Header graphic for print
Marler Blog Providing Commentary on Food Poisoning Outbreaks & Litigation

2006 was a Very Bad Year for E. coli and Leafy Greens

This is another it what will be a long – too long – series of outbreak investigations where we have represented consumers in what I hope will be a cautionary tale, and a learning experience, for manufacturers of food.

It was not that outbreaks linked to leafy greens had not happened before – 2006 was just an especially bad year.

Dole Spinach - Official word of the spinach outbreak broke with the FDA’s announcement, on September 14, 2006, that a number of E. coli O157:H7 illnesses across the country “may be associated with the consumption of produce.”  “Preliminary epidemiological evidence suggests,” the statement continued, “that bagged fresh spinach may be a possible cause of this outbreak.”  By the date of the announcement, fifty cases had been reported to the CDC, including eight cases of hemolytic uremic syndrome (HUS) and one death.  States reporting illness included Connecticut, Idaho, Indiana, Michigan, New Mexico, Oregon, Utah, and Wisconsin.

The much-publicized outbreak grew substantially over the next several days.  By September 15, the FDA had confirmed 94 cases of illness, including fourteen cases of HUS and, sadly, one death.  Recognizing the lethality of the developing outbreak, the FDA’s September 15 release warned people should “not eat fresh spinach or fresh spinach containing products.”

Press Releases over the ensuing days announced steady growth in the number of people sickened, hospitalized, and with HUS as a result of the outbreak—109 cases from nineteen states by September 17, and 131 cases from twenty-one states just two days later.  The latter statistic included 66 hospitalizations and twenty cases of HUS.

Meanwhile, the FDA and CDC, in conjunction with local and state health agencies from across the country, worked feverishly to figure out the brand names associated with illness.  Early statistical analysis suggested that many brands were implicated, but the spinach sold under the several brand names had all come from the Natural Selection Foods processing center in San Juan Batista, California.  Accordingly, Natural Selection recalled all of its spinach products with “use by” dates from August 17 to October 1, 2006.   The recall, of course, included Dole brand spinach.  But further data and study ultimately narrowed the possible sources of the outbreak down to one brand of packaged greens:  Dole.

Though epidemiological evidence had already strongly linked Dole to the outbreak, the FDA found the proverbial “smoking gun” on September 20.  The bag of Dole baby spinach had been purchased and consumed by an Albuquerque, New Mexico woman, and testing by the New Mexico State Health Department had confirmed that the product was contaminated with E. coli O157:H7 bearing the same genetic marker as the outbreak strain.  The FDA announced the critical finding on September 21, 2006—also disclosing the “best by” date on the positive Dole bag of August 30—thereby giving a worried public a bit more information on what spinach products to eat, if any, and what to avoid.

By the date of the FDA’s September 21 announcement, the number of confirmed cases had swelled to 157 people from twenty-three states.  Ultimately, the FDA confirmed 204 outbreak-related cases, with 102 hospitalizations, thirty-one cases of HUS, and three deaths, though the actual number of people affected by the outbreak was certainly much larger.  In addition to the elderly Wisconsin resident, the FDA stated that the outbreak had claimed the lives of two-year-old Kyle Algood, from Chubbuck, Idaho, and also 81-year-old Ruby Trautz, from Bellevue, Nebraska.[1]  The tragedy of this outbreak can hardly be overstated.

Epidemiological and laboratory evidence, which had already proved the link to Natural Selection and Dole, soon revealed that the contaminated spinach had been grown at Paicines Ranch in San Benito County, California.  More specifically, investigators had traced the source of the contaminated spinach to one field on the ranch that had been leased by Mission Organics.

Once identified as the likely source for the outbreak, Mission Organics became host to health officials looking for the outbreak strain of E. coli O157:H7.  State and federal investigators took hundreds of environmental samples and swabs from the vicinity of the implicated spinach field, which was fifty acres in size, including from a nearby cattle pasture and water source.  Investigators also sampled the intestinal lining of feral pigs that had been killed as part of the investigation.[2]  Samples from a variety of sources, including the pigs, the water, and cattle feces, tested positive for the same strain of E. coli O157:H7 that had now been isolated in over 200 people nationally.  Finally, the outbreak strain of E. coli O157:H7 has been isolated in at least thirteen separate bags of Dole baby spinach.

Once the investigation was completed, a final report on the outbreak was prepared by the California Food Emergency Response Team (CalFERT), a team comprised of members from the FDA and the California Department of Health Services.[3]  The Final Report is replete with facts damning of all those involved in the growing, harvesting, processing, distribution, and sale of the implicated spinach products.  For example, speaking of the NSF processing facility, it states:

During the production week from August 14-19, 2006, the NSF South facility had the highest weekly production volume of the month.  Between August 13-20, 2006 production email exchanges revealed a string of personnel shortages, including nine absent employees on Sunday, August 13, the date of the weekly-extended sanitation shift.  Personnel records reveal that a number of absences were due to illness or illness in the family…NSF did not conduct ATP testing on a daily basis as required by the firm’s SOP. No ATP testing was conducted from August 15-25, 2006.  One ATP test collected from a scale vibrator failed on August 10, 2006, and no retest was documented.[4]

The Final Report also faulted with NSF’s procedures for monitoring the quality of processing-water, it’s record keeping, and its inability to demonstrate that harvesting bins were being washed to prevent cross-contamination.[5]

As for the Mission Organics growing operation, the findings were even more disturbing. The Final Report found that the land on the ranch where the spinach was grown “was primarily utilized for cattle grazing.” Moreover:

Investigators observed evidence of wild pigs in and around the cattle pastures as well as in the row crop growing regions of the ranch[6]…. Potential environmental risk factors for E. coli O157:H7 contamination identified during this investigation included the presence of the wild pigs in and around spinach fields and the proximity of irrigation wells used for ready-to-eat produce to surface waterways exposed to feces from cattle and wildlife.[7]

And, despite the right to inspect both NSF’s processing facility, and the Mission Organics fields, the Final Report amply demonstrates that Dole did nothing to protect its customers.

Taco Bell Lettuce - Only weeks after the spinach E. coli crisis in September 2006, Taco Bell’s outbreak occurred amongst residents of Delaware, Pennsylvania, New Jersey, and New York.  The first cases were identified in late November 2006, each victim experiencing onset of illness at some point after November 15.  In its summary of the outbreak, the CDC reported as follows on December 14, 2006:

This outbreak was clearly linked to Taco Bell restaurants in the northeastern United States. As of 12 PM (ET) December 14, 2006, Thursday, 71 persons with illness associated with the Taco Bell restaurant outbreak have been reported to CDC from 5 states: New Jersey (33), New York (22), Pennsylvania (13), Delaware (2), and South Carolina (1). States with Taco Bell restaurants where persons confirmed to have the outbreak strain have eaten are New Jersey, New York, Pennsylvania, and Delaware. (The patient from South Carolina ate at a Taco Bell restaurant in Pennsylvania). Other cases of illness are under investigation by state public health officials[8].

By year’s end, 78 probable and confirmed cases of E. coli O157:H7 had been reported.  Among these, at least 53 victims were hospitalized, and 8 people developed HUS.

The case control investigation led by CDC epidemiologists concluded, “[S]hredded lettuce consumed at Taco Bell restaurants in the northeastern United States was the most likely source of the outbreak”.[9]

The New York Department of Health summarized the relevant findings of the CDC studies and of its own investigation in its summary report on the outbreak, as follows:

The national outbreak investigation summary is not available as of April 1, 2007.  Preliminary findings of the multi-state outbreak investigation, including the case control study, reported by FDA on December 13, 2006 considered lettuce to be the single most likely source of the outbreak.  The review of the food items and ingredients associated with NYS cases was consistent with the preliminary findings of the national investigation.  Also, See Final Report.

Taco John’s Lettuce - As the Taco Bell outbreak was winding down, another one was beginning.  On December 8, 2006 the Minnesota Department of Health (MDH) received a report from a microbiologist with the Albert Lea Medical Center that the laboratory had confirmed E. coli O157:H7 in isolates cultured from stool specimens obtained from 3 patients.  Four days later on December 11 MDH epidemiologists interviewed 5 patients confirmed or suspected to be infected with E. coli O157:H7 in Albert Lea.  All 5 had eaten at a Taco John’s restaurant located in Albert Lea in the week before symptom onset.  Also on December 11, MDH became aware of a concurrent outbreak of E. coli O157:H7 associated with Taco John’s restaurants located in Iowa.

To enhance case finding, the MDH issued a press release and health alert to notify health care providers and consumers that a potential foodborne illness outbreak was underway.  A public hotline was set up for reporting illness.  Potential cases were interviewed about symptoms and food history.  Stool collection kits were sent to anyone reporting bloody diarrhea.  All O157 isolates received at the MDH Public Health Laboratory were subtyped by pulsed-field gel electrophoresis (PFGE) with restriction enzymes Xba1 and Bln1.

After an association between illness and food served at Taco John’s restaurants, investigators were established; MDH investigators conducted a case-control study to identify the food vehicle of transmission.  A confirmed case was defined as an individual with a laboratory-confirmed diagnosis of E. coli O157:H7 with symptom onset between November 24 through December 15 who had eaten in the week preceding illness at one of two Taco John restaurants implicated in the outbreak.  A probable case was defined as a person who developed bloody diarrhea within one week after eating at one of the two implicated Taco John restaurants with a meal date from November 24 through December 15, 2006.  Controls were dining companions of cases who had not become ill.  Cases and controls were asked about Taco John menu items consumed and ingredient specific analyses were conducted.

A total of 124 persons reporting illness were interviewed.  Twelve persons met the confirmed case definition, 20 met the probable case definition, and 92 had mild symptoms that did not meet the case definition.  Confirmed and probable case patients ate food at one of the two implicated Taco John restaurants between November 29 to December 6.  Eight case-patients were hospitalized and 1 person developed HUS.  Eleven of the culture confirmed cases had isolates from stool that were indistinguishable by PFGE using both enzymes.  PulseNet designated patterns were EXHX01.3596 for Xba1 and EXHA26.1417 for Bln1.  These PFGE patterns were indistinguishable to the Taco John associated outbreak in Iowa.

Only one food item, shredded iceberg lettuce, was statistically associated with illness (odds ratio, 41.3%; 95% confidence interval 4.7 to 362.3; p<.001).  A separate analysis in Iowa also found shredded lettuce to be significantly associated with illness.  The Minnesota Department of Agriculture (MDA) obtained invoices from the two Taco John restaurants in Minnesota associated with the outbreak.  The primary distributor was Roma Foods of Rice, Minnesota.  The same distributor delivered food to Taco John restaurants associated with illness in Iowa.  The lettuce would eventually be tracked to California.  See, Final Report.

The good thing is that it appears that the leafy green industry has been paying attention.  Although outbreaks have happened since 2006, they are fewer in number and fewer in victims.  Let’s hope it stays that way.


[1]           Two more deaths are also related: Betty Howard and Ruth Dunning.

[2]           Investigators had found tracks from feral pigs in the spinach field, as well as broken, run-down fences surrounding the field.

[4]           Final Report, supra note 56, at 44-45.

[5]           Id. at 45.

[6]           In a May 7, 2007 addendum to the CalFERT report, the presence of pigs and their possible role as a source of the E. coli O157:H7 was clarified as follows:  “This sentence was not intended to indicate that wild pig tracks were observed in the implicated 50.9 lot (lot 1) on Paicines Ranch.  Investigators observed wild pig tracks in fields and on dirt roads approximately 1 mile south of lot 1.  Additionally, as described in the report, numerous wild pigs were observed on the Paicines Ranch.”

[7]           Id. at 3-4; see also Final Report at 31-33, and 45-47.