Length of stay in Denmark before HIV diagnosis and linkage to care: a population-based study of migrants living with HIV, Denmark, 1995 to 2020

Background Migrants face an increased risk of HIV infection and late presentation for HIV care. Aim To examine delays in HIV diagnosis, linkage to care (LTC), and risk of late presentation for migrants living with HIV in Denmark. Methods We conducted a population-based, nationwide study of adult migrants (n = 2,166) presenting for HIV care between 1 January 1995 and 31 December 2020 in Denmark. Time from immigration to HIV diagnosis and from diagnosis to LTC, and late presentation were assessed, stratified by migrants’ geographical regions of origin, using descriptive statistics. Results The demographics of the migrant population changed over time. Overall, migrants diagnosed with HIV after immigration to Denmark resided a median of 3.7 (IQR: 0.8–10.2) years in Denmark before diagnosis. Median time from HIV diagnosis to LTC was 6 (IQR: 0–24) days. Migrants diagnosed with HIV infection before immigration had a median of 38 (IQR: 0–105) days from arrival in Denmark to LTC. The corresponding median times for 2015–20 alone were 4.1 (IQR: 0.9–13.1) years, 0 (IQR: 0–8) days, and 62 (IQR: 25–152) days, respectively. The overall proportion of late presentation among migrants diagnosed with HIV after immigration was 60%, and highest among migrants from sub-Saharan Africa and East and South Asia. Conclusion HIV diagnosis is still substantially delayed in Danish migrants, while LTC is timely. The proportions with late presentation are high. These results call for targeted interventions to reduce the number of migrants with undiagnosed HIV infections and of late presenters.


Introduction
Migrants in Europe are disproportionately affected by the HIV epidemic [1,2] and are recognised by the United Nations (UN) as one of the most vulnerable populations for HIV acquisition [1]. In Europe, migrants constitute an estimated 10% of the population, a proportion which is growing [3]. In Denmark, which has an estimated prevalence of HIV infection of 0.11%, migrants comprised 52% of incident cases presenting for HIV care in 2019 [4], whereas they constituted 10% of the Danish population [5]. Many migrants originate from countries with a high prevalence of HIV infection. Several other factors, which reflect the socioeconomic status of migrants before, during, and after migration, may also contribute to migrants' excess vulnerability to HIV acquisition [1,6,7].
Migrants in Europe [2,8] and Denmark [9] face not only an increased risk of being infected with HIV, but also of presenting late for HIV care. Several studies have shown how barriers such as legal status and fear of legal repercussions, racism, and stigma may prevent migrants living with HIV (MLWH) from seeking timely HIV care [10][11][12]. Late presenters (LPs) face increased risk of adverse outcomes, including increased shortterm mortality, related to their HIV infection [8,13] and miss the health benefits of the recommended early initiation of antiretroviral therapy (ART) [14]. Thus, the medical consequences of late presentation exacerbate inequality in healthcare. Late presentation also presents a challenge for public health efforts to use treatment as prevention of new HIV infections [15,16].
Timely HIV care starts with timely HIV diagnosis, as described in the HIV Continuum of Care [17]. A diagnosis of HIV infection, however, does not guarantee linkage to care (LTC). The HIV epidemic among MSM (men who have sex with men) in Denmark has been thoroughly mapped by previous research from the Danish HIV Cohort [18,19]. However, there is lack of information about the dynamics of the HIV epidemic and the HIV continuum of care among MLWH, and about how it has changed over the past 25 years, where there have been considerable changes in migration dynamics, policies, diagnostic practices, and access to care.
Internationally, little work has been done to study time from migration to diagnosis and LTC among MLWH [20,21]. Determining this is essential to understand how delays in the HIV continuum of care after immigration may contribute to late presentation among MLWH. Such knowledge could guide targeted interventions to reduce the number of undiagnosed and untreated patients with HIV infections.
We aimed to determine the time from immigration to diagnosis of HIV infection and from HIV diagnosis to LTC and to examine late presentation, stratified by migrant geographical region of origin and calendar period, in Denmark over a 25-year period.

Methods
We conducted a nationwide, population-based, retrospective, observational study of all MLWH registered in the Danish HIV Cohort Study (DHCS) between 1995 and 2020.

Setting and testing strategies
In Denmark, all HIV-infected individuals are referred to one of eight specialised HIV healthcare centres, where they are seen on outpatient visits at proposed 12 to 48-week intervals. HIV testing, care and ART are offered free of charge to all individuals living with HIV who have legal residence in Denmark. The Box shows an overview of the Danish HIV testing strategies [9,22].

Data sources
The DHCS, described in detail elsewhere [23], is a nationwide, population-based cohort study of all individuals living with HIV treated at Danish HIV healthcare centres since 1 January 1995. Individuals are continuously enrolled. Data including demographics, date of HIV diagnosis, first contact with a Danish HIV centre, AIDS-defining events, and ART are updated yearly.
The Danish Civil Registry System (CRS), which contains demographic data and vital status on all Danish residents, was accessed to obtain information on

Box
The four primary HIV testing strategies, Denmark, 1995-2020  AIDS: acquired immunodeficiency syndrome; IQR: interquartile range; LTC: linkage to care; NA: not applicable. All characteristics were collected at date of LTC. a Missing information on sex was ≤ 4 (≤ 0.2%) in SSA and in Other. b Data were missing for some MLWH for the following variables: HIV-RNA (n = 179) and CD4 + T-cell count (n = 93) and were excluded from analyses of HIV-RNA, CD4 + T-cell count and late presentation/advanced HIV disease. c CD4 + T-cell normal range: 500-1,500 cells/µl. d Presenting with CD4 + T-cell count < 350 cells/µl or with an AIDS-defining event regardless of CD4 + T-cell count. e Presenting with CD4 + T-cell count < 200 cells/µl or with an AIDS-defining event regardless of CD4 + T-cell count. f Data were missing for some MLWH for the following variables: HIV-RNA (n = 85) and CD4 + T-cell count (n = 99) and were excluded from analyses of HIV-RNA and CD4 + T-cell count.  immigration and emigration dates and countries of origin. The Danish National Hospital Registry (DNHR), containing information on all non-psychiatric hospital admissions since 1977 and all outpatient visits since 1995, was used to obtain information on dates of first visits to specialised infectious diseases departments. Data were linked through the unique personal identification numbers given to all residents in Denmark.

Study population
We included all individuals with HIV-1 in DHCS who (i) immigrated to Denmark and obtained residence, (ii) were aged 16 years or older at time of first contact with an HIV healthcare centre and (iii) had first contact with a Danish HIV healthcare centre between 1 January 1995 and 31 December 2020. Undocumented migrants are not registered in the CRS and were not included.

Definitions
Data on country of origin of MLWH were obtained from the CRS. We categorised MLWH into five geographical regions of origin based on the United Nations Standard country and Area codes classifications [24] According to the European consensus definition [25], late presentation (LP) was defined as presenting for HIV care with a CD4+ T-cell count < 350 cells/ µl or presenting with an AIDS-defining event regardless of CD4+ T-cell count. Presentation with advanced HIV disease (AHD) was defined as presenting with a CD4+ T-cell count < 200 cells/µl or with an AIDS-defining event regardless of CD4+ T-cell count [25]. Presenting with AIDS was defined as having developed an AIDSdefining event no later than 30 days after LTC. Viral suppression was defined as a VL < 200 copies/ml and was used as a marker for being on ART.

Outcomes
For MLWH diagnosed with HIV infection after immigration to Denmark, we examined time from immigration to HIV diagnosis and time from HIV diagnosis to LTC. For MLWH diagnosed before immigration, we examined time from the first date of immigration following HIV diagnosis to LTC. We examined the proportions of LP and AHD.

Statistics
All results were stratified according to MLWHs' status as either diagnosed after or before immigration to Denmark.
MLWH could have several migrations in and out of Denmark before HIV diagnosis or LTC, and thus more than one period of residence in Denmark. Only time spent residing in Denmark contributed time to the outcomes.
We stratified our analyses according to calendar period and region of origin.

A. Years from immigration to HIV diagnosis for migrants diagnosed after immigration (n = 1,514)
C. Days from immigration to linkage to care for migrants diagnosed before immigration (n = 652)

B. Days from HIV diagnosis to linkage to care for migrants diagnosed after immigration (n = 1,514)
a A total of 212 migrants were registered with a date of immigration which occurred after date of linkage to Danish HIV care. Their time from immigration to linkage to care was defined as 0 days.

Graphs show median time (points), overall median (dashed line) and interquartile range (whiskers). Migrants' region of origin was defined by
United Nations standard country and area codes classifications [24].

Migrants diagnosed with HIV infection after immigration
Among MLWH diagnosed with HIV after immigration, 830/1,514 (55%) were men and the median age at LTC was 33 (IQR: 28-40) years (Table 1). The demographic and HIV-related characteristics changed over time.
From 1995-99, MLWH migrated predominantly from SSA (52%), 50% were women, and a high proportion (70%) reported heterosexual transmission. Over time, the regions of origin became more diverse, with an increasing proportion of MLWH from eastern Europe. More MLWH were men (68%) and MSM was more frequently reported as the route of infection (40%) ( Table  2). In general, MLWH originating from SSA and East and South Asia were more frequently women and heterosexually infected and had lower CD4+ T-cell counts at LTC (< 300 cells/uL), whereas MLWH from Western countries and eastern Europe were predominantly men and MSM and had higher CD4+ T-cell (> 300 cells/uL) counts at LTC (Table 1).

Time from immigration to HIV diagnosis and from HIV diagnosis to LTC
The median time from immigration to HIV diagnosis was 3.7 (IQR: 0.8-10.2) years. There were notable geographical differences: time from immigration to HIV diagnosis was 1.9 (IQR: 0.4-5.7) years for MLWH from SSA and 8.0 (IQR: 2.5-17.8) years for MLWH from Western countries (Figure 1).
The proportion of MLWH diagnosed with HIV less than 1 year after immigration to Denmark ranged from 39% among MLWH from SSA to 14% among MLWH from Western countries (Figure 2).
For the total population (Figure 3), as well as for MLWH from SSA and East and South Asia specifically (Supplementary Figure S1: Years (median, IQR) from immigration to HIV-diagnosis for MLWH from Sub-Saharan Africa and East and South Asia stratified by calendar period), time from immigration to HIV diagnosis was stable throughout the study period. In all analyses, we found a greater range in the duration of time from immigration to HIV diagnosis in the later calendar periods ( Figure 3A).  Table S1: Late presentation and presentation with advanced HIV disease among migrants from Sub-Saharan Africa and East and South Asia stratified by calendar periods).

Time from immigration to HIV diagnosis for late presenters and migrants presenting with advanced HIV disease
For LPs and MLWH with AHD, the median times from immigration to HIV diagnosis were 3.5 (IQR: 0.9-9.3) and 3.9 (IQR: 1.0-9.4) years. Time from immigration to HIV diagnosis was stable over calendar time (Supplementary Figure S2: Years (median, IQR) from immigration to HIV-diagnosis for late presenters, stratified by calendar period).

Migrants diagnosed with HIV infection before immigration
Among MLWH diagnosed with HIV infection before immigration, 361/652 (55%) were men and the median age was 33 (IQR: 28-40) years. Overall, SSA was the predominant region of origin (Table 1), but over time, Western countries became the predominant region of origin ( Table 2). The proportion of men also increased over time. The demographic characteristics for each region resembled those of MLWH diagnosed after immigration, as described above

Discussion
In this nationwide study, we demonstrated that MLWH had resided a median of 3.7 years in Denmark before diagnosis of HIV infection with no change over time from 1995 to 2020. However, regardless of whether migrants were diagnosed with HIV infection in Denmark or arrived in Denmark with a known HIV diagnosis, linkage to care was timely for a high (62%-95%) proportion of MLWH. Consistent with other studies [9,26], we found high proportions of LP and AHD among MLWH diagnosed after immigration. In comparison, the proportion of LP among Danish-born individuals has been estimated to be 42% [9].
Although the median time between immigration and diagnosis remained unchanged, the proportions of LP, AHD and AIDS at time of LTC decreased in later calendar periods. Consistent with this, the CD4+ T-cell counts at LTC were higher in the later calendar periods. These changes mainly reflect that the composition of the migrant population changed over time, with fewer migrants from SSA following increasingly restrictive migration policies in Denmark [27]. Furthermore, the global HIV pandemic has changed following the increased focus on early and regular HIV testing and immediate ART provision. The proportion of MLWHespecially from Western countries -who were already diagnosed with HIV and on effective ART at time of immigration increased over time. This might explain why time from immigration to LTC increased over time, as MLWH already on ART at immigration might bring medicine and have no need of immediate healthcare.
In line with other studies [9], our findings illustrate that MLWH constitute a heterogenous population with different trends across geographical regions of origin. In addition, the interpretation and clinical implications of measuring time from immigration to HIV diagnosis are different for subpopulations of MLWH. We presume that MLWH from the high prevalence regions of SSA and East and South Asia were mainly infected with HIV before immigration to Denmark [7,9,20] and, consequently, most of the time spent in Denmark before HIV diagnosis reflected a diagnostic delay. In both subpopulations, we found consistently high proportions of LP and AHD. This makes it crucial to reduce the diagnostic delays, which even during 2015-20 exceeded 2 years for MLWH from SSA and 4 years for MLWH from East and South Asia. Other studies have also identified migrants from SSA and East and South Asia to be at high risk of late presentation [2,8,9], and modelling studies have identified risk of delays in the HIV continuum of care for these migrant subpopulations [20,28]. We saw a high correlation between geographic region of origin and sex. MLWH from SSA and East and South Asia were more likely to be women, and in line with this, we found high proportions of LP and ADH among women.
Compared with migrants from SSA and East and South, MLWH from Western countries are more likely to become infected with HIV after immigration, especially among MSM [7]. Thus, for migrants from Western countries, the observed large dispersion and sometimes long time periods between immigration and HIV diagnosis probably represent a combination of diagnostic delays and post-migration HIV transmission. As a consequence, healthcare efforts towards this migrant subpopulation should focus on prevention of new infections [7] in addition to continuous HIV testing efforts.

Figure 3
A. MLWH from eastern Europe constituted an increasing proportion of the total population of MLWH over calendar time. Eastern Europe is one of the few regions where the HIV epidemic is growing with a high incidence:prevalence ratio of 10% [29]. Our results underline that the risk of HIV infection among key populations in eastern Europe should be targeted not only in the region but also recognised among migrants from the region.
Over time, the proportion of MSM MLWH increased. This reinforces that HIV prevention and testing efforts should ensure initiatives aimed at MSM migrants.
Healthcare is free of charge in Denmark, and healthcare efforts have been very effective in reducing the HIV epidemic among Danish-born MSM [18,19,30]. Nevertheless, we still face a problem with migrants' access to HIV testing. It is essential to make anonymous, voluntary HIV testing available free of charge to everyone, both at a community level and at primary healthcare providers, and to ensure provider-initiated (voluntary) testing, so that migrants who do not perceive themselves at risk of HIV are still offered tests. However, several studies have shown how structural barriers prevent migrants from seeking timely HIV diagnosis and care [10][11][12]. Furthermore, a recent study has pointed to many missed opportunities for provider-initiated HIV testing at contact with the primary healthcare system for people later diagnosed with HIV in Denmark [31]. Our study demonstrates that the barriers to HIV testing in combination with missed testing opportunities potentially cause diagnostic delays of several years.
Other studies have also shown how migrants are at increased risk of late presentation [2,8,9], our study further elucidates that while some MLWH arrive as LPs, others likely become LPs after immigration because of diagnostic delays. This implies that increased and diverse HIV testing efforts in the migrants' country of arrival, combined with the breakdown of barriers, have the potential to reduce morbidity and mortality associated with presentation with advanced HIV disease.
This population-based study provides new insights into the extend of the delays in the first steps in the HIV continuum of care among migrants. The strengths of our study include the population-based and nationwide design. In contrast to another Danish study, which only included refugees and family-reunified immigrants [9], our study included all migrants with residence in the country. The use of a population-based clinical database linked to Danish health and administrative registers allowed us to trace patients retrospectively from date of LTC without recall bias. Although this work was planned as a descriptive study and we did not include a multivariate analysis, the large population enabled us to stratify our analysis according to region of origin, calendar period and late presentation.
Our study had some limitations. The study design and number of migrants did not allow us to take socioeconomic differences within regions and migrant status into account. CRS registers arrival in Denmark as the date of obtaining official residence, which means we were unable to study undocumented migrants living in Denmark without residence and with limited access to healthcare. This may have led us to underestimate the true delay in HIV diagnosis among MLWH. We were unable to take post-migration HIV transmission fully into account, and the assumption that migrants from high prevalence regions are mostly HIV-infected before immigration may not always be correct [32]. Our study design only allowed us to include MLWH once they were linked to care. The number of MLWH diagnosed with HIV but never linked to care is most likely very small. It is mandatory to report HIV infections, and HIV care is very structured in Denmark. However, presumably, a number of MLWH with short stays in Denmark have maintained contact with the healthcare system in their home country, and consequently their HIV status would not be registered in Danish national registries.

Conclusion
This nationwide study of MLWH in Denmark demonstrates that, while almost every aspect of HIV healthcare has improved greatly over the last decades, we still face the challenge that many MLWH, including those with late presentation and presentation with advanced HIV disease, have resided years in Denmark before being diagnosed with HIV infection. These results point to missed opportunities for both detection and reducing late presentation among MLWH in Denmark. Our findings call for an increased effort to systematically implement voluntary HIV testing, and to make it more accessible for migrants -especially for those arriving in Denmark from high HIV prevalence areas and for women. Work should be done to investigate and combat the barriers preventing migrants from seeking timely HIV testing, to close this crucial gap in the HIV continuum of care.

Ethical statement
According to Danish law, registry-based studies do not require approval from an ethics board. The study was approved by the Danish Data Protection Agency (registration number: P-2021-446). All data were anonymised.
manuscript critically for important intellectual content. All authors approved the final version of the manuscript.