Out of 304 people who met the inclusion criteria, a total of 209 people were included, of whom 67% (n = 141) completed the follow-up at 3 years. Figure 1 shows the flowchart of the included population. The mean age of the included participants was 38 years; 51% (106/209) were Latin American, and 43% (81/190) had a university education.

Fig. 1Sexual Contacts

During the follow-up, the median number of sexual partners in the previous 3 months remained stable in the range of 30–53 contacts. However, the percentage of traceable sexual partners decreased significantly over 3 years, with an IRR of 0.93 (p < 0.018) (Fig. 2). Between 10% (7/68) and 29% (7/24) of the participants reported sex work, with no difference during the study period. Most of them used drugs during sex work. (Further details are provided in the supplementary material.)

Fig. 2

Sexual contacts. M months, Y year

Sexual Practices

Unprotected anal sex was highly prevalent throughout the study period, reported by 88–96% of those engaging in anal sex. Fisting, the colloquial term to refer to brachioproctic insertion practice, remained stable over time, with prevalence ranging from 40% to 57%. The tendency to use protection decreased, with a reduction in the practice of protected brachioproctic insertion (fisting), in year 2 (IRR 1.17, 95% CI 1.02–1.34, p = 0.03) and year 3 (IRR 1.20 95% CI 1.05–1.39, p = 0.01) of the follow-up. (Further details are provided in the supplementary material.)

Drug Use

Figure 3 illustrates drug use-related characteristics. Compared with the baseline visit, the incidence of chemsex decreased significantly over the 3 years of follow-up (IRR 0.88 95% CI 0.83–0.92, p < 0.001), with significant reductions in all drug use except mephedrone use during the third year (IRR 0.70, 95% CI 0.57–0.86). (More details in supplementary material.)

Fig. 3

Drug-related characteristics. a Drug use. b Drug administration. c Frequency of drug use. d Concerns and need for help. M months, Y year, GHB/GBL gamma-hydroxybutyrate/gamma-butyrolactone, MDMA 3,4-methylenedioxymethamphetamine, Meth methamphetamine, Poppers amyl nitrite, Viagra sildenafil and derivatives, STI sexually transmitted infections

Most participants reported practicing chemsex on a weekly basis (between 30% (43/142) and 44% (37/85)) or monthly basis (between 26% (22/85) and 44% (63/142)). An increase in monthly use and a decrease in weekly use were observed, although these decreases were not significant except for a decrease in weekly use in the first year (IRR 0.81 95% CI 0.68–0.98, p = 0.029).

Slamming, the colloquial term to refer to intravenous drug use in the chemsex context, decreased significantly in the second year of follow-up (IRR 0.71 95% CI 0.52–0.98, p = 0.037) with an increasing trend thereafter that was not significant.

During follow-up, between 21% (25/119) and 38% (29/76) of the participants stated that they could not remember the last time they had had sober sex. In the second year of follow-up, there was a decrease in sober sex (IRR 0.90 95% CI 0.81–1.00, p = 0.049).

We found a significant decrease in concern about engaging in chemsex (IRR 0.52 95% CI 0.43–0.63, p < 0.001) during follow-up. However, among those who expressed concern and a need for help, the need for help with respect to chemsex practices increased from the baseline visit to become statistically significant in year 3 (IRR 1.20 95% CI 1.06–1.36, p = 0.004). The need for help related to consumption management increased (although not significantly) during the 3 years of follow-up, whereas the need for help related to possible STIs decreased in years 1 and 2, with an IRR of 0.87 (95% CI 0.76–0.99, p = 0.030) and an IRR of 0.78 (95% CI 0.66–0.92, p = 0.003), respectively. The need for help related to sexuality increased significantly in the third year of follow-up compared with baseline (IRR 1.53, 95% CI 1.20–1.94, p = 0.001).

Sexually Transmitted Infections

No significant changes were observed in the incidence of CT, with a cumulative incidence of 10% (IQR 6.4%, 14.8%) or MG infections throughout the study period. A significant decrease in NG infections was observed only during the second year of follow-up (IRR 0.59, 95% CI 0.41–0.83, p = 0.002) (Fig. 4). (More details are provided in the supplementary material.)

Fig. 4

Sexually transmitted infections. M months, Y year, VHC hepatitis C virus

The cumulative incidence of syphilis at baseline was 19.7% (95% CI 14.3–25.1%), with a significant reduction during the entire study period (IRR 0.40 95% CI 0.26–0.60, p = 0.001).

We do not have data on postexposure prophylaxis with doxycycline (DoxiPEP) because it was not officially approved during the study period and we did not ask specifically about its use. This situation might have an impact on the trends of STI observed in our cohort.

At the first visit of the study, 29% (60/209) of the participants had a positive IgG-HCV; 25% (15/60) of them had active HCV infection with positive HCV-RNA. During the 3-year follow-up, five new cases of acute hepatitis C were diagnosed, none of which were reinfections.

HIV-Associated Characteristics

HIV control during follow-up remained stable with a median CD4 count above 600 cells/mm3 and a cumulative incidence of 14.4% (95% CI 9.6–19.1%) of participants with a detectable viral load at baseline. A significant decrease in the number of participants with detectable viral loads was observed during the first year of the study (IRR 0.62, 95% CI 0.39–0.98, p = 0.042) but was not maintained during the remainder of the study period. All of these participants were on antiretroviral therapy, and the detectable viral load was related to adherence to treatment. No virological failures were confirmed. Supplementary Table S1 provides more information about cumulative incidence and 95% confidence interval (CI) of sexual practice behaviours, substance use, sexually transmitted infections, and HIV viral load.

Referrals Performed

During the 3-year period, 61% (127/209) of the participants agreed and were referred to consumption management. Among these participants, 64% (81/127) underwent follow-up, and 70% (57/81) attended regularly. More information about referrals and follow-up can be found in Fig. 5.

Fig. 5

Referral for consumption management. NGO nongovernmental organization

Evolution of Vulnerability ConditionsLoss to Follow-up

A 21% rate of loss to follow-up was noted throughout the course of the 3-year follow-up period; 77% represented actual losses, whereas the remaining 33% showed erratic follow-up patterns.

Notably, at baseline, participants lost to follow-up had lower levels of university education (10 (26%) vs. 29 (74%); p = 0.023)), were younger (mean age 35.00 (29.00; 44.00) vs. 38.00 (33.00; 43.00); p = 0.042), consumed more GHB (IRR 1.22, 95% CI 1.06–1.41, p = 0.007), were more likely to have detectable HIV viral load (IRR 3.004, 95% CI 1.57–5.76, p = 0.007), and reported a greater need for help with chemsex practices (IRR 1.28, 95% CI 1.000–1.62, p = 0.05).

Throughout the follow-up period, the participants engaged significantly more in intravenous drug use (slamming) (IRR 2.43, 95% CI 1.34–4.42, p = 0.004) and presented with more diagnoses of syphilis (IRR 1.76, 95% CI 1.20–2.60, p = 0.004).

During the follow-up, two deaths were reported, one due to cardiac arrest and the other due to lung-related cancer.

Migrant Participants

Migrant participants were significantly younger than Spanish participants (p < 0.001) and presented significantly more positive syphilis results (IRR 2.772, 95% CI 1.22–6.28, p = 0.02). During follow-up, migrant participants engaged significantly less in intravenous drug use (slamming) (IRR 0.47–95% CI 0.26–0.85, p = 0.01) but consumed more cocaine (IRR 1.46, 95% CI 1.02–2.08, p = 0.04).

Participants Engaged in Intravenous Drug Use (Slamming)

Participants with intravenous drug use were significantly older (p = 0.004), had positive HCV serology (IRR 1.935, 95% CI 1.27–2.95, p = 0.002), and used more mephedrone (IRR 1.76, 95% CI 1–2.27, p = 0.001) and methamphetamine (IRR 1.25, 95% CI 1.118–1.396, p = 0.001) at baseline. During follow-up, these participants used more methamphetamine (IRR 1.42, 95% CI 1.17–1.73, p = 0.001) and had more syphilis diagnoses (IRR 1.87 95% CI 1.26–2.77, p = 0.002).

Sex Workers

Sex workers were significantly younger (p = 0.001), were treatment-naive (p = 0.001) for HIV, had lower education levels, had more diagnoses of CT (IRR 2.60, 95% CI 1.17–5.85, p = 0.020), had fewer HCV diagnoses (IRR 0.24 95% CI 0.08–0.73, p = 0.012), and consumed less GHB/GBL (IRR 0.71, 95% CI 0.52–0.98, p = 0.039).

During follow-up, these participants continued to have more CT diagnoses (IRR 1.66, 95% CI 1.10–2.50, p = 0.016), consumed more cocaine (IRR 1.58, 95% CI 1.14–2.19, p = 0.006), had a higher percentage of detectable viral load (IRR 3.60, 95% CI 1.84–7.04, p = 0.001), and had higher rates of loss to follow-up. However, their rates of polydrug use (IRR 0.62, 95% CI 0.41–0.93, p = 0.007) and intravenous drug use (slamming) (IRR 0.34, 95% CI 0.13–0.92, p = 0.034) were lower.