More than a month after the Hawaii State Department of Health announced an outbreak of E. coli O157:H7 on the island of Oahu, epidemiologists at the agency say they still hope to determine the source of the outbreak, though they acknowledge the likelihood grows slimmer as time goes on. But ending an outbreak investigation without pinpointing a specific food source is much more common than non-epidemiologists may realize, said Michele Nakata, Disease Investigation Branch Chief for the state health department. In fact, Nakata said, for complex investigations such as this cluster of eight illnesses spread across the island, the agency identifies a specific source less than ten percent of the time. At the beginning of April, Food Safety News reported that state health investigators planned to trace back the grocery purchases of each of the confirmed patients using grocery membership card data, the next step in an investigation that began in March. That strategy didn’t reveal any strong leads, and now the investigation appears to be drying up, offering a lesson on the challenges of epidemiology along the way. In epidemiological investigations, Nakata said, the likelihood of success depends on a number of factors: the type of pathogen, the type of source and the number of people affected. For one, the very nature of the E. coli bacteria complicates E. coli investigations. Some pathogens, such as Staphylococcus aureus, often produce symptoms two to four hours after exposure, making it easy for health professionals to identify what the patient ate just prior to falling ill. E. coli, on the other hand, can take up to 10 days to exhibit symptoms in an infected person, if symptoms appear at all. And while the majority of those infected likely do not seek medical care, those who do often spend additional days waiting until the symptoms become too severe to ignore. Even during clinical care, the patient still may not get tested for E. coli specifically. For the select few who are correctly tested for E. coli, test results can take one to two days to return, followed by several more days during which confirmation testing in performed and another one to two days to match the bacteria’s DNA “fingerprint” to an outbreak cluster. By the time a patient matches up with a cluster of E. coli illnesses, investigators may be asking them to recall meals from two to four weeks prior. The type of food source itself can further complicate the investigation, as it may be an ingredient distributed in a variety of food products – and E. coli has been linked to a wide range of foods in recent years. It’s one thing if everyone says they ate hamburgers; it’s another if a contaminated batch of a minor ingredient such as a spice made it into a variety of dishes. This can all get especially complicated in a place like Hawaii. “Hawaii’s ethnic diversity, in addition to posing interview barriers, can complicate things further,” Nakata said. “Cilantro, for example, could be in Mexican salsa, a plate of Chinese noodles, or in a Vietnamese banh mi sandwich.” Food tracebacks can further complicate the investigation, as imports to the islands sometimes go through a complex maze of distributors before ultimately reaching the marketplace. When investigating a Salmonella outbreak linked to Indonesian tuna, Nakata was surprised to find that the fish first skipped over Hawaii and went to the continental U.S. where it was passed to as many as five or six distributors before finally reaching Hawaiian customers. Investigators may face further challenges when only a handful of patients are confirmed as part of the outbreak, as it becomes more crucial that each person provide as much detailed information as possible. Sometimes patients don’t return calls or provide accurate contact information when investigators are trying to figure out what they’ve eaten in recent weeks. “People can be surprisingly disinterested in something that they’ve already recovered from,” Nakata said. In Hawaii, the state health department conducts more than 7,000 disease investigations a year, with some outbreaks taking precedence over others. “Foodborne illness, though,” Nakata said, “because of the potential for widespread illness, always remains a top priority.”