Primum non-nocere: Is it time to stop screening for Neisseria gonorrhoeae and Chlamydia trachomatis in men who have sex with men taking HIV pre-exposure prophylaxis?

A number of authors have recently questioned whether we should be screening asymptomatic men who have sex with men (MSM) for Neisseria gonorrhoeae (Ng) and Chlamydia trachomatis (Ct).1–3 These authors have noted that screening for these infections results in very high levels of antimicrobial consumption with the attendant risks of antimicrobial resistance (AMR) induction.2 This is concerning if one considers that the decision to screen for these infections was not based on evidence from randomised controlled trials (RCTs).1 Screening guidelines typically require high-quality RCTs to show clear evidence of net benefit of screening before a screening programme can be introduced.4

We recently published the results of the first RCT assessing the efficacy of three-site, three monthly screening (3×3 screening) for Ng/Ct versus non-screening in reducing the incidence of these infections among MSM taking HIV pre-exposure prophylaxis (PrEP).4 We found that 3×3 screening was associated with a slightly decreased incidence of Ct infections, but not Ng infections. This finding is commensurate with the longer duration of asymptomatic Ct-infections compared with Ng-infections.5 This longer duration could explain how periodic screening detects more asymptomatic Ct infections than Ng infections, as the latter would have cleared spontaneously before screening. However, the slightly lower Ct incidence could also have been due to a bias in the study methodology where asymptomatic Ng/Ct infections could be double-counted in the non-screening but not the screening arm.4 Interestingly, the effect of 3×3 screening on the incidence of Ct infections disappeared when controlling for this bias. Importantly, non-screening was associated with a lower consumption of antimicrobials. Macrolide consumption decreased by 51%, ceftriaxone consumption by 44% and doxycycline consumption by 45%.

The findings of this RCT are commensurate with other types of evidence reviewed by Williams et al.1 For example, two large RCTs in the Netherlands and Australia found that periodic screening for Ct in the general population did not lead to a reduction in Ct prevalence.6 7 Likewise, a systematic review of observational studies among MSM concluded that the intensity of screening for Ng and Ct had no detectable effect on the prevalence of these infections.8 A study in a Belgian PrEP centre found that three monthly Ng/Ct screening resulted in macrolide consumption four to sevenfold higher than thresholds associated with inducing macrolide resistance in a range of bacterial species.9 This study also found that reducing screening led to a dramatic decline in antimicrobial consumption without noticeable adverse effects.9 Finally, a large study of over 200 000 MSM in European countries found that screening for Ng/Ct increased over time but was not associated with a decrease in the incidence or prevalence of total Ng/Ct infections.3 Furthermore, screening in this study was associated with an increased incidence of Ng/Ct symptomatic infections, suggesting that screening and treatment of asymptomatic infections might abrogate the development of an effective immunity against these infections, and thus be counterproductive.

We conclude that the available evidence suggests that screening has little or no effect on the incidence and prevalence of Ng/Ct in MSM on PrEP. It is, however, costly and leads to a substantial increase in antimicrobial consumption with resultant negative effects on the microbiomes and resistomes.10 These high levels of antimicrobial consumption are a plausible explanation for the high prevalence of AMR in subpopulations of MSM.1 9 In Belgium, the proportion of Ng isolates with resistance to azithromycin increased from 2% to 33% in a decade, and this increase was steeper among MSM.11 Likewise, all Mycoplasma genitalium isolates from MSM in Belgium are macrolide resistant and a quarter are fluoroquinolone resistant, which means a large proportion of these infections could be untreatable.12 Evidence is also mounting that the intensive use of broad-spectrum antimicrobials such as ceftriaxone and azithromycin may have long-term adverse effects on the microbiome.10

A central principle of medical ethics is not to do any harm (primum non nocere).13 While we have good evidence that certain screening programmes are beneficial, this needs to be established by RCTs.1 Regrettably, we started screening MSM for Ng/Ct before performing such RCTs. We now have evidence from RCTs and other types of studies that suggest that screening MSM for Ng/Ct has little or no benefit but rather results in harm. This rationale has led to the dropping of the recommendation to screen MSM for Ng/Ct in the latest Belgian PrEP guidelines. Future surveillance of Belgian PrEP cohorts will hopefully provide further guidance as to the effect of this policy change.