Article Text

Original article
Delayed linkage to care in one-third of HIV-positive individuals in the Netherlands
  1. MG van Veen1,2,
  2. SCM Trienekens1,
  3. T Heijman2,
  4. HM Gotz3,
  5. S Zaheri4,
  6. G Ladbury1,5,
  7. J de Wit6,
  8. JSA Fennema2,
  9. F de Wolf4,
  10. MAB van der Sande1,7
  1. 1Centre for Infectious Disease Control, National Institute for Public Health and the Environment (RIVM), Bilthoven, The Netherlands
  2. 2Public Health Service, Amsterdam, The Netherlands
  3. 3Municipal Public Health Service Rotterdam-Rijnmond, Rotterdam, The Netherlands
  4. 4HIV Monitoring Foundation (SHM), Amsterdam, The Netherlands
  5. 5EPIET, ECDC, Stockholm, Sweden
  6. 6Centre for Social Research in Health, University of New South Wales, Sydney, New South Wales, Australia
  7. 7Julius Center for Health Sciences and Primary Care, University of Utrecht, Utrecht, The Netherlands
  1. Correspondence to Dr MAB van der Sande, Epidemiology and Surveillance Unit, Centre for Infectious Disease Control, National Institute for Public Health and the Environment, PO Box 1—pb 75, 3720 BA Bilthoven, The Netherlands; marianne.van.der.sande{at}


Objectives To determine time to linkage to HIV care following diagnosis and to identify risk factors for delayed linkage.

Methods Patients newly diagnosed with HIV at sexually transmitted infections (STI) clinics in the Netherlands were followed until linkage to care. Data were collected at the time of diagnosis and at first consultation in care, including demographics, behavioural information, CD4+ counts and HIV viral load (VL) measurements. Delayed linkage to care was defined as >4 weeks between HIV diagnosis and first consultation.

Results 310 participants were included; the majority (90%) being men who have sex with men (MSM). For 259 participants (84%), a date of first consultation in care was known; median time to linkage was 9 days (range 0–435). Overall, 95 (31%) of the participants were not linked within 4 weeks of diagnosis; among them, 44 were linked late, and 51 were not linked at all by the end of study follow-up. Being young (<25 years), having non-Western ethnicity or lacking health insurance were independently associated with delayed linkage to care as well as being referred to care indirectly. Baseline CD4+ count, VL, perceived social support and stigma at diagnosis were not associated with delayed linkage. Risk behaviour and CD4+ counts declined between diagnosis and linkage to care.

Conclusions Although most newly diagnosed patients with HIV were linked to care within 4 weeks, delay was observed for one-third, with over half of them not yet linked at the end of follow-up. Vulnerable subpopulations (young, uninsured, ethnic minority) were at risk for delayed linkage. Testing those at risk is not sufficient, timely linkage to care needs to be better assured as well.


Statistics from

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.


To reduce HIV-related morbidity and mortality, it is crucial that patients with HIV infection are tested as early as possible, and that once diagnosed, they receive prompt optimal care. The HIV cascade of care is a graphical way to identify opportunities to improve HIV care and control1: early diagnosis, prompt linkage to care, receipt of combination antiretroviral therapy (cART), adherence to therapy and achieving viral load (VL) suppression (figure 1, reflecting the Dutch cascade of care in 2012). HIV-infected people not in care will not be able to achieve VL reduction through cART and thus can act as a reservoir facilitating ongoing transmission.2 Public health benefits of early linkage to care include access to risk-reduction interventions, decreased HIV transmission due to reduced plasma and anogenital VL as a result of cART and possibly lower healthcare costs.3 Lowering the amount of HIV circulating in an infected person through cART can result in a substantial decrease of the transmission potential of HIV as well as benefit future health outcomes of the infected person.4 ,5

Figure 1

The 2012 Dutch cascade of care: estimated gaps in % of people living with HIV/AIDS (PLWHA) per step. cART, combination antiretroviral therapy.

Despite promotion of regular HIV testing, we estimated that in 2008, 40% of HIV-infected people in the Netherlands were not diagnosed or not linked to care,6 while a significant proportion of HIV-infected people were linked late (CD4+ cells <350/mm3)7 as also observed elsewhere.8–11 In the Netherlands, HIV testing occurs mainly via general practitioners (GPs), antenatal care clinics (pregnant women only) or at sexually transmitted infections (STI) clinics linked to regional public health services. The STI clinics provide free and anonymous services targeted at high-risk populations, regardless of their health insurance status. Over one-third of new HIV infections annually are diagnosed in these clinics, the majority in men who have sex with men (MSM). STI clinics aim to minimise delayed linkage by offering to arrange a first appointment at an HIV treatment centre (HTC), but there is no routine mechanism to confirm a new patient has been linked to care nor a standard protocol to trace people who miss their first appointment. We aimed to identify the extent of delayed linkage upon HIV diagnosis at an STI clinic and to elucidate characteristics of delayed linkage. This information will serve to assess the contribution of delayed linkage on ongoing HIV transmission, to evaluate current counselling and referral protocols and to develop tailored interventions to reduce delayed linkage to care. This may become even more relevant if ongoing promotion of regular HIV testing to enable early diagnosis would increase delayed linkage, as more people may be tested who have not actively chosen for a test and may, therefore, be less prepared to follow up on a positive test result. Increased availability of home-based testing might also necessitate increased efforts to ensure people diagnosed with HIV are linked to care promptly.

Our primary objectives were to determine time between a positive HIV test and being linked to care and to quantify the proportion of people who had delayed linkage, including those who were not linked at all. The secondary objective was to identify risk factors for delayed linkage by comparing demographic, biological, behavioural and social risk factors, as well as perceived social support and HIV-related stigma. Third, we assessed to what extent biological markers and reported risk behaviour changed between diagnosis and linkage to care.

Materials and methods

HIV testing and treatment facilities

All Dutch STI clinics have an opting-out policy for HIV testing since 2010. Whenever an initial HIV test yields a positive result, a second sample is taken for confirmation. If the initial result is confirmed, the patient is counselled and advised to report to an HTC. At the STI clinic, baseline demographic, behavioural and clinical data of all patients are routinely recorded in a dedicated database. Upon linkage to care, information is collected on demographics and testing history, baseline CD4+ and VL counts. Clinical and laboratory follow-up data are added prospectively.

Study population

From February 2009 until January 2012, all patients testing newly HIV-positive at the STI clinics in Amsterdam, Rotterdam (since October 2009) and Arnhem (since January 2011) were eligible. Exclusion criterion was a previous HIV diagnosis. At the moment of the initial diagnosis, when a sample was collected for confirmation, written information on the study was handed out, explaining that the study would link STI clinic records to records at the HTC.

Study design

When patients returned a week later and were confirmed HIV infected, verbal consent was asked. If provided, additional information was collected on demographics, sexual behaviour preceding HIV diagnosis and history of STI testing and diagnoses. Validated sets of statements were used to assess perceived social support, HIV-related stigma and HIV illness perceptions.12 ,13 CD4+ counts and VL at diagnosis were measured in a residual blood sample. Once linked to care, participants were requested to fill in a second study questionnaire, collecting similar data as the study questionnaire administered following diagnosis.

Data management

At the STI clinic, patient data are routinely entered in an electronic patient database, which is synchronised and anonymised to enable linkage to a centralised national surveillance database. At the HTC, data on all patients linked to care are stored in a longitudinal national database for patients in care, developed and maintained by the Dutch HIV monitoring foundation (HMF). As it is not possible to link these two databases for routine monitoring of linkage to care, a unique study number was created at the STI clinic and added to the referral note to enable matching of records upon linkage to care. Demographic, epidemiological, clinical and biological data were extracted from both databases at either time point (diagnosis and linkage), and merged with the two study questionnaires. Where there was a missing value in a study questionnaire, this was treated as unknown.


Delayed linkage to care was defined as a time period of over 4 weeks between confirmed HIV diagnosis and first consultation at an HTC according to the HMF database. This cut-off was selected based on Dutch expert opinion, taking into consideration the excellent communication and transport arrangement in the Netherlands, whereby linkage within 1 week should be the default. Those who could not be linked to the HMF database at all during the study period were defined as not linked to care. Those who were linked more than 4 weeks after diagnosis were defined as late linkage. A VL <50 copies/mL was defined as undetectable.

HIV-treatment beliefs, perceived HIV-related stigma and perceived social support were each scored on a scale from 1 to 7. Internal consistency of responses was assessed using Cronbach's α (values >0.7 considered to indicate good consistency). Mean scores for each statement set were calculated, and three binary variables were created, ensuring that about half of the participants were in each category, reflecting optimistic treatment expectations (yes/no), perceived HIV-related stigma (yes/no) and perceived sufficient social support (yes/no). Cronbach's α was 0.95, 0.89 and 0.71 for HIV-treatment beliefs, perceived HIV-related stigma and perceived social support, respectively.

Statistical analyses

We performed statistical analyses in SPSS V.19.0 and Stata V.12.0. χ2, Student's t test, logistic regression or non-parametric tests, as appropriate, were used. Prevalence ratio's (PR) with 95% CI by demographic characteristics were calculated regarding reported stigma and treatment beliefs. Univariate analysis was performed to assess potential risk factors of delayed linkage and for not being linked by the end of follow-up. Risk factors with a p value ≤0.1 were included in a multivariable analysis. Furthermore, we assessed changes in risk factors between HIV diagnosis and linkage to care.

The Medical Ethics Committee of the Free University (VU) Amsterdam confirmed no specific ethical clearance was required for this study.


Characteristics of the study population

In total, 310/422 eligible participants were included in the study (response 70%). MSM were more likely to consent than heterosexual men and women (78% vs 54%). Participation was highest among people with health insurance who did not originate from Africa (78%) and lowest among people without health insurance originating from Africa (29%).

As shown in table 1, the majority of participants were MSM (n=279; 90%) and the remainder (n=31, 10%) were heterosexual men (n=22) and women (n=9). Heterosexual participants more often had an African or Surinamese/Antillean ethnicity than MSM (39% vs 16%, p<0.01). Almost 10% indicated they had no health insurance (8% of MSM vs 23% of heterosexual participants, p<0.01), and most of these were of non-Dutch origin. Half of the participants currently had a steady partner and 88% reported casual sex partners in the previous 6 months. Paid sexual contacts were more often reported by heterosexual participants (26%) than by MSM (13%, p=0.02). Almost 80% had tested for HIV before and 67% reported a previous STI diagnosis. Median CD4+ count at diagnosis was 540/mL. VL at diagnosis was detectable in 90% of the samples: in 3% of the samples no VL could be detected, and for 7%, the VL test was not successful.

Table 1

Baseline characteristics of the study population at HIV diagnosis at Dutch STI centres, 2009–2012

Treatment beliefs, stigma and social support

Heterosexual participants experienced more stigma than MSM (p=0.06) and reported less optimistic treatment expectations (p=0.04) at diagnosis (table 1). Among MSM, less optimistic expectations of the benefits of HIV treatment were observed among participants of non-Western-European background compared with those with a Western-European background (40.4% vs 61.2%; PR=0.66, 95% CI 0.52 to 0.85), and among participants who had not previously tested for HIV compared with those who had (37.2% vs 50.9%; PR=0.73, 95% CI 0.55 to 0.96). MSM with a Western-European background were less likely to perceive HIV-related stigma than those with a non-Western-European background (40.4% vs 54.0%; PR=0.74, 95% CI 0.57 to 0.94), as were those who had completed higher education compared with those who had not (39.8% vs 53.7%; PR=0.74, 95% CI 0.57 to 0.95) and those with a previous HIV test or not (42.7% vs 62.8%; PR=0.66, 95% CI 0.50 to 0.86). Demographic characteristics were not associated with perceived social support.

Referral to HIV care

Of all 310 participants, 223 (72%) were referred directly from an STI clinic to an HTC. Six patients were referred back to their GP, 16 patients were referred otherwise (eg, to hospitals in other cities), and 55 would make an appointment on their own initiative. For 10 patients, the method of referral to specialist services was unknown (table 2).

Table 2

Referral mode to care and time to linkage in care for MSM and heterosexuals newly diagnosed with HIV at a Dutch STI clinic, 2009–2012

Time to linkage into care

Of the 310 participants, 259 (84%) had been linked to care by the end of study follow-up (May 2012). Median time from diagnosis to linkage was 9 days (range 0–739 days). Overall, 95 (31%) of participants were not linked within 4 weeks of diagnosis; among them, 44 were linked late, and 51 were not linked at all by the end of the follow-up. Figure 2 shows time between diagnosis and linkage to care or, for those who had not been linked to care, duration of delay at end of study.

Figure 2

Time (in weeks) between diagnosis and linkage to care (blue) or for people not linked to care, time between diagnosis and end of study (red). W, week(s); Y, year(s).

Risk factors for delayed linkage to care

In univariate analyses, delayed linkage was associated with younger age (<25 years), non-Western ethnicity, lack of health insurance, CD4+ cell count >500/mL at diagnosis and undetectable VL at diagnosis (table 3). Furthermore, participants referred to an HTC directly were less likely to have delayed linkage than those who sought care on their own initiative or were referred indirectly via their GP. Participants who reported a current steady partner were less likely to have delayed linkage, as were participants who reported having disclosed their diagnosis to their partner, friends and/or relatives. In the adjusted multivariable model, younger age and indirect referral remained associated with delayed linkage to care.

Table 3

Risk factors for not being linked to care within 4 weeks of HIV diagnosis among newly diagnosed patients with HIV at Dutch STI clinics 2009–2012

Risk factors for not being linked to care

Analyses of covariates of not being linked to care showed comparable associations, as for all delayed linkage (table 4). Indirect referral, lack of health insurance and undetectable VL at diagnosis were significant independent risk factors for not being linked to care by the end of study follow-up. Participants who were not linked to care by the end of follow-up were significantly more likely to have health insurance (adjusted OR (aOR)=15.2, 95% CI 2.6 to 87.6), to have less optimistic treatment expectations (aOR=4.5, 95% CI 1.5 to 13.9) and to not have been previously diagnosed with an STI (aOR=9.2, 95% CI 2.5 to 34.4) than participants who were linked late (data not shown).

Table 4

Risk factors for not being linked to care

Changes between diagnoses and linkage to care

For 218/259 (84%) of participants (eventually) linked to care, CD4+counts were available both at diagnosis and at entry into care. Mean CD4+ count decreased by 16%, from 544 (SD 239) cells/mL at diagnosis to 456 (SD 230) cells/mL upon entry into care (p<0.0001). CD4+ counts declined significantly more already with 1-week delay between diagnosis and linkage to care (mean decline 55 cells/mL without any delay vs 106 cells/mL with more than a week's delay, p=0.03).

Among participants who entered care, 33% (85/259) completed the second questionnaire at the HTC (see online supplementary table 3). There were no significant differences with those who did not complete this questionnaire with respect to age and sexual preference, but they were significantly more likely to have a Western-European background. Furthermore, participants who completed both questionnaires were less likely to have late linkage to care than participants who were linked, but did not complete the second questionnaire (7/85=8.2% vs 37/174=21.3%, p<0.01). Of the seven participants with late linkage to care who completed the second questionnaire, three reported a miscommunication between the STI clinic and the HTC, and one participant was abroad during the 4-week period.

Upon linkage to care, 21/85 (25%) of participants who completed both questionnaires reported having an HIV-negative steady partner. While more than half (11/21) of them reported not having had sexual contacts between diagnosis and entry into HIV care, all but one of the participants who had been sexually active in this period (9/21, 43%) reported not always having used a condom during sex. Compared with the time of diagnosis, participants reported fewer casual partners and less inconsistent condom use (both with steady and with casual partners) at linkage to care. Due to small numbers of participants, statistical power was insufficient to test statistical significance of changes in sexual risk behaviour between those linked to care with and without delay.


Less than 70% of participants newly diagnosed with HIV at Dutch STI clinics between 2009 and 2012 were linked to care at an HTC within 4 weeks of diagnosis, while half of the delayers had not been linked to care by the end of follow-up. Between diagnosis and entering care, CD4+ counts decreased by 16%, and those who were linked within 1 week had on average less than half the CD4+ decline as those who were linked later. As lower CD4+ counts at start of cART are associated with poorer treatment outcomes, this underscores the need for prompt linkage upon diagnosis.14 ,15 In particular, those of younger age or not having health insurance were at risk for delayed linkage and may need additional support upon diagnosis. Having health insurance is mandatory in the Netherlands, those still uninsured need to obtain such insurance before entry to care in HTCs.

Direct referral to an HTC by STI clinic staff contributed strongly to earlier linkage and should, therefore, be standard care. The importance of direct referral for effective linkage to care is supported by previous studies where same-day appointments as well as HIV post-test counselling at the time of diagnosis strongly reduced delay of linkage.16 ,17 Also, not having health insurance has been associated with reduced use of HIV healthcare services,18 as well as with delayed linkage.19 ,20 Furthermore, a study among people living with HIV who were never linked to care also observed that younger people were more likely to delay linkage to care.21 Other studies observed that people from minority backgrounds were linked to HIV care later,20–22 and had longer delays between infection and diagnosis.23 The more optimistic treatment expectations and lower perceived stigma reported by MSM from Western-European background, with higher education or with a previous HIV test, also suggest that social inequalities need to be addressed in HIV prevention.24 We did not observe associations between delayed linkage to care and treatment-related beliefs, perceived HIV-related stigma or perceived social support. This may reflect absence of such influences in this sample or among people currently living with HIV in the Netherlands more generally. It may, however, also reflect limitations of the methods used to capture these associations. However, it was encouraging that participants reported fewer casual partners and less inconsistent condom use between diagnosis and linkage, compared with prior to diagnosis.25 ,26 This is in particular relevant in case of a recent infection when VL is usually very high. Better understanding of the characteristics and needs of people living with HIV at critical steps along the HIV cascade of care is pivotal to improve responses promoting engagement with HIV care.27 ,28

This study was unique in combining biological and risk behaviour data and collecting these both at the time of diagnoses and at entry into care, but also has some limitations. First of all, it was limited to participants recruited at STI clinics only, where opt-out testing is practised and approximately one-third of new HIV cases in the Netherlands are diagnosed. It is unknown to what extent similar observations would hold for patients newly diagnosed by a GP,29 where another one-third of new diagnoses is made. Also, we cannot exclude that some patients not recorded into care were in care, for example, outside of the Netherlands. Another limitation is that 90% of the study population were MSM, limiting generalisability to heterosexual people living with HIV, although in many industrialised countries, the HIV epidemic is currently concentrated among MSM. However, heterosexuals, and in particular heterosexual men, might be the population group most at risk for testing late,17 also illustrated by the lower mean CD4+ count of heterosexual men upon reporting into care in the Netherlands.30 Furthermore, although the 70% response rate was encouraging, non-response is unlikely to have been random, which may have affected some of the estimates.

Timely and adequate use of cART for treatment and prevention of HIV has the potential to further curb the HIV epidemic and achieve an AIDS-free generation, but there are as yet no signs of a persistent decline in HIV incidence.4 ,5 ,14 ,15 w1 w2 Increased testing will have limited impact if people are not linked to care promptly.w3 Therefore, to achieve the full prevention potential of VL suppression by cART, it is essential that HIV-infected people are diagnosed as soon as possible, as the first step in the cascade of care, and that the next steps, in particular linkage to and retention in care, are initiated promptly and successfully.w4–7 Although reported risk behaviour tended to decrease, the observed decrease in CD4+ counts between time of first diagnosis and late linkage to care illustrates the negative impact of such delay.

In conclusion, to effectively reduce delay in people newly diagnosed with HIV entering into HIV care, HIV test providers in STI clinics and other settings are strongly recommended to provide direct referral to a specialist HIV service. Providers of HIV testing are recommended to give sufficient attention to individuals at higher risk of delayed entry into care, including young people, people without health insurance and ethnic minorities. To examine the extent to which specific motivators and barriers could reduce delay and how these differ between different risk groups, further research will be needed, including qualitative as well as quantitative research. Long-term follow-up might clarify to what extent delayed linkage to care is associated with non-retention in care and non-adherence to cART.

Key messages

  • Increased access to testing is not enough to ensure HIV-infected people will be linked to care promptly upon testing

  • Delayed linkage to care following diagnosis puts HIV-infected people at risk of poorer health outcomes, and enables ongoing HIV transmission

  • Young people, ethnic minorities and those without health insurance are at greater risk of delayed linkage to care after HIV diagnosis


Supplementary materials

  • Abstract in Dutch

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.

  • Supplementary Data

    This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.


  • Handling editor Jackie A Cassell

  • Acknowledgements We would like to acknowledge the many excellent suggestions of the anonymous reviewers.

  • Contributors MABvdS conceived of the study design together with JSAF and FdW. MGvV coordinated implementation, was responsible for analysis and wrote the first draft. SCMT assisted in analysis and drafting. TH and HMG were responsible for implementation. SZ assisted in linkage and analysis of data of the HMF. GL and JdW were responsible for the design and analysis of the treatment expectations, stigma perceptions and perceived social support. All authors read and approved the final paper. Several other people contributed to the implementation of the study: CIb: Birgit van Benthem, Eline Op de Coul, Fleur van Aar, Silke David. Staff STI clinics Amsterdam (Roelien Beuker, Edwin van Leent, Petra van Leeuwen, Martijn van Rooijen, Antoinette van Roosmalen, Dewi Usmani), Rotterdam (Jolanda Stam), Arnhem (Bas Boogmans, Manon Pelgrim). Staff HTCs (AMC: Michelle Mutschelknauss, Hans-Erik Nobel; Jan van Goyen: Wilma Brokking; Lucas Andreas: Marijke Spelbrink; OLVG: Aafke Bosma, Kees Brinkman, Narde van der Meche; VUmc: Loek Elsenburg; Slotervaart: Hanneke Paap; Erasmus Rotterdam: Annelies Vervong, Laura Zonneveld; Maasstad Rotterdam: Esther Smit, Jennigje Smit; Rijstate Arnhem: Petra van Benthum). Sanquin: Maarten Koot en Christa Homburg. HMF: Bianca Slieker.

  • Funding The study was funded by the Netherlands Organisation of Health Research and Development, and the Netherlands Ministry of Health, Welfare and Sport (grant 125010005). The funders had no role in the implementation, or analysis or interpretation of the study, or in the drafting of the manuscript.

  • Competing interests None declared.

  • Provenance and peer review Not commissioned; externally peer reviewed.