Open access Original research STI epidemic re-­emergence, socio-­ epidemiological clusters characterisation and HIV coinfection in Catalonia, Spain, during 2017–2019: a retrospective population-­based cohort study Alexis Sentís ‍ ‍,1,2,3 Marcos Montoro-­Fernandez,1,4 Evelin Lopez-­Corbeto,1,4,5 Laia Egea-­Cortés,1,4 Daniel K Nomah,1,6 Yesika Díaz,1,4 Patricia Garcia de Olalla,5,7 Lilas Mercuriali,7 Núria Borrell,8 Juliana Reyes-­Urueña ‍ ‍,1,4,5 Jordi Casabona,1,4,5,6 The Catalan HIV and STI Surveillance Group To cite: Sentís A, Montoro-­ ABSTRACT Fernandez M, Lopez-­ Strengths and limitations of this study Objectives To describe the epidemiology of sexually Corbeto E, et al. STI transmitted infections (STIs), identify and characterise epidemic re-­emergence, ►► In this retrospective population-­based cohort study, socio-­epidemiological clusters and determine factors socio-­epidemiological clusters the use of data from the Catalan HIV/STI Registry associated with HIV coinfection. characterisation and HIV allowed us to characterise the re-­ emergence of coinfection in Catalonia, Design Retrospective population-­based cohort. sexually transmitted infections (STIs), perform Spain, during 2017–2019: Setting Catalonia, Spain. socio-­epidemiological clustering and reveal factors a retrospective population-­ Participants 42 283 confirmed syphilis, gonorrhoea, associated with HIV coinfection. based cohort study. BMJ Open chlamydia and lymphogranuloma venereum cases, among ►► To the best of our knowledge, this is the first study to 2021;11:e052817. doi:10.1136/ 34 600 individuals, reported to the Catalan HIV/STI Registry bmjopen-2021-052817 apply the k-­means clustering methodology to iden- in 2017–2019. tify and characterise distinct socio-­epidemiological ►► Prepublication history and Primary and secondary outcomes Descriptive analysis clusters of STI at a small health area level. additional supplemental material of confirmed STI cases and incidence rates. Factors ►► A key limitation of this study is the high proportion for this paper are available associated with HIV coinfection were determined using of missing data around sociodemographic and life- online. To view these files, logistic regression. We identified and characterized socio-­ style characteristics such as education level, sexual please visit the journal online epidemiological STI clusters by Basic Health Area (BHA) (http://​dx.​doi.​org/​10.​1136/​ preference and country of birth. Nonetheless, our using K-­means clustering. findings are consistent with previous analyses. bmjopen-​2021-​052817). Results The incidence rate of STIs increased by 91.3% Received 08 May 2021 from 128.2 to 248.9 cases per 100 000 population Accepted 12 November 2021 between 2017 and 2019 (p<0.001), primarily driven by populations at a small health area level to design effective increase among women (132%) and individuals below 30 interventions. years old (125%). During 2017–2019, 50.1% of STIs were chlamydia and 31.6% gonorrhoea. Reinfections accounted for 10.8% of all cases and 6% of cases affected HIV-­ INTRODUCTION positive individuals. Factors associated with the greatest The epidemic of sexually transmitted infec- likelihood of HIV coinfection were male sex (adjusted OR tions (STIs) continues to be a major concern (aOR) 23.69; 95% CI 16.67 to 35.13), age 30–39 years and threat to global public health. Undi- (versus <20 years, aOR 18.58; 95% CI 8.56 to 52.13), agnosed and untreated STIs can lead to a having 5–7 STI episodes (vs 1 episode, aOR 5.96; 95% CI multitude of complications including HIV 4.26 to 8.24) and living in urban areas (aOR 1.32; 95% CI acquisition, long-­term disabilities, infertility, © Author(s) (or their 1.04 to 1.69). Living in the most deprived BHAs (aOR adverse pregnancy outcomes and death.1 2 employer(s)) 2021. Re-­use 0.60; 95% CI 0.50 to 0.72) was associated with the least Across Europe, incidence of STIs continue to permitted under CC BY-­NC. No likelihood of HIV coinfection. K-­means clustering identified commercial re-­use. See rights be on the rise with confirmed cases reported three distinct clusters, showing that young women in and permissions. Published by rural and more deprived areas were more affected by in national surveillance systems increasing by BMJ. 50% for gonorrhoea, 36% for syphilis, 68% chlamydia, while men who have sex with men in urban For numbered affiliations see and less deprived areas showed higher rates of STI for lymphogranuloma venereum (LGV) and end of article. incidence, multiple STI episodes and HIV coinfection. 0.6% for chlamydia from 2014 to 2018.3–6 Correspondence to Conclusions We recommend socio-­epidemiological This trend is reflected in Spain where new Alexis Sentís; identification and characterisation of STI clusters and STI cases have been reported to increase ​a.​sentis@​epiconcept.​fr factors associated with HIV coinfection to identify at-­risk 10-­fold from 2000 to 2017, with 23 975 cases Sentís A, et al. BMJ Open 2021;11:e052817. doi:10.1136/bmjopen-2021-052817 1 Open access of gonorrhoea, syphilis, chlamydia and LGV reported in questionnaire that records clinical, epidemiological and 2017 alone.7 8 During 2018 to 2019, the region of Catalonia behavioural variables are included along with the manda- in Spain recorded the highest incidence of STIs across the tory notification in REC (online supplemental table country, with a rise of 37% in the number of cases.9 Inci- S1). Case definitions for surveillance reporting are stan- dence rates were highest among men who have sex with dardised according to the European Union definitions men (MSM), women and in young adults, particularly established by the European Centre for Disease Preven- among young women who in recent years have shown a tion and Control (online supplemental table S2).17 18 proportionally higher increase than men.7 9 The surge in STI incidence rates may be explained by improvements Analysis variables in surveillance systems, introduction of new diagnostic We extracted data around epidemiological, sociode- methods with enhanced sensitivity, changes in sexual atti- mographic and clinical variables as detailed in online tudes and behaviours, sociocultural shifts in society and supplemental table S3. All individuals who had experi- the effects of tourism and globalisation.10 enced at least one STI episode during the study period STIs and HIV infections are overlapping epidemics, were linked, through the Spanish healthcare system which, besides from biological synergies, are largely driven personal identification code (CIP), to the Catalan HIV/ by socioeconomic and other contextual factors acting as STI Registry to identify HIV coinfections either before or syndemics. Individuals affected by STIs are at increased after the recorded STI episode. In addition to the CIP, risk of HIV infection and people living with HIV are more the Catalan HIV/STI Registry surveillance team performs vulnerable to STIs.11 12 Some studies have described social checks of duplicates at least twice annually using a unique determinants of health, discrimination and inequalities as STI episode number (assigned to each notification and the main factors associated with the spatiotemporal clus- disease), name and date of birth. For our analysis, a dedu- tering of STI cases.13 14 While spatiotemporal clustering plicated, HIV/STI-­ linked and anonymised version was may be useful in grouping events or cases, other meth- provided. odologies including k-­means clustering allow grouping of A Basic Health Area (BHA; Àrea Bàsica de Salut (ABS), different geographical units by common characteristics in Catalan) is a territorial unit of coverage served by a such as sociodemographic and epidemiological factors.15 primary healthcare team. Each BHA typically serves a The socio-­ epidemiological characterisation of STIs, population of approximately 5000–25 000 people. The including association with HIV coinfection, and identifi- socioeconomic level of the BHAs was classified according cation of distinct clusters are imperative to strengthen the to a deprivation index (calculated by the Agency of integrated surveillance of STIs and HIV. Data from such Health Quality and Assessment of Catalonia) which was an exercise could potentially increase the sensitivity, time- attributed to each individual according to their residential liness and representativeness of surveillance systems, and address. The deprivation index is a composite measure generate information to tailor public health strategies to based on indicators such as proportion of residents with tackle a continuously growing epidemic. low educational level, proportion of manual workers, Therefore, we aimed to describe the epidemiology of proportion of residents with an annual income below a STIs, identify and characterise socio-­ epidemiological specified amount, and rate of premature mortality. Depri- clusters of STI and determine factors associated with HIV vation indices were categorised in quintiles, with the first coinfection in Catalonia, Spain, during 2017–2019. quintile being the least deprived.19 Clinical variables that were extracted included rein- fections, multiple STI episodes and coinfection with METHODS HIV. Reinfection was defined as an episode of the same Study design and data source STI detected after a defined period, which differed for We conducted a retrospective population-­based cohort each STI, following the previously recorded infection in analysis of all confirmed cases of the notifiable STIs, the same individual during the study period. Multiple syphilis, gonorrhoea, chlamydia and LGV, in Catalonia STI episodes were defined as total number of episodes between 1 January 2017 and 31 December 2019. Data of any STI reported for the individual during the study were obtained from the Catalan HIV/STI Registry,16 period (online supplemental table S3). As information which uses information from the Epidemiological Repos- regarding treatment response was not available, episodes itory of Catalonia (REC, in Catalan), an electronic data- occurring outside of the specific time frames for each STI base used by the Epidemiological Surveillance Network were assumed not to be a persistent infection resulting of Catalonia (XVEC, in Catalan). REC collects informa- from treatment failure. tion from two sources: (1) the microbiological notifica- tion system (SNM, in Catalan) of confirmed cases from K-means clustering of STIs microbiological laboratories; and (2) the mandatory We implemented k-­means to define STI clusters by socio- disease notification system (MDO, in Catalan) based on demographic characteristics. Specifically, the k-­ means physician reporting of clinically suspected/probable and clustering methodology is an unsupervised machine laboratory-­confirmed cases as per established case defini- learning approach that seeks to group heterogeneous tions. Information collected through an epidemiological units (in our case BHAs) into clusters based on similarities 2 Sentís A, et al. BMJ Open 2021;11:e052817. doi:10.1136/bmjopen-2021-052817 Open access in characteristics (variable values and categorical distri- multivariable logistic regression model. All analyses were bution among the BHA).15 A clustering algorithm is a performed using R Statistical Software V.3.6.1). procedure for grouping a series of vectors according to a specific criterion, which could be distance or similarity. Ethics approval statement Proximity is defined in terms of a distance function and Data from the Catalan HIV/STI Registry, REC and all uses a k-­means clustering method based on Euclidean aggregated variables used in the study were handled distance to quantify similarities or differences between according to international recommendations, the observations. This method is sensitive to outliers and Helsinki Declaration revised by the World Medical Orga- requires both internal and external validation processes nization in Fortaleza in 2013 and Spanish Law 3/2018 using different combinations of variables in the algo- on Data protection and Public Health 33/2011. Patient rithms. The internal clustering validation considered an information was anonymised and deidentified prior to average Euclidean distance for each cluster and similar- analysis and therefore no informed consent was required. ities between cases according to the correlation matrix of distances to determine the optimum number of clus- Patient and public involvement ters for which intracluster variation is minimum. Our Patients were not directly involved in this study; only data external validation process was based on a description of from the nationally notifiable disease surveillance system the possible socio-­epidemiological variables to determine were used. the most appropriate clustering algorithm for our data set. Based on these validations, the researchers ended up with an exact number of clusters formed by defined vari- RESULTS ables which were included in the algorithm. In our case, STI epidemic and trends the following variables were chosen to identify and build Between 2017 and 2019, a total of 42 283 cases of STIs were the final three socio-­epidemiological clusters of STI by reported among 34 600 individuals in Catalonia (table 1). BHA: incidence rate by each STI, percentage of women, Throughout the study period, half of all reported STIs percentage of people with HIV coinfection, median age were chlamydia (50.1%) and almost a third were gonor- among all STI cases in each BHA and deprivation index rhoea (31.6%). Reinfections accounted for 10.8% of all of each BHA. reported cases. Among the subjects affected by STIs, the events of gonorrhoea had the highest reinfection rate (15.7%), while chlamydia had the lowest occurrence of Statistical analyses reinfection (6.7%). We performed a descriptive analysis to summarise epide- The number of STI cases doubled from 9687 in 2017 miological, clinical, sociodemographic and geographical to 18 872 in 2019 (table 2). The incidence rate of STIs variables for the total confirmed STI cases and by STI clus- increased by 91.3% from 128.2 cases per 100 000 popu- ters. Continuous variables were summarised as median lation in 2017 to 248.9 cases per 100 000 population in and interquartile range (IQR), while categorical variables 2019. The annual incidence rate of STIs for the period were reported as absolute frequencies and percentages. 2017–2019 was 185.5 per 100 000 population. Incidence Annual incidence rates of STIs are described per 100 000 rates increased significantly (p<0.001) from 2017 to 2019 population for the total confirmed STI cases and by STI for all STI types except for syphilis cases which remained cluster, and calculated based on census information from stable over the 3 years, with the highest increase in number the Statistical Institute of Catalonia (IDESCAT) (online of cases seen in chlamydia (188.8%) followed by gonor- supplemental table S4). Incidence trends were analysed rhoea (63.8%) and LGV (56.1%). In 2017, chlamydia and using the χ2 test for linear trend. For identifying and gonorrhoea represented 36.8% and 36.0% of all reported building clusters in k-­ means clustering, STI incidence STIs, respectively, but by 2019 chlamydia accounted rates were described per 1000 population due to the for 55.1% of all cases. Gonorrhoea showed the second small population size per BHA. greatest increase from 2017 to 2019, with 47.5% occurring We assessed risk factors associated with HIV coinfection in individuals under 30 years of age throughout the study among individuals diagnosed with STIs using multivari- period. This increase in the number of confirmed STI able logistic regression models to estimate odds ratios cases from 2017 to 2019 was remarkably higher in women (ORs) and 95% confidence intervals (CIs). Individuals (132% vs 75% in men) and individuals below the age of 30 with more than one STI episode were counted once (first years (125% vs 68% in those ≥30 years). Indeed, women episode), and successive episodes in the same individual under 30 years of age presented the highest decrease in were grouped in a variable that considers the number both number of cases, with an increase of 155.8% versus of episodes, and included in the models. Sexual prefer- 93.6% in men below 30 years and in incidence rates with ence, country of birth and education level were excluded an increase of 154.1% (from 193.6 to 491.9 per 100 000 from the models because more than 50% of values were population) versus 93.6% in men under 30 years (from missing. We used backward stepwise elimination regres- 202.9 to 384.0 per 100 000 population). sion to include all analysed variables that showed statis- The vast majority of reported cases occurred in men tical significance (p<0.05) by the Wald test in the final for all STI types except chlamydia, of which 61.9% Sentís A, et al. BMJ Open 2021;11:e052817. doi:10.1136/bmjopen-2021-052817 3 Open access Table 1 Epidemiological characteristics of reported STI cases in Catalonia, Spain (2017–2019) Characteristic All STIs Chlamydia Gonorrhoea Syphilis LGV Total cases, n (%) 42 283 (100) 21 202 (50.1) 13 362 (31.6) 6975 (16.5) 744 (1.8) Sex, n (%)  Female 16 676 (39.4) 13 125 (61.9) 2667 (20.0) 875 (12.5) 9 (1.2)  Male 25 607 (60.6) 8077 (38.1) 10 695 (80.0) 6100 (87.5) 735 (98.8) Age, n (%)  <20 years 4438 (10.5) 3311 (15.6) 984 (7.4) 137 (2.0) 6 (0.8)  20–29 years 17 691 (41.8) 10 707 (50.5) 5361 (40.1) 1462 (21.0) 161 (21.6)  30–39 years 11 102 (26.3) 4454 (21.0) 4116 (30.8) 2242 (32.1) 290 (39.0)  40–49 years 6092 (14.4) 2087 (9.8) 2032 (15.2) 1757 (25.2) 216 (29.0)  50–59 years 2037 (4.8) 530 (2.5) 658 (4.9) 789 (11.3) 60 (8.1)  >60 years 923 (2.2) 113 (0.5) 211 (1.6) 588 (8.4) 11 (1.5) Sexual preference, n (%)  MSM* 3270 (7.7) 785 (3.7) 1321 (9.9) 993 (14.2) 171 (23.0)  MSW 3149 (7.5) 1863 (8.8) 1040 (7.8) 243 (3.5) 3 (0.4)  WSW† 415 (1.0) 335 (1.6) 69 (0.5) 10 (0.1) 1 (0.1)  WSM 8189 (19.4) 7034 (33.2) 966 (7.2) 186 (2.7) 3 (0.4)  Missing (male) 19 188 (45.4) 5429 (25.6) 8334 (62.4) 4864 (69.7) 561 (75.4)  Missing (female) 8072 (19.1) 5756 (27.2) 1632 (12.2) 679 (9.7) 5 (0.7) STI reinfection, n (%) 4558 (10.8) 1418 (6.7) 2098 (15.7) 955 (13.7) 87 (11.7) HIV coinfection, n (%) 2443 (5.8) 467 (2.2) 897 (6.7) 893 (12.8) 186 (25.0) Deprivation index‡  First quintile (least deprived) 10 271 (24.3) 5185 (24.5) 3040 (22.8) 1757 (25.2) 289 (38.8)  Second quintile 7465 (17.7) 4328 (20.4) 2012 (15.1) 1037 (14.9) 88 (11.8)  Third quintile 4859 (11.5) 2763 (13.0) 1332 (10.0) 716 (10.3) 48 (6.5)  Fourth quintile 5703 (13.5) 3217 (15.2) 1578 (11.8) 827 (11.9) 81 (10.9)  Fifth quintile 7689 (18.2) 4319 (20.4) 2211 (16.6) 1079 (15.5) 80 (10.8)  Missing 6296 (14.9) 1390 (6.6) 3189 (23.9) 1559 (22.4) 158 (21.2) Health region of residence, n (%)  Barcelona 35 215 (83.3) 17 108 (80.7) 11 566 (86.6) 5833 (83.6) 708 (95.2)  Other regions 7068 (16.7) 4094 (19.3) 1796 (13.4) 1142 (16.4) 36 (4.8) BHA setting  Rural 4193 (9.9) 2614 (12.3) 1039 (7.8) 516 (7.4) 24 (3.2)  Urban 29 969 (70.9) 16 347 (77.1) 8566 (64.1) 4516 (64.8) 540 (72.6)  Missing 8121 (19.2) 2241 (10.6) 3757 (28.1) 1943 (27.9) 180 (24.2) *Includes men who have sex with men, bisexual men and transgender men. †Includes women who have sex with women, bisexual women and transgender women. ‡First quintile (31.52%), second quintile (40.09%), third quintile (46.27%), fourth quintile (53.98%) and fifth quintile (100%). BHA, Basic Health Area; LGV, lymphogranuloma venereum; MSM, men who have sex with men; MSW, men who have sex with women; STI, sexually transmitted infection; WSM, women who have sex with men; WSW, women who have sex with women. occurred in women (table 1). Among all STI cases, 78.6% When examining the distribution of STI cases according were reported in individuals below 40 years of age. Chla- to deprivation index, the highest proportion of cases was mydia was reported most frequently among individuals seen in less deprived areas, with 24.3% of all cases reported below 30 years of age (66.1%), while syphilis occurred in the first quintile. In more deprived areas (fifth quin- most in those above 30 years of age (77.1%). Among the tile), chlamydia (56%) and gonorrhoea (29%) occurred 15 023 (35.5%) reported STI cases for which information more frequently than syphilis (14%) and LGV (1%). Data regarding sexual preference was available, half (54.5%) around country of birth and education level were limited were reported in women who have sex with men (WSM), due to high rates of missing data (56.6% for country 21.8% in MSM and 21.0% in MSW (table 1). of birth and 76.5% for education level). Nevertheless, 4 Sentís A, et al. BMJ Open 2021;11:e052817. doi:10.1136/bmjopen-2021-052817 Open access Table 2 Reported STI cases and incidence rates by year in Catalonia, Spain (2017–2019) 2017 2018 2019 Incidence rate Incidence rate Incidence rate per 100 000 per 100 000 per 100 000 Cases, n population Cases, n population Cases, n population P trend Total Total 9687 128.2 13 724 180.6 18 872 245.9 <0.001 Sex  Femalea 3362 87.4 5503 142.2 7811 200.0 <0.001 b  Male 6325 170.5 8221 220.4 11 061 293.4 <0.001 Age, years  <30c 4607 198.4 7181 306.9 10 341 436.5 <0.001  ≥30d 5080 97.1 6543 124.4 8531 160.8 <0.001 Sex and age  Male <30 yearse 2411 202.9 3347 279.5 4668 384.0 <0.001  Male ≥30 yearsf 3914 155.2 4874 192.4 6393 250.3 <0.001 g  Female <30 years 2196 193.6 3834 335.7 5673 491.9 <0.001  Female ≥30 yearsh 1166 43.0 1669 61.2 2138 77.7 <0.001 i STI type (total Catalonia )  Chlamydia 3562 47.1 7240 95.3 10 400 135.5 <0.001  Gonorrhoea 3492 46.2 4088 53.8 5782 75.3 <0.001  Syphilis 2430 32.2 2175 28.6 2370 30.9 0.1572  LGV 203 2.7 221 2.9 320 4.2 <0.001 Health region of residence  Alt pirineu i Aranj 10 13.9 13 18.1 33 45.7 <0.001 k  Barcelona 8205 165.0 11 475 229.5 15 535 307.8 <0.001  Camp de 294 49.1 491 81.3 870 142.2 <0.001 Tarragonal  Catalunya 358 69.4 484 93.1 583 110.7 <0.001 Centralm  Gironan 564 65.7 893 103.2 1327 151.5 <0.001  Lleidao 212 58.9 247 68.5 404 111.5 <0.001 p  Terres de l'Ebre 44 24.5 121 67.7 120 67.2 <0.001 BHA settingq  Rural 827 NA 1375 NA 1991 NA NA  Urban 6475 NA 9868 NA 13 626 NA NA  Missing 2385 NA 2481 NA 3255 NA NA Data source for denominators (a-­q): the Statistical Institute of Catalonia (IDESCAT), online supplemental table S4. BHA, Basic Health Area; LGV, lymphogranuloma venereum; NA, not applicable; STI, sexually transmitted infection. we report that among cases with available information, from 2017 to 2019 in all regions, regardless of STI type. 72.4% were observed among individuals born in Spain Similarly, the large majority of STI cases occurred in and 85.0% in those with secondary or higher education urban BHAs (70.9%) throughout the study period. (online supplemental table S5). The incidence rate of STI cases was disproportionately Factors associated with HIV coinfection among individuals higher in Barcelona (83.3%) compared with the other with STIs six regions combined (table 1). Barcelona reported the In total, 6% of STI episodes affected HIV-­positive indi- highest incidence rate of STIs, while Alt Pirineu i Aran viduals with a higher proportion of HIV coinfection recorded the lowest consistently throughout the study observed with cases of syphilis and LGV (13% and 25%, period (table 2). In 2019, the incidence rate of STIs was respectively) and the lowest with cases of chlamydia (2%) 307.8 cases per 100 000 population in Barcelona and 45.7 (table 1). Factors associated with HIV coinfection among cases per 100 000 population in Alt Pirineu i Aran. Never- individuals with STIs in the multivariable analyses are theless, incidence rates of STIs increased significantly shown in table 3. The likelihood of HIV coinfection was Sentís A, et al. BMJ Open 2021;11:e052817. doi:10.1136/bmjopen-2021-052817 5 Open access Table 3 Factors associated with HIV coinfection among individuals diagnosed with STIs in Catalonia, Spain (2017–2019) Total, n HIV-­positive, n Characteristic (n=34 600) (n=1376) OR 95% CI aOR 95% CI Sex  Female 14 938 29 1 (ref) 1 (ref)  Male 19 662 1347 37.81 26.69 to 55.93 23.69 16.67 to 35.13 Age group, years  <20 3696 5 1 (ref) 1 (ref)  20–29 14 826 328 16.70 7.70 to 46.83 8.33 3.82 to 23.40  30–39 8704 595 54.17 25.05 to 151.57 18.58 8.56 to 52.13  40–49 4759 339 56.62 26.09 to 158.78 17.66 8.10 to 49.65  50–59 1748 89 39.60 17.80 to 112.58 13.06 5.84 to 37.24  >60 867 20 17.43 7.04 to 52.50 6.98 2.80 to 21.09 Deprivation index*  First quintile (least deprived) 7679 501 1 (ref) 1 (ref)  Second quintile 6098 210 0.51 0.43 to 0.60 0.70 0.59 to 0.83  Third quintile 4163 109 0.38 0.31 to 0.47 0.63 0.50 to 0.78  Fourth quintile 4663 186 0.60 0.50 to 0.71 0.83 0.69 to 1.00  Fifth quintile 6347 175 0.41 0.34 to 0.48 0.60 0.50 to 0.72  Missing 5650 195 0.51 0.43 to 0.60 0.51 0.39 to 0.67 STI episodes (total), n  1 29 104 791 1 (ref) 1 (ref)  2–4 5304 529 3.96 3.54 to 4.44 2.69 2.39 to 3.03  5–7 192 56 14.74 10.64 to 20.16 5.96 4.26 to 8.24 BHA setting  Rural 3699 81 1 (ref) 1 (ref)  Urban 23 812 1023 2 1.61 to 2.54 1.32 1.04 to 1.69 *First quintile (31.52%), second quintile (40.09%), third quintile (46.27%), fourth quintile (53.98%) and fifth quintile (100%). aOR, adjusted OR; BHA, Basic Health Area; CI, Confidence Interval; OR, Odds ratio; STI, sexually transmitted infection.; greater among males (adjusted OR (aOR) 23.69; 95% CI clusters (table 4). Among the included BHAs, the inci- 16.67 to 35.13 compared with females) and in urban dence rate of STIs in 2017–2019 was 160.6 per 100 000 BHAs (aOR 1.32; 95% CI 1.04 to 1.69 compared with population per year. rural BHAs). All age groups from 20 years and above and Of the three clusters, the socio-­ e pidemiological having multiple STI episodes were also associated with characteristics of STI-­i nfected individuals in cluster A greater odds of HIV coinfection. BHA deprivation indices most closely resembled that of the total cases reported beyond the first quintile were associated with lower like- in the Catalan surveillance system that were included lihood of HIV coinfection among individuals with STIs. in the cluster analysis. Among the 109 BHAs in cluster A, median age was 31 years compared with 29 years Identification and characterisation of the socio- among all reported STI cases, median deprivation epidemiological clusters of STIs Of the 373 Catalan BHAs, five (Garraf rural, index was 31.9 versus 39.8, the proportion of men Polinyà-Sentmenat, Ribes-­Olivella, Roquetes-­Canyelles was 67.4% versus 58.1% and HIV coinfection rate was and Viladecans 3) were excluded from the K-­ means 8.8% versus 6.1% (table 4). clustering analysis because their delimitations and Cluster B consisted of the largest number of BHAs populations changed during the study period. In these (251) and had the highest deprivation index (44.9) five BHAs, 679 episodes were reported during the of all three clusters and compared with the total. The 3 years of the study period. This fact and having 5773 incidence rate of STIs was lower in cluster B compared episodes with no information available about BHA of with that of the total reported cases included in the residence reduced the sample size for the cluster anal- cluster analysis (136.3 vs 160.6 per 100 000 popula- ysis from 42 283 to 35 831 STI cases. Of the 368 BHAs tion), but represented the majority of all reported included in the analysis, we identified three distinct STI cases (55.7%). STI-­infected individuals in cluster 6 Sentís A, et al. BMJ Open 2021;11:e052817. doi:10.1136/bmjopen-2021-052817 Open access Table 4 Characteristics of socio-­epidemiological STI clusters in Catalonia, Spain (2017–2019) Characteristic Cluster A Cluster B Cluster C Total* Demographics BHAs, n (%) 109 (29.6) 251 (68.2) 8 (2.2) 368 (100) Median age, median years (IQR) 31 (18–60) 26 (17–58) 34 (20–58) 29 (17–59) Median deprivation index (IQR) 31.9 (3.0–58.2) 44.9 (19.2–76.9) 25.6 (10.7–63.6) 39.8 (10.7–72.3) Annual STI incidence rate (per 100 000 162.0 136.3 721.0 160.6 population) Reported STI cases, n (%) Total 11 527 (32.2) 19 945 (55.7) 4359 (12.2) 35 831 (100) Sex  Female 3758 (32.6) 10 566 (53.0) 686 (15.7) 15 010 (41.9)  Male 7769 (67.4) 9379 (47.0) 3673 (84.3) 20 821 (58.1) Country of birth  Spain 3920 (34.0) 6910 (34.7) 1325 (30.4) 12 155 (33.9)  Outside Spain 1279 (11.1) 2819 (14.1) 435 (10.0) 4533 (12.7)  Missing 6328 (54.9) 10 216 (51.2) 2599 (59.6) 19 143 (53.4) Sexual preference  MSM† 1234 (10.7) 1104 (5.5) 655 (15.0) 2993 (8.4)  MSW 751 (6.5) 2164 (10.9) 64 (1.5) 2979 (8.3)  WSM‡ 1440 (12.5) 6066 (30.4) 130 (3.0) 7636 (21.3)  WSW 75 (0.7) 295 (1.5) 6 (0.1) 376 (1.1)  Missing (male) 5784 (50.2) 6111 (30.6) 2954 (67.8) 14 849 (41.4)  Missing (female) 2243 (19.5) 4205 (21.1) 550 (12.6) 6998 (19.5) STI type  Gonorrhoea 3448 (29.9) 5240 (26.3) 1448 (33.2) 10 136 (28.3)  Chlamydia 5739 (49.8) 12 314 (61.7) 1649 (37.8) 19 702 (55.0)  Syphilis 2117 (18.4) 2263 (11.4) 1027 (23.6) 5407 (15.1)  LGV 223 (1.9) 128 (0.6) 235 (5.4) 586 (1.6) Multiple (>1) STI episodes 1600 (13.9) 1524 (7.6) 1048 (24.0) 4172 (11.6) HIV coinfection 1011 (8.8) 495 (2.5) 686 (15.7) 2192 (6.1) STI incidence rate categories of BHAs  First quintile (2.4 per 1000) 710 (6.2) 1820 (9.1) 0 2530 (7.1)  Second quintile (3.6 per 1000) 2136 (18.5) 3359 (16.8) 0 5495 (15.3)  Third quintile (5.2 per 1000) 1377 (12.0) 6588 (33.0) 0 7965 (22.2)  Fourth quintile (9.8 per 1000) 5688 (49.4) 7508 (37.6) 0 13 196 (36.8)  Fifth quintile (42.8 per 1000) 1616 (14.0) 670 (3.4) 4359 (100) 6645 (18.6) BHA setting  Rural 797 (6.9) 3158 (15.8) 0 3955 (11.0)  Urban 9461 (82.1) 15 787 (79.2) 4359 (100) 29 607 (82.6)  Missing 1269 (11.0) 1000 (5.0) 0 2269 (6.3) *Of the 373 Catalan BHAs, five (Garraf rural, Polinyà-Sentmenat, Ribes-­Olivella, Roquetes-­Canyelles nd Viladecans 3) were excluded from the K-­means clustering analysis because their delimitations and populations changed during the study period. †Includes men who have sex with men, bisexual men and transgender men. ‡Includes women who have sex with women, bisexual women and transgender women. BHA, Basic Health Area; IQR, Interquartile range; LGV, lymphogranuloma venereum; MSM, men who have sex with men; MSW, men who have sex with women; STI, sexually transmitted infection; WSM, women who have sex with men; WSW, women who have sex with women. B were the youngest among all groups (26 years) 12% higher) and chlamydia cases (61.7% vs 55.0%). and were predominantly women (53.0%), whereas Rates of multiple STI episodes and HIV coinfection men represented the majority in all other groups. in cluster B were the lowest of all three clusters and Compared with the total, cluster B consisted of more compared with the total (table 4). rural BHAs (15.8% vs 11.0%) and had a higher propor- Cluster C consisted of only eight BHAs and had the lowest tion of heterosexual men and women (approximately deprivation index (25.6) among all groups (table 4). The Sentís A, et al. BMJ Open 2021;11:e052817. doi:10.1136/bmjopen-2021-052817 7 Open access incidence rate of STIs in cluster C was the highest among rates (fourth and fifth quintiles). All 4359 STI cases in cluster all groups (721.0 per 100 000 population), with all cases C were reported in BHAs in the highest quintile of STI inci- reported in urban BHAs. STI-­infected individuals in cluster dence rate. This correlated well with the fact the number C were the oldest among all groups (34 years) and had the of STI cases per BHA was higher in clusters A and C (105.8 highest proportion of MSM. Similar to other clusters and the and 544.9 cases per BHA, respectively) than in the total (97.4 total reported cases, chlamydia remained the most common cases per BHA), which indicates higher proportion of high STI type; however, cluster C was characterised by higher incidence rates (table 4 and figure 1). rates of gonorrhoea (33.2%), syphilis (23.6%), LGV (5.4%), multiple STI episodes (24.0%) and HIV coinfection (15.7%). Almost 60% of STI cases in cluster B occurred in BHAs in DISCUSSION the three lowest quintiles of STI incidence rates, while more Our findings revealed that the incidence of STIs in Catal- than 60% in cluster A occurred in areas of high STI incidence onia almost doubled from 2017 to 2019, primarily driven Figure 1 Incidence rates (per 1000 population) and socio-­epidemiological clusters of STIs by BHA during 2017–2019. (A) STI incidence rates in Catalonia; (B) STI incidence rates in Barcelona city*; (C) STI socio-­epidemiological clusters in Catalonia and (D) STI socio-­epidemiological clusters in Barcelona city*. *Health Regions were used as a bigger unit of analysis than BHA. The municipality of Barcelona is shown to enhance the visualisation of cluster C. From a total of 373 Catalan BHA, five (Garraf rural, Polinyà-Sentmenat, Ribes-­Olivella. Roquetes-­Canyelles and Viladecans 3) were excluded from the K-­means clustering analysis because their delimitations and populations changed during the study period. BHA, Basic Health Area; STIs, sexually transmitted infections. 8 Sentís A, et al. BMJ Open 2021;11:e052817. doi:10.1136/bmjopen-2021-052817 Open access by the increase in cases among young adults (under 30 synergic infections and should be viewed as a syndemic.25 years) and in cases of chlamydia (particularly in women) The WHO and other public health agencies have empha- and gonorrhoea. In 2017–2019, the majority of STI cases sised the importance of integrating surveillance of STIs, occurred also in individuals below the age of 30 years, HIV and even viral hepatitis, and strengthening under- and those living in urban and less deprived areas, with standing of determinants of these infections by linking most cases reported in Barcelona. The identification biological and behavioural surveillance, to enhance the and characterisation of socio-­ epidemiological clusters identification and characterisation of populations at of STI showed that young women living in rural and increased risk of infection.10 25 Sociodemographic and more deprived areas were more likely to be affected by socioeconomic are increasingly being established as more chlamydia. Furthermore, MSM living in urban and less important risk factors of STI acquisition than individual deprived areas showed, more frequently than other popu- behaviours, particularly among women from disadvan- lation groups, higher STI incidence rates, more multiple taged groups.26 27 STI episodes and higher percentages of HIV coinfection. The k-­ means clustering methodology is a machine Similarly, the factors associated with HIV coinfection were learning approach that has proven its utility and poten- being men, older than 20 years old, living in urban and tial in classifying and grouping health-­related outcomes. less deprived areas and having multiple STI episodes. It has been used in the field of bipolar disorder to define After a long period of continuous reduction of STI cluster-­based disease severity using heterogeneous vari- incidence in Western countries, which coincided with ables such as sociodemographic, clinical, cognitive, vital the beginning and hardest times of the HIV epidemic signs and laboratory parameters.28 More recently, its from the 1980s to 2010s, many countries including the potential to monitor and group SARS-­CoV-­2 prevalence USA and European countries are recently reporting an by magnitude and trends (higher, medium and lower) at ongoing re-­emergence of STIs.3–6 The rise in STI cases a regional level in Italy has been described.29 Identifica- has been partially attributed to enhancement of surveil- tion of these ‘clusters of characteristics’ may be useful, lance systems and the introduction of improved diag- in their specific context, to better detect and charac- nostic tools in recent years.10 Other contributing factors, terise case profiles by site or geographical area, which described mostly among MSM, include the use of HIV could ultimately lead to better-­ designed interventions pre-­exposure prophylaxis, the use of recreational drugs to improve health outcomes. In a recent study of STI for sex, substance and/or alcohol abuse and widespread risk among MSMs, hierarchical cluster analysis, another use of the internet and other technologies to seek sexual machine learning methodology, identified factors other partners.20–22 than behaviour, such as sexual networks and risk percep- Chlamydia has been reported more frequently in WSM, tion, that influence the vulnerability to STIs and HIV while syphilis, gonorrhoea and LGV were more common infections.30 To the best of our knowledge, this current in MSW and MSM.3–9 Similarly, in our study, we found study is the first to apply the k-­means clustering method- different epidemiological characteristics for each STI ology to identify and characterise socio-­epidemiological type. Chlamydia was more common in women, mostly in clusters of STI. WSM, with a large majority occurring in individuals below A key limitation of this study is the high proportion the age of 30 years. Gonorrhoea, syphilis and LGV were of missing data around sociodemographic and lifestyle substantially more frequent in men, specifically among characteristics, a common phenomenon in population-­ MSM, and showed higher percentages of reinfections and based epidemiological studies where questionnaires are HIV coinfections than chlamydia. Most STI cases were used. This may have potentially introduced information observed in Spanish-­born individuals and among those bias or inaccurate representation of the true situation with secondary or higher education levels, although these when describing high-­ risk populations. Although not findings should be interpreted with caution because of formally assessed, we classify these missing data as missing the high proportion of missing data for sexual prefer- completely at random due to time constraints in comple- ence and education level. Our findings are consistent tion of the epidemiological questionnaires by surveil- with earlier studies and reports of STIs in Catalonia in lance officers and healthcare professionals who notified 2007–2015,12 2012–201723 and 2018–2019,9 showing a the diseases to the surveillance systems. Nonetheless, our proportionally higher increase in young adults, mostly findings are similar to those reported in previous anal- women, especially for chlamydia but also for gonorrhoea. yses.3–6 9 12 24 The age category above 60 years old may Our findings are also consistent with that of a previous contribute to residual cofounding although the risk is study among residents of Barcelona showing that STIs are minimal because it is the age group with the smallest becoming more prevalent in individuals with favourable sample size and the range is larger than for other age cate- socioeconomic status and education levels.24 gories. Categorisation of the deprivation indices by quin- Consistent with previous data,12 we found that male tiles could have diluted the findings if deprivation was a sex, age above 20 years (particularly 30–60 years), living strong confounder or unevenly distributed, although we in urban or less deprived areas, and having multiple do not believe either event to be the case in our analysis. STI episodes were associated with an increased risk of A strength of our study is the inclusion of ecolog- HIV coinfection. STIs and HIV have been described as ical variables of socioeconomic status which are highly Sentís A, et al. BMJ Open 2021;11:e052817. doi:10.1136/bmjopen-2021-052817 9 Open access relevant and pertinent for describing groups at increased de l’Agència de Salut Pública de Barcelona); M Company, M Danés, N Camps risk of STIs. We believe that the most valuable outcome (Servei de Vigilància Epidemiològica i Resposta a Emergències de Salut Pública a Girona); RM Vileu, G Ferrús, N Borrell, S Minguell (Servei de Vigilància of our study is that it shows the utility of complementing Epidemiològica i Resposta a Emergències de Salut Pública a Tarragona); J Ferràs traditional epidemiological analyses with new method- (Servei de Vigilància Epidemiològica i Resposta a Emergències de Salut Pública ologies, in this case, a machine learning approach, to a Terres de l’Ebre); I Parrón (Servei de Vigilància Epidemiològica i Resposta a combine heterogeneous data sources. This would allow Emergències de Salut Pública al Barcelonès Nord i Maresme); I Mòdol, A Martinez, P Godoy (Servei de Vigilància Epidemiològica i Resposta a Emergències de Salut identification and characterisation of target populations Pública a Lleida); MA Tarrès, J Pérez, M Boldú, I Barrabeig (Servei de Vigilància at increased risk of STIs to design more efficient measures Epidemiològica i Resposta a Emergències de Salut Pública a Barcelona Sud); to prevent and control STIs and HIV infection at a small E Donate, L Clotet, MR Sala (Servei de Vigilància Epidemiològica i Resposta a health area level. Emergències de Salut Pública al Vallès Occidental i Vallès Oriental); M Carol, V Guadalupe-­Fernández (Servei de Vigilància Epidemiològica i Resposta a In conclusion, consistent with other European coun- Emergències de Salut Pública a Catalunya Central) and J Mendioroz, P Ciruela, G tries, our study found that STIs increased at an alarming Carmona, R Mansilla, JL Martínez, S Hernández (Subdirecció General de Vigilància rate during 2017 to 2019 in Catalonia, Spain, and Epidemiològica i Resposta a Emergències de Salut Pública, Agència de Salut continues to be a worrisome public health concern. The Pública de Catalunya). STI epidemic is both an issue of the health sector, and Contributors AS conceptualised and designed the study. MM-­F cleaned the database, MM-­F, LE-­C and YD performed the statistical and cluster analysis. AS, EL-­ it also poses a threat to the broader global development C and DKN reviewed scientific literature. AS, JMR-­U and JC drafted the manuscript framework and agenda. While declines in HIV infection and AS, EL-­C, PGdO, LM, NB, JMR-­U and JC interpreted the results. All authors has been observed in the last decade in Catalonia, as critically reviewed the manuscript and approved the final version to be published. in many other regions in Europe, primarily due to the AS is the author acting as guarantor. success of wider and earlier use of antiretroviral thera- Funding The authors have not declared a specific grant for this research from any pies, STI rates have been increasing dramatically, both funding agency in the public, commercial or not-­for-­profit sectors. among the MSM population, and also in heterosexual Map disclaimer The inclusion of any map (including the depiction of any women and young adults. We found that young women boundaries therein), or of any geographic or locational reference, does not imply the expression of any opinion whatsoever on the part of BMJ concerning the legal living in rural and deprived areas were more likely to be status of any country, territory, jurisdiction or area or of its authorities. Any such affected by chlamydia, while MSM living in urban and less expression remains solely that of the relevant source and is not endorsed by BMJ. deprived areas had higher overall STI incidence rates, Maps are provided without any warranty of any kind, either express or implied. multiple STI episodes and greater likelihood of HIV coin- Competing interests None declared. fection. Preventative strategies must consider these popu- Patient consent for publication Not required. lations priority targets and take into account structural Provenance and peer review Not commissioned; externally peer reviewed. social determinants identified as crucial in our analysis. Data availability statement All data relevant to the study are included in the Our findings suggest that monitoring the STI epidemic article or uploaded as supplementary information. in accordance with determinants of health and designing Supplemental material This content has been supplied by the author(s). It has intervention programmes targeted at the local context not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been would be of paramount importance rather than using peer-­reviewed. Any opinions or recommendations discussed are solely those national or regional prevalence as the key monitoring of the author(s) and are not endorsed by BMJ. BMJ disclaims all liability and responsibility arising from any reliance placed on the content. Where the content variable. includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, Author affiliations terminology, drug names and drug dosages), and is not responsible for any error 1 Centre of Epidemiological Studies of Sexually Transmitted Disease and AIDS in and/or omissions arising from translation and adaptation or otherwise. Catalonia (CEEISCAT), Department of Health, Generalitat of Catalonia, Badalona, Open access This is an open access article distributed in accordance with the Spain 2 Creative Commons Attribution Non Commercial (CC BY-­NC 4.0) license, which Pompeu Fabra University (UPF), Barcelona, Spain 3 permits others to distribute, remix, adapt, build upon this work non-­commercially, Epidemiology Department, Epiconcept, Paris, France 4 and license their derivative works on different terms, provided the original work is Fundació Institut d'Investigació Germans Trias i Pujol (IGTP), Badalona, Spain properly cited, appropriate credit is given, any changes made indicated, and the use 5 Spanish Consortium for Research on Epidemiology and Public Health (CIBERESP), is non-­commercial. See: http://​creativecommons.​org/​licenses/​by-​nc/​4.​0/. Instituto de Salud Carlos III, Madrid, Spain 6 Department of Paediatrics, Obstetrics and Gynecology and Preventive Medicine, ORCID iDs Universitat Autònoma de Barcelona, Barcelona, Spain Alexis Sentís http://​orcid.​org/​0000-​0002-​8132-​5645 7 Epidemiology Service, Public Health Agency of Barcelona, Barcelona, Spain Juliana Reyes-­Urueña http://​orcid.​org/​0000-​0002-​3122-​6518 8 Epidemiological Surveillance and Response to Public Health Emergencies Service in Tarragona, Agency of Public Health of Catalonia, Generalitat of Catalonia, Tarragona, Spain REFERENCES 1 Rowley J, Vander Hoorn S, Korenromp E, et al. Chlamydia, Acknowledgements We thank all healthcare professionals working in STI/ gonorrhoea, trichomoniasis and syphilis: global prevalence and HIV surveillance, prevention and control in Catalonia who enable case detection, incidence estimates, 2016. Bull World Health Organ 2019;97:548–62. diagnosis and treatment, as well as the notification and information gathering 2 WHO Department of Reproductive Health and Research. Report for the epidemiological questionnaires. We thank Stefanie Chuah for her valuable on global sexually transmitted infection surveillance, 2018. support in editing the manuscript. Available: https://​apps.​who.​int/​iris/​bitstream/​handle/​10665/​277258/​ 9789241565691-​eng.​pdf?​ua=1 [Accessed 8 Jan 2021]. Collaborators The Catalan HIV and STI Surveillance Group: A Sentís, E López, 3 European Centre for Disease Prevention and Control. Chlamydia V Gonzalez, R Lugo, MP Bonamusa, J Reyes, J Casabona (Centre d’Estudis infection. annual epidemiological report for 2018. Stockholm: ECDC, Epidemiològics sobre les Infeccions de Transmissió Sexual i Sida de Catalunya); P 2020. Available: https://www.​ecdc.​europa.​eu/​sites/​default/​files/​ Garcia de Olalla, Lilas Mercuriali, E Masdeu, M Ros, C Rius (Servei d’Epidemiologia documents/​AER-​for-​2018-​STI-​chlamydia.​pdf [Accessed 7 Aug 2021]. 10 Sentís A, et al. BMJ Open 2021;11:e052817. doi:10.1136/bmjopen-2021-052817 Open access 4 European Centre for Disease Prevention and Control. Gonorrhoea. Catalunya. Barcelona, Spain, 2015. Available: https://​scientiasalut.​ annual epidemiological report for 2018. Stockholm: ECDC, 2020. gencat.​cat/​bitstream/​handle/​11351/​3418/​informe_​SIVES_​2015_​ Available: https://www.​ecdc.​europa.​eu/​sites/​default/​files/​documents/​ informe_​epidemiologic_​CEEISCAT_​2015.​pdf.​pdf?​sequence=​1&​ gonorrhoea-​annual-​epidemiological-​report-​2018.​pdf [Accessed 7 isAllowed=y [Accessed 27 Mar 2021]. Aug 2021]. 17 Notifiable diseases and epidemic outbreaks. Department of health, 5 European Centre for Disease Prevention and Control. Syphilis. annual public health agency of Catalonia, Generalitat de Catalunya, epidemiological report for 2018. Stockholm: ECDC, 2020. Available: 2016. Available: https://​canalsalut.​gencat.​cat/​ca/​professionals/​ https://www.​ecdc.​europa.​eu/​sites/​default/​files/​documents/​syphilis-​ vigilancia-​epidemiologica/​malalties-​de-​declaracio-​obligatoria-​i-​brots-​ aer-​2018.​pdf [Accessed 7 Aug 2021]. epidemics/ [Accessed 20 Jun 2021]. 6 European Centre for Disease Prevention and Control. 18 European Centre for Disease Prevention and Control. Eu case Lymphogranuloma venereum. annual epidemiological report for definitions, 2018. Available: https://www.​ecdc.​europa.​eu/​en/​ 2018. Stockholm: ECDC, 2020. Available: https://www.​ecdc.​europa.​ surveillance-​and-​disease-​data/​eu-​case-​definitions [Accessed 7 Aug eu/​sites/​default/​files/​documents/​lymphogranuloma-​venereum-​aer-​ 2020]. 2018.​pdf [Accessed 7 Aug 2021]. 19 Agency for Health Quality and Assessment of Catalonia. Nou 7 Área de vigilancia del VIH y conductas de riesgo. Vigilancia indicador socioeconòmic per al finançament de les ABS. Observatori epidemiológica de las infecciones de transmisión sexual, 2017. del Sistema de Salut de Catalunya, 2017. Available: http://​ Madrid: Centro Nacional de Epidemiología, Instituto de Salud Carlos observatorisalut.​gencat.​cat/​ca/​observatori-​desigualtats-​salut/​ III/Plan Nacional sobre el Sida - Dirección General de Salud Pública, indicador_​socioeconomic_​2015/ [Accessed 6 Aug 2020]. 2019. Available: https://www.​mscbs.​gob.​es/​ciudadanos/​enfLesiones/​ 20 Folch C, Esteve A, Zaragoza K, et al. Correlates of intensive alcohol enfTransmisibles/​sida/​vigilancia/​Vigilancia_​ITS_​1995_​2017_​def.​pdf and drug use in men who have sex with men in Catalonia, Spain. Eur [Accessed 27 Mar 2021]. J Public Health 2010;20:139–45. 8 Área de vigilancia del VIH y conductas de riesgo. Vigilancia 21 Saberi P, Neilands TB, Lally MA, et al. The association between epidemiológica de las infecciones de transmisión sexual, 1995-­ use of online social networks to find sex partners and sexually 2010. Madrid: Centro Nacional de Epidemiología, Instituto de Salud transmitted infection diagnosis among young men who have sex with Carlos III/Plan Nacional sobre el Sida - Dirección General de Salud men and transgender women living with HIV. J Int Assoc Provid AIDS Pública, 2012. Available: https://www.​mscbs.​gob.​es/​ciudadanos/​ Care 2019;18:232595821986732–11. enfLesiones/​enfTransmisibles/​sida/​vigilancia/​VigilanciaITS1995_​ 22 Jansen K, Steffen G, Potthoff A, et al. STI in times of PreP: high 2010.​pdf [Accessed 27 Mar 2021]. prevalence of Chlamydia, gonorrhea, and Mycoplasma at different 9 Centre d’Estudis Epidemiològics sobres les Infeccions de anatomic sites in men who have sex with men in Germany. BMC Transmissió Sexual i Sida de Catalunya (CEEISCAT). Vigilància Infect Dis 2020;20:110. epidemiològica de les Infeccions de Transmissió Sexual (ITS) a 23 Vives N, Garcia de Olalla P, González V, et al. Recent trends in Catalunya. Informe anual 2019. Badalona, Spain, 2020. Available: sexually transmitted infections among adolescents, Catalonia, Spain, https://​canalsalut.​gencat.​cat/​web/.​content/_​A-​Z/​S/​sida/​enllasos/​ 2012-­2017. Int J STD AIDS 2020;31:1047–54. anual_​ITS.​pdf [Accessed 27 Mar 2021]. 24 Martí-Pastor M, García de Olalla P, Barberá M-­J, et al. Epidemiology 10 Taylor MM, Wi TE. Transforming and integrating STI surveillance to of infections by HIV, syphilis, gonorrhea and lymphogranuloma enhance global advocacy and investment in STI control. J Int AIDS venereum in Barcelona City: a population-­based incidence study. Soc 2019;22 Suppl 6:25361. BMC Public Health 2015;15:1015. 11 Centers for Disease Control and Prevention. HIV/AIDS & STDs. 25 Murti M, Wong J, Whelan M, et al. The need for integrated public Available: https://www.​cdc.​gov/​std/​hiv/​default.​htm [Accessed 22 Feb health surveillance to address sexually transmitted and blood-­borne 2021]. syndemics. Can Commun Dis Rep 2019;45:63–6. 12 Sentís A, Martin-­Sanchez M, Arando M, et al. Sexually transmitted 26 Hill AV, De Genna NM, Perez-­Patron MJ, et al. Identifying Syndemics infections in young people and factors associated with HIV for sexually transmitted infections among young adults in the United coinfection: an observational study in a large City. BMJ Open States: a latent class analysis. J Adolesc Health 2019;64:319–26. 2019;9:e027245. 27 Jagadesh S, Combe M, Couppié P. Mapping priority neighborhoods: 13 Le Polain De Waroux O, Harris RJ, Hughes G, et al. The a novel approach to cluster identication in HIV/AIDS population. Res epidemiology of gonorrhoea in London: a Bayesian spatial modelling Sq 2020. approach. Epidemiol Infect 2014;142:211–20. 28 Fuente-­Tomas Ldela, Arranz B, Safont G, et al. Classification of 14 Gesink D, Wang S, Norwood T, et al. Spatial epidemiology patients with bipolar disorder using k-­means clustering. PLoS One of the syphilis epidemic in Toronto, Canada. Sex Transm Dis 2019;14:e0210314. 2014;41:637–48. 29 Maugeri A, Barchitta M, Agodi A. A clustering approach to classify 15 Jain AK. Data clustering: 50 years beyond k-­means. Pattern Recognit Italian regions and provinces based on prevalence and trend of sars-­ Lett 2010;31:651–66. cov-­2 cases. Int J Environ Res Public Health 2020;17:5286–14. 16 Centre d’Estudis Epidemiològics sobre les Infeccions de Transmissió 30 Blondeel K, Dias S, Furegato M, et al. Sexual behaviour patterns and Sexual i Sida de Catalunya (CEEISCAT). SIVES 2015 - Sistema STI risk: results of a cluster analysis among men who have sex with Integrat de Vigilància Epidemiològica de la SIDA/VIH/ITS a men in Portugal. BMJ Open 2021;11:33290. Sentís A, et al. BMJ Open 2021;11:e052817. doi:10.1136/bmjopen-2021-052817 11