Article Text
Abstract
Background In many countries, HIV pre-exposure prophylaxis (PrEP) users are screened quarterly for STIs. We assessed the consequences of less frequent STI testing. We also assessed determinants of asymptomatic STI and potential for onward transmission.
Methods Using data from the AMPrEP study, we assessed the proportion of syphilis, and genital, anal, and pharyngeal chlamydia and gonorrhoea diagnoses which would have been delayed with biannual versus quarterly screening. We assessed the potential for onward transmission by examining reported condomless anal sex (CAS) in periods after to-be-omitted visits when screening biannually. We assessed determinants of incident asymptomatic STIs using Poisson regression and calculated individual risk scores on the basis of the coefficients from this model.
Results We included 366 participants. Median follow-up was 47 months (IQR 43–50). 1,183STIs were diagnosed, of which 932(79%) asymptomatic. With biannual screening, 483 asymptomatic STIs (52%) diagnoses would have been delayed at 364 study visits. Of these visits, 129 (35%), 240 (66%) and 265 (73%) were followed by periods of CAS with steady, known casual or unknown casual partners, respectively. Older participants had a lower risk of asymptomatic STI (incidence rate ratio (IRR) 0.86/10-year increase, 95% CI 0.80 to 0.92), while CAS with known (IRR 1.36, 95% CI 1.10 to 1.68) and unknown (IRR 1.86, 95% CI 1.48 to 2.34) casual partners and chemsex (IRR 1.51, 95% CI 1.28 to 1.78) increased the risk. The individual risk scores had limited predictive value (sensitivity=0.70 (95% CI 0.66 to 0.74), specificity=0.50 (95% CI 0.48 to 0.51)).
Conclusion Reducing the STI screening frequency to biannually among PrEP users will likely result in delayed diagnoses, potentially driving onward transmission. Although determinants for asymptomatic STIs were identified, predictive power was low.
- PREP
- Pre-Exposure Prophylaxis
- Chlamydia Infections
- NEISSERIA GONORRHOEAE
- SYPHILIS
Data availability statement
Data are available upon reasonable request. The AMPrEP data are owned by the Public Health Service of Amsterdam. Original data can be requested by submitting a study proposal to the steering committee of AMPrEP. The proposal format can be obtained from amprep@ggd.amsterdam.nl. Request for further information can also be submitted through the same email address. The AMPrEP steering committee verifies each proposal for compatibility with general objectives, ethical approval, and informed consent forms of the AMPrEP study and potential overlap with ongoing studies. There are no restrictions to obtaining the data and all data requests will be processed in a similar way.
This is an open access article distributed in accordance with the Creative Commons Attribution Non Commercial (CC BY-NC 4.0) license, which permits others to distribute, remix, adapt, build upon this work non-commercially, and license their derivative works on different terms, provided the original work is properly cited, appropriate credit is given, any changes made indicated, and the use is non-commercial. See: http://creativecommons.org/licenses/by-nc/4.0/.
Statistics from Altmetric.com
Key messages
HIV pre-exposure prophylaxis (PrEP) users are screened quarterly for STIs. We assessed the consequences of less frequent STI testing.
Reducing the STI screening frequency to biannually among PrEP users will likely result in delayed diagnoses, potentially driving onward transmission.
A targeted STI screening approach may help guide screening and reduce screening frequency and treatment, thereby reducing antibiotic intake.
Introduction
Pre-exposure prophylaxis (PrEP) is highly effective against HIV acquisition,1 2 but does not protect against other STIs. Since the introduction of PrEP, concerns have been raised that PrEP users may be at increased risk of STIs due to engagement in more condomless sex.3 Indeed, the number of condomless anal sex acts since PrEP initiation seems to be increasing among men who have sex with men (MSM)4–6 and STI incidence is high.4 6–9
Considering the high STI incidence among PrEP users, PrEP guidelines (including the Dutch PrEP guidelines10) advise to screen PrEP users for STIs every 3–6 months.11 12 However, the optimal screening frequency for syphilis, chlamydia and gonorrhoea is under debate as (1) screening for chlamydia and gonorrhoea is costly, (2) serious sequelae in men are rare, also for syphilis when testing is done within 6 months, and (3) frequent antibiotic use may lead to antimicrobial resistance (AMR). A Dutch modelling study suggested that quarterly compared with biannual screening for chlamydia and gonorrhoea is not cost-effective under prevailing health economic criteria.13 However, an empirical study showed that screening biannually could lead to delayed diagnosis and treatment of a substantial proportion of asymptomatic STIs, possibly leading to more transmission.14 We previously showed that the incidence of chlamydia and gonorrhoea was higher during daily than event-driven PrEP use,6 15 probably due to fewer condomless sex acts among event-driven PrEP users,4 6 while the incidence of syphilis was similar. Further determination of a profile for PrEP users at high risk of STIs or a risk prediction score may provide opportunities for targeted STI screening.
To optimise the STI screening frequency of PrEP programmes, more data are needed on the effect of biannual versus quarterly screening. We therefore aimed to assess the proportion of PrEP users with STIs for whom diagnosis would be delayed if screening were conducted biannually instead of quarterly. As syphilis screening is relatively inexpensive and blood is routinely drawn every 3 months for HIV screening, we also assessed the proportion of delayed diagnoses for chlamydia and gonorrhoea separately to estimate the consequences if screening for these STIs only was delayed. In addition, we assessed the potential for onward transmission by examining reported condomless anal sex in periods after to-be-omitted visits when screening biannually and determinants of asymptomatic STIs.
Methods
Study design and participants
The Amsterdam PrEP (AMPrEP) study was a demonstration project that assessed the uptake and feasibility of daily and event-driven PrEP among MSM and transgender persons in Amsterdam, the Netherlands.16 In brief, participants were enrolled between 3 August 2015 and 31 May 2016. HIV-negative MSM and transgender persons were eligible for inclusion if they were ≥18 years old and reported any of the following in the preceding 6 months: condomless anal sex with casual partners, ≥1 diagnosed bacterial STI, post-exposure prophylaxis use, or sex with a partner living with HIV with a detectable or unknown viral load. Follow-up for the AMPrEP study ended December 2020. For this analysis, we used data until 29 February 2020, as PrEP care and routine STI testing at the Center for Sexual Health of the Public Health Service of Amsterdam were scaled down from March 2020 onwards due to the SARS-CoV-2 pandemic.
The AMPrEP study was registered with the Netherlands Trial Registry (NL5302).
Procedures
At the baseline visit, participants were offered a choice between daily and event-driven PrEP and they could switch PrEP regimens at each 3-monthly study visit. PrEP was provided free of charge.
Participants were tested every 3 months for chlamydia, gonorrhoea and syphilis6 at scheduled study visits. Additional STI testing in-between the 3-monthly scheduled visits was available for those with STI-related symptoms or for people who were notified for an STI by a sexual partner. All STI testing was done free of charge. Serology testing for syphilis was done using the LIAISON Treponema Screen assay for primary infections (DiaSorin, Saluggia, Italy), and VDRL (RPR-Nosticon II; bioMérieux) for reinfections. Urine, anal and pharyngeal swabs were tested for Chlamydia trachomatis and Neisseria gonorrhoeae using nucleic amplification testing (Aptima Combo 2 assay; Hologic, San Diego, California, USA). If an STI was detected, immediate, free-of-charge treatment was offered.
At baseline and quarterly scheduled study visits, participants completed a questionnaire on sexual behaviour.16 The Alcohol Use Disorder Identification Test,17 Drug Use Disorder Identification Test,18 Mental Health Inventory screening test,19 Sexual Compulsivity Scale,20 and a questionnaire on substance use was completed at baseline and yearly thereafter. Questions on substance use related to use in general and in sexual settings. Chemsex was defined as the use of gamma-hydroxybutyrate/gamma-butyrolactone, mephedrone or crystallised methamphetamine around the time of sex.21
Statistical analysis
The baseline visit was defined as the date of PrEP initiation. Follow-up started at baseline and continued until PrEP discontinuation, loss to follow-up, HIV seroconversion or 29 February 2020, whichever came first. After PrEP discontinuation, participants were allowed to re-enter the study if they were still HIV negative; follow-up then continued from date of re-entry.
We described baseline demographics using count data and proportions for categorical variables and median and IQR for continuous variables. To estimate the number of asymptomatic syphilis (primary, secondary or early latent), chlamydia or gonorrhoea infections with a delayed diagnosis if testing had been done biannually, we subtracted the sum of asymptomatic bacterial STIs diagnosed at biannual scheduled study visits (ie, month 6, 12, 18, etc) from the sum of asymptomatic STIs diagnosed during all scheduled study visits (ie, at month 3, 6, 9, 12, etc). In addition, we estimated the number of STIs with a delayed diagnosis as a proportion of the total number of STIs diagnosed at scheduled and additional visits, to give an indication of the number of STI diagnoses delayed if screening was done biannually or based on symptoms. An STI was defined as asymptomatic if no symptoms were reported during triage and if the participant was not notified for possible exposure to an STI by a sexual contact. If a participant reported any STI-related symptom or was notified of possible exposure, the STI was defined as symptomatic, regardless whether the reported symptoms were due to the diagnosed STI (hereafter, symptomatic STI). We assessed the number of condomless anal sex acts in the 3-month period after scheduled study visits that would have been delayed with biannual testing to obtain a proxy for transmission potential of undiagnosed and untreated STIs to sexual partners if testing would have been done biannually.
We estimated incidence rates of symptomatic and asymptomatic syphilis, chlamydia and gonorrhoea per 100 person-years by dividing the number of incident infections by the person-years of observation. Multiple incident infections at the same time point were counted as one. STI incidence rates were estimated including and excluding the additional STI visits in-between study visits. Repeated incident infections were included and all follow-up time was considered as time at risk.
We assessed incidence rate ratios (IRRs) and 95% CIs of determinants of any asymptomatic STI using a Poisson regression model, using the number of incident asymptomatic infections as the outcome. Determinants with p<0.2 from a Wald χ2 test in univariable analysis were included in an initial multivariable model. Years since PrEP initiation wasforced into the multivariable model to account for a possible increase in STI risk over time due to a lesser perceived need for condoms. Backwards variable selection was performed to obtain the final multivariable model; variables that did not significantly improve the model (p>0.05) based on the likelihood-ratio test were removed from the multivariable model. Determinants for any asymptomatic chlamydia or gonorrhoea were assessed in a similar matter. We calculated individual risk scores on the basis of the coefficients from the multivariable Poisson models (using categorical variables for age groups <35, 35–44 and ≥45 years instead of the continuous variable). We estimated the cut-off point of these prediction scores to predict the occurrence of asymptomatic STIs using the Youden index. We assessed the accuracy of the prediction scores using sensitivity and specificity for the optimal cut-off point, and the area under the curve (AUC).
All analyses were carried out using Stata (V.15.1, StataCorp, College Station, Texas, USA).
Results
Between August 2015 and May 2016, 376 participants were included in the AMPrEP study. Of these, 9 (2.7%) did not have follow-up data and 1 was diagnosed with HIV at an additional STI visit shortly after enrolment; thus 10 participants were excluded from these analyses. Of the 366 included participants, 268 (73%) chose the daily PrEP regimen at baseline, 364 (99%) were male and median age was 40 years (IQR 32–48) (online supplemental table 1). The majority (n=350, 96%) reported condomless anal sex with casual partners in the 6 months before baseline and 154 (43%) indicated chemsex in the 3 months before baseline. At baseline, 5 participants (1%) had a prevalent syphilis infection, 35 (10%) chlamydia and 35 (10%) gonorrhoea.
Supplemental material
Scheduled study visits
Median follow-up time was 47 months (IQR 43–50), during which participants attended a total of 4974 scheduled study visits. Median time between scheduled study visits was 3.0 (IQR 2.8–3.1) months. A total of 1183 bacterial STIs were diagnosed at these study visits, of which 932 (79%) were asymptomatic. Diagnosis of 483 of 932 asymptomatic STIs (52%) would have been delayed at 364 scheduled study visits if screening had been done biannually rather than quarterly (figure 1). For the various STIs, we found that 411 asymptomatic chlamydia (293 anal, 83 urethral, 35 pharyngeal), 442 asymptomatic gonorrhoea (227 anal, 34 urethral, 181 pharyngeal), and 79 asymptomatic syphilis diagnoses were made at any scheduled study visit, of which 192 (47%), 250 (57%) and 41 (52%) would have been delayed if screening had been done biannually. Twenty-eight asymptomatic anal lymphogranuloma venereum infections were diagnosed, of which 13 (46%) would have been delayed with biannual screening. Of the 364 scheduled study visits during which diagnosis of an asymptomatic STI would have been delayed when screening biannually, 129 (35%), 240 (66%) and 265 (73%) were followed by periods during which the participant reported condomless anal sex with a steady, known casual or unknown casual partner, respectively. In total, 193 (53%) of 366 participants could have transmitted one or more undetected and untreated asymptomatic STI during follow-up to a sex partner.
Study visits including additional STI screening visits
The participants attended 776 additional STI screening visits (ie, additional to the scheduled study visits), so that a total of 5750 visits with STI screening were recorded. Median time between these visits was 2.9 (IQR 2.5–3.0) months. A total of 1624 bacterial STIs were diagnosed at the scheduled and additional study visits. If screening had been done biannually or based on symptoms, diagnosis of 483 STIs (30%) would have been delayed (online supplemental table 2). Overall, the incidence rate of asymptomatic STIs was higher than the incidence rate of symptomatic STI (table 1). More incident symptomatic STIs were diagnosed when including the additional STI visits (incidence rate of any symptomatic STI 29.0 per 100 person-years, 95% CI 26.2 to 32.0) than during the scheduled study visits only (incidence rate 12.4, 95% CI 10.7 to 14.5). In contrast, the incidence rate of any asymptomatic STI was similar when including (incidence rate 57.3, 95% CI 53.3 to 61.5) or excluding (incidence rate 53.2, 95% CI 49.4 to 57.3) the additional STI visits. Similar patterns were seen for chlamydia, gonorrhoea and syphilis (figure 2).
Determinants of asymptomatic STIs
In multivariable analysis, older PrEP users had a decreased risk of any incident asymptomatic chlamydia, gonorrhoea or syphilis infection (IRR 0.86 per 10-year increase in age, 95% CI 0.80 to 0.92) (table 2). Condomless anal sex with known (IRR 1.36, 95% CI 1.10 to 1.68) and unknown (IRR 1.86, 95% CI 1.48 to 2.34) casual partners and participating in chemsex (IRR 1.51, 95% CI 1.28 to 1.78) in the 3 months prior to a scheduled study visit increased the risk of any asymptomatic STI. Based on the coefficients of the variables included in the final multivariable model, a risk score was calculated. The theoretical range of the score was 0.16–1.56; the actual range of observed values was 0.16–1.50. Based on the Youden index, the optimal cut-off point to predict the occurrence of asymptomatic STIs was 0.51, for which the sensitivity and specificity were 0.70 (95% CI 0.66 to 0.74) and 0.50 (95% CI 0.48 to 0.51), respectively; the AUC was 0.60. Multivariable determinants of any asymptomatic chlamydia or gonorrhoea were similar (online supplemental table 3). The optimal cut-off point, sensitivity, specificity and AUC were similar.
Discussion
We estimated that diagnosis of over half of asymptomatic STIs would have been delayed if STI screening would have been done biannually instead of quarterly. As condomless anal sex was commonly reported in the 3-month periods after the scheduled study visits at which asymptomatic STIs were detected, there is potential for further STI transmission to sexual partners if STI diagnosis would have been delayed with biannual screening.
The high overall incidence of STIs has been reported extensively in PrEP research,4 7–9 including during the first 2 years of follow-up of AMPrEP,6 and is likely linked to the high number of condomless anal sex acts among PrEP users.5 While incidence of STI is high, it is unclear what the optimal STI screening frequency among PrEP users is. This debate usually focuses on screening costs and the impact of screening on STI prevalence, serious sequelae and high frequency of antibiotic use. Reducing the frequency of bacterial STI screening would reduce costs, avoid treatment of self-limiting infections and reduce overall antibiotic consumption. Infections that would become symptomatic would presumably lead to additional STI visits in-between scheduled study visits with testing and treatment, which would avoid sequelae.
Modelling studies have suggested that more frequent screening could reduce chlamydia, gonorrhoea22 and syphilis incidence,23 24 but empirical studies assessing the effect of opportunistic testing for chlamydia and gonorrhoea have not demonstrated a decrease.25 26 Similar to a previous study,14 we showed that reducing screening frequency to biannually would result in delayed diagnosis of over half of asymptomatic STIs, and 32% of all diagnosed STIs, which provides a window for ongoing transmission, since the majority of individuals engaged in condomless anal sex acts during the time following the to-be-omitted study visits. On the other hand, participants might refrain from condomless sex without the reassurance of a negative test result. The contribution of delayed diagnosis of asymptomatic infections to onward transmission warrants further attention.
While there is a potential for serious sequelae of syphilis if left untreated, the likelihood of these sequelae within a time frame of 6 months is low. However, as per current guidelines, PrEP users will have to be screened for HIV every 3 months and testing for syphilis is relatively inexpensive, quarterly screening for syphilis could still be considered. Similarly, the risk of serious sequelae after chlamydia and gonorrhoea is low in men. In addition, gonorrhoea and chlamydia infections have been found to clear spontaneously27–29 and thus treating all asymptomatic gonorrhoea and chlamydia with antibiotics may not be necessary and contribute to AMR. A recent study showed that macrolide consumption among PrEP users was high (9.46 daily doses per 1000 inhabitants per day),30 but that macrolide consumption was almost 75% lower when STI screening changed from quarterly to biannually.31 Thus, reducing the STI screening frequency may at least be an effective way of reducing macrolide consumption. As we found a relatively high incidence rate of symptomatic infections, especially when including additional STI visits, this could imply that symptomatic STIs are likely to be diagnosed and treated regardless of planned screening frequency. Reducing the screening frequency of gonorrhoea and chlamydia to biannually and encouraging STI testing when needed could reduce the frequency of treatment with antibiotics.
With regard to the high proportion of infections for which treatment would have been delayed with biannual screening, we examined the feasibility of another approach: screening based on personalised risk profiles. We showed that participants who were older and engaged less in risky sexual practices (eg, chemsex or condomless anal sex with unknown casual partners) were at lower risk of asymptomatic STIs. However, the predictive value of the individual risk scores to distinguish participants at low risk of asymptomatic infections was low. In addition, these determinants are known risk factors for STIs among PrEP users8 and are not specific for asymptomatic infections. To optimise PrEP programmes, it needs to be further evaluated whether a targeted quarterly screening approach is possible and (cost-)effective.
Our study is one of the first to evaluate quarterly versus biannual STI screening using longitudinal data of over 4.5 years, in which we were able to differentiate between asymptomatic and symptomatic STIs. Our approach was also subject to some limitations. First, the participants of AMPrEP were early PrEP adopters, were predominantly white and highly educated, and were at high risk of HIV before they entered the study.16 Therefore, they may not be representative of the current PrEP-using population. Other PrEP users in the Netherlands may have, on average, lower rates of STIs and thus a lower number of asymptomatic diagnoses that would be missed with less frequent screening. Second, it was assumed that asymptomatic infections remained asymptomatic and thus untreated until the next scheduled study visit, while in theory asymptomatic STIs could become pre-symptomatic and symptomatic over time. In addition, we did not account for spontaneous clearance of gonorrhoea and chlamydia.27–29 These assumptions may have caused an overestimation of the proportion of infections that would have had delayed treatment. Lastly, we assumed that participants would not change their sexual behaviour if STI screening would have been done biannually.
In conclusion, reducing the screening frequency of STIs from quarterly to biannually among PrEP users may result in many delayed diagnoses of asymptomatic infections, potentially driving onward transmission. Whether this would pose a serious public health threat as far as chlamydia and gonorrhoea are concerned needs further evaluation. Although a risk-based STI screening approach to guide STI testing has potential to mitigate this effect, more research is warranted to evaluate the potential and (cost-)effectiveness of targeted testing in various PrEP-using populations.
Data availability statement
Data are available upon reasonable request. The AMPrEP data are owned by the Public Health Service of Amsterdam. Original data can be requested by submitting a study proposal to the steering committee of AMPrEP. The proposal format can be obtained from amprep@ggd.amsterdam.nl. Request for further information can also be submitted through the same email address. The AMPrEP steering committee verifies each proposal for compatibility with general objectives, ethical approval, and informed consent forms of the AMPrEP study and potential overlap with ongoing studies. There are no restrictions to obtaining the data and all data requests will be processed in a similar way.
Ethics statements
Patient consent for publication
Ethics approval
Written informed consent was obtained from all participants. The AMPrEP Study was approved by the ethics board of the Amsterdam University Medical Centers, location Academic Medical Center, the Netherlands (NL49504.018.14).
Acknowledgments
We would like to thank all AMPrEP participants, and the members of the advisory board and community engagement group of AMPrEP. Additionally, we thank Roel Achterbergh, Ertan Ersan, Dominique Loomans, Amy Matser, Kees de Jong, Ilya Peters, Princella Felipa, Myra van Leeuwen, Kenneth Yap, Yvonne van Duijnhoven, Adriaan Tempert and all members of the H-TEAM (see Annex 1).
References
Supplementary materials
Supplementary Data
This web only file has been produced by the BMJ Publishing Group from an electronic file supplied by the author(s) and has not been edited for content.
Footnotes
VWJ and HM.Z are joint first authors.
Handling editor Jason J Ong
Twitter @LizaCoyer
Contributors EH, MP, HdV, UD and MSvdL conceptualised and designed the AMPrEP demonstration project. VWJ, HMLZ, MG, JB and MSvdL contributed to study design of the current study. All authors were involved in data acquisition, data analysis or interpretation of the data. VWJ and HMLZ drafted the manuscript. All authors critically revised the manuscript. MSvdL had full responsibility for the work and/or the conduct of the study, had access to the data, and controlled the decision to publish.
Funding The AMPrEP Project received funding as part of the H-TEAM Initiative from ZonMw (grant number 522002003), the National Institute for Public Health and the Environment (grant number not applicable) and GGD research funds (grant number not applicable). The study drugs were provided by Gilead Sciences. The H-TEAM Initiative is being supported by the Aidsfonds Netherlands (grant number 2013169), Stichting Amsterdam Diner Foundation (grant number not applicable), Gilead Sciences Europe (grant number PA-HIV-PREP-16-0024), Gilead Sciences (protocol numbers CO-NL-276-4222, CO-US-276-1712), Janssen Pharmaceuticals (reference number PHNL/JAN/0714/0005b/1912fde), MAC AIDS Fund (grant number not applicable) and ViiV Healthcare (PO numbers 3000268822, 3000747780).
Competing interests The Public Health Service of Amsterdam received the drugs for the Amsterdam PrEP Study from Gilead Sciences based on an unconditional grant. UD received unrestricted research grants and speaker fees from Gilead Sciences, paid to his institute. HdV received grants from Medigene, and advisory board and speaker fees from Gilead Sciences, Medigene, Abbvie, Janssen-Cilag and Willpharma, all paid to his institute. MP received unrestricted research grants and speaker fees from Gilead Sciences, Roche, Abbvie and MSD, all paid to her institute. All other authors declare no competing interest.
Provenance and peer review Not commissioned; externally peer reviewed.
Supplemental material This content has been supplied by the author(s). It has not been vetted by BMJ Publishing Group Limited (BMJ) and may not have been peer-reviewed. Any opinions or recommendations discussed are solely those 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 includes any translated material, BMJ does not warrant the accuracy and reliability of the translations (including but not limited to local regulations, clinical guidelines, terminology, drug names and drug dosages), and is not responsible for any error and/or omissions arising from translation and adaptation or otherwise.