Politics Foreign Affairs Culture Fellows Program

DoxyPEP Is a Disaster in Waiting

Fellini’s Satyricon comes to the pages of the New England Journal of Medicine.

Credit: sirtravelalot

There is a very easy way to eliminate death on the roads: prohibit all traffic. Unfortunately, such a policy would have effects other than the elimination of death on the roads, but road safety experts might have difficulty in appreciating this. As we saw during the Covid epidemic, enthusiasts—who may be perfectly honest and sincere—are sometimes blinded to other considerations by their enthusiasm. The elaboration of public policy requires more than attention to narrow scientific findings, important as the latter are. Policy rarely hits only its target. 

Trials of the old and cheap antibiotic, doxycycline, taken “post-exposure,” have been shown in trials to be effective in preventing sexually transmitted diseases, at least among “cisgender men who have unprotected sex with men,” as the scientific literature now delicately puts it, and among “transgender women.” The Centers for Disease Control and Prevention therefore now proposes to recommend the prophylactic use of doxycycline by such persons in such circumstances.


To enter the literature of what is now known as DoxyPEP is to enter a world in which acronymic medical bureaucratese confronts Sodom and Gomorrah. The paper published in April last year in the New England Journal of Medicine is a good example. It describes a trial in which 501 people in California were assigned either to take prophylactic doxycycline after “unprotected sex” or to receive “normal care.” 

The 501 people were of defined types: they were MSMs (men who have sex with men) and transgender women (men who have attempted to become, and live as, women) who were taking PreP (pre-exposure prophylaxis) to HIV (human immunodeficiency virus) or PLWHs (persons living with HIV infection), all of whom had had gonorrhoea, chlamydia, or syphilis infections within the last year and still practiced unprotected sexual relations. (Surely it would have been less judgmental and stigmatizing of the NEJM to have written that they had experienced gonorrhoea, etc., than to write that they had actually had it?)

One might hope that, even in California, the 501 who entered the trial were not representative of the population as a whole. But the results of the trial were clearer and more decisive than the results of clinical trials often are: Those subjects who took the doxycycline immediately after having had (perhaps I should say experienced) unprotected sex contracted considerably fewer sexually transmitted infections (STIs) than those who did not. Among the PrePs, the percentage of those who took doxycycline who developed STIs during follow-up was 10.7 percent compared with 31.9 percent of those who did not. For the PLWHs, the figures were 11.8 percent and 30.5 percent. 

Certain other figures caught my attention, carefully unremarked upon in the text by the authors. The number of sexual partners experienced in the past three months by the 501 varied between 4 and 17, the lifetime figures being between 143 and 491. (This was self-report, admittedly, with its possible under- and overestimation.) The median age was 36 for the PrePs and 43 for the PLWHs, so the first flush of youth cannot account for the continued large number of sexual partners—or what at any rate to me seemed in my bourgeois conventionality a large number. Neither past infection nor the prospect of future infection seems to have moderated behavior very much in this cohort of people, not even to the extent of using condoms. 

As for the consumption of drugs, 30 percent had taken cocaine, methamphetamine, or crack in the past three months, 32 percent had taken ecstasy, gamma-hydroxybutyrate, or ketamine, 45 percent amyl nitrite, and 48 percent marijuana. 


While carefully avoiding all mention of the above, the authors in their acknowledgements thanked the subjects for their “altruism” in taking part in the trial. This was the one value judgment in the whole paper, a judgment that personally I found unctuous, and probably inaccurate to boot. The participants risked nothing, for doxycycline is a safe drug. Self-interest was as strong a motive as altruism, insofar as they hoped to discover a means by which they might continue their way of life without, or at any rate with reduced, medical risk. Altruism, one might have thought, could or would have suggested other things to them than participating in such a trial. 

The paper did address, however, the possible development of resistance to tetracycline of infectious agents if the drug were prescribed frequently for prophylactic use. This is a serious matter, because once bacterial resistance develops, it has a tendency to spread rapidly in a population, and then the usefulness of the antibiotic declines accordingly. Because the numbers were small, the paper could not answer the question fully, but to my mind pointed at least to the likelihood of such resistance developing. For example, the proportion of Staphylococci resistant to tetracycline was double that in those taking it prophylactically than in those not doing so, and serious staphylococcal infections remain a threat to everyone.

Policy recommendations based on controlled trials like the one cited above seem to be “following the science,” and in a narrow sense they are. But there are caveats to be entered: for example, that results obtained in experimental conditions often cannot be immediately extrapolated to real life conditions, where things are much messier and where there are fewer staff, less follow-up, lower morale, carelessness, and so forth. 

Nor is it valid to conclude from the efficacy of prophylaxis in a particular (and one might hope rather unusual) group of people that it would be efficacious in prophylaxis for society as a whole: It might even have the contrary effect if, for example, the idea became general that risky behaviour could be counteracted by simply taking a pill. Lowering the prevalence of infection in one group does not necessarily lower the prevalence in society as a whole.    

One of the causes of the current epidemic of overdose deaths from opioids was an invalid (and corrupt) extrapolation from one group of patients to another. It was found, correctly, that hospital patients given strong painkillers after heart attacks or post-operatively did not become addicted to the painkillers; it was concluded, wrongly, that it was therefore safe to prescribe such drugs to all patients whatever who complained of pain. As an unfortunate partial corollary of this unjustified extrapolation, half a million people have died. 

Such an unjustified extrapolation of the results that I have signalled above is not inevitable in the case of doxycycline, of course, but it is at least possible. Publicity having been given to the results of the trial, patients, who believe that in all circumstances prevention is better than cure, might demand doxycycline of their doctors who, for one reason or another, will oblige.

There is also an intangible cultural effect of research like this. Certainly it gave me the feeling reading it that Fellini’s Satyricon had come to the pages of the NEJM. Doctors must take their patients as they find them, of course, and they perform no moral triage among them such that they give lesser attention to those whom they think are morally reprehensible. In my career, I treated many an abominable criminal, and did so always to the best of my medical ability. 

But insofar as the CDC’s proposed endorsement of DoxyPEP normalizes what to most people is distasteful conduct, and, by passing no explicit moral judgment on it, in effect asserts that it is morally neutral, the CDC may—I put it no higher than that, for the question is an empirical one—be helping it to spread. 

As for general medical journals such as the NEJM, they have not of late refrained from moral judgment, almost always of the woke variety, on all manner of subjects. But they would go to the stake rather than admit that the conduct of the subjects of this trial was wrong or even distasteful in any sense other than that it was medically imprudent, a situation which they hope to improve by means of DoxyPEP. 

Our fear of appearing censorious is now so great that we remain silent in the face of any degradation and praise the grossly licentious for their altruism in seeking their own safety.