certain population groups reflects an ongoing stereotype about at-risk profiles; however, we should not neglect the risk that also exists in other groups, such as women and the elderly, where higher percentages of late diagnosis are observed [42]. In our study, heterosexual men and women were diagnosed with a lower CD4 cell count than MSM, although the small sample size has not made it possible to obtain statistically significant differences. Late diagnosis was lower in our structured testing program than in routine clinical practice, indicating that broad screening could improve current rates of late diagnosis, which reach almost 50% in Europe [5]. Regarding the site of inclusion in the study, the emergency department represents the most common facility for late diagnosis in our cohort, which has been shown in other studies that place the emergency department as the site where diagnostic opportunities are most frequently missed [43]. As such, both the emergency departments and primary care centers are healthcare settings where more effort should be made to improve HIV screening, as indicated by international testing guidelines [36].

Our study has some limitations. It is a prospective observational single-center study, with retrospective control arm, in which external resources were used to screen for HIV infection. This could decrease its external validity, so further research is necessary to assess the applicability of the screening program in daily life.

In conclusion, the development of a simple screening program, through the use of a RE&IC questionnaire an rapid HIV tests, could achieve an increase in the rates of HIV testing, an increase in new HIV diagnoses, as well as an improvement in the current rates of late diagnosis. This strategy might be a useful tool for decreasing undiagnosed HIV infection and ultimately mitigating the epidemic.