The eight-year-old girl died on a Saturday, a day after her birthday, in a state that had not seen a case like hers in over a decade 67. Lillian Smart likely contracted Naegleria fowleri while swimming in Lake Claiborne, Louisiana—a rare, brain-eating amoeba that is almost always fatal. The Louisiana Department of Health confirmed the infection, the state’s first since 2013 7. There is no vaccine, no treatment protocol that reliably saves these patients, and no way to screen a freshwater lake for a pathogen that kills roughly 97 percent of those it infects. The tragedy is real, and it is also, statistically, vanishingly rare.
That tension—between the visceral and the probabilistic—is the through-line of today’s health news. In the Democratic Republic of Congo, an Ebola outbreak declared on May 15 has become the deadliest in the country’s history, with 2,642 deaths and 5,514 confirmed cases as of August 22 12. The culprit is the Bundibugyo strain, a rare variant for which no approved vaccine or treatment exists 12. The numbers are stark, but they describe a regional catastrophe, not a global one. Spain, meanwhile, is shipping 45,000 vaccine doses to Ceuta—15,000 MMR and 10,000 each for diphtheria-tetanus, varicella, and polio—to prevent outbreaks among migrants who arrived en masse on July 30 3. That is a policy response to a foreseeable risk, executed before the worst case materialized.
The distinction matters because our attention is finite, and our fear is easily misallocated. Consider the new FDA-authorized blood test for Alzheimer’s disease, Elecsys pTau217, developed by Roche with Eli Lilly 10. It measures the p-tau217 biomarker, which predicts beta-amyloid accumulation in people aged 55 and older with cognitive impairment 10. This is a genuine advance—a blood test replacing invasive, costly PET scans—but it is not a diagnosis of dementia. It identifies biological signs, not clinical outcomes. The risk of over-interpretation is real, and the agency’s authorization is not a green light for home testing.
The same caution applies to the Phase 3 trial results from Moderna and Merck, announced Aug. 19, for intismeran autogene, a personalized mRNA melanoma vaccine combined with Keytruda 45. The trial enrolled 1,137 high-risk patients and showed a statistically significant reduction in recurrence and distant metastasis compared with Keytruda alone 45. That is a meaningful result, but it is a treatment for a specific cancer in a specific population, not a general cure. The word “breakthrough” should be used sparingly, if at all.
Then there are the studies that quantify risk in ways that demand careful reading. A Barcelona-based analysis of nearly 30 million births in Brazil found that drought exposure during pregnancy raises preterm birth risk by an average of 2 percent, rising to 5 percent during severe drought 9. A Florida Atlantic University study of 17,024 U.S. men found a 30 percent higher prostate cancer risk among those consuming the most ultraprocessed foods 11. Both are statistically significant, both are observational, and neither proves causation. A 30 percent relative increase sounds alarming until you consider the absolute baseline risk, which the study’s design cannot fully isolate from other dietary and lifestyle factors.
What we can control is more mundane: Chile’s health ministry issued an alert for aflatoxin-contaminated roasted peanuts, a specific lot with a specific production date 12. A psychotherapist observes that family conflicts with preteens now center on screen time rather than going out at night 8. These are the small, actionable items. The amoeba and the Ebola outbreak are not.
The unresolved question is whether our public health systems can hold two truths at once: that rare, terrifying events deserve vigilance, and that the greatest aggregate harm often comes from the slow, unglamorous exposures—drought, diet, delayed diagnosis. The reader’s decision is not whether to fear the lake, but whether to demand the same urgency for the risks that do not make headlines.