Dairy workers across the United States face a growing occupational health risk as the number of confirmed human bird flu cases climbs to 71 since February 2024, according to the CDC. The infections trace largely to contact with cattle on farms where highly pathogenic avian influenza A(H5N1) has been spreading since the multistate dairy-cow outbreak was first reported on March 25, 2024. With federal agencies expanding herd surveillance and retail milk testing, the question is whether those tools will catch infections fast enough to protect the people working closest to sick animals.
Why 71 human bird flu cases signal a widening gap in farm-level detection
The rising human case count is not simply a bookkeeping update. Each new infection represents a failure to break the chain between sick cattle and exposed workers before the virus jumps species. The CDC’s situation summary lists 71 total reported human cases of A(H5) bird flu in the United States since February 2024, detected through both national flu surveillance and targeted monitoring on affected farms. Early in the outbreak, a CDC Morbidity and Mortality Weekly Report documented just two confirmed human cases of H5N1 during the initial phase of the dairy-associated spread, relying on RT-PCR confirmation and genomic sequencing to verify them. The gap between two early cases and 71 total cases reflects how rapidly the virus has moved through cattle herds and, by extension, into the people who milk, feed, and handle those animals.
The tension is straightforward. Federal agencies have built new surveillance tools, including a coordinated program to track herd status and bulk-tank sampling results state by state. Retail milk monitoring conducted during 2024 and 2025 added another layer of detection intended to reassure consumers that pasteurization inactivates the virus. Yet the human case count keeps rising, which suggests that surveillance is finding infections after exposure has already occurred rather than preventing them. If expanded herd sampling under these strategies identifies subclinical cattle infections in states not yet listed as affected, targeted human surveillance in those areas could produce another measurable jump in confirmed cases within weeks.
CDC scientists have also emphasized that the virus is circulating widely in animals beyond poultry and cattle. A detailed overview of H5N1 in mammal populations underscores that repeated spillovers into species such as foxes, seals, and farmed mink have already occurred globally. That broader pattern matters for dairy workers because it demonstrates the virus’s ability to adapt to mammalian hosts and raises concern that sustained transmission in cattle could increase opportunities for viral change. Each additional human infection, even if clinically mild, is another chance for the virus to encounter the human respiratory tract and test its evolutionary options.
Federal surveillance data and the Wisconsin spillover finding
Three federal agencies-the CDC, USDA’s Animal and Plant Health Inspection Service (APHIS), and the Food and Drug Administration-have built overlapping monitoring systems to track the virus through cattle, milk, and people. APHIS maintains a continuously updated dashboard of confirmed livestock detections by state, providing the clearest public picture of how many herds have tested positive and where the virus has moved geographically. Those data show a pattern of spread that began in a handful of states and steadily expanded outward, often following the movement of cattle and shared labor across regional dairy networks.
The National Milk Testing Strategy, outlined by APHIS as part of its response, sets out how states collect and analyze bulk-tank samples for viral genetic material. The NMTS framework is designed to catch infected herds before large numbers of cows show clinical signs, using lab-based PCR assays on pooled milk to flag farms for follow-up testing. In theory, this approach should allow animal health officials to move faster than the virus, isolating affected herds and tightening biosecurity before workers experience prolonged, unprotected exposure.
One finding from APHIS genetic sequencing work stands out. When the agency analyzed virus samples from a Wisconsin dairy herd detection, it determined the infection represented a new wildlife-to-cattle spillover event, genetically distinct from the lineage circulating in herds elsewhere. That result matters because it shows the virus is not spreading through a single chain of cow-to-cow transmission. Wild birds are introducing H5N1 into cattle independently and repeatedly, which means biosecurity measures focused only on preventing movement between farms will not stop new introductions. Each spillover creates a fresh opportunity for the virus to reach workers who have no reason to suspect their herd is infected.
The CDC’s early outbreak investigation, published in its Morbidity and Mortality Weekly Report, described how the first two confirmed human cases were identified through RT-PCR testing and sequencing as part of a coordinated One Health response involving animal and human health agencies. That methodological detail is relevant now because the same confirmation pipeline must process a much larger volume of suspected cases as the outbreak expands into new states and new herds. The more farms that enter monitoring, the more workers will qualify for testing after eye irritation, respiratory symptoms, or other nonspecific signs that could reflect H5N1 infection.
Gaps in case-level data and the next 60 days of detection
Several questions remain unanswered in the public record. No primary CDC or APHIS dataset breaks down the 71 human cases by exact exposure source with individual confirmation dates. That means it is not possible to determine from public data alone how many cases involved dairy workers versus poultry workers or other contacts, nor how many infections were linked to specific herd outbreaks. Full genomic sequences for all livestock and human isolates have not been released through APHIS or CDC portals; only selected announcements, like the Wisconsin spillover finding, are public. Without a comprehensive genetic dataset, outside researchers cannot fully map how many independent introductions have occurred or whether any human infections show early markers of adaptation.
State-by-state NMTS sampling volumes and negative-result counts are summarized at a high level without raw data files that would allow independent verification of how thoroughly each state has been screened. That lack of granularity makes it difficult to assess whether low numbers of confirmed infected herds in some regions reflect true absence of virus or simply limited sampling. For dairy workers, the distinction is critical: a state that has tested a large share of its herds and found no virus presents a very different risk profile than one with minimal testing and no confirmed positives.
Direct accounts from affected dairy workers or farm operators about exposure conditions are also absent from federal agency materials. Without those accounts, it is difficult to assess whether recommended biosecurity measures-such as consistent use of personal protective equipment, prompt reporting of sick animals, and restrictions on moving cattle between farms-are being followed on the ground or whether practical barriers on working farms are leaving workers exposed. Long shifts, heat, and the physical demands of milking and feeding cattle can make respirators and eye protection hard to use continuously, even when they are available.
The practical question for dairy workers and farm communities is whether expanded NMTS sampling will identify infected herds in states that currently show no confirmed cases. If it does, public health agencies will likely deploy targeted human surveillance in those areas, and the 71 documented infections could rise quickly as more exposed workers are tested. Over the next 60 days, the trajectory of the human case count will serve as a rough but important indicator of whether surveillance and biosecurity measures are catching up with the virus or continuing to lag behind it.
For now, the evidence points to a widening gap. Federal dashboards and testing strategies demonstrate that agencies can find the virus in cattle and milk with increasing precision. But each new human case underscores that detection is still arriving after exposure, not before it. Closing that gap will require not only more sampling and sequencing, but also practical support for farm-level protections that keep the people closest to infected animals from becoming the next data point in a growing outbreak.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
