Intention and preference to use long-acting injectable PrEP among MSM in the Netherlands: a diffusion of innovation approach

ABSTRACT

Long-acting injectable pre-exposure prophylaxis (LAI-PrEP) is efficacious in preventing HIV among men-who-have-sex-with-men (MSM) and will be soon available in Europe. This study investigated the intention and preference to use LAI-PrEP among MSM in the Netherlands by employing a diffusion of innovation approach. This study had a cross-sectional design nested within a cohort study established in 2017 to understand oral PrEP use among MSM. 309 MSM completed the survey on their awareness, interest, intention, and preference for LAI-PrEP in June 2022. Among them, 83% showed high/very-high interest in, and 63% showed high/very-high intention to use LAI-PrEP. A repeated innovator effect from the early adopters to LAI-PrEP was not observed. Early adopters did not show increased intention to use LAI-PrEP compared to other MSM subgroups, but neither did PrEP-naïve nor PrEP-discontinued MSM. However, among the 218 current oral PrEP users, suboptimal adherence was associated with preference for LAI-PrEP but not with intention to use it. In conclusion, our findings indicated that an effective, available, and affordable LAI-PrEP would be welcomed in the Netherlands, but that its introduction may not significantly expand PrEP coverage. However, the introduction of LAI-PrEP in the Netherlands could prove beneficial to MSM with suboptimal adherence to oral PrEP.

KEYWORDS:

• Men who have sex with men
• long-acting injectable PrEP
• diffusion of innovation
• suboptimal PrEP adherence
• intention and preference

Introduction

Long-acting injectable (LAI) pre-exposure prophylaxis (PrEP), such as cabotegravir, is superior to tenofovir disoproxil fumarate/emtricitabine (TDF/FTC) as a daily oral regimen (Landovitz et al., 2021). Previous studies have suggested that 60–78% of men-who-have-sex-with-men (MSM) would be willing to use LAI-PrEP (Chan et al., 2022; Greene et al., 2017; Meyers et al., 2014; Meyers et al., 2018; Tolley et al., 2020). However, a European perspective including the Netherlands is missing and warranted, as findings from other contexts do not always directly translate, and especially since cabotegravir is expected to be available in Europe soon (ViiV, 2022). Given its enhanced efficacy and alternative administration route, it is crucial to understand the specific subpopulations of MSM that would be most interested in and would benefit the most from LAI-PrEP. In turn, LAI-PrEP can be targeted and prioritised more efficiently to reach a higher PrEP coverage along the oral regimens.

Adopting a diffusion of innovation (DOI) approach can provide valuable insights by leveraging the experience gained from the introduction of oral PrEP, thereby facilitating the successful implementation of LAI-PrEP. DOI theory explores the diffusion of new ideas, practices, and technologies within social systems, shedding light on their acceptance and normativity (Rogers et al., 2014), and has identified subpopulations in terms of the timepoint of their acceptance or adoption of a novel behaviour (e.g., early adopters for those that start to use an innovation prior to a majority). DOI has been used to describe the acceptance of novel medication in the past but mostly from a healthcare provider (HCP) perspective (Heinrich & Cummings, 2014; Ruof et al., 2002; Steffensen et al., 1999). Yet, it is also possible to apply DOI to clients or patients. In the context of PrEP, one can classify the onset of oral regimen use among MSM into DOI subpopulations based on their history of use, subsequently allowing for a detailed analysis of the MSM subpopulations and their intention to use LAI-PrEP.

Therefore, we turn to the Dutch context. In the Netherlands, TDF/FTC was formally introduced and included in the national PrEP programme via the Public Health Services in 2019 (GGD-Amsterdam, 2019). Via this access, 8,500 individuals can be served, mainly targeting MSM with a higher risk for an HIV infection. Prior to 2019, TDF/FTC could be procured informally (e.g., through buyers’ clubs, online pharmacies), via a pilot study in Amsterdam (AmPrEP), or via general practitioners that prescribed it at full cost for patients. Consequently, a substantial number of MSM had been using TDF/FTC in the Netherlands since its approval in 2015 and prior to its formal availability in 2019 (Coyer et al., 2018; Hoornenborg et al., 2017; van Dijk, et al. 2021).

When LAI-PrEP becomes formally available in the Netherlands, these current TDF/FTC users should not be overlooked and neglected. Specifically, the differentiation of the early adopter and the majority users allows for investigating whether these early adopters of oral PrEP would be at the forefront again of adopting LAI-PrEP. Furthermore, the current users’ adherence to the oral regimens may further determine their intention and preference to switch to LAI-PrEP. HIV seroconversions can still occur at any time along the PrEP care continuum (Nunn et al., 2017). This is most likely due to suboptimal adherence to the oral regimens (Jourdain et al., 2022). As LAI-PrEP only needs to be administrated once every eight weeks, it was suggested to have the potential to alleviate the adherence challenges (Levy et al., 2021). Consequently, current TDF/FTC users, particularly those with suboptimal adherence, could benefit from using LAI-PrEP to overcome their struggles of adhering to oral regimens.

Additionally, to reach a greater reduction in HIV incidence, recent studies have highlighted the need for improved PrEP coverage (Mitchell et al., 2023; Stansfield et al., 2023; Wang et al., 2023). Consequently, novel groups that can be expected to benefit from LAI-PrEP to reach a higher PrEP coverage: MSM who had never used PrEP (hereinafter “PrEP-naïve”) due to a lack of interest in oral medication uptake (Krist et al., 2022; van Dijk, Duken, et al. 2020); and MSM who discontinued PrEP but remain at risk (Jonas & Yaemim, 2018; Jongen et al., 2021). It is important to investigate whether there is a match between a higher intention to use LAI-PrEP and the expected benefits for current non-oral PrEP users.

Taken together, this study aimed to provide unique insights into the awareness, interest, intention, and preference for using LAI-PrEP by fusing a DOI approach, the PrEP use status of MSM, and their adherence. More specifically, by comparing different MSM subpopulations based on their history of PrEP use, this study aimed to understand who would be most interested in using LAI-PrEP and who would potentially benefit the most from using LAI-PrEP.

Methods

Participants and procedures

This study had a cross-sectional study design nested within a cohort study. All MSM living in the Netherlands who participated in this study were recruited from an ongoing cohort which was established in 2017 (T0, prior to formal PrEP access in the Netherlands) and received follow-up questionnaires at three (T1) and six months (T2) to understand oral PrEP use among MSM (van Dijk et al., 2020), and in 2020 during the COVID-19 pandemic (T3) to understand their sexual behaviours and experiences with PrEP during the pandemic (Krist et al., 2022). Participants were contacted again in June 2022 to complete an online survey on their awareness, interest, intention, and preference to use LAI-PrEP (T4). Only MSM without a diagnosed HIV status and currently living in the Netherlands were included in this study. This study was approved by the ERCPN of Maastricht University (ERCPN-174_10_12_2016). Informed consent was provided by all participants.

Measures

Participants were asked whether they were aware of LAI-PrEP (“Yes”/ “No”) (hereinafter LAI-PrEP awareness). After providing the answer regardless of their awareness, all participants were provided with a detailed introduction on what is LAI-PrEP, how to take LAI-PrEP, and the differences between LAI-PrEP and oral PrEP including mechanisms, effective components, effectiveness in HIV prevention, and potential side effects. Then, participants were asked whether they are interested in using LAI-PrEP when it becomes formally available (hereinafter LAI-PrEP interest), measured on a 1–5 Likert scale (with 1 = “definitely no” and 5 = “definitely yes”). We dichotomised LAI-PrEP interest as “Definitely Yes/Yes” and “Not sure/No/Definitely No”.

Participants were further asked how likely is it that they will start using LAI-PrEP if it becomes formally available (hereinafter LAI-PrEP intention), measured on a 1–5 Likert scale (with 1 = “definitely no” and 5 = “definitely yes”). We dichotomised LAI-PrEP intention as “High/very high” and “Not sure/Low/very low”. Participants were additionally asked about their intention to use LAI-PrEP after being given a scenario of “Due to the concerns of the potential drug resistance after stopping with LAI-PrEP, you may be asked to take oral PrEP daily for 180 days after you decide to no longer take the LAI-PrEP”. This was measured as same as the LAI-PrEP intention.

Next, participants who indicated that they were currently using TDF/FTC were asked which regimen option they will prefer to use if they are all available (hereinafter LAI-PrEP preference), given the options of “LAI-PrEP”, “Daily oral PrEP”, “On-demand oral PrEP”, “Mixed-use of LAI-PrEP and oral PrEP”, and “Do not want to use PrEP at all”. For LAI-PrEP preference as the endpoint, following the same approach, we also dichotomised it as “LAI-PrEP or mixed use of LAI-PrEP and oral PrEP” vs. “Oral PrEP”, given no current oral PrEP user was found to not want to use PrEP.

Socio-demographic variables included age, sexual orientation (categorised as “Gay”, “Bisexual” or “Other”), education attainment (“Low” as who did not have a high-school diploma, or “High” as who had high school and higher than a high-school education, following the definition of education level in the Netherlands (CBS, 2020)), financial status (self-reported during T4 assessment, dichotomised as “Low” or “Living comfortable/very comfortable with the current income”), migration background (“Dutch native”, “European” or “Non-European”), place of residence location (“Main urban cities” or “The rest of the country”, based on the provided 4-digit postal codes), and relationship status (“single, not dating”, “single but dating”, “in a monogamous relationship”, or “in an open or polyamorous relationship”). In addition, we tested whether socio-demographics differed between participants included in the original intake cohort (T0) and the MSM included in this study (T4) to interpret the representativeness of our sample. Given the different measurements of the socio-demographic characteristics between T0 and T4, we were only able to compare age, education attainment, financial status, and relationship status.

Sexual behavioural variables included oral PrEP use status following DOI approach (“early adopter” (initiation before 2019), “majority user” (initiation after 2019), “PrEP naïve” or “MSM who discontinued oral PrEP” (any self-reported discontinuation of using PrEP during T4 assessment), based on our longitudinal data), having a steady partner with diagnosed HIV (“Yes” or “No”), having a non-steady partner with diagnosed HIV (“Yes”, “No” or “Not sure”), ever had any type of sexually transmitted infection (STI) (“Yes, within the previous 12 months”, “Yes, more than 12 months ago” or “Never”), ever engaged in chemsex (“Yes” or “No”), ever paid for sex (“Yes, within the previous 12 months”, “Yes, more than 12 months ago” or “Never”), and ever got paid for sex (“Yes, within the previous 12 months”, “Yes, more than 12 months ago” or “Never”).

Additionally, adherence to TDF/FTC was also assessed via self-report among current oral PrEP users and their oral PrEP use modes: Daily, on-demand, and Tuesday-Thursday-Saturday-Sunday (TTSS) scheme. For daily and TTSS scheme users, we measured their adherence by asking “How often do you miss a pill”. We defined “Never” or “Miss using pill once or twice a month” as “Good adherence”. For on-demand users, we measured their adherence by asking (1) “How often do you just take 1 pill in advance”, (2) “How often did you have sex less than 2hrs after the first 2 pills”, and (3) “How often do you forget to take a pill 24 and/or 48hrs after”. We defined on-demand users who “Always” or “Almost all the time” take 2 pills in advance and have no sex less than 2 hrs after the first 2 pills, and who “Never” or “Occasionally” forget to take a pill 24 and 48 hrs after as “Good adherence”. An overall oral PrEP adherence was re-coded and dichotomised as “Good adherence” or “Suboptimal adherence”.

Statistical analysis

We first conducted descriptive analysis for all included variables for all participants and by their LAI-PrEP intention. We then conducted another descriptive analysis for the current TDF/FTC users and by their LAI-PrEP intention. Fisher's exact test and chi-square test were used to test the differences in the included variables between LAI-PrEP intention.

We conducted three multivariable logistic regression analyses to identify socio-demographic, and behavioural determinants of LAI-PrEP intention among MSM, and LAI-PrEP intention and LAI-PrEP preference among current oral PrEP users. For each regression analysis, we first utilised a univariable logistic model with each of the selected socio-demographic, and sexual behavioural determinants to investigate the potential correlation with each endpoint. Next, we performed a multivariable logistic model using a manual stepwise backward selection approach including all variables with p < 0.10 in the univariate analysis. We considered all variables with p < 0.05 statistically significant. For the general MSM analysis, current oral PrEP use status was retained regardless of the significance; and for the current TDF/FTC user analyses, both current oral PrEP use status and their adherence to oral PrEP were retained regardless of the significance, given they were the key variables of interest in this study. All analyses were conducted in R (version R 4.2.1).

Results

Study characteristics

All 758 cohort participants from T0 since 2017 were invited to participate in this study, of which 327 completed this survey. We excluded 18 participants from this analysis due to having received an HIV diagnosis (n = 6) and not currently living in the Netherlands (n = 12), resulting in 309 participants in total included in this study (Figure 1).

Figure 1. Flowchart of participants per data point.

Note: LAI-PrEP = long-acting injectable pre-exposure prophylaxis. Detailed information for each data point see our previous study (Krist et al., 2022).

Among all MSM included in this study, the median age was 47 (interquartile range (IQR) 38–54), 15% (45/309) were PrEP-naïve, 15% (46/309) discontinued oral regimens, 45% (128/309) were early adopters (Table 1). Among the current TDF/FTC users, the median age was 48 (IQR 39–55), 59% (128/218) were early adopters and 10% (22/218) reported suboptimal adherence (Table 2). Compared to the original intake cohort, MSM included in this study were older, higher educated, more financially disadvantaged, and more often in a relationship. Differences in socio-demographics between the participants included in this study and in the original intake cohort are reported in Supplementary S1 table.

Table 1. Study characteristics of all included MSM.

Variables		Total sample (n = 309)		High or very high intention to use LAI-PrEP (n = 194)		Not sure, low or very low intention to use LAI-PrEP (n = 115)		P-value (Fisher's exact test or chi-square test)
		n/median	%/IQR	n/median	%/IQR	n/median	%/IQR
Age		47	38–54	46	37–53	50	42–56	0.002
Sexual orientation
	Gay	287	93.2	181	93.3	106	92.2	0.67
	Bisexual	13	4.2	9	4.6	4	3.5
	Other	9	2.6	4	2.1	5	4.3
Education
	Low	33	10.7	23	11.9	10	8.7	0.45
	High	276	89.3	171	88.1	105	91.3
Employment
	Employed	279	90.3	178	91.8	101	87.8	0.53
	Unemployed	5	1.6	3	1.5	2	1.7
	Retired	19	6.1	9	4.6	10	8.7
	Student	6	1.9	4	2.1	2	1.7
Finance
	Low	93	30.1	64	33.0	29	25.2	0.16
	Living comfortable/very comfortable with current income	216	69.9	130	67.0	86	74.8
Migrant status
	Dutch	249	80.6	148	76.3	101	87.8	0.04
	European	20	6.5	16	8.2	4	3.5
	Non-European	40	12.9	30	15.5	10	8.7
Geo-location
	Main urban cities	211	68.3	128	66.0	83	72.2	0.31
	The rest of the country	98	31.7	66	34.0	32	27.8
TDF/FTC use status
	Early adopter	128	41.4	86	44.3	42	36.5	0.03
	Majority oral PrEP user	90	29.1	62	32.0	28	24.3
	PrEP naïve	45	14.6	21	10.8	24	20.9
	MSM who discontinued oral PrEP	46	14.9	25	12.9	21	18.3
LAI-PrEP awareness
	Yes	124	40.1	82	42.3	42	36.5	0.16
	No	185	59.9	112	57.7	73	63.5
LAI-PrEP interest
	Definitely yes/Yes	256	82.8	192	99.0	64	55.7	<0.001
	Definitely no/No/Not sure	53	17.2	2	1.0	51	44.3
LAI-PrEP intention
	High/very high	194	62.8	NA	NA	NA	NA	NA
	Not sure/Low/very low	115	37.2	NA	NA	NA	NA
LAI-PrEP intention with 180 days obligation
	Yes	152	57.1	131	70.1	21	26.6	<0.001
	No	72	27.1	31	16.6	41	51.9
	Not sure	42	15.8	25	13.4	17	21.5
Relationship status
	Single	118	38.2	84	43.3	34	29.6	0.06
	Single but dating	12	3.9	9	4.6	3	2.6
	In a relationship	78	25.2	43	22.2	35	30.4
	In an open/polyamorous relationship	101	32.7	58	29.9	43	37.4
Having steady partner with diagnosed HIV
	Yes	159	51.5	96	49.5	63	54.8	0.41
	No	150	48.5	98	50.5	52	45.2
Having non-steady partner with diagnosed HIV
	Yes	114	36.9	76	39.2	38	33.0
	No	110	35.6	68	35.1	42	36.5	0.56
	Not sure	85	27.5	50	25.8	35	30.4
STIs diagnosis
	Yes, within the previous 12 months	102	33.0	74	38.1	28	24.3	0.04
	Yes, more than 12 months ago	157	50.8	90	46.4	67	58.3
	Never	50	16.2	30	15.5	20	17.4
Ever engage in Chemsex
	Yes	132	42.7	93	47.9	39	33.9	0.02
	No	177	57.3	101	52.1	76	66.1
Buy sex
	Yes, within the previous 12 months	34	11.0	18	9.3	16	13.9	0.32
	Yes, more than 12 months ago	16	5.2	12	6.2	4	3.5
	Never	259	83.8	164	84.5	95	82.6
Paid sex (n = 305)
	Yes, within the previous 12 months	31	10.0	22	11.3	9	7.8	0.14
	Yes. more than 12 months ago	9	2.9	8	4.1	1	0.9
	Never	256	85.8	161	83.0	104	90.4

Note: LAI-PrEP = long-acting pre-exposure prophylaxis. NA = not applicable.

Table 2. Study characteristics of MSM who were currently using PrEP.

Variables		Total sample (N = 218)		High or very high intention to use LAI-PrEP (n = 148)		Not sure, Low or very low intention to use LAI-PrEP (n = 70)		P-value (Fisher's exact test or chi-square test)
		n/Median	%/IQR	n/Median	%/IQR	n/Median	%/IQR
Age		48	39–55	46	37–53	52	44–59	<0.001
TDF/FTC use status
	Majority oral PrEP user	90	41.3	62	41.9	28	40.0	0.88
	Early adopter	128	58.7	86	58.1	42	60.0
TDF/FTC adherence
	Perfect/good adherence	196	89.9	135	91.2	61	87.1	0.65
	Poor adherence	22	10.1	13	8.8	9	12.9
Finance
	Low	52	28.4	49	33.1	13	18.6	0.03
	Living comfortable/very comfortable with current income	156	71.6	99	66.9	67	95.7
Migrant status
	Dutch	174	79.8	111	75.0	63	90.0	0.02
	European	14	6.4	13	8.8	1	1.4
	Non-European	30	13.8	24	16.2	6	8.6
Relationship status
	Single	84	38.5	63	42.6	21	30.0	0.15
	Single but dating	9	4.1	7	4.7	2	2.9
	In a relationship	37	17.0	26	17.6	11	15.7
	In an open/polyamorous relationship	88	40.4	52	35.1	36	51.4
STI diagnosis
	Yes, within the previous 12 months	93	42.7	69	46.6	24	34.3	0.04
	Yes, more than 12 months ago	18	49.1	14	9.5	42	60.0
	Never	107	8.3	65	43.9	4	5.7
Ever engage in Chemsex
	Yes	112	51.4	79	53.4	33	47.1	0.47
	No	106	48.6	69	46.6	69	98.6
Buy sex
	Yes, within the previous 12 months	13	6.0	10	6.8	3	4.3	0.16
	Yes, more than 12 months ago	22	10.1	11	7.4	11	15.7
	Never	183	83.9	127	85.8	56	80.0
Paid sex (n = 216)
	Yes, within the previous 12 months	7	3.2	6	4.1	1	1.4	0.11
	Yes, more than 12 months ago	20	9.2	17	11.5	3	4.3
	Never	189	87.5	123	83.1	66	94.3

Note: LAI-PrEP = long-acting pre-exposure prophylaxis.

LAI-PrEP awareness, interest, intention, and preference

Among all MSM, 40% (124/309) were aware of LAI-PrEP, 83% (256/309) showed high/very high interest in, and 63% (194/309) showed high/very high intention to use LAI-PrEP. After being informed of the potential 180-day obligation of using oral PrEP daily after discontinuing using LAI-PrEP, still, 57% (152/309) reported high/very high intention to use LAI-PrEP (Table 1). Despite the overall moderate LAI-PrEP awareness and high LAI-PrEP interest and intention, early adopters of oral PrEP were found to have the highest LAI-PrEP awareness, interest, and intention, while PrEP naïve MSM was found to have the lowest (Figure 2). Among the current TDF/FTC users, half of them (104/218) were aware of LAI-PrEP, 82% (179/218) showed high/very high interest in, and 68% (148/218) showed high/very high intention to use LAI-PrEP. Half of them (110/218) showed a preference for LAI-PrEP and 18% (39/218) would prefer to use LAI-regimens and oral regimens at different times over the oral regimen preference (Table 2).

Figure 2. Distribution of (a) LAI-PrEP awareness among MSM by oral PrEP use status, (b) LAI-PrEP interest among MSM by oral PrEP use status, and (c) LAI-PrEP intention among MSM by oral PrEP use status.

Note: LAI-PrEP = long-acting injectable pre-exposure prophylaxis.

Determinants of LAI-PrEP intention among MSM

For high/very high LAI-PrEP intention as the endpoint, being an early adopter did not show a higher LAI-PrEP intention compared to the majority users and MSM who discontinued PrEP. Yet, PrEP-naïve MSM had a statistically lower LAI-PrEP intention (adjusted odds ratio (aOR) = 0.39, 95%CI = 0.16–0.92, p = 0.03) compared to early adopters. In addition, MSM who were older (aOR = 0.97, 95%CI = 0.95–0.99, p = 0.02), and who never got paid for sex (aOR = 0.03, 95%CI = 0.00–0.75, p = 0.03) were less likely to have an LAI-PrEP intention (Table 3).

Table 3. Determinants of LAI-PrEP intention (high/very high vs. the rest of the scale) among MSM.

Variables		Univariable model			Multivariable model
		OR	95%CI	P-value	aOR	95%CI	P-value
Age		0.97	0.95–0.99	0.02	0.97	0.94–0.99	0.01
Sexual orientation
	Gay	Ref
	Bisexual	1.32	0.42–4.96	0.65
	Other	0.59	0.14–2.52	0.46
Education
	Low	Ref
	High	0.70	0.31–1.51	0.39
Employment
	Employed	Ref
	Unemployed	0.85	0.14–6.54	0.51
	Retired	0.51	0.20,1.30	0.88
	Student	1.53	0.22–8.29	0.86
Finance
	Low	Ref
	Living comfortable/very comfortable with current income	0.68	0.40–1.14	0.15
Migrant status
	Dutch	Ref
	European	1.00	0.969.74	0.08
	Non-European	0.71	0.99–4.58	0.06
Geo-location
	Main urban cities	Ref
	The rest of the country	1.34	0.81–2.23	0.25
TDF/FTC use status
	Early adopter	Ref			Ref
	Majority oral PrEP user	1.08	0.61–1.49	0.79	0.70	0.34–1.42	0.33
	PrEP naïve	0.42	0.21–0.85	0.01	0.39	0.16–0.92	0.03
	MSM who discontinued oral PrEP	0.58	0.29–1.16	0.12	0.46	0.18–1.14	0.09
Relationship status
	Single	Ref
	Single but dating	1.21	0.34–5.72	0.78
	In a relationship	0.50	0.27–0.90	0.02
	In an open/polyamorous relationship	0.55	0.31–0.95	0.03
Having steady partner with diagnosed HIV
	Yes	Ref
	No	1.24	0.78–1.97	0.36
Having non-steady partner with diagnosed HIV
	Yes	Ref
	No	0.81	0.47–1.40	0.45
	Not sure	0.74	0.41–1.32	0.30
STI diagnosis
	Yes, within the previous 12 months	Ref
	Yes, more than 12 months ago	0.51	0.29–0.86	0.01
	Never	0.57	0.28–1.16	0.12
Ever engage in Chemsex
	Yes	Ref
	No	0.56	0.34–0.89	0.02
Buy sex
	Yes, within the previous 12 months	Ref
	Yes, more than 12 months ago	0.38	0.09–1.32	0.14
	Never	0.58	0.16–1.70	0.35
Paid sex (n = 305)
	Yes, within the previous 12 months	Ref			Ref
	Yes, more than 12 months ago	0.31	0.02–2.04	0.30	0.05	0.00–1.31	0.07
	Never	0.19	0.01–1.08	0.12	0.03	0.00–0.75	0.03

Note: § indicate variables were measured on a 1–5 agreeable scale (with 1 = “Disagree strongly” and 5 “Agree strongly”). aOR = adjusted odds ratio. CI = confident interval. OR = odds ratio.

Determinants of LAI-PrEP intention and preference among current oral PrEP users

For high/very high LAI-PrEP intention as the endpoint among current TDF/FTC users, being an early adopter (aOR = 1.47, 95%CI = 0.72–3.02, p = 0.29) or reporting suboptimal adherence (aOR = 0.91, 95%CI = 0.42–1.96, p = 0.78) were not associated with a high LAI-PrEP intention. Yet, those who were older (aOR = 0.95, 95%CI = 0.92–0.98, p < 0.01) were less likely to have higher LAI-PrEP intention (Table 4).

Table 4. Determinants of LAI-PrEP intention and preference among current oral PrEP users.

Variables		LAI-PrEP intention (High/very high vs. the rest of the scales)						LAI-PrEP preference (LAI-PrEP or mixed use of LAI-PrEP and oral PrEP vs. Oral PrEP)
		Univariable model			Multivariable model			Univariable model			Multivariable model
		OR	95%CI	P	aOR	95%CI	P-value	OR	95%CI	P-value	aOR	95%CI	P-value
Age		0.95	0.92–0.98	<0.01	0.95	0.92–0.98	<0.01	1.02	0.99–1.05	0.07	1.02	0.99–1.05	0.11
TDF/FTC use status
	Majority PrEP user	Ref			Ref			Ref
	Early adopter	0.92	0.52–1.65	0.791	1.47	0.72–3.02	0.29	0.94	0.53–1.68	0.84	0.91	0.50–1.66	0.76
TDF/FTC adherence
	Perfect/Good adherence	Ref			Ref			Ref			Ref
	Poor adherence	0.65	0.26–1.65	0.35	0.91	0.42–1.96	0.78	2.27	0.92–5.57	0.07	2.39	0.94–6.04	0.06
Finance
	Low	Ref			Ref			Ref
	Living comfortable/very comfortable with current income	0.46	0.22–0.90	0.02	0.39	0.15–1.03	0.06	1.13	0.61–2.17	0.70
Migrant status
	Dutch	Ref						Ref			Ref
	European	1.99	0.94–57,73	0.06				0.14	0.02–1.08	0.06	0.14	0.02–1.12	0.06
	Non-European	0.82	0.88–4.84	0.09				0.65	0.27–1.56	0.34	0.85	0.33–2.18	0.74
Relationship status
	Single	Ref						Ref
	Single but dating	1.17	0.25,5.03	0.85				2.00	0.46–8.19	0.33
	In a relationship	0.79	0.33–1.90	0.58				1.90	0.84–4.27	0.11
	In an open/polyamorous relationship	0.48	0.25–0.92	0.03				1.10	0.57–2.13	0.76
STI diagnosis
	Yes, within the previous 12 months	Ref						Ref
	Yes, more than 12 months ago	0.53	0.29–0.97	0.04				1.78	0.97–3.30	0.51
	Never	1.21	0.39–4.61	0.74				1.43	0.46–4.14	0.06
Ever engage in Chemsex
	Yes	Ref						Ref
	No	0.78	0.44–1.37	0.39				1.04	0.58–1.83	0.89
Buy sex
	Yes, within the previous 12 months	Ref						Ref
	Yes, more than 12 months ago	0.30	0.05–1.29	0.12				1.86	0.45–8.66	0.39
	Never	0.68	0.14–2.32	0.57				1.02	0.31–3.88	0.98
Paid sex (n = 216)
	Yes, within the previous 12 months	Ref						Ref
	Yes, more than 12 months ago	0.31	0.02–1.87	0.28				1.50	0.17–32.55	0.73
	Never	0.94	0.04–9.17	0.96				3.21	0.53–61.48	0.28

Note: § indicate variables were measured on a 1–5 agreeable scale (with 1 = “Disagree strongly” and 5 “Agree strongly”). aOR = adjusted odds ratio. CI = confident interval. OR = odds ratio.

For having an LAI-PrEP preference as the endpoint among the current TDF/FTC users, being an early adopter was not associated with a higher odd of LAI-PrEP preference. Yet, there was a trend for current TDF/FTC users with suboptimal adherence (aOR = 2.39, 95%CI = 0.94–6.04, p = 0.06) to be more likely to have an LAI-PrEP preference (Table 4).

Discussion

Our study, based on 309 MSM from a longitudinal cohort in the Netherlands, showed that the LAI-PrEP intention and preference among MSM are similar to other intention and preference studies among the MSM population outside Europe (Chan et al., 2022; Greene et al., 2017; Meyers et al., 2014; Meyers et al., 2018; Tolley et al., 2020). Our findings indicated that an effective, available, and affordable LAI-PrEP would be welcomed in the Netherlands.

However, not aligning with our expectations, our findings showed that PrEP-naïve and MSM who discontinued oral PrEP, as the potentially targeted group for a higher PrEP coverage, were not more likely to have an LAI-PrEP intention compared to current TDF/FTC users. Notably, our findings showed that PrEP-naïve MSM had a significantly lower intention for LAI-PrEP compared to other subgroups. These findings indicated that the introduction of LAI-PrEP may not motivate the non-current PrEP users to start using LAI-PrEP, as hypothesised in other contexts such as New York City and China (Meyers et al., 2014; Meyers et al., 2018). Consequently, LAI-PrEP may not seem complementary and may not serve a larger population to reach higher levels of PrEP coverage among MSM in the Netherlands.

In addition, surprisingly, our study did not observe a repeated innovator effect on the intention and preference for LAI-PrEP by the early adopters. Our analyses indicated that being an early adopter of oral PrEP was neither associated with a higher intention, nor with a preference for LAI-PrEP compared to the majority users. One possible reason could be the similarly identified determinants of LAI-PrEP and oral PrEP intention. Our analyses identified that MSM who were younger and who got paid for sex were more likely to use LAI-PrEP. These findings aligned with the previous investigations for oral PrEP intention that younger MSM (Holt et al., 2019), and MSM who were involved in transactional sex (Wang et al., 2022) were more likely to use oral PrEP. In turn, these findings may indicate a similar perception of LAI-PrEP and oral regimens, suggesting that LAI-PrEP may be considered as just a new format or regimen instead of a completely novel biomedical HIV prevention tool. Consequently, this conceptualisation may not motivate those early adopters to be at the forefront of the LAI-PrEP again. Another possible reason may be the positive experience with oral regimens among MSM in the past. Using oral regimens can be seen as sufficient to maintain a high quality of sex life (van Dijk et al., 2020). In turn, even though using LAI-PrEP can help alleviate the adherence challenges of oral regimens, it may not be sufficiently motivating for current TDF/FTC users, especially when they have good adherence (90% of our samples), to switch to LAI-PrEP, regardless of their oral regimens using status.

Yet, our analyses showed that the current TDF/FTC users with suboptimal adherence, as a subgroup of whom are not well-served by PrEP, were more likely to prefer to use LAI-PrEP if it becomes formally available. Even though this effect was not statistically significant, we considered this effect relevant for public health and clinical context, given how TDF/FTC adherence can influence its effectiveness (Jourdain et al., 2022). With a higher preference for LAI-PrEP and without the concerns of medication adherence, this sub-population may benefit the most from using LAI regimens. This finding thus pointed out the alternative targeted population for LAI-PrEP: Current TDF/FTC users with suboptimal adherence.

The major strength of this study was the application of the DOI theory, with a focus on PrEP users instead of HCPs, to investigate the LAI-PrEP intention and preference among MSM and oral PrEP users for the first time. This approach ensures that the unique needs of different subpopulations are considered. Consequently, HCPs and community activists can tailor implementation strategies and identify key influencers for successful adoption. Another strength of this study is the cross-sectional design nested in a cohort study. Given the known and validated PrEP use status since 2017, our data on the PrEP use status was thus less likely subjected to information bias for the estimation following the DOI approach.

Our study has several limitations. First, due to insufficient data on the adoption history of oral regimens, our study could only categorise our samples into four of the five DOI categories (Rogers et al., 2014). A furthermore detailed breakdown of majority users into “early majority” and “late majority” was unfortunately not possible. Second, the cost of LAI regimens, such as cabotegravir, remains unknown in Europe. The findings from our study were hence based on the scenario of formal introduction (available and affordable). Our finding may have resulted in an overestimation in the case of an unaffordable LAI-PrEP, given the established impact of the cost on oral PrEP intention in the Netherlands (van Dijk et al., 2020). Therefore, this leaves future studies to assess the intention of LAI-PrEP in different scenarios of the costs of the LAI-PrEP to further understand how the costs of LAI regimens may influence MSM's intention and preference. Third, given this study's relatively small sample size, our relevant findings may be biased to the limited power. Therefore, future studies should investigate with a larger sample size to provide more concise estimations. Furthermore, we recognise the presence of socio-demographic differences among MSM participants in this study compared to the original intake cohort, as outlined in Supplementary S1 table. It is essential to note that the cohort's inception was in 2017 (T0), while our study data (T4) was collected in 2022. Therefore, we deem this 7-year age difference as irrelevant. Additionally, our logistic regression analyses did not reveal any substantial impact of the other socio-demographic characteristics. Consequently, we believe that differences in these characteristics among MSM participants in our study and the original intake cohort are unlikely to introduce bias into our findings. Lastly, given the original study cohort was to understand PrEP use among MSM in the Netherlands, the attitude towards PrEP use among our participants in the cohort may be more positive. The generalisability of our findings may thus be limited to the Dutch general MSM population.

In conclusion, we observed a high interest and intention to use LAI-PrEP among MSM and a high preference for LAI-PrEP among the current TDF/FTC users in the Netherlands. Early adopters of past oral PrEP use did not show increased intention to use LAI regimens and neither did PrEP-naïve nor PrEP-discontinued MSM. Suboptimal adherence determines LAI-PrEP preference but did not determine LAI-PrEP intention. Therefore, the introduction of LAI-PrEP in the Netherlands could prove beneficial for a subgroup of MSM who experience challenges with adherence to oral regimens, but may not significantly expand PrEP coverage. To fully leverage the potential of LAI-PrEP for HIV prevention, current non-oral PrEP using MSM need to be targeted for a higher PrEP coverage, but MSM who currently exhibit suboptimal PrEP adherence should be prioritised for LAI-PrEP when an intention is indicated.

Acknowledgements

The authors would like to thank all study participants, who generously gave their time and efforts for this research.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

The author(s) reported there is no funding associated with the work featured in this article.