A public health ethics analysis of Doxy-PEP: arguments for and against – the ‘yes’ case

Introduction

The resurgence of syphilis globally is a major public health concern.¹ In Australia, notifications of syphilis have risen over the past decade, from 6.7 per 100,000 population in 2012, to 22.7 per 100,000 population in 2021.² Overall, notifications remain over four times higher among men compared with women,² with gay, bisexual and other men who have sex with men (GBMSM) particularly at risk, especially those who are HIV-positive or have other risk factors.³ Aboriginal and Torres Strait Islander peoples also suffer a significantly higher burden than the non-Indigenous population, with 122.6 notifications per 100,000 population, including rising rates of congenital syphilis.²

Using the Kass public health ethics framework (Box 1),⁴ to analyse the ethical justification for doxycycline post-exposure prophylaxis (Doxy-PEP), I will argue that the intervention is justified.

What are the public health goals, and how effective is the intervention at achieving them?

Although Doxy-PEP is broadly effective in reducing bacterial STIs, including chlamydia and gonorrhoea, its primary goal in GBMSM populations would be to reduce the incidence of syphilis.⁵ Preventing syphilis has a range of beneficial impacts, both at the public health level – reduced disease burden, reduced cost, less exposure to the antibiotics in treatment and reduced mortality from congenital syphilis – and at the individual level, including reduced anxiety and an increased sense of control over sexual health.⁶ Evidence to date indicates that the proposed intervention would be effective as a population health measure in GBMSM. An open-label randomised controlled trial with 501 participants conducted in the USA demonstrated a 62–66% reduction in bacterial STIs in participants using Doxy-PEP.⁷ Trials from France have also shown efficacy. An open-label randomised controlled trial with 232 participants found a 73% reduction in syphilis infections (the IPERGAY study),⁸ while DOXYVAC, an open-label randomised factorial design trial, showed 84% reduction in first episode of chlamydia or syphilis.⁹ A modelling study using a hypothetical form of antibiotic prophylaxis showed that if 50% of GBMSM in Australia used antibiotic prophylaxis at 70% effectiveness, syphilis notifications could be halved within 12 months.¹⁰ Together, these results demonstrate reasonable effectiveness at the individual level and potential public health impact. Reducing syphilis infections is important, as untreated syphilis can have serious health impacts, including congenital abnormalities, and people may not recognise primary symptoms.¹¹ Notably, however, there is no evidence of Doxy-PEP efficacy in cisgender women, as the only study in women to date produced a null result, likely attributable to low product use.¹²

What are the known or potential burdens?

In her public health ethics framework, Kass identifies three categories of risk to consider: to privacy and confidentiality, to liberty, and to justice.⁴ Using this categorisation, the most significant risk of Doxy-PEP is to justice. Burdens – in the form of antimicrobial resistance (AMR) – are distributed unevenly. However, the severity of this risk remains unclear. For example, Langelier et al. from Luetkemeyer’s group reported that in 2024, Doxy-PEP use over a 6-month period was associated with an increase in the proportion of tetracycline antimicrobial resistance genes comprising the gut resistome, and an increase in the expression of tetracycline antimicrobial resistance genes, but did not significantly alter the composition of gut microbiome.¹³ This adds to the 2023 finding of Luetkemeyer et al., who reported an 11% increase in tetracycline-resistant gonorrhoea cultures at 12 months. This was the measured increase in resistance among gonorrohea isolates from participants using Doxy-PEP through the clinical trial, who had quarterly STI testing performed and who had a high adherence to Doxy-PEP during follow-up, and absolute numbers were small (4/15–5/13).⁷ It is not yet understood what, if any, clinical impacts these findings might have, nor how they might contribute to the generalised risk posed by AMR. For example, doxycycline is not used to treat gonorrhoea infections in Australia.¹⁴ Resistance to doxycycline has not been demonstrated for syphilis or chlamydia,¹⁵ although studies have found resistance generated to other pathogens.¹⁶

The risk of AMR forms the basis of the utilitarian argument against Doxy-PEP: that the utility gained from syphilis prevention is outweighed by utility lost through AMR. As can be seen above, the scope and magnitude of this risk is not clear, and it may not be clinically significant. Furthermore, the utility calculus alone does not determine whether a public health intervention is ethically acceptable: there should also be an attempt to eliminate socially driven health disparities.^17,18 According to the principle of positive liberty, the state also has a responsibility to enable people to flourish.¹⁸ This does not suggest that people freely undertake high-risk behaviours at society’s cost, but that people should be enabled to live free from illness.

How can the burdens and benefits be fairly balanced?

Burdens can be weighed and minimised through more research into AMR. Policies on testing and expanding molecular testing for AMR⁵ will better characterise and monitor the burdens on the wider population. Targeting populations at increased risk for STIs keeps the number of people taking antibiotics low. GBMSM will most benefit from the intervention (and trans women too are likely key beneficiaries); however, benefits will likely be shared, should GBMSM be a bridging population for syphilis transmission to others¹⁹ (noting that men who have sex with both men and women have been implicated as a ‘bridging’ population with respect to gonorrhoea).²⁰

The burden of AMR will be shared more widely; however, not evenly. Those taking antibiotics are more likely to harbour resistant organisms and to suffer resistant infections.²¹ Another population strongly impacted by syphilis is Aboriginal and Torres Strait Islander peoples.² The proposed intervention in GBMSM does not include this group for two key reasons: Doxy-PEP has not been tested for acceptability or effectiveness in this population, and the is as yet no evidence of Doxy-PEP efficacy in cisgender women.¹² In the interests of fairness then, community consultation and research into the acceptability and effectiveness of Doxy-PEP in Aboriginal and Torres Strait Islander populations should be an urgent priority.

Citizens and stakeholders must be included in deciding how this intervention can be implemented fairly.⁴ ASHM discussed its proposal with specialists from numerous scientific disciplines and community representatives; however, broader consultation could be conducted; for example, with the Australian Multicultural Health Collaborative. Benefits and burdens could be better balanced through expanding the research agenda to Aboriginal and Torres Strait Islander peoples. This may increase the likelihood of AMR emerging; however, this is a fair trade-off given high rates of syphilis within this population.

Critics of Doxy-PEP may argue that STI infection results from individual choices, yet STI risk is related to one’s population pool, as well as individual behaviour,²² and many structural factors influence choices; for example, a study in GBMSM in Australia found an association between sexuality-related discrimination and condomless anal intercourse, a sexual behaviour that is associated with increased STI risk.²³ The sequelae of individual choice is a factor in many illnesses. Basing decisions on this, whether for STIs or diabetes, is impractical and exposes care provision to the individual bias of care providers. From how it is financed to how it is delivered, public health care is – and should be – inherently about majorities supporting minorities.

Opponents of Doxy-PEP may also highlight the precautionary principle – when scientific evidence is uncertain and consequences are high, action should not be taken until more is learnt.²⁴ Although persuasive, GBMSM might question why this principle was not applied to prophylactic doxycycline for malaria, the treatment of acne and for growth promotion in livestock.¹⁴ These indications should not take priority over Doxy-PEP, and scale up of doxycycline for PEP could be offset by scaled down use for other indications.

Doxy-PEP is effective in reducing syphilis incidence; however, it carries risks in propagating AMR. The utilitarian case argues that the loss of utility from AMR outweighs the gain in utility from preventing STIs. However, public health should not be governed solely by utilitarianism, but should invoke the principle of justice, seeking to reduce health disparities, particularly among vulnerable populations. The state should not punish people for their choices, but should focus on enabling people to live healthy lives. Doxy-PEP should be implemented in Australia in GBMSM, in line with current data, and research should be undertaken in Aboriginal and Torres Strait Islander peoples to assess whether the intervention is acceptable and effective in these populations too.