How the Surveillance System Was Slowed

A foodborne outbreak is a race between detection and distribution. In this summer’s Cyclospora outbreak, the system designed to detect anomalies was deliberately slowed, and the result is a preventable documentation exercise: 1,645 confirmed cases, 141 hospitalizations, 5,100 more suspected cases sitting unanalyzed, and a single state — Michigan — reporting more than 3,700 cases against a normal annual baseline of 40 to 50. A 74- to 93-fold surge that would have triggered an immediate response under active surveillance instead crept past a downgraded system and reached consumers before anyone knew it was happening.

The July 1, 2025 decision by the CDC to make Cyclospora tracking optional at FoodNet sites did not contaminate anyone’s lettuce, but it eliminated the one instrument capable of flagging an anomaly against a consistent baseline. The agency reclassified Cyclospora, listeria, campylobacter, shigella, vibrio, and yersinia from mandatory to optional at the 10 state sites that make up the network; only salmonella and E. coli remained mandatory. The change arrived with budget cuts and was not publicly announced at the time. It surfaced nearly two months later only after a journalist inquired, according to Dr. Robert B. Shpiner, a clinical professor of medicine in pulmonary and critical care at the David Geffen School of Medicine at UCLA. FoodNet, the active surveillance network the CDC has operated jointly with the FDA, the USDA, and 10 state health departments since 1995, was the system built to produce a consistent national baseline against which an unusual cluster stands out. After the downgrade, that guarantee was gone. An outbreak that builds gradually across 34 states — as this one has — is exactly the kind of pattern a mandatory baseline catches and an optional one misses. As Shpiner wrote: “A count that lands after the food has been eaten can document an outbreak. It cannot stop one.”

The Workforce Erosion Compound

The downgrade was not an isolated administrative move. More than 3,000 public health workers had already left the CDC through firings, forced retirements, and attrition — roughly a quarter of the agency’s workforce, according to an analysis by KFF Health News. This was not merely a staffing problem; it was a structural erosion of the capacity that state and local health departments depend on for food tracebacks and patient interviews. Shpiner again: “Much of what the CDC does is push money and expertise down to the state and local departments that conduct the interviews and the food tracebacks, and those are the people who will find whatever is doing this.” When fewer workers are processing fewer data points, the system does not just run more slowly — it misses entire dimensions of the outbreak.

These two factors are not independent. They degrade the same detection pipeline through different mechanisms: the procedural downgrade reduced the data collected at the source; the workforce loss reduced the personnel available to process what little data remained. Together they produce a deficit larger than either would alone. The clinical-suspicion dependency illustrates the compound effect. Routine stool tests do not include Cyclospora; clinicians must specifically suspect the parasite before a lab will test for it. Under active surveillance, FoodNet’s anomaly signal would have alerted clinicians to start looking for Cyclospora during the June peak, triggering tests before cases multiplied. Without that signal, the trigger disappears before the first diagnosis occurs. Early-week cases go undiagnosed, and the system loses a critical lead-time advantage.

A Pathogen That Punishes Passivity

Cyclospora’s biology makes this downgrade disproportionately consequential. The parasite is not transmitted person-to-person; its oocysts must mature in the environment for days before they become infectious. The incubation period runs roughly a week, so by the time a patient is sick enough to seek care, the contaminated meal is a distant memory and the produce has been eaten or shipped onward. The pathogens that remained mandatory at FoodNet sites — salmonella and E. coli — present faster, appear in standard diagnostic panels, and are more likely to be caught by passive surveillance. The downgrade created an asymmetric gap: the pathogens most likely to produce large, diffuse outbreaks requiring network-level detection were exactly the ones removed from mandatory status.

Two Structural Absences

Two structural absences explain why the system failed and will fail again unless corrected. First, no statutory minimum-surveillance standard exists — no floor below which pathogen tracking cannot be reduced without congressional notification and public comment. The downgrade was possible because no mechanism required the agency to maintain a baseline or seek approval before cutting it. Second, no institutional guardrail required public notice, congressional consultation, or a notice-and-comment period before FoodNet’s scope could be narrowed. The change was reported nearly two months later only because a journalist inquired. These are not technical failures; they are deliberate omissions in the architecture of oversight.

Who Bears the Cost, Who Holds the Leverage

The political accountability is similarly asymmetrical. The party with the power to restore mandatory surveillance — CDC leadership — faces moderate urgency to do so: the outbreak is active, but the decision preceded it by a year, and institutional momentum favors inaction. The parties bearing the consequences — state health departments, local epidemiologists, infected patients — have no leverage to compel a reversal. Michigan cannot unilaterally restore national active surveillance. Patients cannot force institutional change while fighting a parasitic infection. The departed workers are gone; their institutional knowledge is irrecoverable. The result is a system that self-protects at the expense of those it is supposed to serve.

Shpiner, who began practicing intensive care medicine in Los Angeles in 1981 during the early months of the AIDS epidemic, drew a direct institutional parallel: “What a health system does not measure can spread in plain sight, and by the time the measurement catches up, the argument is no longer about prevention.” The same unmeasured-harm pattern recurs across decades and pathogens. The names change; the mechanism does not.

The Timing Failure

This is fundamentally a timing failure, and timing is what separates a preventable outbreak from a documented one. Restoring mandatory surveillance, establishing a statutory minimum-surveillance floor, and requiring public notice for any future scope changes would directly address the two root causes. The question is whether the surveillance system will be repaired before the next outbreak — or whether the cycle of downgrade, detection delay, and documentation will repeat with another pathogen that was made optional.

Analytical techniques used in this piece

This analysis applies the methods below. Each links to a short, plain-English explainer you can read and reuse.

Relationship Mapping
Extracts the network of ties among people, institutions, and entities.
Root-Cause Analysis
Traces a symptom back along its causal chain to the conditions that actually generated it.
Stakeholder Mapping
Charts the parties to a situation — their interests, power, and alignments.