Table of Contents
Published: May 25, 2026
Read Time: 5.9 Mins
Total Views: 38
Two pieces of clinical data landed this month that look on their surface like unrelated news from different corners of infectious disease. A study in Clinical Infectious Diseases found that doxyPEP, a single dose of doxycycline taken within 72 hours of unprotected sex, reduced syphilis cases in King County, Washington by 52.3 percent over two years. The CDC’s provisional tuberculosis data for 2025 showed cases falling by only one percent, from 10,395 to 10,260, after four consecutive years of increases. The first looks like a triumph. The second looks like an inflection point at best. Both findings, read together, reveal something important about how clinical interventions actually function as public health tools, and about how fragile they are when the infrastructure underneath them is removed.
Why clinical interventions sit where they do on the pyramid
Medical care is often described, in political rhetoric and in popular imagination, as the most important determinant of health. The evidence does not support that view. Clinical interventions sit in the fourth tier of the health impact pyramid, below socioeconomic policy, default-context interventions like clean water and smoke-free laws, and long-lasting protective measures like vaccines. They are powerful tools, but their power is conditional. To benefit from a clinical intervention, a person has to reach a clinic, receive a correct diagnosis, obtain the right treatment, and adhere to that treatment for as long as it takes to work. Every one of those steps is a place where the intervention can fail, and the populations most likely to need clinical care are the populations for whom each step is the hardest.
This is why public health’s role in clinical medicine is not to deliver the care itself but to ensure that the people who need it most actually complete the cascade. Diagnosing, treating, and following up are the visible parts of medicine. Finding the patients in the first place, helping them navigate the system, paying for the treatment, supporting adherence over months or years, those are the parts that determine whether a clinical intervention works at the population level. They are also the parts that have been most aggressively cut.
How doxyPEP actually worked
Syphilis rates in the United States have been climbing for more than a decade, a trend that has prompted a lot of hand-wringing about individual behavior and very little serious investment in the public health response. In 2023, King County, Washington began actively promoting doxyPEP and offering it to the populations at highest risk, men who have sex with men and transgender women who have sex with men. The intervention is biologically straightforward: doxycycline taken within 72 hours of exposure prevents bacterial sexually transmitted infections, including syphilis. What made the King County program work was not the pharmacology alone but the public health system that wrapped around it. Clinics promoted the intervention. Providers were trained to offer it. Patients were supported in using it consistently.
The results are striking. Syphilis cases fell by 52.3 percent overall, an absolute reduction of more than 3,000 cases. Cases fell among cisgender women as well, almost certainly through an indirect effect: as transmission declined in the highest-risk networks, it declined in the broader sexual networks those individuals were connected to. This is how public health is supposed to work. A clinical intervention, deployed thoughtfully and supported by a functioning local health department, produced a measurable population-level reduction in disease in two years. It is the kind of result that should be replicated everywhere syphilis is rising, which is to say almost everywhere.
Why tuberculosis is harder, and more fragile
Tuberculosis demands more from the health system than almost any other infectious disease. The diagnosis requires specialized testing that most primary care providers do not perform. The treatment requires months of strict adherence to multiple antibiotics, often with significant side effects. The symptoms are burdensome enough to disrupt a person’s life but rarely debilitating in the early stages, which means diagnosis is often delayed. Most cases in the United States occur in immigrants and other populations with limited access to high-quality medical care, populations that the health system was already failing before recent policy changes made things worse. For all of these reasons, most TB care in the United States is either delivered directly by public health agencies or monitored closely by them. There is no commercial model for treating tuberculosis. Without public health, there is no TB control.
After four straight years of increases, the one percent drop in 2025 cases is a modest reversal, and it would be wrong to read it as more than that. The progress the United States made toward eliminating domestic TB transmission was built on decades of sustained investment in domestic public health services and in global TB control. Both of those investments have been cut dramatically in the past year, and the effects of those cuts will not appear in surveillance data immediately. Tuberculosis incubates for years. The cases that show up in 2027 and 2028 will reflect the transmission that is happening now, in the gaps left by clinics that have closed and outreach workers who have been laid off. The 2025 numbers are not yet the bad news. The bad news is being written now and will arrive on schedule.
The common feature
DoxyPEP works because a public health system promotes it, distributes it, and supports the people using it. TB control works because a public health system finds cases, delivers treatment, and stays with patients for the months it takes to cure them. The drugs themselves are nothing without the infrastructure that delivers them. When that infrastructure is cut, the clinical intervention does not become 30 percent less effective in some abstract sense. It stops reaching the people who need it, and the disease comes back.
The contrast between King County’s 52 percent reduction in syphilis and the United States’ barely-moving tuberculosis numbers is not a story about which intervention is better. It is a story about what happens when you fund the public health system that wraps around a clinical intervention, and what happens when you do not.
