Estimating the HIV PrEP care continuum among gay, bisexual and other men who have sex with men in Amsterdam and the Netherlands

Abstract

Objective:

HIV-preexposure prophylaxis (PrEP) implementation can be examined by the PrEP care continuum that defines steps toward effective HIV prevention. We examined the PrEP continuum among gay, bisexual and other men who have sex with men (GBMSM) in the Netherlands and in Amsterdam.

Design:

Cross-sectional study combining multiple data sources.

Methods:

The following steps of the PrEP care continuum were estimated for 2023: GBMSM not living with HIV, accessed sexual health services, PrEP-eligible, PrEP-intention, current PrEP use, and PrEP continuation. The number of GBMSM progressing through each step were estimated and percentages were calculated using the nominator of the previous step.

Results:

An estimated 746 061 [uncertainty interval (UI) 634 503–856 079] GBMSM were not living with HIV in the Netherlands. Of those, 44 911 (6%, UI 5–7%) accessed sexual health services, of whom 34 895 (78%, UI 71–86%) were PrEP-eligible. We estimate that 22 002 (63%, UI 49–78%) of those had intention to use PrEP. Of those with PrEP intention, 12 341 (56%, UI 47–69%) currently used PrEP and 10 097 (82%) of those continued PrEP. Of those who accessed sexual health services, 23% continued PrEP in the Netherlands, and 30% in Amsterdam.

Conclusions:

When having accessed sexual health services, retention in the PrEP care continuum was high, except for ‘PrEP-intention’ and ‘current use’ on the national level. To optimize PrEP implementation, targeted interventions to close this gap and annual monitoring of the PrEP care continuum using standardized resources are needed.

Introduction

The use of oral preexposure prophylaxis (PrEP) to prevent human immunodeficiency virus (HIV) acquisition plays a crucial role in achieving the Joint United Nations Programme on HIV and AIDS (UNAIDS) target of zero new HIV infections by 2030 [1]. In 2015, both the World Health Organization (WHO) and the European Centre for Disease Prevention and Control (ECDC) recommended to integrate PrEP in HIV prevention services for individuals vulnerable for HIV infection [2,3]. By 2023, PrEP has been implemented in 23 European Union (EU)/European Economic Area (EEA) countries and PrEP implementation is still evolving [1,4].

Although PrEP has become an essential part of HIV prevention, it is known that significant barriers prevent uptake and continued use [5–7]. To examine gaps in PrEP uptake and use, the PrEP care continuum can be used, as done in previous studies [8,9]. The PrEP care continuum defines steps towards effective HIV prevention by PrEP, like the HIV care continuum defines steps towards effective HIV treatment by achieving viral suppression. Steps of the PrEP care continuum may include the number of people who can benefit from and are eligible for PrEP, who are aware of PrEP, willing to use it, have access to PrEP services, and start and continue using PrEP correctly. Between each step, PrEP-eligible individuals have the potential to drop out, resulting in a gap. As these gaps highlight opportunities to increase the prevention potential of PrEP, the WHO and ECDC have developed PrEP monitoring tools based on the PrEP care continuum [10,11].

PrEP continuum-based monitoring has been documented for some high-income countries [7,9,12,13]. In the Dutch context, studies have examined PrEP eligibility, awareness, intention, use and continuation among key populations in early study settings [14–17], but the PrEP care continuum has not been evaluated since nationwide PrEP implementation in 2019. Moreover, region-specific data are lacking, while comparing regional and national-level data can help identify tailored points of action [1]. Amsterdam is the capital city of the Netherlands, with a significant HIV burden [18] and a relatively high number of PrEP prescriptions compared to other cities – mainly among gay, bisexual and other men who have sex with men (GBMSM). Therefore, we aim to provide insights to optimize PrEP implementation by examining the PrEP care continuum among GBMSM, overall in the Netherlands and in the municipality of Amsterdam in 2023.

Methods

Dutch preexposure prophylaxis care context

In the Netherlands, PrEP is primarily prescribed by Centers for Sexual Health (CSHs) and by general practitioners (GPs), though online provision is increasingly common. At CSHs, PrEP care is free for all individuals at substantial risk of HIV, defined as men and transgender people who have sex with men or transgender people who had at least one of the following in the past six months: insertive and/or receptive anal sex without a condom, anal sexually transmitted infection (STI) or infectious syphilis, postexposure prophylaxis (PEP) use or chemsex [19]. Others at substantial HIV risk are eligible based on indication (e.g., those engaging in sex with MSM, or individuals from high-prevalence regions). In 2023, PrEP pills were available for a co-payment of €7.50 for 30 tablets at CSHs. PrEP care provided by GPs is free-of-charge, but costs related to HIV and STI laboratory testing are subject to the patient's deductible. Prices for PrEP tablets vary between €15 and €65 per 30 pills. In August 2019, PrEP was officially implemented through a national PrEP pilot. Before 2019, PrEP was only available through participation in studies, while people also used PrEP informally [15,20–22].

Study design

We conducted a cross-sectional study that combined multiple data sources to produce estimates of individuals progressing through each step in the continuum. The target population for this study was GBMSM not living with HIV, as this group is considered a key population at risk of HIV acquisition in the Netherlands [18]. We did not include other key populations for HIV (such as transgender women) due to a lack of available data on all PrEP continuum steps in these groups. We estimated the PrEP care continuum for the Netherlands overall and for the municipality of Amsterdam separately for the year 2023.

The steps of the PrEP care continuum in this study were based on ECDC [10] and WHO [11] PrEP monitoring tools and definitions from previous research on the PrEP care continuum [7–9,12,14]. Six steps were distinguished, being:

• (1)Not living with HIV;
• (2)Accessed sexual health services;
• (3)PrEP eligible according to the Dutch PrEP guidelines;
• (4)Intention to use PrEP;
• (5)Current PrEP use;
• (6)PrEP continuation.

We estimated the number and percentages of individuals in each step.

Measures

Data sources for estimating the PrEP care continuum and their characteristics are shown in Table S1, Supplemental Digital Content and File S2, Supplemental Digital Content. We used 2023 data for our estimates; if unavailable, data from the most recent year (2022–2024) were used. When estimating the care continuum for the Netherlands and Amsterdam, we used country- or Amsterdam region-specific data sources where available. Calculation methods for each step are summarized below; see Tables 1 and 2 for details.

Table 1.

Calculation methods used to estimate the PrEP care continuum for GBMSM, the Netherlands.

Step in PrEP care continuum
	Sourcea		N/% (UI/range/95%CI)b	Formulac
a	GBMSM not living with HIV in the Netherlands		746 061 (UI 643 503-856 079)	$\frac{\left(s1.1*s3.1\right)+\left(s1.2*s4.1\right)}{2}-s5$
	s1.1	Statistics Netherlands (CBS), registered inhabitants	15 083 914
	s1.2	Statistics Netherlands (CBS), registered male inhabitants	7 458 801
	s3.1	Safety Monitor, percentage of GBMSM of total inhabitants	6.9%
	s4.1	Sexual Health Monitor, average percentage of GBMSM of male inhabitantsd	6.45% (range 3.7–9.4)
	s5	HIV Monitoring Foundation (SHM), estimated number of GBMSM with HIV	14 880
b	GBMSM not living with HIV who accessed sexual health services		44 911 (UI 42 470–47 623)	b1+b2
b1	At CSH		37 393	s6.1
	s6.1	SOAP, number of GBMSM not living with HIV at CSHs	37 393
b2	At GP		7518 (UI 5077–10 230)	$\frac{\left(s8*s4.1\right)+\left(s6.1*s4.2\right)}{2}$
	s8	Netherlands Institute for Health Services Research (NIVEL), extrapolated number of STI-related episodes registered among male	82 400
	s4.2	Sexual Health Monitor, GP/CSH ratioe	0.26 (95% CI 0.19–0.34)
c	GBMSM eligible for PrEP use		34 895 (UI 33 633–36 297)	c1 + c2
c1	At CSH		31 008	s6.2
	s6.2	SOAP, number of GBMSM not living with HIV and PrEP-eligible at CSH	31 008
c2	At GP		3887 (UI 2625–5289)	b2 * s10.1
	s10.1	Amsterdam Cohort Study (ACS), percentage of PrEP-eligible GBMSM among those not living with HIV	51.7%
d	GBMSM with the intention to use PrEP		22 002 (UI 17 938–26 067)f	$\frac{\left(c*s11.1\right)+\left(\frac{e}{s11.2}\right)}{2}$
	s11.1	PROTECT, percentage of GBMSM with PrEP intention among PrEP-eligible in the Netherlands	74.7%
	s11.2	PROTECT, percentage of GBMSM who use PrEP among those with PrEP intention	68.8%
e	Current PrEP use		12 341	s6.3 + ((s12.1*s6.4)-(s13.1*s6.4))
	s6.3	SOAP, number of GBMSM with at least one PrEP consultation at CSH	9908
	s12.1	Foundation for Pharmaceutical Statistics (SFK), number of individuals with at least one PrEP distribution	3437
	s13.1	National Institute for Public Health and the Environment (RIVM), number of current PrEP users receiving PrEP through municipal programs	763
	s6.4	SOAP, percentage of GBMSM among PrEP users at CSH	91%
f	PrEP continuation		10 097	s6.5 + ((s12.2*s6.4)-((s13.2*s6.4))
	s6.5	SOAP, number of GBMSM with at least two PrEP consultations	9050
	s12.2	Foundation for Pharmaceutical Statistics (SFK), number of individuals with at least two PrEP distributions	1848
	s13.2	RIVM, number of PrEP users receiving PrEP through municipal programs	697	$s13.1*\frac{s6.5}{s6.3}$
	s6.4	SOAP, percentage of GBMSM among PrEP users at CSH	91%

UI, uncertainty interval; 95% CI, confidence interval; GBMSM, gay, bisexual and other men who have sex with men; CSH, Center for Sexual Health; SOAP, “Seksueel Overdraagbare Aandoening Peilstation”; GP, general practitioner.

^aFor supplementary information on characteristics, specifics and references of data sources used, see Table S1, Supplemental Digital Content and File S2, Supplemental Digital Content.

^bUIs in step a and b were calculated based on source ranges (s4.1) and 95% CI (s4.2), UIs in steps c and d were calculated based on UIs of the previous step. When a range/95%CI was reported in a data source, the UI for that step was calculated using the lower and upper bounds. For example, in the formula of step a (GBMSM not living with HIV in the Netherlands), the lowest bound of the UI was calculated using 3.7% as value for s4.1 in the formula and the highest bound was calculated using 9.4% as value for s4.1.

^cEntries beginning with ‘s’ (e.g., s1, s2) refer to sources, whereas plain numbers (e.g., 2, 5) represent actual numerical values.

^dAdditional information underlying this calculation; this is a combination of sexual behavior, sexual attraction and self-identification. This is calculated as: (9.4% (men who ever had sex with same-sex male) + 3.7% (men who had sex with same-sex partner in past 6 months) + (3.6% + 0.5% + 0.7% + 2.9% (men who are primarily attracted to women + equally attracted to women and men + primarily attracted to men + exclusively attracted to men)) + (2.9% + 2.1% (homo- + bisexual men)))/4.

^eAdditional information underlying this calculation; this is a combination of the self-reported locations of STI test in the last year (18.4% at GP (95% CI 11.5–27), 70.4% at CSH (95% CI 61.2–79.4)). This is calculated as: 18.4/70.4 = 0.26, lowest bound 95%CI: 11.5/61.2 = 0.19, highest bound 95%CI: 27/79.4 = 0.34.

^fAdditional information underlying this calculation; this is a combination of percentage PrEP intention among PrEP-eligible and percentage PrEP-use among intention, to include potential overestimation of PrEP intention among PROTECT data.

Table 2.

Calculation methods used to estimate the PrEP care continuum for GBMSM, Amsterdam, the Netherlands.

	Step in PrEP care continuum
	Sourcea		N/% (UI/95% CI)b	Formulac
a	GBMSM not living with HIV in Amsterdam		74 728 (UI 74 725–74 732)	$\frac{\left(s1.1*s3.2\right)+\left(s2*s3.2\right)}{2}$
	s1.1	Statistics Netherlands (CBS), registered inhabitants	793 847
	s2	Base File of Areas Amsterdam (BBGA), obtained through Research and Statistics, Municipality of Amsterdam, registered inhabitants	793 916
	s3.2	Safety Monitor, percentage of GBMSM of total inhabitants	10%
	s5	HIV Monitoring Foundation (SHM), estimated number of GBMSM with HIV	4660
b	GBMSM not living with HIV who accessed sexual health services		13 863 (UI 12 770–14 380)	b1 + b2
b1	At CSH		10 731	s7.1
	s7.1	Aphrodite, number of GBMSM not living with HIV at CSH of Amsterdam	10 731
b2	At GP		3132 (UI 2039–3649)	$\frac{\left(\left(\frac{s9.1}{35}*100\right)*s3.2\right)+\left(s7.1*s4.2\right)}{2}$
	s9.1	Academic Network for General Practice at Amsterdam UMC (ANHA), number of individuals not living with HIV with registered sexual healthcare episodes.	12 160
	s4.2	Sexual Health Monitor, GP/CSH ratiod	0.26 (95% CI 0.19–0.34)
c	GBMSM eligible for PrEP use		10 702 (UI 10 137–10 969)	c1 + c2
c1	At CSH		9083	s7.2
	s7.2	Aphrodite, number of GBMSM not living with HIV and PrEP-eligible at CSHA	9083
c2	At GP		1619 (UI 1054–1886)	b2 * s10.1
	s10.1	Amsterdam Cohort Study (ACS), percentage of PrEP-eligible GBMSM among those not living with HIV	51.7%
d	GBMSM who have the intention to use PrEPe		7647 (UI 6310–8985)e,f	$\frac{\left(c*s10.2\right)+\left(\frac{e}{s10.3}\right)+\left(c*s11.1\right)+\left(\frac{e}{s11.2}\right)}{4}$
	s10.2	Amsterdam Cohort Study (ACS), percentage of GBMSM who have the intention to use PrEP or remain using PrEP of all PrEP-eligible GBMSM	84.1%
	s10.3	Amsterdam Cohort Study (ACS), percentage of GBMSM who have used PrEP in the last 6 months of all GBMSM who have the intention to use PrEP or remain using PrEP	88.9%
	s11.1	PROTECT, percentage of GBMSM with PrEP intention among PrEP-eligible in Amsterdam	83.8%
	s11.2	PROTECT, percentage of GBMSM who use PrEP among those with PrEP intention	81.3%
e	Current PrEP use		5359 (UI 4816–5901)f	$s7.3+\left(\frac{\left(\left(\frac{s9.2}{35}*100\right)*s6.4\right)+\left(\left(s12.1*s\mathrm{.6.4}\right)-\left(s13.1*s6.4\right)\right)}{2}\right)$
	s7.3	Aphrodite, number of GBMSM with at least one PrEP consultation at CSHA	3192
	s9.2	Academic Network for General Practice at Amsterdam UMC (ANHA), number men who had at least one PrEP prescription	1042
	s12.1	Foundation for Pharmaceutical Statistics (SFK), number of individuals with at least one PrEP distribution	2208
	s13.1	National Institute for Public Health and the Environment (RIVM), number of PrEP users receiving PrEP through municipal program in Amsterdam	423
	s6.4	SOAP, percentage of GBMSM among PrEP users at CSH	91%
f	PrEP continuation		4122 (UI 3628–4615)f	$s7.4+\left(\frac{\left(\left(\frac{s9.3}{35}*100\right)*s6.4\right)+\left(\left(s12.2*s6.4\right)-\left(s13.2*s6.4\right)\right)}{2}\right)$
	s7.4	Aphrodite, number of GBMSM with at least two PrEP consultations	2907
	s9.3	Academic Network for General Practice at Amsterdam UMC (ANHA), number men who had at least two PrEP prescriptions	657
	s12.2	Foundation for Pharmaceutical Statistics (SFK), number of individuals with at least two PrEP distributions	1178
	s13.2	RIVM, number of PrEP users receiving PrEP through municipal program in Amsterdam	385	$s13.1*\frac{s7.4}{s7.3}$
	s6.4	SOAP, percentage of GBMSM among PrEP users at CSH	91%

UI, uncertainty interval; 95% CI, confidence interval; GBMSM, gay, bisexual and other men who have sex with men; CSH, Center for Sexual Health; GP, general practitioner.

^aFor Supplementary Information on characteristics, specifics and references of data sources used, see Table S1, Supplemental Digital Content and File S2, Supplemental Digital Content.

^bUIs in step b were calculated based on source 95% CI (s4.2), UIs in steps c, d, e, and f were calculated based on UIs of the previous step. When a 95% CI was reported in a data source, the UI for that step was calculated using the lower and upper bounds. For example, in the formula of step b2 (GBMSM not living with HIV who accessed sexual health services – at GP), the lowest bound of the UI was calculated using 0.19 as value for s4.2 in the formula and the highest bound was calculated using 0.34 as value for s4.2.

^cEntries beginning with ‘s’ (e.g., s1, s2) refer to sources, whereas plain numbers (e.g., 2, 5) represent actual numerical values.

^dAdditional information underlying this calculation; this is a combination of the self-reported locations of STI test in the last year (18.4% at GP (95% CI 11.5–27), 70.4% at CSH (95% CI 61.2–79.4)). This is calculated as: 18.4/70.4 = 0.26, lowest bound 95% CI: 11.5/61.2 = 0.19, highest bound 95%CI: 27/79.4 = 0.34.

^eAdditional information underlying this calculation; this is a combination of percentage PrEP intention among PrEP-eligible and percentage PrEP-use among intention, to include potentional overestimation of PrEP intention among PROTECT and ACS data. The lowest bound of the UI was calculated by taking the average percentage of GBMSM with PrEP intention among PrEP-eligible (((e/s10.3) + (e/s11.2))/2) and the highest bound was calculated as taking the average percentage of GBMSM who have the intention to use PrEP (((c*s10.2) + (c*s11.1))/2).

^fIn case of an average estimate used of multiple sources, the source with the lowest value was used for the lowest bound of the UI and the source with the highest value was used for the highest bound of the UI. For example, in the formula of step e (current use), the lowest bound of the UI was calculated by using SFK data only for the calculation (s7.3 + ((s9.2/35)*100)*s6.4) and the highest bound was calculated by using ANHA data only for the calculation (s7.3 + (s12.1*s6.4)).

GBMSM not living with HIV

We calculated the number of GBMSM not living with HIV as the estimated number of GBMSM minus the estimated number of GBMSM with HIV. We used two data sources for the total number of GBMSM in the Netherlands and Amsterdam: the Safety Monitor (covering both) and the Sexual Health Monitor (Netherlands only). The Safety Monitor provides the weighted proportion of GBMSM (based on self-reported sexual orientation) among a random sample of the population (aged ≥15 years) and the Sexual Health Monitor provides the weighted proportion of GBMSM (based on self-reported sexual behavior, attraction and orientation) among a random sample of the total male population (18–80 years). Combined with population data from Statistics Netherlands (CBS), this provides an estimate of the total GBMSM population. The estimated number GBMSM with HIV (aged ≥15 years) was obtained from the HIV Monitoring Foundation (SHM).

Accessed sexual health services

We calculated the number of GBMSM who accessed sexual health services by summing the registered number of GBMSM without HIV with ≥1 sexual health visit at a CSH and the estimated number of GBMSM without HIV with ≥1 sexual health visit at a GP. This served as the denominator for the subsequent steps in the care continuum, as it is the only population for which data were available. All CSH consultations are systematically registered in the Netherlands, GP data are not. GP estimates were based on two sources: GP surveillance registries: Nivel-PCD for the Netherlands [23] and ANHA for Amsterdam; Sexual Health Monitor for both regions. NIVEL data is based on an extrapolation from approximately 10% of general practices and uses International Classification of Primary Care (ICPC) codes to identify relevant consultations. The reported number of consultations among men with ‘fear of STI/HIV’ in the STI year report [23] were included as accessed sexual health services. ANHA data covers 35% of all GP practices in Amsterdam. To obtain city-wide estimates, data were extrapolated to 100% to represent all practices. We included the number of individuals (≥15 years) with at least one STI/sex/PrEP-prevention-consultation (based on ICPC codes) Combined with the estimated proportion of GBMSM among the total population in the previous step, the number of GBMSM who accessed sexual health services at GPs were calculated. No de-duplication of GP and CSH consultations was possible as GP data were aggregated (i.e., individuals with consultations at both providers may have been counted more than once). Additionally, the Sexual Health Monitor includes national data on STI testing locations (i.e. GP, CSH, hospital/clinic, self-test/self-sample, elsewhere) among GBMSM. We used this information to calculate a GP/CSH-ratio (Netherlands and Amsterdam), which, combined with CSH data, provided a second estimate of the total number of GBMSM in sexual healthcare at GPs.

Eligible for preexposure prophylaxis

The number of PrEP-eligible GBMSM was calculated as the sum of the registered number of PrEP-eligible GBMSM at CSHs and the estimated number of PrEP-eligible GBMSM at GPs. To estimate the number of PrEP-eligible GBMSM at GPs, we multiplied the number of GBMSM accessing sexual health services at GPs by the estimated percentage PrEP-eligible. For both the Netherlands and Amsterdam, this was based on the percentage of PrEP-eligible GBMSM (≥18 years) in the Amsterdam Cohort Studies (ACS), as data on PrEP-eligibility among GBMSM visiting GPs is lacking, and the ACS is more representative of the GBMSM population seeking sexual health services than the CSH population in terms of sexual behavior. The ACS is an ongoing open prospective cohort study; inclusion criteria are male, aged ≥16 years, having had sex with men in the past 6 months preceding enrolment, and residing in or around Amsterdam or frequently engaging in MSM-related activities in the region [24]. PrEP eligibility was defined by the Dutch guidelines in effect in 2023 [25]. Details of the operationalization of the eligibility criteria are in Table S3, Supplemental Digital Content.

Intention to use preexposure prophylaxis

To estimate the number of GBMSM intending to use PrEP, we used survey data from nonprobability convenience sampling (aged ≥18 years). PROTECT [26] data (Netherlands and Amsterdam subsamples) were used to estimate intention. Eligible participants were MSM and trans people aged ≥18. No weighting to approximate nationally representative data was applied. For Amsterdam, the average intention, based on PROTECT and ACS data, was calculated. Both data sources included PrEP-eligible GBMSM who completed questions on PrEP-intention. At ACS, intention was measured with a 7-point Likert scale (>4 indicated intention [14,17]) by asking: “Are you planning to start or continue using PrEP in the next 6 months?”. In PROTECT, intention was defined as agreeing (≥4 on 5-point Likert scale) with “I am going to take PrEP as soon as possible”. Those who reported to currently use PrEP were also defined as having PrEP intention.

Current preexposure prophylaxis use and preexposure prophylaxis continuation

Current PrEP use was defined as individuals with at least one PrEP consultation (with PrEP prescription unless a person still had a sufficient number of tablets in case of on-demand PrEP use), PrEP prescription (with or without PrEP consultation) or distribution in 2023. Continued use was defined as individuals with at least two PrEP consultations/prescriptions/distributions in 2023. Given the 3-monthly consultation advice in 2023, most PrEP users would have at least two consultations annually. This aligns with the ECDC PrEP monitoring tool [10], which defines continuous use as at least one PrEP follow-up visit within 12 months after initiation, as individuals may use PrEP on-demand or in ‘cycles’ of HIV risk. In line with the Dutch standard of care, we did not assess PrEP eligibility across multiple visits. We measured PrEP use by adding the number of PrEP consultations at CSHs to the number of PrEP distributions at GPs, specialists and other providers (including online) using pharmaceutical data (Foundation for Pharmaceutical Statistics (SFK) (Netherlands and Amsterdam). In 2023, PrEP provided directly by CSHs (not dispensed via pharmacies) was recorded only at CSHs and not duplicated in pharmaceutical data. In some municipalities, in additional programs, PrEP care is provided through CSHs, but the medication is dispensed via pharmacies. To avoid double counting, these cases were subtracted from the pharmaceutical data. For continuous use among municipality PrEP users, we multiplied the number of users by the ratio continuous vs. current users from CSH data. For Amsterdam, extrapolated GP data from ANHA was used as a second data source to calculate the average number of PrEP prescriptions at GPs. Pharmaceutical and ANHA data were combined with national CSH data on the percentage of GBMSM among all PrEP users to assess the number of GBMSM PrEP users.

Analyses

The number of GBMSM progressing through each step was estimated, with percentages calculated using the nominator of the previous step. Additionally, the percentage of GBMSM not living with HIV who accessed sexual health services, and PrEP-eligible GBMSM progressing through each step were calculated. A sensitivity analysis included chemsex as a PrEP-eligibility-criterion, based on the 2024 Dutch PrEP guidelines.

When multiple sources were available, we calculated the average of the different estimates and defined the uncertainty interval (UI) using the range between the highest and lowest values of these estimates. If only one data source was available, no UI was used.

Data management, analysis, and visualization were done in R version 4.4.2 (packages: dplyr, tidyr, ggplot2, cowplot, scales, janitor).

Results

The Netherlands

We estimated that 746 061 (UI 634 503–856 079) GBMSM not living with HIV were living in the Netherlands as of January 1, 2023 (Tables 1 and 3, Fig. 1). Of those not living with HIV, we estimated that 44 911 (UI 42 470–47 623) GBMSM accessed sexual health services, corresponding to 6.0% (UI 5.0–7.4), of whom 34 895 (UI 33 633–36 297) were estimated to be PrEP eligible (77.7%, UI 70.6–85.5). We estimate that 22 002 (UI 17 938–26 067) of those eligible had an intention to use PrEP (63.1%, UI 49.4–77.5). 12 341 GBMSM used PrEP in 2023 (56.1% of the estimated number with PrEP-intention, UI 47.3–68.8) and 10 097 continued PrEP (81.8% among current users).

Table 3.

Estimates of the number and percentages of GBMSM per step along the PrEP care continuum, the Netherlands and Amsterdam, the Netherlands, 2023.

Step in PrEP care continuum	N (UI)	Percentage of previous step % (UI)a	Percentage of those who accessed sexual health services % (UI)a	Percentage of PrEP-eligible % (UI)a
Without HIV
The Netherlands	746 061 (643 503–856 079)	–	–	–
Amsterdam	74 728 (74 725–74 732)	–	–	–
Accessed sexual health services
The Netherlands	44 911 (42 470–47 623)	6.0 (5.0–7.4)	–	–
Amsterdam	13 863 (12 770–14 380)	18.6 (17.1–19.2)	–	–
Eligible for PrEP
The Netherlands	34 895 (33 633–36 297)	77.7 (70.6–85.5)	–	–
Amsterdam	10 702 (10 137–10 969)	77.2 (70.5–85.9)	–	–
PrEP intention
The Netherlands	22 002 (17 938–26 067)	63.1 (49.4–77.5)	49.0 (37.7–61.4)	–
Amsterdam	7647 (6310–8985)	71.5 (57.5–88.6)	55.2 (43.9–70.4)	–
Current PrEP useb
The Netherlands	12 341	56.1 (47.3–68.8)	27.5 (25.9–29.1)	35.4 (34.0–36.7)
Amsterdam	5359 (4816–5901)	70.1 (53.6–93.5)	38.7 (33.5–46.2)	50.1 (43.9–58.2)
PrEP continuationb
The Netherlands	10 097	81.8	22.5 (21.2–23.8)	28.9 (27.8–30.0)
Amsterdam	4122 (3628–4615)	76.9 (61.5–95.8)	29.7 (25.2–36.1)	38.5 (33.1–45.5)

UI, uncertainty interval. For supplementary information on calculation methods of PrEP care continuum steps and characteristics, specifics and references of data sources used, see Tables 1 and 2 and Table S1, Supplemental Digital Content and File S2, Supplemental Digital Content.

^aUIs of percentages were calculated by dividing the lower bound of the numerator by the upper bound of the denominator, and vice versa. For example, the UI for PrEP-eligible GBMSM in the Netherlands among the estimated number of whom accessed sexual health services in the Netherlands (77.7 (70.6–85.5)), was calculated as follows: (33 633/47 623)*100 (lower bound) and (36 297/42 470)*100 (upper bound).

^bNo UI is provided for the number of current and continuous PrEP users in the Netherlands, as these numbers are based on exact counts based on the original data.

Fig. 1.

The PrEP care continuum among gay, bisexual and other men who have sex with men who accessed sexual health services, the Netherlands, 2023. PrEP, postexposure prophylaxis.

Of all GBMSM without HIV who accessed sexual health services 49.0% (UI 37.7–61.4) had PrEP intention, 27.5% (UI 25.9–29.1) used PrEP and 22.5% (UI 21.2–23.8) used PrEP continuously in 2023. Of all GBMSM who were PrEP-eligible, 35.4% (UI 34.0–36.7) used PrEP and 28.9% (UI 27.8–30.0) continued PrEP use in 2023.

Amsterdam

We estimated that 74 728 (UI 74 725–74 732) GBMSM not living with HIV were living in Amsterdam as of January 1, 2023 (Tables 2 and 3, Fig. 2). Of those, we estimated that 13 863 (UI 12 770–14 380) individuals accessed sexual health services, corresponding to 18.6% (UI 17.1–19.2), of whom 10 702 (UI 10 137–10 969) were estimated to be PrEP eligible (77.2%, UI 70.5–85.9). We estimated that 7647 (UI 6310–8985 of those eligible had the intention to use PrEP (71.5%, UI 57.5–88.6). 5359 (UI 4816–5901) used PrEP in 2023 (70.1%, UI 53.6–93.5) and 4122 (UI 3628–4615) of those using PrEP continued PrEP (76.9%, UI 61.5–95.8).

Fig. 2.

The PrEP care continuum among gay, bisexual and other men who have sex with men who accessed sexual health services, Amsterdam, the Netherlands, 2023.

The percentages in the bar indicate the percentages of all GBMSM without HIV who accessed sexual health services, with uncertainty intervals (UIs) when applicable. The percentages above the arrows indicate the percentages of the previous step, UIs of those percentages are not shown in this figure. (1) Persons without HIV who accessed sexual health services; (2) based on criteria used in the Dutch multidisciplinary guideline for PrEP [i.e., having had in the past 6 months: insertive and/or receptive anal sex without a condom, anal STI or infectious syphilis, or used postexposure prophylaxis (PrEP)]; (3) intention to start or continue using PrEP; (4) defined as the number of unique individuals who had at least one PrEP consultation in 2023; (5) defined as the number of unique individuals who had at least two PrEP consultations in 2023.

Of all GBMSM without HIV who accessed sexual health services, 55.2% (UI 43.9–70.4) had PrEP intention, 38.7% (UI 33.5–46.2) used PrEP and 29.7% (UI 25.2–36.1) continued PrEP in 2023. Of all GBMSM who were eligible for PrEP, 50.1% (UI 43.9–58.2) used PrEP and 38.5% (UI 33.1–45.5) continued PrEP use in 2023.

Sensitivity analysis

Adding chemsex as a PrEP eligibility criterion resulted in an estimated increase of 715 eligible individuals in the Netherlands and 153 in Amsterdam. This had a very small effect on the percentages of the subsequent steps (i.e. max two percentage points difference, Table S4–S6, Supplemental Digital Content).

Discussion

This study employed a multisource data approach to estimate the PrEP care continuum among GBMSM in the Netherlands and in Amsterdam. We showed that the proportion of PrEP eligibility among GBMSM who accessed sexual health services and the proportion of continuous PrEP use among those with current PrEP use are high compared to the other steps in the continuum for the Netherlands. However, the proportions of intention and current PrEP use relative to the previous step are much lower compared to the other steps in the continuum. Amsterdam showed higher percentages at those two steps of the continuum compared to the national level.

This is the first study to examine all steps of the PrEP care continuum in the Netherlands. A strength of this study was the use of multiple validated data sources. However, the use of multiple data sources also has limitations. First, heterogeneity exists between sources in data quality, data type, inclusion criteria and definitions. the Safety Monitor GBMSM data were, for example, based on sexual orientation, rather than actual sexual behavior as in other data sources. While pooled proportions using meta-analytic methods could adjust for variation, substantial heterogeneity between data sources limited their use. To ensure a reliable representation of the estimate range, we consistently used minimum and maximum values, increasing confidence that the true estimate lies within this interval. Second, some individuals may have appeared in multiple datasets, for example when they used PrEP via their GP and CSH. This may have led to an overrepresentation of PrEP users as de-duplication methods could not be applied with the provided aggregated data. However, if the true number of PrEP users is lower than our estimate, the gap between PrEP intention and use is even larger. Finally, the denominator for our analyses in the care continuum was limited to those who accessed sexual health services. By focusing solely on GBMSM linked to care, we miss capturing eligibility and intention among those not engaged in care, which is an important area for future research.

Although the definitions of each step of the PrEP care continuum in our study were informed by ECDC [10] and WHO [11], PrEP monitoring tools as well as previous research [7–9,12,14], the specific operationalization of these steps (i.e., in definitions and denominators used) varies across studies. This makes direct comparisons with other studies difficult. Our percentages at each step of the care continuum are in general higher than those of earlier Dutch and international studies among GBMSM. Previous Dutch studies, both at the initiation and during the implementation of PrEP, estimated 32% eligibility (among GBMSM not living with HIV in ACS) [14], 45–51% intention (among eligible GBMSM in ACS and an European MSM Internet Survey (EMIS)-2017) [14,27] and 4–10% PrEP use (among PrEP-eligible GBMSM in EMIS-2017) [27], compared to 77–78%, 63–71%, and 35–50% in our study (both Amsterdam and Netherlands). Moreover, more recent international data also reported lower percentages with 33–54% eligibility [7,12] (among GBMSM not living with HIV included in GBMSM surveys), 26% intention [28] (among PrEP-eligible adolescent GBMSM in a cohort), 18–67% current use [12,29,30] (among PrEP-eligible GBMSM in community-based surveys and among those classified as PrEP-eligible in pooled estimates across studies) and 16–92% continuation at 12  months after initiation (among GBMSM across studies in a systematic review) [31] (77–82% in our study). Our relatively higher estimates likely reflect the timing of our study compared to others conducted during earlier phases of PrEP implementation, when awareness and availability were more limited. Moreover, differences in denominators should be considered: PrEP eligibility was estimated among GBMSM attending sexual health services, and PrEP intention among survey participants who are likely more PrEP-aware than the general GBMSM population (e.g., as in a national cohort). Although corrected for this potential overestimation in the PrEP intention estimates, the use of data from sexual health services and surveys may still partly explain the higher estimates observed compared to other studies.

Our study estimates a substantial drop between individuals eligible to actual PrEP use. At the national level, this gap is particularly seen between intention and use, commonly referred to as the ‘intention-behavior gap’. This gap is concerning, as delayed uptake is a key factor for HIV acquisition among GBMSM in the Netherlands; an increasing number of individuals is diagnosed with HIV at their initial PrEP intake visit or shortly after initiation [6,18]. The drop off between eligibility and use in Amsterdam is markedly lower compared to the Netherlands as a whole. Possible explanations are the earlier availability and awareness of PrEP in Amsterdam due to demonstration projects [15,21], greater accessibility of PrEP in a large urban setting compared to smaller or rural areas, and a long-standing city-based approach to reach zero new HIV infections [32]. The lower intention and use percentages for the Netherlands may be because people prefer alternative prevention strategies (e.g., condom use or U=U). Long-acting PrEP, when available, may help bridge this intention-behavior gap, as it could increase uptake among those who are eligible but hesitant to start daily oral PrEP due to e.g. adherence issues. However, injectable PrEP was not available in the Netherlands during the study period in 2023. There are also other explanations such as waiting lists for PrEP use at CSHs or experienced barriers in PrEP access at GPs [1,6]. In addition, individual barriers (e.g., (fear of) stigma/healthcare discrimination [1], side effects [6,33], costs [1], and underestimated [33] or shifting HIV-risk [34]) may be reasons for PrEP nonuse. However, it is important to acknowledge that PrEP eligibility does not necessarily equal PrEP need, as some individuals may meet eligibility criteria while still effectively relying on other HIV prevention methods. Also, disparities in PrEP use and continuation exist. In Amsterdam, doing sex work, being uninsured and being a trans person were associated with lower PrEP continuation rates among GBMSM and transgender persons who started PrEP at CSH Amsterdam [35]. Further research should identify vulnerable groups to disengage along the PrEP care continuum, ways to support their uptake and continuation, and the factors contributing to the intention-behavior gap. Some initiatives to improve the PrEP continuum in Amsterdam include recurring campaigns to reduce stigma.

Beyond the specific gaps identified within the PrEP care continuum in this study, a broader interpretation of the PrEP care continuum is needed. To ensure comprehensive and consistent monitoring using the PrEP care continuum and enable meaningful comparisons over time, it is essential to establish a standardized framework with uniform datasets to harmonize data within countries and between countries [13,36]. Continued evaluation is especially important during times of PrEP policy changes, as was the case in the Netherlands in August 2024, when PrEP care was formally integrated into CSHs [37]. While the HIV care continuum is widely used in Europe [38], routine data for monitoring each step of the PrEP care continuum remain limited, despite existing ECDC indicator proposals [1,10,36]. A European data collection tool would require the implementation of additional data collection instruments at national or regional (EU) level capturing data on PrEP eligibility and ideally covering both the general population and different key populations (e.g., GBMSM, transgender people).

Conclusion

Starting from GBMSM who accessed sexual health services, approximately 70–80% of individuals are retained at each step of the PrEP care continuum, with larger drops from the ‘PrEP-eligible’ to ‘PrEP-intention’, and ‘PrEP intention’ to ‘current use’ step in the Netherlands. As the decline in new HIV diagnoses has stalled in the Netherlands in 2023 [18], PrEP access and continued use is even more crucial. Going forward to reaching zero new HIV infections, interventions are needed to remove barriers in order to increase PrEP uptake among those eligible and intending to use it. Peer-led programs and tailored, technology-based approaches (e.g., apps or social media content) may help overcome barriers to PrEP uptake [39]. In the long term, we recommend regular monitoring of the PrEP continuum at both national and regional levels, including information on underserved groups, to inform ongoing optimization of PrEP.

Acknowledgements

We gratefully acknowledge the numerous persons who provided data and offered context to the data that could be included in this study. Specifically, we thank Sara de Wilde and Farah Makrani from Research and Statistics of the Municipality of Amsterdam for providing data from the Safety Monitor Amsterdam-Amstelland, Hanneke de Graaf from Rutgers for providing unpublished data from the Sexual Health Monitor, Maartje Visser from the National Institute for Public Health and the Environment (RIVM) for providing national data from the Centers for Sexual Health, Steffen de Groot from the Public Health Service Amsterdam for providing data from the Center for Sexual Health of Amsterdam, Coralie Arends and Hanna Joosten from Amsterdam UMC for providing regional general practitioner data from Amsterdam (ANHA) and Sander Hogewoning from the Foundation for Pharmaceutical Statistics (SFK) for providing PrEP distribution data from pharmacies.

Authors’ contributions: I.J.M.W., E.H., J.C.M.H., and M.P. conceived and designed the study. I.J.M.W., J.C.H.M., A.S., V.W.J., B.H.B.B., J.B., E.H., and H.W. contributed to the data collection. I.J.M.W. performed data analyses under supervision of J.C.M.H. and drafted the manuscript. All authors contributed to critically revising the manuscript; the final version of the manuscript has been read and approved by all authors.

Note: Data presented previously at 16th AIDSImpact Conference Casablanca, Morocco 2025.

Ethical statement: This study utilized secondary data from publicly available online sources and reports added with pseudonymized routinely collected surveillance data from Dutch CSHs (SOAP), primary care (ANHA) and HIV Monitoring Foundation and therefore no additional ethical approval is required. Data from survey studies (Safety Monitor, Sexual Health Monitor, PROTECT) were approved by ethical committees, separate from our use of the data. The Amsterdam Cohort Studies was approved by the Medical Ethical Committee of the Academic Medical Center of Amsterdam, the Netherlands (2007_182 – NL18679.018.07 – A2007_182.0001). Written informed consent was obtained from all ACS participants at enrolment. No additional ethical approval was required, as the analysis did not involve new data collection or was conducted within the institution responsible for the original data collection.

Funding statement: Not applicable.

Conflicts of interest

There are no conflicts of interest.

Data availability

Data is available upon reasonable request from each separate data source.