I track a lot of variables. Blood panels, HRV, glucose curves, sleep architecture. I am used to watching numbers move slowly. So when I saw that XDR Shigella went from 0% of US isolates in 2011 to 8.5% in 2023, I stopped. That is not a slow drift. That is a system failing in real time, and almost no one outside infectious disease circles is talking about it.
XDR means extensively drug-resistant: resistant to ampicillin, azithromycin, ceftriaxone, ciprofloxacin, and trimethoprim-sulfamethoxazole simultaneously. Those are not obscure backup antibiotics. Those are the standard treatment stack. When all 5 fail, there are no FDA-approved oral options left. You are managing symptoms and hoping the patient's immune system closes the gap.
The Resistance Curve Nobody Is Watching
The CDC PulseNet data covering January 2011 through October 2023 shows 510 XDR isolates out of 16,788 total. That sounds small until you clock the trajectory: zero for the first 4 years, then a steady climb to 8.5% by 2023. Among the 116 US XDR cases where HIV status was recorded, 46.6% had HIV co-infection. That is not a coincidence. Immunocompromised patients are both more vulnerable to severe shigellosis and more likely to be on the antibiotic regimens that select for resistance. The biology here is not subtle.
A University of Washington study published in early 2026 added the global dimension. By age 2, 25% of children in low- and middle-income countries have experienced severe Shigella-related diarrhea requiring medical care, and a significant portion of those infections already resist common antibiotics. Patricia Pavlinac, the associate professor who led the research, put it plainly: the burden is high among children in lower-resourced settings, a population without a platform. That sentence explains the entire funding gap.
Dr. Alex Chen would point out that 8.5% is still a minority of cases and that most healthy adults clear Shigella without antibiotics anyway. Fair. But resistance percentages do not plateau on their own, and the population most at risk, children under 5 and HIV-positive adults, cannot wait out a self-limited infection the way a healthy 35-year-old can.
The Newborn Data That Should Alarm Everyone
Research presented at ESCMID Global 2026 in April found antibiotic resistance genes in the meconium of 105 NICU newborns within 72 hours of birth. Median of 8 resistance genes per sample. Carbapenem-resistance genes in 21% of samples. These are infants who have never received an antibiotic. The resistance is arriving pre-loaded, through maternal or environmental exposure, before the child has taken a single breath outside the womb. That is not a future problem. That is the current baseline.
The optimization framing I usually apply to personal health protocols does not map cleanly here, and I want to be honest about that tension. I cannot run an n=1 experiment on a bacterial resistance curve. What I can do is read the trajectory data the same way I read a worsening biomarker: as a signal that the current protocol is failing and that waiting for more data before acting is itself a choice with consequences.
A Phase 3 vaccine trial is underway in Bangladesh. That is the right investment. The CDC needs to publish updated XDR surveillance data faster than every 2-plus years. And Congress needs to stop treating enteric disease funding as a line item for other people's problems. Shigella does not need a celebrity diagnosis or a wealthy-country outbreak to deserve serious money. The resistance curve is already telling us where this goes.
