Late Diagnosis and Entry to Care after Diagnosis of Human Immunodeficiency Virus Infection: A Country Comparison

Background Testing for HIV infection and entry to care are the first steps in the continuum of care that benefit individual health and may reduce onward transmission of HIV. We determined the percentage of people with HIV who were diagnosed late and the percentage linked into care overall and by demographic and risk characteristics by country. Methods Data were analyzed from national HIV surveillance systems. Six countries, where available, provided data on two late diagnosis indicators (AIDS diagnosis within 3 months of HIV diagnosis, and AIDS diagnosis within 12 months before HIV diagnosis) and linkage to care (≥1 CD4 or viral load test result within 3 months of HIV diagnosis) for people diagnosed with HIV in 2009 or 2010 (most recent year data were available). Principal Findings The percentage of people presenting with late stage disease at HIV diagnosis varied by country, overall with a range from 28.7% (United States) to 8.8% (Canada), and by transmission categories. The percentage of people diagnosed with AIDS who had their initial HIV diagnosis within 12 months before AIDS diagnosis varied little among countries, except the percentages were somewhat lower in Spain and the United States. Overall, the majority of people diagnosed with HIV were linked to HIV care within 3 months of diagnosis (more than 70%), but varied by age and transmission category. Conclusions Differences in patterns of late presentation at HIV diagnosis among countries may reflect differences in screening practices by providers, public health agencies, and people with HIV. The percentage of people who received assessments of immune status and viral load within 3 months of diagnosis was generally high.


Introduction
Diagnosis of human immunodeficiency virus (HIV) infection soon after acquisition and prompt entry into care are part of national strategies in most high-income countries for minimizing ongoing HIV transmission. People who present with advanced HIV disease have higher morbidity and mortality than those diagnosed early and can unknowingly transmit HIV for a longer period of time [1,2]. HIV-infected people who engage in care soon after initial HIV diagnosis have a lower risk for premature mortality and are more likely to achieve viral suppression and lower viral load burden [3,4]. People who do not engage in care may have relative poor treatment adherence [5,6], resulting in reduced viral suppression and increased secondary transmission [7][8][9][10][11][12]. Early initiation of care is also essential to provide opportunities for additional activities for HIV prevention through health care, including screening and counseling for risk behaviors [13].
In high-income countries, about a quarter to a third of people with HIV are diagnosed late, with a CD4+ T-lymphocyte count of ,200 cells/mL or an illness that defines the acquired immunodeficiency syndrome (AIDS). However, this may vary by country and by population subgroup. In Europe, overall about 29% of people are diagnosed with late stage HIV disease [14]; with a range of 23-45% of people diagnosed late across countries [15,16]. Estimates from other high-income countries are 20% in Australia [17], 28% in the United States [18], 17% in British Columbia, Canada [19], 30%-37% in Spain [20,21], 30% in France [22], and 28-39% in Alberta, Canada [23,24]. Late presentation differs by risk group, generally with a lower percentage diagnosed late among men who have sex with men (MSM) than people who inject drugs (IDU) or heterosexuals [16][17][18]21,22,[25][26][27], and by age, with higher percentages presenting late among older age groups [15,18,21,26,27]. The percentage of people diagnosed with HIV who are linked to care may also vary by country and population subgroups. However, little population-based information is available on linkage to care, which is generally defined as evidence of having at least one CD4+ T-lymphocyte count or viral load test result within 3 months of HIV diagnosis. In the United States, an estimated 77% of people diagnosed with HIV are linked to care; however, this varies by race/ethnicity, age, and transmission category [18,28]. A study in Ontario, Canada, found that 64.1% of people were linked to care within 3 months following diagnosis, and 75.2% were linked to care within 12 months [29]. A study in Northern Alberta, Canada, found that 92% of individuals diagnosed with HIV infection accessed care following their diagnosis; 70% within 2 months of diagnosis [24]. Both Canadian studies also found that time to care varied by race/ethnicity, age, exposure category, and geography.
Differences in late diagnosis with HIV disease or linkage to care between countries may be due to differences in individual and system level factors such as health care delivery and access to care. Differences may also result from lack of uniformity in definitions for late diagnosis or linkage to care, or from data based on particular subpopulations. We analyzed data from the national HIV surveillance systems of Australia, Canada, France, Italy, Spain, and the United States and used uniform definitions to determine the percentage of people with HIV who were diagnosed late and the percentage linked into care overall and by demographic and risk characteristics.

Methods
HIV surveillance data can be used to monitor stage of disease at HIV diagnosis and care visits by examining CD4 count and viral load data for HIV-infected people. We used data from HIV surveillance systems in Australia, Canada, France, Italy, Spain, and the United States to determine the percentage of people diagnosed late with HIV and who were linked to care after HIV diagnosis as measured by CD4 count and viral load (VL) tests overall and by demographic and risk characteristics. The countries included in the analyses agreed to participate among member countries from Europe, North America, and Australia that participate in an informal annual meeting to discuss HIV surveillance. Each country maintains a national surveillance system with country-specific reporting rules and representation of the population.
National HIV surveillance is coordinated in Australia by the Kirby Institute, in collaboration with the State and Territory health authorities, the Australian Government Department of  [23,26] or a combination of a CD4 value of ,200 cells/mL or an opportunistic illness [18]. Other definitions have been proposed, such as a definition of very late diagnosis for people with CD4,200 cells/mL and late   diagnosis with CD4,350 cells/mL based on treatment recommendations in 2010 [1,34]. In the United States, recent guidelines recommend all people with HIV be offered treatment [35,36]. For this analysis, we defined late diagnosis as a CD4,200 cells/mL or presence of an opportunistic illness within 3 months of diagnosis among people newly diagnosed with HIV. We also assessed late diagnosis with a second definition because not all countries had data available to meet the first definition: percentage of people who had an HIV diagnosis within 12 months before AIDS diagnosis, among people diagnosed with AIDS in 2010. Entry to care was defined as $1 CD4 or viral load test within 3 months of diagnosis; in Spain only data on CD4 test were used since information on viral load test within 3 months of diagnosis is not available.
We determined late diagnosis and linkage to care overall and by age, sex, and transmission category. Analyses were conducted by the respective country HIV surveillance program and data were provided where available and for the most recent year available (2009 or 2010).

Results
Late Diagnosis (CD4,200 Cells/mL or Presence of an Opportunistic Illness within 3 Months of Diagnosis) among People Newly Diagnosed with HIV (Table 1) Information on late diagnosis was available from Australia, Canada, France, Italy, and the United States. In each country, the majority of diagnoses of HIV infection were among men and among people aged 20-49 years old (Table 1). In Australia, Canada and the United States, the largest number of people diagnosed with HIV reported their exposure to HIV as male-tomale sexual contact; while in France and Italy a higher number of infections was attributed to heterosexual contact. The highest percentage of people diagnosed with late stage disease was among people in the United States (28.7%), followed by Australia (18.8%), France (15.2%), Italy (14.5%), and Canada (8.8%). Late diagnosis was generally higher among older age groups. Late diagnosis was either similar for males and females (Italy) or higher among males (Australia, Canada, France, United States). The pattern of late diagnosis by transmission category differed between countries: compared to MSM, the percentage diagnosed late was higher for people with infection attributed to heterosexual contact in Australia, France, and Italy (absolute difference $5%), and among IDU in France and the United States.
Late Diagnosis (HIV Diagnosis within 12 Months before AIDS Diagnosis) among People with AIDS (Tables 2 and  3) Data were available from all 6 countries on the percentage of people diagnosed with HIV within 12 months of AIDS diagnosis (Tables 2 and 3). The majority (53-65%) of people diagnosed with AIDS were diagnosed with HIV in the 12 months prior to their AIDS diagnosis. There were no clear trends by age and compared to females, males had higher percentages diagnosed within 12 months in all countries except Australia. Compared to MSM, a higher percentage of people diagnosed with AIDS whose infection was attributed to heterosexual contact had their first HIV diagnosis in the 12 months prior in Australia, Canada, France, and Italy and percentages were lower among IDU compared to MSM in all 6 countries.
Linkage to Care (Table 4) Data were available on linkage to care from Australia, Canada, Italy, Spain, and the United States (Table 4). Overall linkage was high, with 89.9% linked to care in Australia, 89.6% in Italy, 80.3% in the United States, 76.0% in Spain, and 72.6% in Canada. Linkage to care was lower among adolescents and younger adults compared to older adults in Australia, Canada, and the United States. Compared to MSM, a greater percentage of IDU were linked to care in Australia; while fewer were linked in Spain, Italy, and Canada. The percentage linked to care was lower among people with HIV infection attributed to heterosexual contact in Australia and Canada, compared to MSM.

Discussion
The percentage of people presenting with late stage disease at HIV diagnosis varied by country, overall with a range from 28.7% (United States) to 8.8% (Canada), and by transmission categories. The percentage of people diagnosed late was lower in several countries than previously reported (e.g., Canada) [24,23]. The percentage of people diagnosed with AIDS who had their initial HIV diagnosis within 12 months before AIDS diagnosis varied little among countries, except the percentages were somewhat lower in Spain and the United States. Overall, the majority of people diagnosed with HIV were linked to HIV care within 3 months of diagnosis (more than 70%), but varied by age and transmission category. Improvements in detecting HIV earlier and differences in patterns of late presentation at HIV diagnosis among countries may reflect differences in screening practices by providers, public health agencies, and people with HIV.
Screening recommendations, screening patterns independent of screening recommendations, and access to health care influence early detection of HIV infection. Some countries have shifted from HIV testing based on risk assessment to HIV screening in certain settings. For example, in 2006 the U.S. CDC recommended routine HIV screening in clinical settings where HIV prevalence is .0.1 percent and expanded support to state health departments for HIV testing [37]. However, the percentage diagnosed late was higher in the United States compared to the other countries included in this analysis. Another factor that can impact access to care is lack of health insurance, with 32% of Latinos, 21% of blacks, and 16% of whites being uninsured in the United States [38]. Generally, MSM test more frequently and have lower percentages of late presentation. The higher percentage presenting late for people with infection attributed to heterosexual contact in Australia, France, and Italy may be related to lack of perceived risk or a denial of risk and therefore failure to test for HIV, or a higher percentage of foreign-born people. The higher percentage presenting late among IDU in the United States may be due to not accessing health care due to lack of insurance (United States), social disadvantage, lack of social/familial support, or stigma associated with drug use behaviors. The situation is different in France compared with the United States. Most people who use intravenous drugs in France have been infected with HIV in the 1990s and most have been diagnosed many years ago [39]. Very few people who use intravenous drugs were newly diagnosed with HIV, that is less than 100 cases were diagnosed during the analysis year, representing less than 1% of all new HIV cases, and half of these cases were immigrants. Moreover, the second indicator used in this analysis (HIV diagnosis within 12 months before AIDS) shows that only 28% of IDU in France were diagnosed with HIV within 12 months before AIDS diagnosis, a low percentage compared to other categories (70% among heterosexuals, 64% among MSM).
Similar factors may affect linkage to care among IDU (e.g., in Canada and Spain, fewer IDU were linked to care compared with MSM). Effective drug treatment programs may include HIV prevention services, such as HIV testing and linkage to care. Similar to earlier findings from North America [22,28,29], in most countries a lower percentages of younger people was linked to care soon after diagnosis compared to older people. Younger people may feel healthy and therefore may not feel the need to engage in care soon after diagnosis. In general, people diagnosed with HIV should receive an initial assessment of immune system status and viral load promptly after diagnosis [35,36], and with guidelines now recommending treatment be offered to people with less severe disease (e.g., WHO guidelines recommend treatment at CD4 count #350 cells/mm 3 ) [40] prompt linkage to care could improve worldwide.
Because not all countries could provide data on stage of disease at diagnosis, we also assessed AIDS diagnosis in relation to the time of HIV diagnosis (Tables 2 and 3). This measure reflects people with AIDS who may have been recently diagnosed or people who were diagnosed with HIV years before their AIDS diagnosis and who may or may not have been on treatment. While this measure is not directly comparable to the late diagnosis measure presented in Table 1, and does not reflect people who may have developed AIDS but were promptly treated, it can indicate a need for earlier detection and prompt treatment.
Our analyses are subject to several limitations. Reporting of HIV diagnoses and HIV-related laboratory test results to HIV surveillance differ between countries. For some countries information was not available for the entire population. Classification of severe stage of HIV (i.e., AIDS) may also differ between countries, depending on whether only CD4 test results are considered and/or opportunistic illnesses. We assumed that people who did not have a CD4 count did not have AIDS since if they had severe disease, they would have been diagnosed with AIDS. Our results need to be interpreted in light of the treatment guidelines and health care systems of the respective countries. We measured linkage to care based on CD4 and viral load test results; we did not have information on clinic visits that may not have resulted in such testing. Incomplete reporting of test results may have underestimated linkage to care.
In summary, the percentage with advanced disease at HIV diagnosis and linkage to care vary between the high-income countries included in this analysis. In general, MSM are diagnosed earlier than people with infection attributed to other risk factors. The percentage of people who received assessments of immune status and viral load within 3 months of diagnosis was generally high. Future analyses may need to explore factors within countries that promote higher benchmarks on these indicators and support early diagnosis and prompt linkage to care.