The observed reduction in gonorrhea ranged from 0% to 55%, with variability largely explained by local patterns of gonococcal tetracycline resistance. The forum says these findings support doxy-PEP while making continued evaluation essential because widespread antibiotic use can also select for resistance beyond the intended STI targets. Doxy-PEP differs from HIV pre-exposure prophylaxis because doxycycline acts against a broad range of organisms rather than a single virus.

Its expected selection pressure may therefore extend beyond Treponema pallidum, Chlamydia trachomatis and Neisseria gonorrhoeae to organisms in the gut, skin and other microbiomes. The clinical consequences of these changes remain incompletely understood, and the policy forum warns against assuming either that resistance will have negligible effects or that every observed change is caused by doxy-PEP. Its proposed approach is to gather epidemiological, prescribing and resistance data and assess them against criteria agreed in advance.

That framework would make it possible to modify guidance when the balance between infections prevented, patient-centered benefits, adverse effects and ecological harms changes across settings. Resistance patterns vary among the pathogens targeted by doxy-PEP. T. pallidum, which causes syphilis, has no known acquired resistance to doxycycline, and laboratory efforts to develop resistance have suggested apparent evolutionary roadblocks.

C. trachomatis, which causes chlamydia, is highly susceptible to doxycycline, and no resistance has been observed. The emergence of doxycycline resistance in T. pallidum would nevertheless be a major concern because doxycycline is one of the few alternatives to penicillin. Gonorrhea presents a different situation because N. gonorrhoeae develops resistance to tetracyclines, and some areas of the world already have resistance rates exceeding 80%.

High-level resistance, defined as a minimum inhibitory concentration of at least 16 micrograms per milliliter, can be conferred by the plasmid-borne tetM gene. Evidence summarized in the forum is consistent with selection for resistant gonococci as doxy-PEP becomes more widely used. The authors cite selection for tetM in the DOXYVAC study, increasing phenotypic resistance among doxy-PEP users in Seattle, rising tetM prevalence in the United States, and an association between lower doxy-PEP efficacy against gonorrhea and higher tetM prevalence.

These observations support an expectation of an initial reduction where resistant strains initially account for a relatively small share, followed by reduced effectiveness as resistant strains spread. In a model without changes in behavior or transmission networks, prevalence could then return toward its earlier level, driven by resistant strains. The forum also notes that restructuring the gonococcal population could favor strains with reduced susceptibility to cephalosporins, although the size of that potential long-term effect remains uncertain.

Signals in bystander organisms broaden the monitoring challenge. Among doxy-PEP users, Staphylococcus aureus colonization was reduced, but those who remained colonized had higher rates of tetracycline-resistant S. aureus, raising concern about selection for multidrug-resistant strains. An analysis of gut-microbiome samples from the DoxyPEP trial found minimal changes in overall microbial composition but significant expansion and increased expression of tetracycline-resistance genes over six months, including a dose-response pattern.

A separate study of 100 milligrams of daily doxycycline for 56 days found persistent changes to the skin microbiome more than 200 days after the antibiotic course ended. These findings demonstrate biological resistance signals, but they do not yet establish the frequency of subsequent infection, treatment failure or other clinical consequences. Monitoring therefore must connect laboratory and genomic findings with outcomes relevant to patients and health systems.

The cost-benefit balance can change as an intervention succeeds. When bacterial STI prevalence falls, the risk of acquisition after a sexual encounter declines, fewer infections are averted per dose and the number needed to treat to prevent one infection rises. At the same time, if increases in antimicrobial resistance result in treatment failures or harder-to-treat infections, the external cost of each dose may grow.

Current users receive the preventive benefit, while resistance-related costs could emerge later, affect other patients and fall on healthcare systems. The forum also argues that a full accounting must extend beyond infections averted and resistance prevalence. Potential benefits include reduced anxiety about STI acquisition and avoidance of repeated treatment visits, while harms include gastrointestinal adverse effects such as gastritis and concerns about doxycycline use in pregnancy.

These calculations will not be uniform across populations or locations. The balance may differ where baseline gonorrhea resistance is already high, where syphilis transmission is driven by heterosexual networks, or where the burden of S. aureus infection is elevated. Evidence of efficacy comes chiefly from men who have sex with men and transgender women, with the original studies including people living with HIV or taking HIV PrEP; efficacy in cis-gender women has not yet been established.

A study in cis-gender women in Kenya did not significantly reduce bacterial STIs, although limited use of doxycycline in the intervention arm may have contributed to that result. Ongoing and planned trials are evaluating doxycycline prophylaxis in cis-gender women. Even if future trials establish efficacy in cis-gender women, a lower average STI incidence could mean a higher number needed to treat, while severe consequences such as congenital syphilis, pelvic inflammatory disease and infertility could make a smaller absolute benefit more consequential.

Individual effects also complicate population-level assessment. If recommendations expand to cis-gender women, safety concerns surrounding early pregnancy will become more prominent. Any formal analysis should therefore combine population estimates with patient-centered outcomes rather than treating reduced bacterial STI incidence as the sole measure of success or harm.

Attributing shifts in disease or resistance to doxy-PEP is methodologically difficult because relevant trends predated broad implementation. Before widespread US use, syphilis rates were increasing, chlamydia rates were relatively stable, and gonorrhea rates were declining in some populations for reasons that remain incompletely understood. The prevalence of tetM in N. gonorrhoeae was also rising, plausibly partly because therapeutic doxycycline was used for chlamydia.

Before-and-after comparisons can therefore misattribute pre-existing trends to the intervention. Surveillance adds uncertainty through shifts in testing volume, representativeness, testing-seeking behavior, reporting delays and heterogeneous sub-epidemics. The forum proposes comparisons among settings with different uptake levels and integrated surveillance that links individual doxy-PEP use with STI testing outcomes and resistance data for pathogens and commensal organisms.

Such designs are needed to distinguish intervention effects from background trends. None of these systems is currently routine in most locations. National and international systems provide some N. gonorrhoeae resistance data, but their geographic coverage is limited and reporting can be delayed.

Resistance in S. aureus and other bystander organisms is usually sampled opportunistically from clinical cases rather than measured across populations. Clinical consequences of resistance, including changes in prescribing practices, empirical treatment failures and infection complications, are even less consistently tracked. Integrating antimicrobial-resistance surveillance with STI surveillance and prescribing data is technically feasible but largely unoperationalized, especially in low- and middle-income settings.

Doxycycline’s low cost and broad availability add another difficulty: informal use through telehealth, standing prescriptions or direct acquisition is likely to continue regardless of official guidance, making both stewardship and evaluation harder. The forum recommends a formal mechanism that reviews the population-level case for doxy-PEP at defined intervals against current epidemiological and resistance data, with community advocates participating in governance. Pre-agreed thresholds could trigger review if tetracycline resistance in S. aureus, tetM prevalence in N. gonorrhoeae or co-selection for cephalosporin resistance crosses a specified level, while individual eligibility could be tightened toward those at greatest risk.

Conversely, expansion might be reconsidered in light of improved STI testing, earlier treatment or bacterial STI vaccines. The central proposal is neither automatic expansion nor abrupt withdrawal, but continuing justification through transparent criteria as the evidence changes.

Subject of Research: Monitoring and revising doxy-PEP guidance as effectiveness, antimicrobial resistance and population-level benefits and harms change

Article Title: How do we monitor doxy-PEP impact and when should we consider revisiting guidelines?

Article References: Grad, Y. H., Mittelstaedt, R., & Peters, R. P. H. (2026). How do we monitor doxy-PEP impact and when should we consider revisiting guidelines?. PLOS Medicine, 23(10), e1005276. https://doi.org/10.1371/journal.pmed.1005276

Image Credits: AI Generated

DOI: 10.1371/journal.pmed.1005276

Keywords: doxy-PEP, antimicrobial resistance, bacterial STIs, syphilis, chlamydia, gonorrhea, WHO guidelines, surveillance, adaptive policy, public health, monitor, impact