From: TSS ()
Subject: ICEID 2008 Foodborne Disease Outbreaks Leafy Greens and Meats
Date: March 19, 2008 at 1:21 pm PST
Page 1 of 262
International Conference on Emerging Infectious Diseases 2008
Slide Sessions and Poster Abstracts
Foodborne Disease Outbreaks Associated with Leafy Greens, 1973–2006
K. M. Herman, T. L. Ayers, M. Lynch; Centers for Disease Control and Prevention, Atlanta, GA.
Background: Several recent large outbreaks have been associated with leafy green foods in the
United States, such as Escherichia coli O157:H7 infections due to spinach; however, the characteristics
of all reported outbreaks due to leafy greens have not been described. The Centers for Disease Control
and Prevention conducts surveillance for foodborne disease outbreaks (FBDO) investigated by local and
state health departments in the United States. Methods: We reviewed data from the FBDO surveillance
system for 1973-2006. A leafy green-associated FBDO is defined as two or more illnesses due to the
consumption of a single leafy green food item (lettuce, cabbage, mesclun mix, spinach) or a salad item
containing one or more leafy greens. These data were compared with U.S. leafy greens per capita
availability, a proxy for leafy green consumption. Results: Among 10,421 FBDO reported during 1973-
2006, 502 (4.8%) outbreaks, 18,242 (6.5%) illnesses, and 15 (4.0%) deaths were associated with leafy
greens. Among leafy green-associated FBDO with a confirmed etiology, Norovirus was responsible for
196 (58.3%) outbreaks, followed by Salmonella, 35 (10.4%) outbreaks, and Escherichia coli O157:H7, 30
(8.9%). The median size of leafy green-associated outbreaks (18 illnesses) was twice the median size of
non-leafy green-associated outbreaks (9). During 1986-1995, U.S. leafy green consumption increased
17.2% from the previous decade. During the same period, the proportion of all FBDO due to leafy greens
increased 59.6%. Likewise, during 1996-2005 leafy green consumption increased 9.0% and leafy greenassociated
outbreaks increased 38.6%. In 296 (69.4%) outbreaks, leafy greens were served at a
restaurant; 11 (2.2%) involved cases in multiple states. Conclusions: Leafy greens are an important
cause of FBDO and may transmit pathogens with human or animal reservoirs. The proportion of FBDO
due to leafy greens has increased, and cannot be accounted for completely by an increase in leafy green
consumption. Contaminated leafy greens may cause restaurant-associated or widespread outbreaks.
Efforts by local, state, and federal agencies to control leafy green outbreaks should span from the point of
harvest to the point of preparation.
Board 21. Fresh produce outbreaks in Australia, 2001-2006
M. D. Kirk1, K. Fullerton1, J. Gregory2; 1OzFoodNet, Canberra, AUSTRALIA, 2OzFoodNet, Department Human
Services, Victoria, AUSTRALIA.
Background: Recent outbreaks in Australia and abroad have highlighted the role of fresh produce in
foodborne disease outbreaks. Since fresh produce is often eaten without cooking, its outbreak potential can be
significant. Methods: Data from the OzFoodNet Outbreak Register from January 2001 to June 2005 were reviewed.
OzFoodNet Quarterly and Annual Reports (published and unpublished) from July 2005 through December 2006 were
also reviewed. Produce-related outbreaks were defined as outbreaks of foodborne or suspected foodborne
transmission where the confirmed or suspected vehicle included fresh, uncooked produce. Results: From January
2001 through June 2005 there were 1767 reported gastroenteritis outbreaks recorded. Of these outbreaks, 426 (24%)
were either foodborne (157, 37%) or suspected foodborne (269, 63%). Sixteen (4%) of these outbreaks were
identified as produce-related. An additional 9 produce-related outbreaks were identified in the review of quarterly and
annual reports. These 25 outbreaks affected at least 686 people, with 51 people hospitalized and no fatalities. The
mean number of people affected in these outbreaks was 30 people, with a range from 2 to 125 people. These
outbreaks occurred in association with food served at restaurants (44%), primary produce (20%), and fastfood/
takeaway food (20%). These outbreaks were caused by Salmonella (60%), followed by unknown aetiology
(20%), norovirus (12%), and Campylobacter (8%). Conclusions: Fresh produce causes considerable foodborne
disease in Australia. Fresh produce is particularly vulnerable to causing outbreaks due to the lack of an adequate killstep
Board 95. Tomato Handling Policies and Practices in Restaurants
E. G. Kirkland, V. J. Radke, C. A. Selman; CDC, EHS-Net, Atlanta, GA.
Background: Recently, several foodborne illness outbreaks have been associated with tomatoes served in
restaurants. Prevention of foodborne illness outbreaks requires proper food storage and preparation practices; these
tomato outbreaks suggest that tomatoes are being stored and prepared improperly in restaurants. Yet relatively little
data exists on tomato storage and preparation policies and practices in restaurants. Thus, the purpose of this study is
to gain a better understanding of how tomatoes are stored and prepared in restaurants. Methods: Data were
collected in 450 restaurants through interviews with restaurant managers and observations of tomato storage,
washing and cutting in restaurant kitchen environments. This study was conducted by the Environmental Health
Page 127 of 262
Specialists Network (EHS-Net). EHS-Net is a collaboration involving the Centers for Disease Control and Prevention
(CDC), the U.S. Food and Drug Administration (FDA), and nine Emerging Infections Program sites (California,
Connecticut, Georgia, Iowa, Minnesota, New York, Rhode Island and Tennessee); these partners have come
together in an effort to better understand the environmental causes of foodborne illness. Results: The median
temperature of prepared tomatoes in storage and holding was 44°F degrees; the median temperature of cut tomatoes
in storage and holding was 43°F degrees. In 94% of restaurants, managers reported that tomatoes were washed; in
83% of washing observations, tomatoes were rinsed or held under running water and in 18% of washing
observations, tomatoes were soaked or immersed in water. Produce-only cutting boards were used in 51% of tomatocutting
observations, and gloves were used in 64% of observations. Conclusions: Some good tomato handling
practices were observed. For example, most restaurants reported washing their tomatoes, and gloves were used in
the majority of tomato-cutting observations. However, some practices did not meet FDA recommendations for
preventing pathogen contamination and proliferation on tomatoes- cut tomatoes were stored above 41 degrees,
tomatoes were washed by soaking in water rather than by placing under running water, and produce only cutting
boards were not always used. These data indicate that education concerning safe tomato handling practices is
needed in restaurants.
Foodborne & Waterborne Infections
Board 87. Foodborne Disease Outbreak Trends, and Sources and Timeliness of Detection in Connecticut,
D. Mlynarski, Q. Phan, T. Rabatsky-Ehr, K. Purviance, J. Brockmeyer, J. Krasnitski, A. Nepaul, L. LoBianco, K.
Frenette, P. Mshar, J. L. Hadler; Connecticut Department of Public Health, Hartford, CT.
Background: It is critical to public health preparedness (PHP) to understand the epidemiology of outbreaks,
and means and timeliness of their detection. In Connecticut (CT), foodborne disease outbreaks (FOs) are considered
public health emergencies reportable by telephone to state and local public health authorities. We examined 3.5
years of FO surveillance in CT as part of a broader effort at enhancing PHP-related surveillance. Methods:CT data
reported to CDC by the electronic Foodborne Outbreak Reporting System was reviewed for January 2004-June 2007.
The number, causative agents, settings and food vehicles implicated were examined. Methods of detection were
summarized; median time delay from first illness onset to public health notification was calculated. Results: From
1/1/04-6/30/07, 58 FOs were reported. The number increased annually from 13 to 19 between 2004 and 2006 with 10
FOs in the first 6 months of 2007. Causative agents included norovirus (62%), bacterial pathogens (28%), toxins (3%)
and parasites (2%). The increase from 2004 to 2006 was mainly due to norovirus (7 to 12). Overall, 64% of FOs were
associated with contamination or mishandling of food at food service establishments and 19% with widely distributed
contaminated food items. Private citizens reported 62% (median 3-day delay) and clinicians reported 12% (median 1-
day delay) of FOs. Public health surveillance, including routine Pulsed Field Gel Electrophoresis (PFGE) of selected
bacterial pathogens, detected 21% (median 26-day delay). Most (71%) PFGE-identified outbreaks were associated
with widely distributed food items. Conclusions: Reported FOs have been increasing, in part driven by an increase
in those due to norovirus. Astute citizens and clinicians are the most important sources of rapid detection and
reporting. Clinicians have a responsibility to report suspected outbreaks and assist in their detection by ordering
appropriate diagnostic testing on persons with acute gastrointestinal illness. Active public health surveillance,
including routine PFGE typing, is important to detection of outbreaks caused by widely distributed contaminated food
items; however, it is not very timely. Efforts are needed to improve timeliness of detection of outbreaks, particularly
those using molecular subtyping methods such as PFGE.
Board 88. Indiana Outbreak of Salmonella I 4,,12:i:- monophasic at a Supermarket Deli – 2006
L. Granzow1, S. Gorsuch1, B. Swearingen2; 1Indiana State Department of Health, Indpls, IN, 2Johnson County Health
Department, Greenwood, IN.
Background: On July 11, 2006, the Indiana State Department of Health (ISDH) initiated an investigation in
response to an increase in salmonellosis in two adjacent counties. Geographic information system (GIS) mapping
confirmed the clustering of cases at the respective north-south borders of the two counties. The most common
exposure (76.2%) among cases was a supermarket (SM) location, with 78.9% of cases having purchased items from
the deli. Pulse-field gel electrophoresis (PFGE) confirmed 199 cases in a 2-enzyme matched outbreak of Salmonella
I 4,,12:i:- monophasic, an emerging serotype. The PFGE pattern was unique during the outbreak period on the
national PulseNet database with a rare Bln I pattern (0.35%). The multiple-locus variable-number tandem repeat
analysis (MLVA) pattern was also unique (N of I 4,,12:i:- database = 450). The outbreak ended after a near 4-
month period, including 15 counties and 2 out-of state residents. A knife block at the SM was a 2-enzyme match for
the outbreak strain. Methods: The ISDH conducted a case-control study (1:1, N=32), matched by zip code, using a
reverse digit dialing system. The study case definition included those with onset dates in July and met the clinical
definition of salmonellosis. The Fisher’s exact test (SAS 9.1) was used to evaluate the association between illness
and exposure. Results: Of the five supermarkets reported, a statistically significant relationship (α = 0.05) was found
between illness and shopping at the SM deli (OR=21.21, p<0.0001). No food categories or items were significant. An
environmental sample from a knife block located in the deli/bakery area was a 2-enzyme match for the for Salmonella
I 4,,12:i:- monophasic. All food samples provided by cases tested negative. Conclusions: The outbreak ceased
shortly after the removal of the knife block from the SM deli. Raw chicken, cooked on the rotisserie, was the only raw
food item in the deli. The likely mode of transmission was environmental cross-contamination; the knife block was
likely originally contaminated by a knife used to open bagged raw chicken. Knives housed in the contaminated knife
block were then used to open packages of ready-to-eat foods, contaminating a small surface area.
Board 94. Rural Exposure to Shiga Toxin-Producing E. coli _ South Dakota, 1998-2007
N. S. Chappelle1, N. Tarkhashvili2, N. Hill1, L. Schaefer1, L. Kightlinger1; 1South Dakota Department of Health, Pierre,
SD, 2Center for Disease Control and Prevention, Pierre, SD.
Background: Escherichia coli O157:H7 causes an estimated 70,000 infections annually in the United States
and can result in bloody diarrhea, hemolytic uremic syndrome (HUS), and death. South Dakota has the highest
reported rate of Shiga toxin-producing E. coli in the US averaging 5.2/100,000 population per year (annual national
average 1.2/100,000 population 1998-2005). We describe characteristics of South Dakota’s Shiga toxin-producing E.
coli cases reported during 1998-2007. Methods: We abstracted demographic, clinical, and exposure data from
investigation records gathered at the time of disease during 1998-2007 and archived at the South Dakota Department
of Health (SD DOH). Results: A total of 432 cases was reported to SD DOH. E. coli serotype O157:H7 was identified
in 91% of cases, O111 was the primary other serotype identified. Of all cases, 55% were female and 48% were
children 10 years old, median age was 11 (range 0-97 years). Outbreak-associated cases were 26% of cases, and
50% were hospitalized. Bloody diarrhea was reported in 81% of cases and 10% developed HUS. The case-fatality
rate of patients infected with laboratory-confirmed Shiga toxin-producing E. coli was 0.7%. Exposure to animals was
reported in 53% of cases and 35% reported visiting locations with animals. Direct exposure to cattle was reported in
34% of cases, with direct exposure to animal dung reported in 29% of cases, and 61% of cases lived in rural counties
where beef cattle or dairy production is common. Yet, 63% of urban cases reported recent exposure to cattle. Overall,
85% of cases either lived in rural counties or reported recent exposure to cattle. Conclusions: Our data are
consistent with previously reported rates of non-O157 serotypes, case-fatality, and HUS. Exposure to animals,
particularly cattle, emerged as the most often reported risk. This differs from predominantly urban states, where
exposure to contaminated food is more common. Agricultural data showing that South Dakota has the highest cattle
per person ratio in the nation: 4.8 (national average 0.33) support our results. Understanding these variables is
important for devising prevention strategies related to cattle-to-human transmission in rural agricultural states.
Board 152. Use of Antibiotics among Dairy Veterinarians in Mid-Atlantic States
N. M. M'ikanatha1, S. C. Rankin2, D. O. Morris3; 1Pennsylvania Department of Health, Harrisburg, PA, 2University of
Pennsylvania School of Veterinary Medicine, Philadelphia, Philadelphia, PA, 3University of Pennsylvania School of
Veterinary Medicine, Philadelphia, PA.
Background: Public health concerns about antimicrobial resistance in zoonotic pathogens of animal origin
have prompted a call for judicious use of antibiotics on the farm. However, current management practices among
food-animal veterinarians are largely unknown. Methods: During 2007, we administered an online survey of all dairy
veterinarians in Delaware, Maryland, New Jersey and Pennsylvania who were members of the American Association
of Bovine Practitioners (AABP). The survey sought to assess practices and opinions related to antibiotic use among
dairy veterinarians in the Mid-Atlantic States and to optimize future educational interventions. The use of written
management protocols, recommendations regarding milk replacement, and antibiotics prescribing practices were
documented. Veterinarians’ opinions regarding antimicrobial use by non-veterinary staff on dairy farms and the
emergence of antimicrobial resistance were also assessed. Results: A total of 35 AABP members (20%) participated
in the study. The majority (83%) spend most of their time on dairy calls. Eight-nine percent spend less than 10
minutes per visit discussing antimicrobial use with farmers. Forty percent of the respondents indicated they had
provided written disease management protocols to clients and 48% of those thought farmers followed these
protocols. Twenty-nine percent recommend colostrum replacer for newborn heifer calves in the first 24 hours and
38% reported the milk replacer contained oxtetracycline with neomycin. While respondents’ choice for treating
common conditions included a range of antibiotics, they most commonly cited ceftiofur and oxytetracycline. Forty-nine
percent of the respondents thought that more than 50% of their clients used unprescribed antibiotics. Over sixty
percent of the respondents were concerned about antimicrobial resistance. The majority of the respondents thought
that availability of antibiotics for use by non-veterinarians was among the contributing factors to antimicrobial
resistance. Conclusion: Most veterinarians surveyed were concerned about antimicrobial resistance. Most
veterinarians surveyed use ceftiofur as their preferred antibiotic. Continuing education is necessary to address
prescribing patterns in the face of emerging antimicrobial resistance.
Board 153. Prevalence and Antimicrobial Resistance of Campylobacter from Retail Meats: Results: of the
National Antimicrobial Resistance Monitoring System (NARMS): 2002-2006
P. McDermott, L. English, E. Hall-Robinson, S. Friedman, J. Abbott, S. Zhao; Food & Drug Administration, Laurel,
Background: Campylobacter is a leading cause of foodborne bacterial pathogen in the United States. Since
2002, the U.S. National Antimicrobial Resistance Monitoring System (NARMS) has investigated the prevalence and
the antimicrobial susceptibility of Campylobacter present in the retail meat supply. Methods: Campylobacter were
recovered from a monthly sampling of chicken breasts, ground turkey, ground beef, and pork chops purchased from
grocery stores in 10 states (CA,CO, CT, GA, MD, MN, NM, NY, OR, TN). A total of 20,294 meat samples taken from
2002-2006 were screened for presence of Campylobacter. In 2002-2003, susceptibility testing included four
antimicrobials (CIP, ERY, GEN, and DOX) using agar dilution. After 2004, nine antimicrobials were tested (CIP, NAL,
ERY, GEN, TET, FFN, AZI, TEL, and CLI) using broth micro dilution. Campylobacter isolates were also analyzed by
PFGE. Results: Chicken showed the highest Campylobacter contamination rate: 47% in 2002, 52% in 2003, 60% in
2004, and 46% in 2005 and 49% in 2006. Other meats had contamination rates < 2.5%. Overall, C. jejuni constituted
71.1% of strains, C. coli 28.6% and C. lari 0.3%. Overall CipR (MIC > 4 µg/mL) increased significantly (p=0.0158)
from 2002 (13.8%) to 2005 (19.6%), and dropped slightly in 2006 (18.9%) following the ban of fluoroquinolone use in
poultry production. In C. jejuni from chicken breast, however, CipR was present in 15.1% of isolates from 2005 and
16.9% in 2006. During the five year testing, TetR increased from 27.6% to 47.8% overall. EryR each year was present
in <1% C. jejuni, and decreased steadily in C. coli from 18.9% in 2002 to 5.7% in 2006. The most common MDR
pattern among strains resistant to > 5 antimicrobials was AZI-CLI-ERY-TEL-TET. C. coli showed higher proportion of
MDR than C. jejuni. The PFGE results showed that Campylobacter genetically were very diverse, however certain
clones were widely dispersed in different meat brands from different store chains in all five years. Conclusions:
Campylobacter, including antimicrobial resistant strains, persist in retail chicken meats and provide a reservoir of
resistant strains in the food supply.
Board 6. Salmonella enterica serovar Heidelberg from Retail Meats: Results: of the National Antimicrobial
Resistance Monitoring System (NARMS): 2002-2006.
S. Zhao, E. Hall-Robinson, A. Glenn, S. Friedman, J. Abbott, S. Ayers, P. McDermott; Food & Drug Administration,
Background: Salmonella enterica serovar Heidelberg is frequently associated with foodborne illness in
humans, and is commonly isolated from poultry and their derived meats. A recent upsurge in antimicrobial resistance
in this serovar has been recognized. There are few data on the prevalence, antimicrobial susceptibility, and genetic
diversity of S. Heidelberg isolates in retail meats. Methods: We compared the prevalence of S. Heidelberg in a
sampling of 20,294 meats, including chicken breast, ground turkey, ground beef and pork chops collected during
2002-2006 for the National Antimicrobial Resistance Monitoring System (NARMS). Isolates were analyzed for
Page 84 of 262
antimicrobial susceptibility and compared genetically using pulsed-field gel electrophoresis (PFGE). Results: A total
of 297 S. Heidelberg isolates were recovered, representing 22% (297/1372) of all Salmonella serovars from retail
meats. Among the 297 isolates, 178 (60%) from ground turkey, 109 (37%) from chicken breast and 10 (2%) from pork
chop; no S. Heidelberg was found in ground beef. A total of 197 (66%) of the isolates were resistant to at least one of
the 15 antimicrobial agents tested and 49 (16%) of the isolates were resistant to >5 antimicrobials. Five isolates
(1.7%) were resistant to >9 antimicrobials, all of which were recovered form ground turkey. The proportion of
resistance to different antimicrobials were: tetracycline (40%), streptomycin (38%), sulfamethoxazole (30%),
gentamicin (27%), and kanamycin (21%), ampicillin (19%), amoxicillin-clavulanic acid (10%), cefoxitin (9%), ceftiofur
(9%), chloramphenicol (1%), and nalidixic acid (1%). Resistance was consistently more prevalent in S. Heidelberg
from ground turkey than from chicken breast. All isolates were susceptible to amikacin, ceftriaxone, ciprofloxacin and
trimethoprim/ sulfamethoxazole. PFGE using XbaI and BlnI generated 107 patterns. Certain clones were widely
dispersed in different types of meats and meat brands from different store chains in all three sampling years.
Conclusions: These date indicate that S. Heidelberg is a common serovar in retail poultry meats, and includes
clones of multidrug-resistant strains.
Board 18. Human Health Burden of Acute Diarrheal Illness in the United States, FoodNet Population Survey,
L. B. Moyer1, P. Clogher2, C. Fuller3, T. F. Jones4, A. Lasher5, D. M. Norton6, S. Solghan7, M. Tobin-D'Angelo8, O.
Henao9; 1CDC and VA Medical Center, Atlanta, GA, 2CT EIP, New Haven, CT, 3MN Dept of Health, St Paul, MN, 4TN
Dept of Health, Nashville, TN, 5FDA, College Park, MD, 6CA EIP, Oakland, CA, 7NY Dept of Health, Albany, NY, 8GA
Div of Publ Health, Atlanta, GA, 9CDC, Atlanta, GA.
Background: Foodborne pathogens cause an estimated 76 million illnesses each year. Although most
illnesses are self-limiting, the human health burden is substantial and includes healthcare provider visits, medications,
laboratory tests, hospitalizations, and time missed from work. We describe the human health burden of acute
diarrheal illness using data from the 2006-2007 FoodNet Population Survey. Methods: During 2006-2007, FoodNet
conducted a 12-month population-based survey in 10 sites using a standard random-digit dialing methodology.
Demographic information was collected and respondents were asked about illness and activities in the month before
the interview. Acute diarrheal illness was defined as >3 loose stools in 24-hours lasting >1 day or resulting in
impairment of daily activities. Persons with a chronic illness in which diarrhea was a major symptom were excluded.
Weighted proportions were calculated to adjust for study design and age and sex. Results: The weighted prevalence
of acute diarrheal illness in the month prior to interview was 6.9% (95% CI 6.6-7.2). Acute diarrheal illness prevalence
was greatest among those <5 years old (12.9%, 95% CI 9.7-16.1) and least among those >65 years old (4.4%, 95%
CI 3.5-5.3). The mean duration of diarrhea was 4 days (median 2) and 1.7% reported bloody diarrhea. Of the 20.0%
who visited a medical care provider, 26.7% went to a doctor’s office or clinic more than once, 13.0% went to an
emergency department, 1.0% were admitted to a hospital overnight and an additional 2.3% spent more than one
night in the hospital. Stool specimens were submitted from 4.0% of persons reporting acute diarrheal illness, 32.9%
took anti-diarrheal medications and 10.6% used antibiotics. Of those who had a job during this time period (52.0%),
33.2% missed time from work because of their or their child’s illness (median 2 days). Of those attending school
(7.2%), 66.3% missed time from school because of the illness (median 2 days). Conclusion: Acute diarrheal illness
remains an important human health burden causing substantial personal and societal costs from multiple healthcare
contacts and days missed from work or school. Continued efforts are needed to identify the causes and risk factors
for illness; thus, helping to direct intervention and prevention efforts to reduce the burden of illness.
Board 19. Salmonella Bacteriuria in New York State FoodNet Counties, 2002-2006
G. L. Smith1, S. M. Zansky2, D. L. Hoefer2, J. B. Karr3, N. L. Spina2, S. M. Solghan2, T. P. Root4; 1NYSDOH, Geneva,
NY, 2NYSDOH, Albany, NY, 3NYSDOH, Rochester, NY, 4Wadsworth Center Laboratories-NYSDOH, Albany, NY.
Background: Salmonellosis is a major cause of foodborne illness in the U.S. Commonly associated with
gastrointestinal illness, it can also cause extra-intestinal illness including urinary tract infections. Recent national
trends indicate Salmonella bacteriuria increasing in incidence and as a proportion of all Salmonella, especially among
elderly women. Analysis of Salmonella bacteriuria in 34 New York State (NYS) Emerging Infections Program (EIP)
FoodNet counties was conducted to look for similar trends. Methods: The NYS FoodNet program has conducted
population-based active surveillance for laboratory-confirmed cases of infection of Salmonella. Using the NYS
Department of Health (NYSDOH) Communicable Disease Electronic Surveillance System (CDESS) and the Clinical
Laboratory Information System (CLIMS), confirmed cases of Salmonella identified between 2002-2006 were included
in this analysis. The most invasive source for each case is recorded. Data was examined to compare incidence rates
of Salmonella isolated from urine and proportion of urine isolates by gender and age group. Results: 2337 confirmed
cases of Salmonella were identified between 2002-2006. Of 2333 cases with a known specimen source, 189 cases
were isolated from urine (8.1%). The annual proportion of urine isolates ranged from 5.1% (2002) to 11.1% (2005).
The incidence rate of Salmonella bacteriuria increased from 0.8 cases per 100,000 persons in 2002 to 1.1/100,000 in
2006 with the highest rate of 1.3/100,000 in 2005. Most cases isolated from urine occurred among women (88.4%).
For both sexes, incidence from urine increased with age, beginning around 50 years of age with the greatest
percentage of cases attributable to those >70 years. Conclusions: Salmonella bacteriuria has long been identified in
Page 90 of 262
NYS residents but recently represent a greater proportion of the total cases. The majority of cases occur in elderly
females similar to national reports. Collecting and reporting specimen source is an important component of foodborne
disease surveillance, helping to identify changes in the epidemiology of those illnesses. Possible change in test
practices, role of urine catheters, and urine predilection of some Salmonella serotypes may contribute to the
increased incidence of Salmonella bacteriuria in NYS and warrant further study.
Board 239. Antimicrobial Resistant Salmonella from Retail Chicken in Pennsylvania 2006-2007
S. C. Rankin1, N. M. Mikanatha2, D. Tewari3, A. Russo3, R. Localio1, S. Altekruse4, C. Sandt5, E. Lautenbach1, J.
Tait2, M. Hydock3, S. Reynolds5, N. G. Warren5, T. M. Chiller6, D. G. White7; 1University of Pennsylvania,
Philadelphia, PA, 2Pennsylvania Department of Health, Harrisburg, PA, 3Pennsylvania Department of Agriculture,
Harrisburg, PA, 4National Institutes of Health, Bethesda, MD, 5Pennsylvania Department of Health, Lionville, PA,
6Centers for Disease Control and Prevention, Atlanta, GA, 7Food and Drug Administration, Laurel, MD.
Background: Infections caused by antibiotic-resistant Salmonella strains are associated with more severe
Page 196 of 262
illnesse, and higher mortality than those due to susceptible strains. Consumption of chicken is a known risk factor for
salmonellosis. However, the prevalence of resistant Salmonella in poultry meat and their relationship to human
Salmonella isolates has not been well-characterized. Methods: Chicken was purchased from a stratified random
sample of retail outlets in a three county region of Central Pennsylvania for 12 months during 2006-2007. Each month
30 samples were purchased from retail outlets. The samples included prepackaged and open display poultry.
Information was obtained on USDA establishment numbers and organic/antibiotic-free status from package labels,
where available. Isolates were characterized by serotyping, antibiotic susceptibility testing and pulsed-field gel
electrophoresis (PFGE). Antibiotic resistant strains were analyzed for the presence of resistance genes by PCR.
PFGE profiles of antibiotic resistant isolates were compared with human Salmonella isolates from the state public
health laboratory during the study period. Results: Salmonella was isolated from 84 (22%) of 378 samples. The most
common serotypes were Typhimurium 28 (33%), Kentucky 24 (29%), and Enteritidis 22 (26%). 45 isolates showed
resistance to one or more drugs; 40% demonstrated resistance to at least five drugs. 8/45 (18%) resistant isolates
had a blaCMY ß-lactamase gene. The packaged chicken originated from 20 different establishments. Salmonella
isolation was associated with poultry from particular establishments (p=0.007, chi sq=37.6, df=19). In one
establishment that processed only organic poultry, 10 (53%) of 19 samples were positive. PFGE patterns of
Typhimurium and Kentucky isolates from chicken matched patterns in the human database. Conclusion: The
occurrence of drug-resistant Salmonella in retail chicken is a public health concern. This study identified strains with
reduced susceptibility to expanded-spectrum cephalosporins and isolates from humans and chickens with the same
PFGE profiles. The study also identified an association of Salmonella contamination with poultry from specific
establishments. These results indicate the need for enhanced inter-agency surveillance.
Board 225. Longitudinal Study of Antimicrobial Resistance among Escherichia coli Isolated from Integrated
Multi-site Cohorts of Humans and Swine
W. Alali1, H. M. Scott1, R. Harvey2, B. Norby1, B. Lawhorn1, S. Pillai1; 1Texas A&M University, College Station, TX,
2USDA-ARS-SPARC, College Station, TX.
Background: Many studies have attempted to link antimicrobial use in food animal agriculture with an
increased risk of antimicrobial-resistant (AR) bacterial levels in humans. Methods: We examined the relationship
between the prevalence of AR E. coli isolated from human wastewater and swine fecal samples and the risk factors:
(host species, production type (swine), vocation (human swine workers, non-workers, and slaughter-plant workers),
and season) in a multi-site housing, vertically integrated swine and human population agri-food system. Human and
swine E. coli (N = 4048 and 3485, respectively) isolated from wastewater and fecal samples were tested for
antimicrobial susceptibility using the SensititreTM broth microdilution system. Results: There were significant (P <
0.05) differences in AR isolates: 1) between host-species with swine at higher risk for tetracycline, kanamycin,
ceftiofur, gentamicin, streptomycin, chloramphenicol, sulfisoxazole, and ampicillin, 2) swine production group was
significantly associated with AR with purchased boars, nursery piglets, and breeding boars at a higher risk of
Page 190 of 262
resistance to streptomycin and tetracycline, and 3) human swine worker cohorts exhibited lowered sulfisoxazole and
cefoxitin prevalence compared to non-workers, while slaughter-plant workers exhibited elevated cefoxitin prevalence
compared to non-workers. High variability among seasonal samples over the 3-year period was observed. There
were significant differences in multiple resistance isolates between host species, with swine at higher risk than
humans of carrying multi-resistant strains, slaughter-plant workers at higher risk than swine non-workers; however,
there were no significant differences in multiple resistance isolates within swine by production group. Conclusions:
Occupational exposure to slaughter facilities appeared to be associated with an increased relative odds for the
prevalence of cefoxitin resistance and multiple resistance compared to swine non-workers.
A Veterinary Curriculum in the Appropriate Use of Antibiotics
P. C. Bartlett; Michigan State University, E. Lansing, MI.
Background: Debate still continues regarding the relative importance of agricultural antibiotic
usage in fostering antimicrobial resistance (AR) among human pathogens. This debate may have delayed
the development of educational materials aimed at promoting the prudent and judicious use of antibiotics
in veterinary medicine. However, a consensus exists that unnecessary or wasteful usage should be
curtailed whenever possible. Methods: The U.S. Centers for Disease Control and Prevention (CDC)
funded the development of an interactive and multi-media educational web site to aid in teaching
veterinary students about their responsibilities and obligations to curtail the unnecessary use of
antibiotics. The overall purpose is to preserve antibiotic efficacy for both humans and animals. Audio,
video, interactive questions and animation are used to make the presentation varied and entertaining.
Written for veterinary students, the site also has applicability to others in the food animal industries.
Results: The web site is designed to supplement existing courses in public health, epidemiology,
pharmacology and species-specific veterinary medicine. The introductory module emphasizes the
interconnectedness of animal and human health as it follows an outbreak investigation scenario to teach
basic microbiologic and epidemiologic principles of AR. Also taught are methods for determining AR and
the importance of AR to public health. Species-specific modules regard international issues and multiple
clinical situations for dairy cattle, beef cattle, exotic animals and swine. Future modules will address
Page 38 of 262
companion animals and poultry. Emphasis is on therapeutic situations when antimicrobial agents are
often used unnecessarily when alternative treatment methods are available, when the use of
antibiograms (susceptibility profiles) is possible or when antibiotic treatment may be less efficacious and
economical than are available preventive medicine procedures. Conclusions: This web site is being
written for training veterinary students throughout the USA, but it will be freely available to the
international public. The web site is due for release in 2008. Search: CDC Get Smart on the Farm or find
the link at the CDC web site: http://www.cdc.gov/narms/get_smart.htm or www.cvm.msu.edu
Board 79. Antimicrobial Resistance in Salmonella Isolates Recovered from Cattle at Slaughter
P. J. Fedorka-Cray1, J. G. Frye1, M. Rose2, N. Anandaraman3, J. Haro1; 1USDA-ARS, Athens, GA, 2Intervet
Innovation GmbH, Schwabenheim, GERMANY, 3USDA-FSIS, Washington, DC.
Background: Since 1997, the animal arm of the National Antimicrobial Resistance Monitoring System
(NARMS) has monitored changes in antimicrobial susceptibilities of Salmonella isolates from animal origin.
Additionally, since 2000, susceptibility of bovine Salmonella isolates collected in the US has been monitored against
the 4th generation cephalosporins (4-GC) cefquinome, exclusively developed for veterinary medicine, and cefepime.
Cephalosporins are used extensively to treat human and cattle diseases. To identify emerging resistance patterns,
resistance trends to the cephalosporins (ceftriaxone, ceftiofur, cefoxitin and cefquinome) in Salmonella isolates
collected from cattle at slaughter were analyzed. Methods: Salmonella enterica isolates (n=7,199) obtained from
cattle at federally inspected slaughter/processing plants during 1997 - 2006 and submitted to NARMS were tested for
minimum inhibitory concentrations (MICs) using a custom panel of antimicrobials. Isolates collected during 2000-2006
(n=4,685) were also tested on a second panel with cefquinome and cefepime. Results: Resistance to ceftriaxone
remained below 1%, except in 2005 when resistance increased to 2.1%. From 1997 to 2005, resistance to ceftiofur
increased from 0% to 21.6%, with the exception of 2004 when it decreased to 13.3%. Resistance decreased again in
2006 to 18.9%. A similar pattern was observed for cefoxitin (testing started in 2000). Cefoxitin resistance increased
from 2000 to 2005 from 9.1% to 19.8%, except in 2004 when it decreased to 13.2%. As with ceftiofur, a decrease
was observed in 2006 when resistance was 17.9%. From 2000 to 2006 the MIC50 of cefquinome remained at 0.06
µg/ml, except in 2002 when the MIC50 increased by one dilution to 0.12 µg/ml. The highest MIC90 for cefquinome
was 1.0 µg/ml in 2002 and 2005. MICs of cefepime were generally about one dilution step below those of
cefquinome. Changes in resistance for all drugs were in large part driven by serotype, particularly S. Newport,
Reading, Typhimurium and Agona. Conclusions: Salmonella enterica isolates remained highly susceptible to the
human 3-GC ceftriaxone and the 4-GCs cefquinome and cefepime. Overall, an increase in both veterinary 3-GC
ceftiofur and human 2-GC cefoxitin has been observed at similar levels and appears to be serotype dependent.
Board 80. Prevalence of Antibiotic Use, Knowledge and Attitudes Toward Antibiotic-free Meat
S. C. Rankin1, N. M'ikanatha2, R. Dewar2, E. Lautenbach1; 1University of Pennsylvania, Philadelphia, PA,
2Pennsylvania Department of Health, Harrisburg, PA.
Background: Inappropriate use of antibiotics is a major public health concern as it contributes to
antimicrobial resistance. Understanding consumers’ knowledge and attitudes can inform multifaceted actions to
achieve judicious use of antibiotics. Methods: In 2006 questions regarding antibiotic prescriptions and attitudes
toward antibiotic-free meat were asked as part of Pennsylvania Behavioral Risk Factor Surveillance System
(BRFSS). Results: Among Pennsylvania adults, 38.1% (95% CI, 36.4-39.8%) responded they had been prescribed
antibiotics in the previous 12 months and 43.9% (95%CI, 41.1-46.7%) of these had received at least two
prescriptions. Among adults taking antibiotics in the past year, 11.4% (95%CI, 9.6-13.6%) had not completed the
course of treatment. Prescriptions for self-identified diagnoses usually not caused by bacteria (e.g., cough or cold)
were reported by 30.3% (95%CI, 27.7-33.1%). Knowledge about use of antibiotics in animal husbandry varied by
demographic characteristics (Figure): 53.6% (95% CI, 52.8%-55.4%) of non-Hispanic whites were aware of
antibiotics in feed, whereas 42% (95% CI, 34.8-50.1%) of African Americans and 39.8% (95%CI, 29.0-51.7%) of
Hispanics were aware. Knowledge also increased with age, education, and income. Overall 26.2% (95% CI, 24.6-
31.9%) of the survey participants reported that that they try to purchase antibiotics-free meat. Conclusion:
Inappropriate prescriptions and noncompliance with treatment regimens call for simultaneous interventions among
clinicians and patients. To promote judicious use of antibiotics in food animals, additional measures to engage
consumers are needed.
see full text of these abstracts 262 pages ;
Post a Followup