An Outbreak of Escherichia Coli O157:H7 Infection Linked to Unpasteurized Apple Cider in Oklahoma, 1999 Mamadou O
Total Page:16
File Type:pdf, Size:1020Kb
Food Protection Trends, Vol. 31, No. 2, Pages 88–92 Copyright© 2011, International Association for Food Protection ARTICLES 6200 Aurora Ave., Suite 200W, Des Moines, IA 50322-2864 An Outbreak of Escherichia coli O157:H7 Infection Linked to Unpasteurized Apple Cider in Oklahoma, 1999 MAMADOU O. DIALLO1*, KRISTY K. BRADLEY2, JAMES M. CrutcHER3, MIKE LytLE4, ANTHONY LEE2 and RONALD L. MOOLENAAR5 1Epidemic Intelligence Service, Centers for Disease Control and Prevention, Atlanta, GA 30333, USA; 2Communi- cable Disease Division, Oklahoma State Dept. of Health, Oklahoma City, OK 73117, USA; 3State Epidemiologist, Oklahoma State Dept. of Health, Oklahoma City, OK 73117, USA; 4Public Health Laboratory, Oklahoma State Dept. of Health, Oklahoma City, OK 73117, USA; 5Epidemiology Program Office, Center for Disease Control and Prevention, Atlanta, GA 30333, USA INTRODUCTION ABSTRACT In 1982, Riley et al. (22) investi- During the fall of 1999, an outbreak with 16 cases (11 gated two outbreaks of an unusual gas- trointestinal illness that affected 47 or confirmed and five probable) of Escherichia coli O157:H7 more persons. They determined that infections was identified in Oklahoma. Nine persons (82%) the outbreak was caused by a rare E. coli experienced bloody diarrhea, six (54%) required hospitalization, serotype, O157:H7, apparently trans- and three (27%) developed hemolytic uremic syndrome (HUS). mitted by undercooked meat. During Twelve of the sixteen (75%) cases were children from 2 to the following decades, multiple E. coli O157:H7 outbreaks were associated with 13 years old. All nine available E. coli O157:H7 isolates had an consumption of raw or undercooked identical pulsed-field gel electrophoresis pattern. Ten of the 11 ground meat, meat products, and unpas- confirmed patients and none of the 24 control subjects had teurized milk (2, 3, 15, 16, 17). After the drunk unpasteurized apple cider from Orchard A (matched 1990s, an increased number of reported odds ratio undefined;P < 0.00001). All environmental specimen outbreaks were associated with fresh veg- etables (e.g., lettuce, spinach, sprouts) or cultures were negative. Orchard A was in compliance with unpasteurized apple juice (1, 4, 5, 6, 8, U.S. Food and Drug Administration requirements for product 11, 12, 13, 19, 24). warning labels on unpasteurized products, although only one of In 1999, Mead et al. (18), from the eight patients (or surrogates) interviewed recalled having read Centers for Disease Control and Pre- the label. This outbreak raises questions about whether the vention (CDC), estimated that E. coli O157:H7 accounted for 73,480 illnesses current practice of requiring warning labels for unpasteurized annually in the United States, including cider provides sufficient protection for vulnerable populations, 2,000 hospitalizations and 60 deaths, as especially children. a direct result of E. coli O157:H7 in- fections and its complications. Despite progress in making food safer in the United States, E. coli O157:H7 associ- A peer-reviewed article ated with animal products and fresh pro- *Author for correspondence: +1 347.488.4881; Fax: +1 347.920.4771 duce remains a public health challenge E-mail: [email protected] putting a large number of persons at risk 88 FOOD PROTECTION TRENDS | FEBRUARY 2011 for E. coli O157:H7 infection and its without diarrhea history from October 1 The Oklahoma Dairy, Food, and complications (7, 11, 14, 21). to the date of interview. Water Laboratory and the OSDH Public In Oklahoma, from 1994 through During October 12–25, a total of Health Laboratory tested the environ- 1998, an average of 17 cases of E. coli 16 case-patients (11 confirmed and five mental swabs and samples of the apple O157:H7 infections were reported to the probable) and 24 control subjects were cider collected during the environmen- Oklahoma State Department of Health interviewed by use of a standard ques- tal investigation. They performed total (OSDH) annually. On October 11, tionnaire. If the age of the case-patient and fecal coliform testing and testing for 1999, a hospital laboratory in northeast or control subject was less than 14 years, E. coli O157:H7 antigen, using the visual Oklahoma notified OSDH of four pa- a parent/guardian responded to the ques- immunoprecipitation assay (VIP™: Bio- tient specimens that had tested positive tions. Case-patients were asked questions Control System, Bellevue, WA, USA) for E. coli O157:H7 by a fecal antigen- regarding food consumption within 7 and direct plating of samples from eight screening assay. In response to these re- days of illness onset. Control subjects unopened containers of apple cider (mul- ports, OSDH immediately initiated an were asked to report on foods consumed tiple lot numbers) and one cider sample epidemiologic investigation to (1) de- during the 7-day period before the inter- collected from a case-patient’s household. termine if these represented an outbreak view. Bacterial culture and antigen testing were and if so, the magnitude of the outbreak, also performed on 20 environmental (2) identify risk factors for illness and the swabs collected during the investigation Environmental investigation possible source or vehicle for infection, (unopened jugs of cider, environmental and (3) implement preventive measures. After epidemiologically linking the swabs of cider production equipment, outbreak to consumption of unpasteur- and samples of tap water from the cider MATERIALS AND METHODS ized apple cider from a single orchard mill). Additionally, coliform counts and (defined here as Orchard A), staff from bacterial culture were performed on tap- Case finding the OSDH Consumer Protection Di- water samples collected from the cider After the initial case report, OSDH vision, the Wagoner County Health processing room and packing shed. initiated active surveillance. All hospitals Department, and the OSDH Com- and clinical laboratories in Tulsa and municable Disease Division conducted Statistical analysis neighboring counties in Oklahoma were separate site visits and inspections of the Data were entered into Epi-Info™ contacted by telephone and encouraged suspected orchard. Unopened contain- 6.04b (CDC, Atlanta, Georgia). A de- to report suspected cases of E. coli infec- ers of cider, environmental swabs of the scriptive analysis was performed to calcu- tion or postdiarrheal hemolytic uremic cider production equipment, and sam- late frequency of selected variables. Only syndrome (HUS). A confirmed case ples of tap water from the cider mill were confirmed cases were included in the was defined as diarrheal illness (more collected for laboratory analyses. matched case-control analysis of possible than three loose stools per 24 hours) risk factors to determine matched odds occurring among persons of Northeast- Laboratory investigation ratios (ORs) and their 95% confidence ern Oklahoma (Tulsa and neighboring intervals (95% CIs). counties) during October 1999, and a All available patients’ stools were positive fecal antigen test or culture for cultured at the county laboratory for E. coli O157:H7. A probable case was diarrheal pathogen identification. Test- RESULTS defined as diarrhea (three or more loose ing included direct culture attempts, ® Active surveillance stools per 24 hours) occurring in a fam- Shiga-toxin testing (ImmunoCard STAT ily member of a patient with confirmed EHEC, Meridian Diagnostics, Inc., Cin- Sixteen cases of E. coli O157:H7 in- illness who was ill at approximately the cinnati, Ohio), and E. coli O157:H7 fection (11 confirmed and five probable) same time as that patient, but without isolation attempts by using immuno- were identified during October 12–25, laboratory confirmation. magnetic beads provided by the Food- 1999, with diarrhea onsets during borne Diseases Laboratory, CDC. Posi- October 2–17. Reported clinical symp- tive E. coli O157:H7 tests were referred Case-control study toms of the 11 confirmed cases included to the OSDH Public Health Laboratory abdominal cramping (91%), bloody di- OSDH conducted a matched case- for confirmation. Isolates cultured from arrhea (82%), fever (73%), and vomit- control study to determine possible risk available case-patient stools and con- ing (55%). Duration of diarrheal illness factors and sources for acquiring infect- firmed as E. coli O157:H7 by testing at ranged from 2 days to 13 days (mean = ion with E. coli O157:H7. For each case- the OSDH laboratory were forwarded to 5 days). Six persons (55%) were hospi- patient, two to three acquaintance con- the Texas Department of Health Labo- talized, including three children with trol subjects were selected and matched ratory for pulsed-field gel electrophore- hemolytic uremic syndrome requiring by age and neighborhood. The control sis (PFGE) subtyping. PFGE subtypes dialysis treatment. Twelve (75%) were subjects’ names and contact information were compared among outbreak cases children from 2 to 13 years old. No oth- were provided by the index case-patient. and with previously identified cases that er vulnerable population was identified All case-patients and control subjects had occurred earlier in the year. All Xba- during this outbreak. were asked questions regarding clinical I patterns were forwarded to the PFGE symptoms and risk factors for illness, in- E. coli national database at the national Case-control study cluding animal contact, food exposures, network of public health and food regu- and recreational water exposure. All latory agency laboratories coordinated by All 11 confirmed case-patients