By then, investigators had been searching for the source of a nationwide salmonella outbreak for more than two months; tomatoes were the prime suspect. Grubbs told three nurses her story and was transferred to the voice mail of a fourth. After hearing nothing for three days, she called the state health department in Denver. The next day, a nurse four hours away in Grand Junction called and instructed her to drop off her peppers with health officials in Cortez, who would send them to Denver for testing.
Grubbs didn't know that federal investigators had begun suspecting that jalapeño peppers might be responsible for at least some of the outbreak's 1,200-plus cases.
On Monday, Colorado's state health department announced it had isolated the outbreak strain in the samples she provided. It was the first time contaminated peppers had been linked to a case of illness. Investigators had announced the discovery of another contaminated pepper a few days earlier at a small distribution center in Texas.
Grubbs is still angry about the slow response. "Nobody seemed to be flipping concerned," she said.
Although the nation's ability to detect food-borne illness outbreaks has improved sharply in the past 20 years thanks to advances such as DNA fingerprinting, it still rests on the efforts of local and state public health departments. And not all states have the same resources or expertise. The uneven response can have repercussions far beyond a single jurisdiction's borders, slowing multi-state investigations where speed is critical to preventing illness and limiting financial losses for growers.
That disparity is one of the issues expected to be examined in two congressional hearings scheduled to begin today. Lawmakers plan to look at why it took so long -- from mid-May when the outbreak was detected to last week's discovery of the first contaminated pepper -- for investigators to figure out what was making people sick. Before the Food and Drug Administration changed its warning to include jalapeños on July 9, scores of consumers including Grubbs bought jalapeño peppers that made them or others ill. Tomato growers had incurred losses of more than $100 million.
Officials with the FDA and the Centers for Disease Control and Prevention have blamed much of the delay on the challenge of tracing raw tomatoes through a complicated distribution system. Although the 2002 Bioterrorism Act was supposed to improve traceability in case of an attack on the food supply, the salmonella investigators had to rifle through paper invoices instead of computer records. If they had been able to trace tomatoes more quickly, the FDA's top food safety official David Acheson said, they might have realized sooner that they needed to expand their investigation. They didn't begin pursuing peppers until early July, after the tomato warning failed to prevent more illnesses.
The salmonella outbreak became a real-time test of the abilities of the FDA, CDC and their state and local partners. The sooner they could find the source, the sooner they could stop the illnesses.
But public health officials are much better at detecting outbreaks than they are at determining what happened, said Michael Osterholm, the former top epidemiologist for Minnesota and now director of the Center for Infectious Disease Research and Policy at the University of Minnesota. He is scheduled to testify today.
Because it can take weeks to verify a food-borne illness and identify other cases with the same genetic fingerprint, many outbreaks are over by the time they are investigated. Frequently, public health officials never know what made people sick, let alone the source of the contamination. In 2005, the CDC took part in at least 49 multi-state investigations but in only 12 did investigators implicate a food, CDC data show.
Osterholm said lack of coordination among local, state and federal public health agencies was partly to blame. Each jurisdiction has its own procedures. Some have more funding than others. Responsibility for monitoring gastrointestinal disease, which could alert health officials to an outbreak, also varies. In half the states, it falls to local agencies; in a quarter, to a central state office; and in a fifth, to regional offices that can encompass several counties.
The experience of front-line staff can also make a difference. In the 2006 spinach E. coli outbreak, Cathy Powers, a longtime nurse epidemiologist for New Mexico, discovered a leftover bag of spinach during a routine interview with a victim. The spinach turned out to be tainted -- and a major break.
The CDC keeps an eye on national disease trends and coordinates multi-state outbreak investigations, making sure state investigators ask the same questions and sending epidemiologists to help.
"In states that were heavily affected, this was an enormous challenge just to keep up with the cases . . . to do lab work and interviews. This strained the resources available in state health departments for food-borne illness investigations," said Robert Tauxe, a CDC deputy director.
Health officials at all levels faced a host of other obstacles, big and small.
Texas didn't immediately recognize the outbreak until the CDC informed officials there that they had salmonella cases with the same genetic fingerprint as ones in New Mexico. Once the CDC notified the state, its lab began requesting specimens. The state was soon backlogged and was still reporting illnesses from May well into June.
By contrast, New Mexico officials issued their first warning about a possible outbreak on May 23, two days after confirming they had three Salmonella saintpaul cases with the same genetic fingerprint.
Many of the early victims lived in remote parts of the Navajo reservation that stretches across New Mexico, Arizona and Colorado, which required more travel than the usual phone interviews, said John Redd, a top epidemiologist with the Indian Health Service. Health officials from New Mexico and the IHS spent more than a week at the end of May, including Memorial Day, driving around the high desert east of Monument Valley to locate victims.
The first reported cases didn't give epidemiologists much to work with. Initially, there were no clusters, in which two or more people get sick after eating at the same location, such as a restaurant. Some of the cases were separated by hundreds of miles, and it's hard to draw conclusions from a single case.
The first cluster wouldn't be identified by Chicago public health officials until mid-June. Another cluster that emerged later involving a Minnesota restaurant provided an important lead. Once Minnesota's health department identified the cluster on June 29, department officials said, it took 10 days for epidemiologists there to complete a statistical analysis that implicated raw jalapeños and for the agriculture department to trace the jalapeños to a distribution center. That center shares an address with Agricola Zaragoza, a small produce distributor in McAllen, Tex., where FDA inspectors found a contaminated pepper last week.
Minnesota has a centralized system of food-borne illness surveillance. All specimens go to the state, not local, health departments. There is a toll-free phone number for reporting food-borne illness. Whereas in many states it takes five or more weeks to confirm and DNA fingerprint a case and get the results to epidemiologists, in Minnesota it usually takes two to three days.
In the Grubbs case, the nurse at the county health department in Cortez who handles communicable diseases was out of town.
"I am generally the back-up. I happened to be out of the office that week as well," said Lori Cooper, director of the Montezuma County Health Department. "Any public health nurse who is here should follow up on it."
Grubbs said she didn't see any sense of urgency.
"I didn't know I could go to my local grocery store and buy something approved by the FDA that almost killed my husband. . . . Someone should have told me it was the jalapeño," she said. "I was so disappointed in how our system was working."