E. coli O157:H7 Outbreak in Washington State: Lessons Learned

By Sally Fallon Morell

In late November of 2005, an outbreak of illness attributed to virulent E. coli O157:H7 afflicted eight Washington State individuals who had consumed raw milk, sparked a flurry of news reports, and vaulted the subject of raw milk into the national media.

The milk came from Dee Creek Farm, a small family farm near the town of Woodland in Cowlitz County, in southwestern Washington State, just north of the town of Vancouver in southwest Washington. The owners, Anita and Michael Puckett, operate a diverse pasture-based operation while raising their large family—11 children, of which seven still live with them on the farm. A married daughter, Summer Steenbarger, also lives on the farm with her husband and two small children.

In June of 2004, the Pucketts purchased a cow to provide milk for their large family. Their customers for wanted milk also—were clamorous for raw milk—so in January, 2005, the family began a cow share program. Theirs was one of several cowshare programs in Washington State, which later became loosely tied together under the auspices of WASDO—the Washington Association of Shareholder Dairy Owners.

Under the guidance of the Weston A. Price Foundation’s Spokane, Washington chapter leader George Calvert, and Chrys Ostrander, the organizer of the Raw Dairy Choice Campaign, the WASDO group had met in October, 2005, to formulate operating and safety guidelines for share-holder operations. Calvert’s vision embraced the share-holder model as a way of allowing extremely small producers—such as home schooling mothers or retired individuals on social security—to generate a decent income with a small number of cows on just a few acres.

Concurrent with the increase in share-holder operations, the state of Washington had modified the dairy regulations in July, 2005, to make it easier and less expensive for small dairies to become licensed. The principal change was the allowance of hand-capping, so small dairies could become licensed without the burdensome expense of a bottling machine. The architect of this laudatory revision was Claudia Coles, head of the Department of Agriculture’s safety program. While favoring regulations that made it easier to become licensed, Coles has consistently opposed the shareholder model, even expressing a refusal to share the stage with George Calvert at a conference on local value-added products held at Green River College in April, 2005. (However, at the conference, she sat with the Puckett family extensively and even had lunch with them.) Calvert and others maintained that the Washington State licensing requirements were still too expensive for very small operations.

By November, Dee Creek Farm was milking five cows for 45 families, who together had purchased 72 shares. All the families regularly visited the farm and knew how their cows were cared for and milked. Because of conversations held with the Pucketts at the Green River conference several months earlier, the Department of Agriculture was familiar with the Dee Creek operation and knew that the Pucketts’ goal was to become a licensed dairy.

The Pucketts were new to dairying. They did not have fancy facilities, but were careful to follow good sanitary practices as recommended by several other dairymen in the area. The cows grazed on pasture but came into a barn for milking. The Pucketts washed the teats with iodine and milked with a milking machine into a closed stainless steel bucket which they carefully washed before milking. They then transported the buckets into their kitchen where they transferred the milk into gallon-sized glass jars owned by the shareholders. It was the shareholders’ duty to provide clean jars, but the Pucketts then washed the jars again in the dishwasher, just to be safe. The filled jars went immediately into the freezer for 90 minutes and then into the refrigerator.

The various shareholder families had organized themselves into groups for milk pick-up. Each day of the week, one family picked up milk at the farm for a dozen or so other shareholders. When the pickup person arrived at the farm, the Pucketts placed the glass jars into a large cooler with blue ice. Each group of families had a pickup location, such as a front porch, where the cooler was left unattended. Families for each group had a window of 3 to 9 pm to come by and retrieve their milk from the cooler. The Pucketts had explained the pickup system to agents from the Department of Health and the Department of Agriculture on one of their visits to the farm; the officials offered no feedback, neither positive or negative.

Michael Puckett often picked up the empty cooler early the following morning, on the way home from his night job. About two weeks before the incident, he was somewhat unnerved by the presence of a car parked across the street of the pickup house. A driver sat behind the steering wheel. Puckett stalled to see how long the car would remain. Twenty minutes passed, at which time a police car arrived and the driver drove off.

In late November, an article by journalist Francis Robinson, originally published in Mother Earth News (October/November 2005), was carried by the Associated Press and appeared in several local newspapers. Robinson described the heartwarming story of Kelsey Kozak of Vashon Island who, at age 16, had set up a share-holder program for Iris, a six-year-old Jersey cow. The article described in detail how the cowshare worked, and the delight of shareholders at the delicious cream, butter, cheese and yoghurt that Kelsey produced from Iris’s milk.

The Outbreak

The first Dee Creek shareholder to become sick was a child from a family that picked up milk on Monday from the milking of November 26 or 27, at the pickup house where Michael had noticed the parked car. Soon other children fell ill and several were hospitalized. Health officials quickly confirmed the culprit as a strain of virulent E. coli O157:H7.

The newspapers reported 17 or 18 confirmed cases of food borne illness from E. coli in Dee Creek Farm shareholders—the first of many exaggerations to appear in the media. According to the Washington Department of Health there are only seven; according to Summer Steenbarger, who has acted as spokesperson for the farm since the incident, there were actually eight confirmed cases, six people (from four families) from the Monday pickup group of 13 families and two from the Tuesday pickup group of six families. In all, five were hospitalized, from three families, all from the Monday pickup group. One was released after treatment with IV for dehydration, one was released after three days, one after five days, but two of the children, from two different families, remained hospitalized in serious condition for around a month. One member of the Puckett family tested positive but did not get sick, and three other individuals among the shareholder families had symptoms but did not test positive. In fact, of the other 11 “cases,” at least eight of them specifically tested negative. One important point: Dee Creek Farm stopped distribution and advised their shareholders to cease consumption of their milk days before the Department of Health even acknowledged that they knew about the issue and contacted the farm.

The Washington State Department of Agriculture (WSDA) visited the farm three times during the week of December 12 to conduct a thorough investigation, taking samples of milk and swabs from all five cows on the farm. At that point, the Pucketts had put away the equipment they used for their share holders and were milking only for their pigs. This they explained to the inspectors, noting how they did things differently when they milked for human consumption. Mrs. Puckett had to remind the inspectors to hose down and sterilize their boots before they entered the milking barn—the farm was very muddy at the time, as were all farms in the region, because of a record-breaking rainy season. The family was present during the entire investigation and was able to capture most of it on video.

Predictably, the media coverage of the incident leaned heavily on health department reports in what Chrys Ostrander described as “a showcase of sensationalism and unprofessional journalism.” However, TV interviews did include many consumers who expressed passionate support of raw milk. In one, Anna Olson, a shareholder, poured a glass of Dee Creek Farm’s milk that she still had in her refrigerator for her family to drink.

The newspapers reported that the law offices of William D. Marler had contacted two of the three families. Bill Marler is a high-profile attorney who specializes in food illness litigation. His law firm won the Jack-in-the-Box and Odwalla juice cases.

On December 21, the Clark County Health Department issued a report stating that the Washington State Department of Agriculture had confirmed E. coli O157:H7 in Dee Creek Farm milk. (Clark county lies across the river from Cowlitz County and was the scene of an outbreak of E. coli O157:H7 at a county fair earlier in the year.) Claudia Coles left two telephone messages, one to the Pucketts and one to their daughter Summer, apologizing for the Clark County report and stating that at that point, no specific pathogens had been found.

Newspaper reports also stated that Dee Creek Farms had received a Cease and Desist letter from the state. What Dee Creek had received was a letter, friendly in tone, dated August 11, 2005, from the Washington State Department of Agriculture (WSDA) stating that providing milk through cowshare operations constituted a sale.

The State requested a reply within 15 days and Dee Creek complied with a letter dated September 9, 2005, that is, 15 days within receipt of the Department’s query. In their reply, the Pucketts stated that 1) the farm did not sell milk, 2) the farm was in the process of obtaining a Grade A permit and 3) they requested more information about obtaining the permit. George Calvert had received a similar letter, written in a less friendly tone, which was answered by his lawyer, who responded that cowshare operations did not constitute a sale. The first to receive such a letter responded via an attorney in March, 2005, requesting more information. They had not heard from WSDA since their response.

It was at a press conference on January 16, that Washington State officials announced a positive finding of E. coli O157:H7 in two samples of milk and five environmental samples from Dee Creek farm, of a strain that matched those cultured from the stool of the sick children. The report stated that “Raw milk bought and distributed from Dee Creek Farm was consumed by all affected people prior to their becoming ill,” and that “Dee Creek Farm illegally sold raw milk.” The state report concluded: “The WSDA FSP investigation along with the epidemiological work by the County Health Departments demonstrates that the illegal raw milk provided by Dee Creek Farms was the source of the E. coli O157:H7 that sickened at least 18 people in Washington and Oregon.”

Many Questions

While state officials express confidence that the outbreak was caused by raw milk, they have ignored many facts that call their conclusions into question. First, and most importantly, is the fact that independent labs testing Dee Creek’s milk, following the same testing protocol used by the state, found no E. coli O157:H7, not even in the same samples the state claims tested positive. It needs to be emphasized that no tests were done on the milk consumed by those who fell ill.

Another item: there was at least one concurrent outbreak of infection from E. coli O157:H7, affecting several family members in the Tualatin area. One boy ended up in the same hospital as the Dee Creek kids, in critical condition, and is still there after a month, possibly with long-term brain, nerve and kidney damage. Several of his family members were also confirmed with E. coli O157:H7. Neither the state nor the media has seen fit to report this important fact. We can only guess whether there were other unreported cases of E. coli O157:H7 at the same time.

It should be noted that Washington State has been plagued with E. coli O157:H7 outbreaks during the last 15 years. Between 1990-1999, 23 such outbreaks were reported, afflicting at least 288 individuals, from sources as diverse as fish, lettuce and lasagna, with a large number attributed to ground beef or hamburger. The worst outbreak occurred in October 1996, when 70 individuals became sickened from contaminated apple juice. This highly publicized outbreak led to federal regulations requiring the pasteurization of all fruit juice sold in retail outlets.

Officials attributed a June, 2000 incident on Vancouver Island, in which several children fell ill, to raw milk—eight children according to one report posted by the Center for Science in the Public Interest’s 2002 “Outbreak Alert,” five according to another (http://www.phac-aspc.gc.ca/publicat/ccdr-rmtc/02vol28/dr2801eb.html). The children were diagnosed with an infection of E. coli O157:H7 after drinking raw goat’s milk. However, official reports do not reflect other possible sources of infection—all the children lived on a cooperative farm and one had visited a petting zoo just prior to the outbreak. Petting zoos are common sources of infection. Furthermore, investigators had to work hard to find E. coli in the milk. One milk sample was found to be “presumptively” positive after “enrichment” with a testing substance; no E. coli was found in samples before “enrichment” and no E. coli was found in a second sample.

(The Pucketts believe this is how WSDA found E. coli O157:H7 in their milk—if they actually found it—they had a “presumptive” hit and then used an “enrichment” substance to find what they were looking for.)

Just two months before the Dee Creek incident, in September 2005, E. coli O157:H7 was found in water samples in a north Spokane water district, prompting a health alert.

According to information posted at the website of attorney Bill Marler, packaged lettuce and other vegetables are often sources of E. coli because they are sprayed with contaminated irrigation water. Contaminated water from intensive livestock operations also finds its ways into wells and water systems. Such water may be the underlying cause of the many E. coli outbreaks in this farming state.

One theory holds that the ultimate source of the virulent E. coli strains is genetically engineered soy, created by using fragments of E. coli micro-organisms as a vector for gene insertion, fragments which can mutate easily and which find the perfect environment for proliferation in the acidic guts of grain-fed confinement cows. Only feedlot cows headed for slaughter and fed pure grain have been found to harbor E. coli O157:H7 in their guts and manure. Runoff from large farms then carries the organism into water supplies. How convenient for agribusiness that the finger of blame points to milk from a tiny family farm where animals are fed on pasture.


Then there are the tactics of the Washington State Department of Agriculture—reminiscent of those used against Alta Dena, a raw milk dairy in California, back in the 1970s and 1980s, and more recently during a campylobacter outbreak in Wisconsin (where raw milk was accused of an outbreak that affected hundreds of people, most of whom did not drink raw milk). The state kept its final report secret until the day before the introduction of new raw milk legislation, and announced “proof positive” at a press conference. Dee Creek Farm and others involved learned of the final report and the press release from the media, not the state.

To make their case for new legislation, officials claimed that Dee Creek had barred entry to the farm. In fact, the Pucketts had never barred entry to the farm, only requested that inspections take place when they did not have other obligations (such as meeting with their attorney or visiting the doctor), to which officials agreed. (When confronted with these lies, Michael Tokos of WSDA apologized and agreed that the Pucketts did not ever bar entry.) The new legislation requires a customer list, based on claims that Dee Creek refused to supply such a list. This, too, is based on lies. The Pucketts did not ever deny the list, but first took steps to obtain authorized signatures from the owners. (They had 18 within 12 hours, with more on the way.)

The Senate bill, number 6377, sponsored by Washington Senator Mark Coumit, D-Cathlamet, would make cowshares illegal without a Grade A permit and imposes stiff penalties for farmers operating without a license. The bill grants the Department of Agriculture authority to inspect dairies suspected of operating outside health regulations and to obtain a search warrant if owners bar inspectors.

Under the capable leadership of Emmy McAllister, Snohomish, Washington Weston A. Price Foundation chapter leader, local activists quickly formed a group called the Washington State Raw Milk Micro-Dairy Task Force to work out alternative legislation that would make the licensing requirements less difficult and less expensive (such as a separate kitchen structure) for very small dairies selling directly to the public, while adding regulations that would provide a greater assurance of safety. The committee includes farmers from both licensed and share-holder dairies as well as Mark McAfee from Organic Pastures Dairy in California and officials from the Weston A. Price Foundation.

McAfee traveled from California to join the group for a meeting in Olympia on January 18 and testified at hearings before the Agriculture Committee the next day. Specifically, the task force recommended that small licensed dairies perform routine testing for E. coli O157:H7, keep frozen milk samples, maintain a list of customers and use containers with tamper-evident seals—none of these provisions is currently required for licensed dairies in the state of Washington. According to McAfee, the hearing seemed to be “just a formality” and that “no one on the committee was listening.”

The Current Situation

All three hospitalized children are on the road to recovery and leading normal lives. (Was it the raw milk they drank that saved them?) Dee Creek Farm expects to be in full operation this season. The farm has been testing their milk weekly—without finding E. coli O157:H7.

Dee Creek plans to obtain a Grade A license and continue its cowshare operation, aiming to begin milking cows again by May. Many of the share holders have requested that they participate in building the Grade A facility—they all have skills in different areas. Nevertheless, their estimated expenses are expected to reach $20,000.

The family remains confused about reports of a lawsuit. Two of the most critically afflicted families have assured the family that they have no plans of suing and indeed have never talked to a lawyer, and the other family is on friendly terms. Yet the following statement remains posted at billmarler.com: “Marler Clark plans to pursue litigation on behalf of at least two families.”

The Pucketts have been obliged to take on legal help for potential issues, however, which is being provided at a reduced rate but not gratis. The family’s financial resources are minimal and the state may impose fines. Donations for legal expenses are tax deductible and may be sent to WAPF/Dee Creek Farm Support, PO Box 1936, Woodland, WA 98674.

Senate Bill 6377 has passed out of the Agriculture Committee and on to the Rules Committee, which schedules legislation for votes by the full Senate. Language making distribution of unlicensed milk a felony has been downgraded to a gross misdemeanor, although repeat offenses would be felonies. The bill allows the Department authority to order a dairy to stop distributing milk—and also to embargo and dispose of milk from a dairy suspected of health violations—which could be an open door to harassment unless protections for the farmer are included.

Chrys Ostrander has set up a list serve for the committee at raw-milk-taskforce@lists.onenw.org and is posting updates at www.shareholderdairies.org. The committee is optimistic that separate, less onerous provisions for micro-dairies will be incorporated into the final legislation.

In deference to the current situation, the Weston A. Price Foundation has removed listings of unlicensed cowshare programs from the Washington State section of realmilk.com. But licensed dairies are posted—there are now five licensed dairies selling raw cow or goat milk, both on farm and in retail outlets throughout the state, with more small dairies are in the process of certification. However, there are no available sources in southwest Washington. One licensed dairy nearby milks four goats, but they are all dry over the winter. Families who need raw milk must obtain it through food co-ops or by mail order.

Given the anomalies contained in the WSDA final report, the unattended cooler and the suspicious behavior observed by Michael Puckett, an investigation is in order. Sabotage and entrapment cannot be ruled out. Indeed, media reports that “Agriculture officials were unable to get onto the farm or stop the distribution of milk until the matter became a public health issue (Daily News, Longview, WA, January 27, 2006),” indicate a strong motive for the state to create an incident as a means of obtaining changes in the dairy regulations.

Lessons Learned

Several important lessons emerge from the Dee Creek incident—lessons that all small dairy operations should apply, whether licensed or unlicensed. First is the importance of storing the milk in disposable plastic containers with tamper-evident seals. This will help protect milk from deliberate sabotage, or contamination from poorly washed jars. Several companies supply these containers at a very reasonable price. The plastic used in these containers is very stable and will not leach chemicals into the milk as long as the milk is kept cold. Even in states that require customers to supply their own containers, farmers should insist that they use these disposable containers with tamper-evident seals.

Second, milk producers should keep three samples of milk drawn every day. The samples should be clearly marked with a date, frozen and kept for several weeks, until there is no possibility of any reports of illness from milk drawn that particular day. If the state asks for a sample after an illness, the farmer can provide one sample to the state and send out the other two to two different independent laboratories—and the farmer should not tell the state which labs he will be using. We have a report from a Florida farmer whom the state accused of having high bacteria counts. He sent his milk to two independent labs and came back with cell counts under 10,000. The state lab claimed a finding of 320,000, but did not harass the farmer any further.

Third, farmers should have their milk tested periodically for pathogens, including E. coli O157:H7. Fortunately, an inexpensive test for E. coli O157:H7 is now available from Strategic Diagnostics, Inc. (www.sdix.com, 559-217-8909), which can even be carried out on the farm.

Finally, some advice from Kathryn Russell, who runs a cowshare program in Virginia. Dairy farmers should not allow government agents to obtain and test samples of milk or cow’s manure after an incident of food borne illness until they have first identified the offending strain of E. coli. Very precise identification is now possible using a technique called Pulsed Field Gel Electrophoresis, which relies on genetic fingerprinting to identify specific strains of E. coli. In order to keep state agencies and their laboratories honest, farmers should only allow samples to be taken after the offending organism taken from human stool specimens has been definitively identified.


It is important to be positive. In spite of the Dee Creek incident, raw milk continues to be available in Washington state. In fact, it is becoming more available as more small dairies become licensed, even available in retail stores. The outbreak has actually galvanized and brought together many talented individuals to work on guidelines that protect both the farmer and the public. The lessons we have learned may be instrumental in nurturing the raw milk movement as it grows, not only in Washington state but also throughout the country.


Anomalies, Errors, and Lies

The Washington State Department of Agriculture’s (WSDA) final report on the Dee Creek Farm incident is filled with anomalies, errors and downright lies. Here are a few out of many:

  • The Report claimed to find E.coli O157:H7 in the milk, whereas independent labs found none.
  • WSDA claimed that Dee Creek Farm did not “meet the standards” of the dairy regulations. Yet from recent hearings emerged quotes from WSDA admitting that they previously had no regulation authority over shareholder dairies. Before the incident, WSDA agreed that shareholder dairies were not required to “meet standards.”
  • The Report claims that during the WSDA investigation, a cow was “munching” on a tube while in the milking parlor. Nothing of the sort is recorded on the Dee Creek video.
  • During the inspection, Dee Creek consistently explained that they were not milking that day for human consumption. This was never noted in the report.
  • The Pucketts pointed out the large numbers of elk seen on the farm, and the strange consistency of their droppings. Referring to the elk, the Report states, “The inspectors were not able to collect any additional information or observe any other conditions at that time,” insinuating that Dee Creek did not allow them to make further investigations. WSDA was free to inspect whatever they wanted, but they chose not to pursue this avenue of research.
  • According to the Report, “the firm also delivered to the health department a whey sample and a skimmed cream sample.” This is untrue.
  • The report states that “The inspectors sprayed their boots with a 100 ppm chlorine solution and used a hand sanitizing gel.” They only did this after the Pucketts pointed out that they were required to sanitize.
  • The Report contained photos of the entrance to the barn, which was muddy, and stated that the cows lived in a “mudpack/pasture.” The Report did not show photographs of the actual pasture, which was not a mudpack. And no dairy at the time could have passed the barn-entrance test as Washington state had received record-breaking rains for the previous month.
  • The report stated that someone in Woodland reported to WSDA that Dee Creek had traded chicken for milk in the summer of 2004. The Pucketts were not doing milk at all until late summer and as they produced both chicken and milk, the claim that they traded chicken for milk does not make sense.

Sidebar: Cowshares vs. Licensed Dairies

This article appeared in the Winter 2005/Spring 2006 edition of Wise Traditions, the quarterly journal of the Weston A. Price Foundation.

Sally Fallon Morell

Sally Fallon Morell is the author of Nourishing Traditions: The Cookbook that Challenges Politically Correct Nutrition and the Diet Dictocrats (with Mary G. Enig, PhD), a well-researched, thought-provoking guide to traditional foods with a startling message: Animal fats and cholesterol are not villains but vital factors in the diet, necessary for normal growth, proper function of the brain and nervous system, protection from disease and optimum energy levels. She joined forces with Enig again to write Eat Fat, Lose Fat, and has authored numerous articles on the subject of diet and health. The President of the Weston A. Price Foundation and founder of A Campaign for Real Milk, Sally is also a journalist, chef, nutrition researcher, homemaker, and community activist. Her four healthy children were raised on whole foods including butter, cream, eggs and meat.

Leave a Reply

Your email address will not be published. Required fields are marked *

This site uses Akismet to reduce spam. Learn how your comment data is processed.