894 Journal o f Food Protection, Vol. 77, No. 6, 2014, Pages 894-902 doi: 10.4315/0362-028X. JFP-13-439 Copyright © , International Association for Food Protection
Counts, Serotypes, and Antimicrobial Resistance of Salmonella Isolates on Retail Raw Poultry in the People’s Republic of China BAOWEI YANG,1 YUE CUI,12 CHAO SHI, 1 JIAQI WANG, 1 XIAODONG XIA, 1 MEILI XI,1 XIN WANG,1 JIANGHONG MENG, 13 WALID Q. ALALI,2* ISABEL WALLS,4 a n d MICHAEL P. DOYLE2 1College o f Food Science, Northwest A&F University, Yangling 712100, People’s Republic o f China; 2Center for Food Safety, University o f Georgia, Griffin, Georgia 30223, USA; 3Joint Institute fo r Food Safety and Applied Nutrition and Department o f Nutrition and Food Science, University o f Maryland, College Park, Maryland 20742, USA; and 4U.S. Department o f Agriculture, National Institute o f Food and Agriculture, Washington, D.C. 20250, USA MS 13-439: Received 14 October 2013/Accepted 31 January 2014
A BSTR A CT The objective of this study was to determine Salmonella counts, serotypes, and antimicrobial resistance profiles in retail raw chicken meat in the People’s Republic of China. Salmonella counts were determined according to the most-probable-number (MPN) method for 300 whole chicken carcasses. These samples were collected from large, small, and wet (open) markets in Guangdong, Shaanxi, and Sichuan provinces. Salmonella isolates were serotyped and tested for antimicrobial susceptibility. Of the 300 chicken carcasses, 43.3% were positive for Salmonella, with an overall mean of 1.7 log MPN per carcass (95% confidence interval, 1.6 to 1.8 log MPN per carcass). No significant differences (P > 0.05) were detected for storage temperature (i.e., chilled, frozen, or ambient), market type (large, small, or wet), province, or location (capital or noncapital city). Seventyeight serotypes were identified among the 1,094 Salmonella isolates. The top five most common Salmonella serotypes on raw chicken carcasses were Enteritidis (19.2%), Indiana (15.2%), Typhimurium (14.6%), Agona (7.1%), and Thompson (6.6%). Salmonella isolates (n = 779) were most frequently resistant to sulfisoxazole (74.1%) and tetracycline (71.1%) and least resistant to ceftriaxone (22.5%) and cefoxitin (19%). Only 4% of the isolates were susceptible to all 15 antimicrobial agents, 45% were resistant to 1 to 5 agents, 29% were resistant to 6 to 10 agents, and 22% were resistant to 11 to 15 agents. Our findings revealed that Salmonella contamination was common in retail raw poultry in China, and the counts on contaminated carcasses were mostly low. Salmonella isolates were diverse in their serotype distribution and antimicrobial susceptibility profiles, with more than half of the isolates resistant to more than five antimicrobial agents. These data may be used in risk assessment models to reduce the transmission of Salmonella via chicken meat to humans in China.
Salmonella is an important zoonotic pathogen that has a significant economic and health impact on animals and humans worldwide (31). Approximately 70 to 80% of foodbome bacterial outbreaks in the People’s Republic of China are caused by Salmonella (32). In the United States, an estimated 1 million cases of foodbome salmonellosis with 378 deaths occur annually (25). In Europe, 95,548 confirmed cases of human salmonellosis occurred in 2011 (12). Poultry (meat and eggs) have been recognized as significant sources of human salmonellosis (23). Currently, antimicrobial agents such as fluoroquino lones and cephalosporin class drugs are the first choice of treatment for human salmonellosis; however, in recent years an increased number of multidrug-resistant Salmonella strains have been isolated from human cases (19). Although more than 2,500 Salmonella enterica serotypes have been identified, most cases of human salmonellosis are caused by a small number of serotypes, e.g., Typhimurium, Enteritidis, and Derby (13, 16, 33). Salmonella can be transmitted to humans via direct contact with animals or indirectly via * Author for correspondence. Tel: 770-467-6066; Fax: 770-229-3216; E-mail: [email protected].
consumption of contaminated animal products, cross contamination, and undercooking of contaminated foods ( 22).
Quantitative data on Salmonella on raw poultry in China are very limited. Because risk to human health is based on a dose-response relationship, low numbers of pathogen cells might not pose a serious risk, even though the prevalence is high. Information on frequency of common and unique Salmonella serotypes in raw poultry in China also is unavailable. Data on antimicrobial-resistant Salmonella strains are needed to assess the possible impact on public health of multidrag-resistant isolates from raw poultry meat. We recently completed a study in which we determined the baseline Salmonella prevalence on raw poultry in six provinces and two cities in China (35). In the present study, our goals were to determine (i) Salmonella counts on raw poultry from retail markets in Guangdong, Shaanxi, and Sichuan provinces, which had high, medium, and low Salmonella prevalences, respectively, based on our previous study (35), and (ii) the serotypes and antimicrobial susceptibility of Salmonella isolates from the previous study (35) and the present study to obtain a wider
J. Food Prot., Vol. 77, No. 6
understanding of the geographical distribution of these isolates. MATERIALS AND METHODS Sample collection. A total of 300 whole chicken carcasses were collected from retail establishments, i.e., wet (open) markets and large and small supermarkets, in Guangdong Province (with the highest historic Salmonella prevalence; 64.6%), Shaanxi Province (with a medium historic Salmonella prevalence; 50%), and Sichuan Province (with the lowest historic Salmonella prevalence; 38.9%) (35). The sample collection months by province were similar to those in the previous study (i.e., December in Guangdong and April in Shaanxi and Sichuan). In each province, the three cities (one capital and two noncapital) in which samples were collected were the same as those visited in our previous study. Within each city, four large markets, four small markets, and three wet markets were visited for sample collection. One hundred poultry samples were collected in each province: 35 samples were from large markets, 35 samples from small markets, and 30 samples from wet markets. Samples were collected from the same markets as visited in our previous study. Large markets (“ hypermarkets” ) had areas of 14,000 to 21,000 m2 with a variety of departments. Chicken meat was sold either refrigerated or frozen in the meat department. Smaller supermarkets had areas of less than 14,000 m2, and chicken meat was sold in refrigerated or frozen condition. Wet markets were open (mostly meat) markets, including butcher shops for live poultry, fish stalls, and stands for fruits and vegetables. Refrigeration of chickens was limited. Chicken products were stored for only short periods to be sold fresh. After collection, samples were transported on ice and processed immediately at the microbiology laboratories of South China Agricultural University, Northwest A&F University, and the University for Nationalities in Guangdong, Sichuan, and Shaanxi provinces, respectively. Quantitative Salmonella analysis using the three-tube MPN method. Each chicken carcass was aseptically placed in a sterile plastic bag, and 400 ml of buffered peptone water (BPW; Difco, BD, Sparks, MD) was added. The carcass was rinsed by manually massaging with a rocking motion for 1 min, and the rinsate was used to quantify Salmonella counts using the threetube, three-dilution most-probable-number (MPN) method accord ing to U.S. Department of Agriculture (USDA) protocols (29). For each carcass rinse sample, nine tubes were used with the first threetube set containing 10 ml of the rinsate and the remaining second and third sets of three tubes each containing 9 ml of BPW. One milliliter of the chicken rinsate was added to the second set and vortexed for 30 s, and a 10-fold serial dilution was performed with the third three-tube set. All nine tubes were incubated at 35 °C for 24 h. One portion (0.5 + 0.05 ml) of the preenrichment culture was transferred to 10 ml of tetrathionate (TT; Difco, BD) broth, and another portion (0.1 ± 0.02 ml) was transferred into 10 ml of modified Rappaport Vassiliadis (mRV; Difco, BD) broth. Both cultures were incubated at 42 + 0.5°C with shaking at 100 rpm (Difco, BD) for 22 to 24 h. A loopful of each TT and mRV broth culture was then streaked onto xylose lysine Tergitol 4 (XLT4; Difco, BD) and brilliant green sulfa (BGS; Difco, BD) agar and incubated at 35 + 2°C for 22 to 24 h. Up to three presumptive Salmonella colonies on XLT4 and BGS plates were selected and inoculated onto triple sugar iron (TSI; Difco, BD) agar and lysine iron agar (LIA; Difco, BD) slants and incubated at 35 + 2°C for 24 h. Isolates with typical Salmonella reactions on TSI agar and
SALMONELLA COUNTS ON RAW POULTRY IN CHINA
895
LIA were then confirmed by poly-O and poly-H antisera (Difco, BD). The MPN per milliliter of each sample was determined using the USDA MPN table (29). The limit of detection was estimated at 12 MPN of Salmonella cells per carcass. Serotyping. Of the 1,094 Salmonella isolates serotyped, 670 isolates were recovered from our previous study (35): 71 isolates from Shaanxi Province, 72 from Sichuan, 77 from Henan, 90 from Beijing, 113 from Guangdong, 100 from Guangxi, 76 from Fujian, and 71 from Shanghai. The remaining 424 isolates were recovered from the 300 chicken carcasses analyzed for Salmonella counts via the MPN method in the present study: 115 from Guangdong, 84 from Sichuan, and 225 from Shaanxi. Salmonella isolates were serotyped at the Henan Center for Disease Control and Preven tion (Zhengzhou, Henan, China). Salmonella O and H antigens were characterized using slide agglutination with hyperimmune sera (S&A Company, Bangkok, Thailand), and the serotype was assigned following the Kauffmann-White scheme and the manufacturer’s instructions (14). Antimicrobial susceptibility testing. A subset of 779 of the 1,094 isolates was tested for antimicrobial susceptibility. All 671 Salmonella isolates from the previous study and 108 of the 424 isolates found in the present study were tested for antimicrobial susceptibility. The 108 isolates were selected to represent the range of provinces, cities, market type, and storage temperatures. Antimicrobial MICs were determined by the agar dilution method using Mueller-Hinton agar (Beijing Land Bridge Technology Co., Ltd., Beijing, China) according to the guidelines recommended by the Clinical and Laboratory Standards Institute (CLSI) (6). The antimicrobial agents were selected according to the National Antimicrobial Resistance Monitoring System (NARMS) managed by the U.S. Food and Drug Administration, the Centers for Disease Control and Prevention, and the USDA. The antimicrobial susceptibility of the Salmonella isolates was tested against the following 15 antimicrobial agents; ampicillin, amoxicillin-clavulanic acid, cefoxitin, ceftriaxone, ceftiofur, chloramphenicol, nalidixic acid, ciprofloxacin, gentamicin, kanamycin, streptomy cin, amikacin, tetracycline, sulfisoxazole, and trimethoprimsulfamethoxazole (Table 1). Escherichia coli strains ATCC 25922 and ATCC 35218 and Enterococcus faecalis strain ATCC 29212 were used as quality control microorganisms in MIC determinations. The breakpoints used for the interpretation of resistance and susceptibility were the interpretive standards provided by CLSI (6) except for streptomycin, for which we used provisional breakpoints from the NARMS (30). Isolates with intermediate MICs were classified as susceptible. Statistical analysis. The outcomes of the study were Salmonella prevalence and counts, frequency of Salmonella serotypes, and frequency and distribution of antimicrobial-resistant Salmonella (resistant or susceptible) and multidrug-resistant isolates (resistant to three or more antimicrobial classes). The Salmonella prevalence data were cross-tabulated with each of the study variables: market type (large, small, and wet), storage temperature (frozen, chilled, and ambient), province (Guangdong, Shaanxi, and Sichuan), and city type (capital and noncapital). Fisher’s exact test or a 2 x n likelihood ratio chisquare test were used where appropriate with STATA software version 10.1 (Stata Corp., College Station, TX). The MPN per milliliter was adjusted to the original rinse volume (400 ml) and log transformed using the approximate normality. The MPN values per carcass that were higher than the limit of detection were used in the analysis. The relationship
896
J. Food Prot., Vol. 77, No. 6
YANG ET AL.
TABLE 1. MIC ranges tested and interpretation fo r antimicrobials tested in the study MIC interpretive standard (pg/ml)a Antimicrobial agent
Abbreviation
MIC range tested (pg/ml)
AMK GEN KAN STR
8-64 2-16 4—64 32-64
S
I
R
64
32
Aminoglycosides Amikacin Gentamicin Kanamycin Streptomycin6 Penicillin P-lactamase inhibitor combinations Amoxicillin-clavulanic acid Ampicillin
AMC AMP
4, 2-32, 16 4-32
CTX FOX CRO
4-64 4-32 2-64
64
NAL
4-32
32
CIP
1-8
8
TCY CHL
2-16 4-32
32
4 , 76
Cephalosporins (third and fourth generations) Ceftiofur Cefoxitin Ceftriaxone Quinolone Nalidixic acid Fluoroquinolones Ciprofloxacin Tetracyclines Tetracyclines Chloramphenicol Folate pathway inhibitors Sulfisoxazole Trimethoprim-sulfamethoxazole
SUL SXT
64-512 0.5, 9.5-4, 76
a S, sensitive; I, intermediate; R, resistant. 6 No CLSI interpretative criteria are available for streptomycin; therefore, provisional breakpoints from NARMS were used (30).
between the log MPN per carcass and the study variables was assessed using the generalized linear model with the identity link function and adjustment for dependency within city using generalized estimated equations in STATA. A difference was considered significant at P < 0.05. Salmonella serotypes were cross-tabulated by market type and storage temperature. For data analysis, serotypes that were not in the top five most frequent serotypes were collapsed into one category (other serotypes). The proportion of isolates resistant to each of the antimicrobial agents was compared by serotype, market type, and storage temperature using either Fisher’s exact test or a 2 x n likelihood ratio chi-square test as appropriate in STATA. Multidrug resistance (based on the 15 antimicrobial agents) was compared by serotype, market type, and storage temperature using an m x n likelihood ratio chi-square test.
RESULTS Salmonella prevalence and counts. Of the 300 chicken carcasses collected to determine Salmonella counts using the MPN method, 130 (43.3%) were Salmonella positive with an overall mean of 1.7 log MPN per carcass (95% confidence interval [Cl], 1.6 to 1.8 log MPN per carcass). Salmonella prevalence and mean log MPN data by the study variables are shown in Table 2. There were no significant differences (P > 0.05) in either Salmonella prevalence or counts by the study variables (i.e., market type, storage temperature,
province, and city type) (Table 2). The overall distribution of the log MPN of Salmonella in carcass rinses is shown in Figure 1. The largest percentage of Salmonella-positive samples (i.e., 34%) had 1 to 2 log MPN per carcass. Fewer samples (7.3 and 2.3%) had higher Salmonella counts of 2.1 to 3.0 and 3.1 to 3.8 log MPN per carcass, respectively. Distribution and frequency of Salmonella serotypes. Seventy-eight serotypes were identified among 1,094 Salmo nella isolates, and 26, 30, 18, 21, 30, 20, 16, and 11 serotypes were identified in Shaanxi, Sichuan, Henan, Beijing, Guang dong, Guangxi, Fujian, and Shanghai, respectively. The top five most common Salmonella serotypes on raw chicken carcasses at retail were Enteritidis (19.2%), Indiana (15.2%), Typhimurium (14.6%), Agona (7.1%), and Thompson (6.6%), representing 62.7% of the total isolates. Salmonella Enteritidis was the most prevalent serotype among isolates recovered in Shaanxi Province (29.7%, n = 296), Henan Province (19.5%, n = 77), Guangxi Province (23.0%, n = 100), and Fujian Province (29.9%, n = 77). Salmonella Indiana was the most prevalent serotype in Sichuan Province (12.9%, n = 155) and Shanghai City (31.0%, n = 71), Salmonella Typhimurium was the most prevalent serotype in Guangdong Province (28.5%, n = 228), and Salmonella Bsilla was the most prevalent serotype in Beijing City (18.9%, n = 90).
J. Food Prot., Vol. 77, No. 6
SALMONELLA COUNTS ON RAW POULTRY IN CHINA
897
TABLE 2. Salmonella prevalence by study variables and MPN
values from broiler chicken carcasses obtained from retail stores in Chinaa Variable
% (no.) of Salmonella Mean log MPN/carcass positive samples (95% Cl)6
Storage temp Chilled Frozen Ambient
43.8 (160) 43.4 (53) 45.9 (87)
1.6 (1.47-1.76) 1.8 (1.58-2.08) 1.8 (1.53-2.02)
46.7 (90) 41.5 (106) 45.2 (104)
1.8 (1.58-2.08) 1.7 (1.50-1.90) 1.6 (1.43-1.73)
47.0 (100) 50.0 (100) 36.0 (100)
1.6 (1.45-1.76) 2.0 (1.77-2.22) 1.4 (1.28-1.56)
39.8 (108) 46.9 (192)
1.6 (1.42-1.78) 1.7 (1.60-1.90)
Market type Wet Large Small Province Guangdong Shaanxi Sichuan City type Capital city Not capital city
a Within each column for each variable, none of the values (percentages or mean log MPN) were significantly different (P > 0.05). h Mean log most probable number (MPN) of Salmonella per carcass and its 95% confidence interval (Cl), n = 300.
For the distribution of serotypes by market type and storage temperature, 50, 51, and 48 serotypes were detected among isolates recovered from chickens collected from wet markets, small markets, and large markets, respectively (Table 3). Forty-five serotypes were detected among isolates recovered from chickens stored at ambient conditions, 65 serotypes from chilled chickens, and 29 serotypes from frozen chickens. No significant differences (P > 0.05) were found by market type and storage temperature among the top three Salmonella serotypes (Enteritidis, Indiana, and Typhimurium). Salmonella Agona was significantly more preva lent among isolates recovered from chickens collected in wet markets compared with chickens from other market types and in chickens stored frozen compared with those stored at other temperatures. Salmonella Thompson was significantly more prevalent among isolates from chickens collected from small markets than from other types of markets and from chickens stored under frozen and chilled conditions than from those stored at ambient temperature (Table 3). Antimicrobial resistance profiles. In general, Salmo nella isolates (n = 779) were most frequently resistant to sulfisoxazole (74.1%) and tetracycline (71.1%) and least resistant to ceftriaxone (23.4%), ciprofloxacin (22.5%), and cefoxitin (19.1%). Salmonella isolates recovered from samples collected in Sichuan, Guangdong, and Fujian provinces were highly resistant to sulfisoxazole at 87.1, 92.7, and 88.3%, respectively. Isolates from Beijing City, Shanghai City, and Henan Province were most frequently resistant to tetracycline, ampicillin, and trimethoprimsulfamethoxazole at 72.2, 95.8, and 94.8%, respectively.
Log M P N /c a rc a s s
FIGURE 1. Percentage bar chart illustrating the log mostprobable-number (MPN) distribution of Salmonella isolates on broiler chicken carcasses at retail stores in China.
The individual Salmonella antimicrobial resistance profiles by market type and market temperature are shown in Table 4. No significant differences (.P > 0.05) were found among the percentages of isolates resistant to individual antimicrobial agents except for kanamycin compared by market type, i.e., wet market isolates were less resistant (32.5%) than were isolates from small (42.7%) and large (43.1%) markets. However, significant differences by storage temperature were found for isolates resistant to kanamycin, i.e., higher resistance in frozen chicken isolates (51.6%) than in isolates from chilled (40.4%) and ambient (30.6%) chickens, and to ceftiofur, i.e., higher resistance in frozen chicken isolates (38.1%) than in isolates from chilled (24.2%) and ambient (21.3%) chickens (Table 4). The percentage of resistant isolates (« = 779) to each antimicrobial agent by serotype (five most common serotypes and others) are shown in Table 5. These other serotypes represented 14.8% of the total isolates analyzed for antimicrobial susceptibility. Salmonella Indiana isolates were significantly more resistant to each of the 15 antimicrobial agents than were all other serotypes. In general, the five most common serotypes had relatively high resistance to tetracycline, sulfisoxazole, nalidixic acid, ampicillin, trimethoprim-sulfamethoxazole, and amoxicillin-clavulanic acid. The overall distributions of multidrag-resistant Salmo nella phenotypes are shown in Figure 2. Of the 779 isolates tested, approximately 4% were susceptible to all 15 antimicrobial agents, 45.1% were resistant to 1 to 5 agents, 28.8% were resistant to 6 to 10 agents, and 22.1% were resistant to 11 to 15 agents. Overall, no significant differences (P > 0.05) were found among multidragresistant isolates by retail market and storage temperature. Among the five most common Salmonella serotypes, all (100%) Indiana and Thompson isolates were resistant to at least one antimicrobial agent, and 97.9, 94.1, 97.8, and 93% of Enteritidis, Typhimurium, Agona, and other serotypes, respectively, were resistant to at least one antimicrobial agent. Also among these five most common Salmonella serotypes, approximately 63, 59, 57, 18, and 9% of Agona, Enteritidis, Typhimurium, Thompson, and Indiana isolates, respectively, were resistant to 1 to 5
898
J. Food Prot., Vol. 77, No. 6
YANG ET AL.
TABLE 3. Salmonella serotypes isolated from raw chicken carcasses collected at three types of retail markets under three storage conditions in China No. (%) of positive samples Storage temp"
Market type Total
Wet
Small
Serotype
(n = 1,094)
(n = 400)
(n = 342)
Enteritidis Indiana Typhimurium Agona Thompson Derby Rissen Bsilla Heidelberg Blockley Edinburg Infantis Gueuletapee Shubra Bovismorbificans Corvallis Saintpaul Duesseldorf Albany Potsdam II Hadar Baiboukoum Lockleaze Weltevreden Chartres Haardt Kentucky Meleagridis Stanley Gatuni Newlands Tshiongwe Uppsala Abony Athinai Ball Djugu Dublin Essen Hillingdon Kingston Paratyphi A Anatum Arapahoe Blegdam Bonn Hidalgo Kallo Kouka London Remo Nitra Sangera Schwarzengrund
210 (19.2) 166 (15.2) 160 (14.6) 78 (7.1) 72 (6.6) 52 (4.8) 25 (2.3) 24 (2.2) 22 (2.0) 20 (1.8) 19 (1.7) 19 (1.7) 15 (1.4) 11 (1.0) 10 (0.9) 9 (0.8) 9 (0.8) 9 (0.8) 8 (0.7) 8 (0.7) 8 (0.7) 7 (0.6) 6 (0.5) 6 (0.5) 6 (0.5) 5 (0.5) 5 (0.5) 5 (0.5) 5 (0.5) 5 (0.5) 4 (0.4) 4 (0.4) 4 (0.4) 4 (0.4) 3 (0.3) 3 (0.3) 3 (0.3) 3 (0.3) 3 (0.3) 3 (0.3) 3 (0.3) 3 (0.3) 3 (0.3) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2)
81 (20.3) 57 (14.3) 60 (15.0) 38 (9.5) 11 (2.8) 18 (4.5) 8 (2.0) 15 (3.8) 6 (1.5) 3 (0.8) 4 (TO) 11 (2.8) 6(1.5) 4(1.0) 1 (0.3) 6(1.5) 2 (0.5) 6 (1.5) 7 (1.8) 3 (0.8) 4(1.0) 5 (1.3) 2 (0.5) 5 (1.3) 1 (0.3)
59 (17.3) 60 (17.5) 47 (13.7) 16 (4.7) 36 (10.5) 15 (4.4) 14 (4.1) 2 (0.6) 8 (2.3) 2 (0.6) 6 (1.8) 6(1.8) 2 (0.6) 2 (0.6) 5 (1.5) 1 (0.3) 2 (0.6) 2 (0.6) 1 (0.3) 4(1.2) 3 (0.9) 2 (0.6) 3 (0.9)
3 (0.8) 3 (0.8) 1 (0.3) 4 (1.0) 1 (0.3) 1 (0.3) 2 (0.5) 1 (0.3) 3 (0.8) 2 (0.5)
4 (1.2) 5 (1.5) 2 (0.6) 2 (0.6) 4 (1.2) 1 (0.3) 4 (1.2) 1 (0.3) 2 (0.6) 1 (0.3)
Ambient (n = 346)
Chilled (n = 555)
70 (19.9)* 49 (14.0)* 53 (15.1)* 24 (6.8)c 25 (7.1)c 19 (5.4) 3 (0.9) 7 (2.0) 8 (2.3) 15 (4.3) 9 (2.6) 2 (0.6) 7 (2.0) 5 (1.4) 4(1.1) 2 (0.6) 5 (1.4) 1 (0.3)
69 (19.9) 56 (16.2) 58 (16.8) 20 (5.8) 12 (3.5) 20 (5.8) 9 (2.6) 14 (4.0) 2 (0.6) 3 (0.9) 4 (1.2) 11 (1.0) 3 (0.9) 4 (1.2) 1 (0.3) 6(1.7) 1 (0.3) 6 (1.7) 7 (2.0) 2 (0.6) 4 (1.2) 2 (0.6)
109 (19.6) 76 (13.7) 82 (14.8) 27 (4.9) 44 (7.9) 29 (5.2) 11 (2.0) 10 (1.8) 12 (2.2) 9(1.6) 7 (L3) 7 (1.3) 11 (2.0) 5 (0.9) 5 (0.9) 3 (0.5) 8 (1.4) 3 (0.5) 1 (0.2) 5 (0.9) 4 (0.7) 5 (0.9) 6(1.1) 3 (0.5) 5 (0.9) 3 (0.5) 5 (0.9) 2 (0.4) 2 (0.4) 5 (0.9) 3 (0.5)
1 (0.3) 1 (0.3) 1 (0.3) 1 (0.3) 1 (0.3) 3 (0.9) 2 (0.6) 1 (0.3) 2 (0.6)
2 (0.6) 1 (0.3) 1 (0.3)
2 (0.6) 1 (0.3)
2 (0.6) 3 (0.9) 1 (0.3) 4(1.2) 1 (0.3) 1 (0.3) 2 (0.6) 1 (0.3)
1 (0.3) 1 (0.3) 1 (0.3)
3 (0.9) 2 (0.6) 2 (0.6) 1 (0.3)
1 (0.3)
1 (0.3) 2 (0.6)
1 (0.3)
1 (0.3) 2 (0.6)
1 (0.3)
1 (0.3)
1 (0.3) 2 (0.6) 2 (0.6)
1 (0.3) 1 (0.3)
32 (16.6)* 34 (17.6)* 20 (10.4)* 31 (16.1)c 16 (8.3)c 3 (L6) 5 (2.6) B (4.1) 8 (4.1) 8 (4.1) 1 (0.5) 1 (0.5) 2(1.0) 4(2.1)
1 (0.5)
1 (0.5) 2 (LO) 1 (0.5)
1 (0.5) 1 (0.5)
1 (0.3) 1 (0.3) 1 (0.3) 1 (0.3)
1 (0.2) 1 (0.2) 1 (0.2) 3 (0.5) 3 (0.5) 1 (0.2) 2 (0.4) 1 (0.2) 2 (0.4) 1 (0.2) 2 (0.2)
2 (LO) 1 (0.5)
1 (0.5)
2(1.0) 1 (0.3)
1 (0.3) 1 (0.3) 2 (0.6)
1 (0.3)
4 (0.7) 2 (0.4) 1 (0.2) 1 (0.2) 2 (0.4)
Frozen
(n = 193)
3 (1.6) 2 (0.6)
1 (0.3)
1 (0.3) 1 (0.3) 1 (0.3)
Large (n = 352)
1 (0.2) 2 (0.4) 1 (0.2) 2 (0.2) 1 (0.2)
1 (0.5)
J. Food Prot., Vol. 77, No. 6
SALMONELLA COUNTS ON RAW POULTRY IN CHINA
899
TABLE 3. Continued No. (%) of positive samples Market type
Serotype Aberdeen Augustenborg Bellevue Bessi Chinco Choleraesuis Col indale Fillmore Gallinarum Glostrup Kaapstad Leer Lindenburg Massenya Montevideo Nchanga Newport Orion Pomona Schwerin Tinda Ughelli Wangat
Storage temp"
Total
Wet
Small
Large
Ambient
(n = 1,094)
Chilled
(n = 400)
Frozen
(n = 342)
(n = 352)
(n = 346)
(n = 555)
in = 193)
1 1 1 1
( 0 . 1) ( 0 . 1)
(0 . 1) (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1)
1 ( 0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1)
1 (0.3) 1 (0.3) 1 (0.3)
1 (0.3) 1 (0.3) 1 (0.3) 1 (0.3)
1 (0 .2)
1 (0.3) 1 (0.3) 1 (0.3)
1 (0 .2) 1 (0 .2)
1 (0.3)
1 (0.5)
1 ( 0 .2 )
1 (0.3) 1 (0.3) 1 (0.3) 1 (0.3)
1 (0 .2) 1 (0 .2) 1 (0 .2) 1 ( 0 .2 ) 1 (0.3)
1 (0.2)
1 (0.3) 1 (0.3)
1 ( 0 .2 ) 1 (0.5)
1 (0.3) 1 (0.3)
1 (0.3) 1 (0.3) 1 (0.3)
1 ( 0 .2) 1 ( 0 .2) 1 ( 0 .2)
1 (0.3) 1 (0.3) 1 (0.3)
1 (0.5)
1 (0.3) 1 (0.3)
1 (0.3) 1 (0.3)
“ Ambient chickens were stored at 20 to 30°C, chilled chickens were stored at 4 to 10°C, and frozen chickens were stored at - 1 0 to -2 0 °C . Differences among percentage of Salmonella -positive samples by market type or storage temperature were not significant IP > 0.05). ‘ Differences among percentage of Salmonella -positive samples by market type or storage temperature were significant (P < 0.05). These differences were based on a likelihood ratio chi-square test of the differences in risk by market type and storage temperature for the top five most prevalent serotypes.
antimicrobial agents; 44, 33, 30, 22, and 18% of Thompson, Enteritidis, Typhimurium, Agona, and Indiana isolates, re spectively, were resistant to 6 to 10 agents; and 74, 38, 13, 7, and 6% of Indiana, Thompson, Agona, Typhimurium, and Enteritidis isolates, respectively, were resistant to 11 to 15 agents. High percentages (91 and 82%) of isolates of two serotypes (Indiana and Thompson, respectively) were resistant to 6 to 15 antimicrobial agents. DISCUSSION In our previous study (35), 52.2% of 1,152 retail chickens were Salmonella positive, whereas in the present study only 43.3% of 300 retail chickens were Salmonella positive. The overall mean Salmonella counts were 1.7 log MPN per carcass (95% Cl, 1.6 to 1.8 log MPN per carcass). This mean is similar that reported on postchilled chickens at processing plants in the United States (1.7 log MPN per carcass) (28). When comparing Salmonella count data from China (Fig. 1) and the United States (28), Salmonella counts of 1.0 to 2.0 log MPN per carcass were found in 34% of carcasses in the present study compared with 3.7% of U.S. carcasses (28); counts of 2.1 to 3.0 log MPN per carcass were found in 7.3% of our samples compared with 1.2% of U.S. carcasses; and counts of 3.1 to
3.8 log MPN per carcass were found in 2.3% of our samples compared with 0.27% of U.S. samples. Compared with other countries, the mean MPN for Salmonella found on the chickens in China was slightly higher than that on fresh retail chicken in The Netherlands where 89% of fresh (nonfrozen) chickens had less than 1 log MPN per carcass (10) and in Cambodia where Salmonella counts were 3 to 4 log CFU/g in 22.4% of the poultry samples (n = 152) (21). Differences in type of chicken meat sample, sampling times, sample number, slaughterhouse and retail store sani tation, and Salmonella quantification and isolation methods should be taken into consideration when comparing findings among studies. Salmonella Enteritidis was the most frequently identi fied serotype in China overall and in Beijing City and Shaanxi, Henan, Guangxi, and Fujian provinces. This serotype was found on chickens from various market types (wet, large, and small) and on chickens stored at various temperatures (ambient, chilled, and frozen). Salmonella Enteritidis was also identified as the predominant serotype in poultry products in other studies (1, 2, 16, 19, 34) and as one of the most common serotypes that cause human salmonellosis in many countries (7, 17, 18, 26). According to a laboratory-based survey of nontyphoidal Salmonella
900
J. Food Prot., Vol. 77, No. 6
YANG ET AL.
TABLE 4. Anitimicrobial resistance of Salmonella isolates from raw poultry collected at three types of retail markets under three storage conditions in Chinaa No. (%) of resistant isolates Market type
Sulftsoxazole Tetracycline Ampicillin Nalidixic acid Trimethoprimsulfamethoxazole Amoxicillin-clavulanic acid Chloramphenicol Kanamycin Streptomycin Gentamicin Ceftiofur Amikacin Ceftriaxone Ciprofloxacin Cefoxitin
Wet
(n = 779)
Large (n = 248)
Small (n = 260)
(n = 271)
577 554 481 479
185 173 154 154
191 196 169 158
201 185 158 167
Total Antimicrobial agent
Storage temp
P value"
Ambient (.n = 235)
Frozen (n = 126)
Chilled (n = 418)
P value
96 86 82 88
307 305 262 254
(74.2) (68.3) (58.3) (61.6)
0.982 0.489 0.693 0.937
174 163 137 137
(73.4) (73.0) (62.7) (60.8)
0.822 0.841 0.886 0.868
441 (56.6) 139 (56.0) 162 (62.3) 140 (51.7)
0.113
125 (53.2) 107 (84.9) 209 (50.0)
0.492
116 (46.8) 126 (48.5) 117 (43.2) 100 (40.3) 127 (48.8) 102 (37.6) 107 (43.1) 111 (42.7) 88 (32.5) 101 (40.7) 98 (37.7) 85 (31.4) 90 (36.3) 97 (37.3) 86 (31.7) 69 (27.8) 71 (27.3) 59 (21.8) 69 (27.8) 67 (25.8) 49 (18.1) 65 (26.2) 58 (22.3) 59 (21.8) 62 (25.0) 55 (21.2) 58 (21.4) 49 (19.8) 52 (20.0) 48 (17.7)
0.638 0.054 0.049 0.360 0.447 0.120 0.181 0.472 0.495 0.670
106 (45.1) 94 (40.0) 72 (30.6) 72 (30.6) 71 (30.2) 50 (21.3) 46 (19.6) 50 (21.3) 51 (21.7) 42 (17.9)
(74.1) (71.1) (61.7) (61.5)
359 (46.1) 329 (42.2) 306 (39.3) 284 (36.5) 273 (35.0) 199 (25.5) 185 (23.7) 182 (23.4) 175 (22.5) 149 (19.1)
(74.6) (69.8) (62.1) (62.1)
(73.5) (75.4) (65.0) (60.8)
(74.0) (69.4) (58.3) (58.3)
69 82 65 55 51 48 37 43 35 31
(76.2) (68.3) (65.1) (69.8)
(54.8) (65.1) (51.6) (43.7) (40.5) (38.1) (29.4) (34.1) (27.8) (24.6)
184 153 169 157 151 101 102 89 89 76
(44.0) (36.6) (40.4) (37.6) (36.1) (24.2) (24.4) (21.3) (21.3) (18.2)
0.460 0.524 0.020 0.298 0.227 0.008 0.483 0.052 0.665 0.249
a p values are based on a likelihood ratio chi-square test of the differences in risk by market type and storage conditions.
infections in humans in China conducted by the Chinese Center for Disease Control and Prevention, Salmonella Enteritidis was the most frequently identified serotype (31%, n = 662 isolates) (24). Data from other regions have revealed that Salmonella Enteritidis is the most common serotype among human salmonellosis cases in Africa, Asia,
Europe, Latin America, and the Caribbean and is the second common serotype in North America (13). Other Salmonella serotypes, including Indiana, Typhimurium, Agona, Thompson, and Derby, were repeatedly recovered from chicken meat in the present study. Our findings are in agreement with those described previously, indicating that
TABLE 5. Antimicrobial resistance by most common serotype of Salmonella isolates from raw chicken collected in China No. (%) of resistant Salmonella isolates
Antimicrobial agent
Tetracycline Sulftsoxazole Nalidixic acid Ampicillin Trimethoprimsulfamethoxazole Amoxicillinclavulanic acid Chloramphenicol Kanamycin Gentamicin Streptomycin Ceftiofur Amikacin Ciprofloxacin Ceftriaxone Cefoxitin
Typhimurium (n = 102)
Agona (« = 46)
Thompson (n = 55)
30 37 23 22
70 66 39 66
Enteritidis
Indiana
(n = 146)
(n = 114)
178 189 197 154
(62.7) (66.6) (69.4) (54.2)
104 (91.2) 102 (90.3) 106 (93.0) 105 (92.1)
90 95 79 68
119 (41.9)
105 (92.9)
65 (48.5)
23 (50.0)
111 (39.1) 84 (29.6) 82 (28.9) 64 (22.5) 76 (26.8) 36 (12.7) 42 (14.8) 23 (8.1) 44 (17.9) 44 (15.5)
91 (79.8) 97 (85.1) 93 (81.6) 93 (81.6) 83 (73.5) 79 (69.9) 61 (53.5) 93 (81.6) 73 (64.0) 42 (36.8)
44 (32.8) 40 (29.9) 42 (31.3) 38 (28.4) 57 (42.5) 19 (14.2) 12 (9.0) 22 (16.4) 12 (10.4) 9 (6.7)
16 (34.8) 14 (30.4) 11 (23.9) 6 (13.0) 14 (30.4) 7 (15.2) 4 (8.7) 10 (21.7) 6 (15.4) 8 (17.4)
(67.2) (70.9) (59.0) (50.8)
(65.2) (80.4) (50.0) (47.8)
Other serotypes Cn = 316)"
P value*
(71.3) (76.5) (58.3) (57.4)
Counts, Serotypes, and Antimicrobial Resistance of Salmonella Isolates on Retail Raw Poultry in the People’s Republic of China BAOWEI YANG,1 YUE CUI,12 CHAO SHI, 1 JIAQI WANG, 1 XIAODONG XIA, 1 MEILI XI,1 XIN WANG,1 JIANGHONG MENG, 13 WALID Q. ALALI,2* ISABEL WALLS,4 a n d MICHAEL P. DOYLE2 1College o f Food Science, Northwest A&F University, Yangling 712100, People’s Republic o f China; 2Center for Food Safety, University o f Georgia, Griffin, Georgia 30223, USA; 3Joint Institute fo r Food Safety and Applied Nutrition and Department o f Nutrition and Food Science, University o f Maryland, College Park, Maryland 20742, USA; and 4U.S. Department o f Agriculture, National Institute o f Food and Agriculture, Washington, D.C. 20250, USA MS 13-439: Received 14 October 2013/Accepted 31 January 2014
A BSTR A CT The objective of this study was to determine Salmonella counts, serotypes, and antimicrobial resistance profiles in retail raw chicken meat in the People’s Republic of China. Salmonella counts were determined according to the most-probable-number (MPN) method for 300 whole chicken carcasses. These samples were collected from large, small, and wet (open) markets in Guangdong, Shaanxi, and Sichuan provinces. Salmonella isolates were serotyped and tested for antimicrobial susceptibility. Of the 300 chicken carcasses, 43.3% were positive for Salmonella, with an overall mean of 1.7 log MPN per carcass (95% confidence interval, 1.6 to 1.8 log MPN per carcass). No significant differences (P > 0.05) were detected for storage temperature (i.e., chilled, frozen, or ambient), market type (large, small, or wet), province, or location (capital or noncapital city). Seventyeight serotypes were identified among the 1,094 Salmonella isolates. The top five most common Salmonella serotypes on raw chicken carcasses were Enteritidis (19.2%), Indiana (15.2%), Typhimurium (14.6%), Agona (7.1%), and Thompson (6.6%). Salmonella isolates (n = 779) were most frequently resistant to sulfisoxazole (74.1%) and tetracycline (71.1%) and least resistant to ceftriaxone (22.5%) and cefoxitin (19%). Only 4% of the isolates were susceptible to all 15 antimicrobial agents, 45% were resistant to 1 to 5 agents, 29% were resistant to 6 to 10 agents, and 22% were resistant to 11 to 15 agents. Our findings revealed that Salmonella contamination was common in retail raw poultry in China, and the counts on contaminated carcasses were mostly low. Salmonella isolates were diverse in their serotype distribution and antimicrobial susceptibility profiles, with more than half of the isolates resistant to more than five antimicrobial agents. These data may be used in risk assessment models to reduce the transmission of Salmonella via chicken meat to humans in China.
Salmonella is an important zoonotic pathogen that has a significant economic and health impact on animals and humans worldwide (31). Approximately 70 to 80% of foodbome bacterial outbreaks in the People’s Republic of China are caused by Salmonella (32). In the United States, an estimated 1 million cases of foodbome salmonellosis with 378 deaths occur annually (25). In Europe, 95,548 confirmed cases of human salmonellosis occurred in 2011 (12). Poultry (meat and eggs) have been recognized as significant sources of human salmonellosis (23). Currently, antimicrobial agents such as fluoroquino lones and cephalosporin class drugs are the first choice of treatment for human salmonellosis; however, in recent years an increased number of multidrug-resistant Salmonella strains have been isolated from human cases (19). Although more than 2,500 Salmonella enterica serotypes have been identified, most cases of human salmonellosis are caused by a small number of serotypes, e.g., Typhimurium, Enteritidis, and Derby (13, 16, 33). Salmonella can be transmitted to humans via direct contact with animals or indirectly via * Author for correspondence. Tel: 770-467-6066; Fax: 770-229-3216; E-mail: [email protected].
consumption of contaminated animal products, cross contamination, and undercooking of contaminated foods ( 22).
Quantitative data on Salmonella on raw poultry in China are very limited. Because risk to human health is based on a dose-response relationship, low numbers of pathogen cells might not pose a serious risk, even though the prevalence is high. Information on frequency of common and unique Salmonella serotypes in raw poultry in China also is unavailable. Data on antimicrobial-resistant Salmonella strains are needed to assess the possible impact on public health of multidrag-resistant isolates from raw poultry meat. We recently completed a study in which we determined the baseline Salmonella prevalence on raw poultry in six provinces and two cities in China (35). In the present study, our goals were to determine (i) Salmonella counts on raw poultry from retail markets in Guangdong, Shaanxi, and Sichuan provinces, which had high, medium, and low Salmonella prevalences, respectively, based on our previous study (35), and (ii) the serotypes and antimicrobial susceptibility of Salmonella isolates from the previous study (35) and the present study to obtain a wider
J. Food Prot., Vol. 77, No. 6
understanding of the geographical distribution of these isolates. MATERIALS AND METHODS Sample collection. A total of 300 whole chicken carcasses were collected from retail establishments, i.e., wet (open) markets and large and small supermarkets, in Guangdong Province (with the highest historic Salmonella prevalence; 64.6%), Shaanxi Province (with a medium historic Salmonella prevalence; 50%), and Sichuan Province (with the lowest historic Salmonella prevalence; 38.9%) (35). The sample collection months by province were similar to those in the previous study (i.e., December in Guangdong and April in Shaanxi and Sichuan). In each province, the three cities (one capital and two noncapital) in which samples were collected were the same as those visited in our previous study. Within each city, four large markets, four small markets, and three wet markets were visited for sample collection. One hundred poultry samples were collected in each province: 35 samples were from large markets, 35 samples from small markets, and 30 samples from wet markets. Samples were collected from the same markets as visited in our previous study. Large markets (“ hypermarkets” ) had areas of 14,000 to 21,000 m2 with a variety of departments. Chicken meat was sold either refrigerated or frozen in the meat department. Smaller supermarkets had areas of less than 14,000 m2, and chicken meat was sold in refrigerated or frozen condition. Wet markets were open (mostly meat) markets, including butcher shops for live poultry, fish stalls, and stands for fruits and vegetables. Refrigeration of chickens was limited. Chicken products were stored for only short periods to be sold fresh. After collection, samples were transported on ice and processed immediately at the microbiology laboratories of South China Agricultural University, Northwest A&F University, and the University for Nationalities in Guangdong, Sichuan, and Shaanxi provinces, respectively. Quantitative Salmonella analysis using the three-tube MPN method. Each chicken carcass was aseptically placed in a sterile plastic bag, and 400 ml of buffered peptone water (BPW; Difco, BD, Sparks, MD) was added. The carcass was rinsed by manually massaging with a rocking motion for 1 min, and the rinsate was used to quantify Salmonella counts using the threetube, three-dilution most-probable-number (MPN) method accord ing to U.S. Department of Agriculture (USDA) protocols (29). For each carcass rinse sample, nine tubes were used with the first threetube set containing 10 ml of the rinsate and the remaining second and third sets of three tubes each containing 9 ml of BPW. One milliliter of the chicken rinsate was added to the second set and vortexed for 30 s, and a 10-fold serial dilution was performed with the third three-tube set. All nine tubes were incubated at 35 °C for 24 h. One portion (0.5 + 0.05 ml) of the preenrichment culture was transferred to 10 ml of tetrathionate (TT; Difco, BD) broth, and another portion (0.1 ± 0.02 ml) was transferred into 10 ml of modified Rappaport Vassiliadis (mRV; Difco, BD) broth. Both cultures were incubated at 42 + 0.5°C with shaking at 100 rpm (Difco, BD) for 22 to 24 h. A loopful of each TT and mRV broth culture was then streaked onto xylose lysine Tergitol 4 (XLT4; Difco, BD) and brilliant green sulfa (BGS; Difco, BD) agar and incubated at 35 + 2°C for 22 to 24 h. Up to three presumptive Salmonella colonies on XLT4 and BGS plates were selected and inoculated onto triple sugar iron (TSI; Difco, BD) agar and lysine iron agar (LIA; Difco, BD) slants and incubated at 35 + 2°C for 24 h. Isolates with typical Salmonella reactions on TSI agar and
SALMONELLA COUNTS ON RAW POULTRY IN CHINA
895
LIA were then confirmed by poly-O and poly-H antisera (Difco, BD). The MPN per milliliter of each sample was determined using the USDA MPN table (29). The limit of detection was estimated at 12 MPN of Salmonella cells per carcass. Serotyping. Of the 1,094 Salmonella isolates serotyped, 670 isolates were recovered from our previous study (35): 71 isolates from Shaanxi Province, 72 from Sichuan, 77 from Henan, 90 from Beijing, 113 from Guangdong, 100 from Guangxi, 76 from Fujian, and 71 from Shanghai. The remaining 424 isolates were recovered from the 300 chicken carcasses analyzed for Salmonella counts via the MPN method in the present study: 115 from Guangdong, 84 from Sichuan, and 225 from Shaanxi. Salmonella isolates were serotyped at the Henan Center for Disease Control and Preven tion (Zhengzhou, Henan, China). Salmonella O and H antigens were characterized using slide agglutination with hyperimmune sera (S&A Company, Bangkok, Thailand), and the serotype was assigned following the Kauffmann-White scheme and the manufacturer’s instructions (14). Antimicrobial susceptibility testing. A subset of 779 of the 1,094 isolates was tested for antimicrobial susceptibility. All 671 Salmonella isolates from the previous study and 108 of the 424 isolates found in the present study were tested for antimicrobial susceptibility. The 108 isolates were selected to represent the range of provinces, cities, market type, and storage temperatures. Antimicrobial MICs were determined by the agar dilution method using Mueller-Hinton agar (Beijing Land Bridge Technology Co., Ltd., Beijing, China) according to the guidelines recommended by the Clinical and Laboratory Standards Institute (CLSI) (6). The antimicrobial agents were selected according to the National Antimicrobial Resistance Monitoring System (NARMS) managed by the U.S. Food and Drug Administration, the Centers for Disease Control and Prevention, and the USDA. The antimicrobial susceptibility of the Salmonella isolates was tested against the following 15 antimicrobial agents; ampicillin, amoxicillin-clavulanic acid, cefoxitin, ceftriaxone, ceftiofur, chloramphenicol, nalidixic acid, ciprofloxacin, gentamicin, kanamycin, streptomy cin, amikacin, tetracycline, sulfisoxazole, and trimethoprimsulfamethoxazole (Table 1). Escherichia coli strains ATCC 25922 and ATCC 35218 and Enterococcus faecalis strain ATCC 29212 were used as quality control microorganisms in MIC determinations. The breakpoints used for the interpretation of resistance and susceptibility were the interpretive standards provided by CLSI (6) except for streptomycin, for which we used provisional breakpoints from the NARMS (30). Isolates with intermediate MICs were classified as susceptible. Statistical analysis. The outcomes of the study were Salmonella prevalence and counts, frequency of Salmonella serotypes, and frequency and distribution of antimicrobial-resistant Salmonella (resistant or susceptible) and multidrug-resistant isolates (resistant to three or more antimicrobial classes). The Salmonella prevalence data were cross-tabulated with each of the study variables: market type (large, small, and wet), storage temperature (frozen, chilled, and ambient), province (Guangdong, Shaanxi, and Sichuan), and city type (capital and noncapital). Fisher’s exact test or a 2 x n likelihood ratio chisquare test were used where appropriate with STATA software version 10.1 (Stata Corp., College Station, TX). The MPN per milliliter was adjusted to the original rinse volume (400 ml) and log transformed using the approximate normality. The MPN values per carcass that were higher than the limit of detection were used in the analysis. The relationship
896
J. Food Prot., Vol. 77, No. 6
YANG ET AL.
TABLE 1. MIC ranges tested and interpretation fo r antimicrobials tested in the study MIC interpretive standard (pg/ml)a Antimicrobial agent
Abbreviation
MIC range tested (pg/ml)
AMK GEN KAN STR
8-64 2-16 4—64 32-64
S
I
R
64
32
Aminoglycosides Amikacin Gentamicin Kanamycin Streptomycin6 Penicillin P-lactamase inhibitor combinations Amoxicillin-clavulanic acid Ampicillin
AMC AMP
4, 2-32, 16 4-32
CTX FOX CRO
4-64 4-32 2-64
64
NAL
4-32
32
CIP
1-8
8
TCY CHL
2-16 4-32
32
4 , 76
Cephalosporins (third and fourth generations) Ceftiofur Cefoxitin Ceftriaxone Quinolone Nalidixic acid Fluoroquinolones Ciprofloxacin Tetracyclines Tetracyclines Chloramphenicol Folate pathway inhibitors Sulfisoxazole Trimethoprim-sulfamethoxazole
SUL SXT
64-512 0.5, 9.5-4, 76
a S, sensitive; I, intermediate; R, resistant. 6 No CLSI interpretative criteria are available for streptomycin; therefore, provisional breakpoints from NARMS were used (30).
between the log MPN per carcass and the study variables was assessed using the generalized linear model with the identity link function and adjustment for dependency within city using generalized estimated equations in STATA. A difference was considered significant at P < 0.05. Salmonella serotypes were cross-tabulated by market type and storage temperature. For data analysis, serotypes that were not in the top five most frequent serotypes were collapsed into one category (other serotypes). The proportion of isolates resistant to each of the antimicrobial agents was compared by serotype, market type, and storage temperature using either Fisher’s exact test or a 2 x n likelihood ratio chi-square test as appropriate in STATA. Multidrug resistance (based on the 15 antimicrobial agents) was compared by serotype, market type, and storage temperature using an m x n likelihood ratio chi-square test.
RESULTS Salmonella prevalence and counts. Of the 300 chicken carcasses collected to determine Salmonella counts using the MPN method, 130 (43.3%) were Salmonella positive with an overall mean of 1.7 log MPN per carcass (95% confidence interval [Cl], 1.6 to 1.8 log MPN per carcass). Salmonella prevalence and mean log MPN data by the study variables are shown in Table 2. There were no significant differences (P > 0.05) in either Salmonella prevalence or counts by the study variables (i.e., market type, storage temperature,
province, and city type) (Table 2). The overall distribution of the log MPN of Salmonella in carcass rinses is shown in Figure 1. The largest percentage of Salmonella-positive samples (i.e., 34%) had 1 to 2 log MPN per carcass. Fewer samples (7.3 and 2.3%) had higher Salmonella counts of 2.1 to 3.0 and 3.1 to 3.8 log MPN per carcass, respectively. Distribution and frequency of Salmonella serotypes. Seventy-eight serotypes were identified among 1,094 Salmo nella isolates, and 26, 30, 18, 21, 30, 20, 16, and 11 serotypes were identified in Shaanxi, Sichuan, Henan, Beijing, Guang dong, Guangxi, Fujian, and Shanghai, respectively. The top five most common Salmonella serotypes on raw chicken carcasses at retail were Enteritidis (19.2%), Indiana (15.2%), Typhimurium (14.6%), Agona (7.1%), and Thompson (6.6%), representing 62.7% of the total isolates. Salmonella Enteritidis was the most prevalent serotype among isolates recovered in Shaanxi Province (29.7%, n = 296), Henan Province (19.5%, n = 77), Guangxi Province (23.0%, n = 100), and Fujian Province (29.9%, n = 77). Salmonella Indiana was the most prevalent serotype in Sichuan Province (12.9%, n = 155) and Shanghai City (31.0%, n = 71), Salmonella Typhimurium was the most prevalent serotype in Guangdong Province (28.5%, n = 228), and Salmonella Bsilla was the most prevalent serotype in Beijing City (18.9%, n = 90).
J. Food Prot., Vol. 77, No. 6
SALMONELLA COUNTS ON RAW POULTRY IN CHINA
897
TABLE 2. Salmonella prevalence by study variables and MPN
values from broiler chicken carcasses obtained from retail stores in Chinaa Variable
% (no.) of Salmonella Mean log MPN/carcass positive samples (95% Cl)6
Storage temp Chilled Frozen Ambient
43.8 (160) 43.4 (53) 45.9 (87)
1.6 (1.47-1.76) 1.8 (1.58-2.08) 1.8 (1.53-2.02)
46.7 (90) 41.5 (106) 45.2 (104)
1.8 (1.58-2.08) 1.7 (1.50-1.90) 1.6 (1.43-1.73)
47.0 (100) 50.0 (100) 36.0 (100)
1.6 (1.45-1.76) 2.0 (1.77-2.22) 1.4 (1.28-1.56)
39.8 (108) 46.9 (192)
1.6 (1.42-1.78) 1.7 (1.60-1.90)
Market type Wet Large Small Province Guangdong Shaanxi Sichuan City type Capital city Not capital city
a Within each column for each variable, none of the values (percentages or mean log MPN) were significantly different (P > 0.05). h Mean log most probable number (MPN) of Salmonella per carcass and its 95% confidence interval (Cl), n = 300.
For the distribution of serotypes by market type and storage temperature, 50, 51, and 48 serotypes were detected among isolates recovered from chickens collected from wet markets, small markets, and large markets, respectively (Table 3). Forty-five serotypes were detected among isolates recovered from chickens stored at ambient conditions, 65 serotypes from chilled chickens, and 29 serotypes from frozen chickens. No significant differences (P > 0.05) were found by market type and storage temperature among the top three Salmonella serotypes (Enteritidis, Indiana, and Typhimurium). Salmonella Agona was significantly more preva lent among isolates recovered from chickens collected in wet markets compared with chickens from other market types and in chickens stored frozen compared with those stored at other temperatures. Salmonella Thompson was significantly more prevalent among isolates from chickens collected from small markets than from other types of markets and from chickens stored under frozen and chilled conditions than from those stored at ambient temperature (Table 3). Antimicrobial resistance profiles. In general, Salmo nella isolates (n = 779) were most frequently resistant to sulfisoxazole (74.1%) and tetracycline (71.1%) and least resistant to ceftriaxone (23.4%), ciprofloxacin (22.5%), and cefoxitin (19.1%). Salmonella isolates recovered from samples collected in Sichuan, Guangdong, and Fujian provinces were highly resistant to sulfisoxazole at 87.1, 92.7, and 88.3%, respectively. Isolates from Beijing City, Shanghai City, and Henan Province were most frequently resistant to tetracycline, ampicillin, and trimethoprimsulfamethoxazole at 72.2, 95.8, and 94.8%, respectively.
Log M P N /c a rc a s s
FIGURE 1. Percentage bar chart illustrating the log mostprobable-number (MPN) distribution of Salmonella isolates on broiler chicken carcasses at retail stores in China.
The individual Salmonella antimicrobial resistance profiles by market type and market temperature are shown in Table 4. No significant differences (.P > 0.05) were found among the percentages of isolates resistant to individual antimicrobial agents except for kanamycin compared by market type, i.e., wet market isolates were less resistant (32.5%) than were isolates from small (42.7%) and large (43.1%) markets. However, significant differences by storage temperature were found for isolates resistant to kanamycin, i.e., higher resistance in frozen chicken isolates (51.6%) than in isolates from chilled (40.4%) and ambient (30.6%) chickens, and to ceftiofur, i.e., higher resistance in frozen chicken isolates (38.1%) than in isolates from chilled (24.2%) and ambient (21.3%) chickens (Table 4). The percentage of resistant isolates (« = 779) to each antimicrobial agent by serotype (five most common serotypes and others) are shown in Table 5. These other serotypes represented 14.8% of the total isolates analyzed for antimicrobial susceptibility. Salmonella Indiana isolates were significantly more resistant to each of the 15 antimicrobial agents than were all other serotypes. In general, the five most common serotypes had relatively high resistance to tetracycline, sulfisoxazole, nalidixic acid, ampicillin, trimethoprim-sulfamethoxazole, and amoxicillin-clavulanic acid. The overall distributions of multidrag-resistant Salmo nella phenotypes are shown in Figure 2. Of the 779 isolates tested, approximately 4% were susceptible to all 15 antimicrobial agents, 45.1% were resistant to 1 to 5 agents, 28.8% were resistant to 6 to 10 agents, and 22.1% were resistant to 11 to 15 agents. Overall, no significant differences (P > 0.05) were found among multidragresistant isolates by retail market and storage temperature. Among the five most common Salmonella serotypes, all (100%) Indiana and Thompson isolates were resistant to at least one antimicrobial agent, and 97.9, 94.1, 97.8, and 93% of Enteritidis, Typhimurium, Agona, and other serotypes, respectively, were resistant to at least one antimicrobial agent. Also among these five most common Salmonella serotypes, approximately 63, 59, 57, 18, and 9% of Agona, Enteritidis, Typhimurium, Thompson, and Indiana isolates, respectively, were resistant to 1 to 5
898
J. Food Prot., Vol. 77, No. 6
YANG ET AL.
TABLE 3. Salmonella serotypes isolated from raw chicken carcasses collected at three types of retail markets under three storage conditions in China No. (%) of positive samples Storage temp"
Market type Total
Wet
Small
Serotype
(n = 1,094)
(n = 400)
(n = 342)
Enteritidis Indiana Typhimurium Agona Thompson Derby Rissen Bsilla Heidelberg Blockley Edinburg Infantis Gueuletapee Shubra Bovismorbificans Corvallis Saintpaul Duesseldorf Albany Potsdam II Hadar Baiboukoum Lockleaze Weltevreden Chartres Haardt Kentucky Meleagridis Stanley Gatuni Newlands Tshiongwe Uppsala Abony Athinai Ball Djugu Dublin Essen Hillingdon Kingston Paratyphi A Anatum Arapahoe Blegdam Bonn Hidalgo Kallo Kouka London Remo Nitra Sangera Schwarzengrund
210 (19.2) 166 (15.2) 160 (14.6) 78 (7.1) 72 (6.6) 52 (4.8) 25 (2.3) 24 (2.2) 22 (2.0) 20 (1.8) 19 (1.7) 19 (1.7) 15 (1.4) 11 (1.0) 10 (0.9) 9 (0.8) 9 (0.8) 9 (0.8) 8 (0.7) 8 (0.7) 8 (0.7) 7 (0.6) 6 (0.5) 6 (0.5) 6 (0.5) 5 (0.5) 5 (0.5) 5 (0.5) 5 (0.5) 5 (0.5) 4 (0.4) 4 (0.4) 4 (0.4) 4 (0.4) 3 (0.3) 3 (0.3) 3 (0.3) 3 (0.3) 3 (0.3) 3 (0.3) 3 (0.3) 3 (0.3) 3 (0.3) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2) 2 (0.2)
81 (20.3) 57 (14.3) 60 (15.0) 38 (9.5) 11 (2.8) 18 (4.5) 8 (2.0) 15 (3.8) 6 (1.5) 3 (0.8) 4 (TO) 11 (2.8) 6(1.5) 4(1.0) 1 (0.3) 6(1.5) 2 (0.5) 6 (1.5) 7 (1.8) 3 (0.8) 4(1.0) 5 (1.3) 2 (0.5) 5 (1.3) 1 (0.3)
59 (17.3) 60 (17.5) 47 (13.7) 16 (4.7) 36 (10.5) 15 (4.4) 14 (4.1) 2 (0.6) 8 (2.3) 2 (0.6) 6 (1.8) 6(1.8) 2 (0.6) 2 (0.6) 5 (1.5) 1 (0.3) 2 (0.6) 2 (0.6) 1 (0.3) 4(1.2) 3 (0.9) 2 (0.6) 3 (0.9)
3 (0.8) 3 (0.8) 1 (0.3) 4 (1.0) 1 (0.3) 1 (0.3) 2 (0.5) 1 (0.3) 3 (0.8) 2 (0.5)
4 (1.2) 5 (1.5) 2 (0.6) 2 (0.6) 4 (1.2) 1 (0.3) 4 (1.2) 1 (0.3) 2 (0.6) 1 (0.3)
Ambient (n = 346)
Chilled (n = 555)
70 (19.9)* 49 (14.0)* 53 (15.1)* 24 (6.8)c 25 (7.1)c 19 (5.4) 3 (0.9) 7 (2.0) 8 (2.3) 15 (4.3) 9 (2.6) 2 (0.6) 7 (2.0) 5 (1.4) 4(1.1) 2 (0.6) 5 (1.4) 1 (0.3)
69 (19.9) 56 (16.2) 58 (16.8) 20 (5.8) 12 (3.5) 20 (5.8) 9 (2.6) 14 (4.0) 2 (0.6) 3 (0.9) 4 (1.2) 11 (1.0) 3 (0.9) 4 (1.2) 1 (0.3) 6(1.7) 1 (0.3) 6 (1.7) 7 (2.0) 2 (0.6) 4 (1.2) 2 (0.6)
109 (19.6) 76 (13.7) 82 (14.8) 27 (4.9) 44 (7.9) 29 (5.2) 11 (2.0) 10 (1.8) 12 (2.2) 9(1.6) 7 (L3) 7 (1.3) 11 (2.0) 5 (0.9) 5 (0.9) 3 (0.5) 8 (1.4) 3 (0.5) 1 (0.2) 5 (0.9) 4 (0.7) 5 (0.9) 6(1.1) 3 (0.5) 5 (0.9) 3 (0.5) 5 (0.9) 2 (0.4) 2 (0.4) 5 (0.9) 3 (0.5)
1 (0.3) 1 (0.3) 1 (0.3) 1 (0.3) 1 (0.3) 3 (0.9) 2 (0.6) 1 (0.3) 2 (0.6)
2 (0.6) 1 (0.3) 1 (0.3)
2 (0.6) 1 (0.3)
2 (0.6) 3 (0.9) 1 (0.3) 4(1.2) 1 (0.3) 1 (0.3) 2 (0.6) 1 (0.3)
1 (0.3) 1 (0.3) 1 (0.3)
3 (0.9) 2 (0.6) 2 (0.6) 1 (0.3)
1 (0.3)
1 (0.3) 2 (0.6)
1 (0.3)
1 (0.3) 2 (0.6)
1 (0.3)
1 (0.3)
1 (0.3) 2 (0.6) 2 (0.6)
1 (0.3) 1 (0.3)
32 (16.6)* 34 (17.6)* 20 (10.4)* 31 (16.1)c 16 (8.3)c 3 (L6) 5 (2.6) B (4.1) 8 (4.1) 8 (4.1) 1 (0.5) 1 (0.5) 2(1.0) 4(2.1)
1 (0.5)
1 (0.5) 2 (LO) 1 (0.5)
1 (0.5) 1 (0.5)
1 (0.3) 1 (0.3) 1 (0.3) 1 (0.3)
1 (0.2) 1 (0.2) 1 (0.2) 3 (0.5) 3 (0.5) 1 (0.2) 2 (0.4) 1 (0.2) 2 (0.4) 1 (0.2) 2 (0.2)
2 (LO) 1 (0.5)
1 (0.5)
2(1.0) 1 (0.3)
1 (0.3) 1 (0.3) 2 (0.6)
1 (0.3)
4 (0.7) 2 (0.4) 1 (0.2) 1 (0.2) 2 (0.4)
Frozen
(n = 193)
3 (1.6) 2 (0.6)
1 (0.3)
1 (0.3) 1 (0.3) 1 (0.3)
Large (n = 352)
1 (0.2) 2 (0.4) 1 (0.2) 2 (0.2) 1 (0.2)
1 (0.5)
J. Food Prot., Vol. 77, No. 6
SALMONELLA COUNTS ON RAW POULTRY IN CHINA
899
TABLE 3. Continued No. (%) of positive samples Market type
Serotype Aberdeen Augustenborg Bellevue Bessi Chinco Choleraesuis Col indale Fillmore Gallinarum Glostrup Kaapstad Leer Lindenburg Massenya Montevideo Nchanga Newport Orion Pomona Schwerin Tinda Ughelli Wangat
Storage temp"
Total
Wet
Small
Large
Ambient
(n = 1,094)
Chilled
(n = 400)
Frozen
(n = 342)
(n = 352)
(n = 346)
(n = 555)
in = 193)
1 1 1 1
( 0 . 1) ( 0 . 1)
(0 . 1) (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1)
1 ( 0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1) 1 (0 . 1)
1 (0.3) 1 (0.3) 1 (0.3)
1 (0.3) 1 (0.3) 1 (0.3) 1 (0.3)
1 (0 .2)
1 (0.3) 1 (0.3) 1 (0.3)
1 (0 .2) 1 (0 .2)
1 (0.3)
1 (0.5)
1 ( 0 .2 )
1 (0.3) 1 (0.3) 1 (0.3) 1 (0.3)
1 (0 .2) 1 (0 .2) 1 (0 .2) 1 ( 0 .2 ) 1 (0.3)
1 (0.2)
1 (0.3) 1 (0.3)
1 ( 0 .2 ) 1 (0.5)
1 (0.3) 1 (0.3)
1 (0.3) 1 (0.3) 1 (0.3)
1 ( 0 .2) 1 ( 0 .2) 1 ( 0 .2)
1 (0.3) 1 (0.3) 1 (0.3)
1 (0.5)
1 (0.3) 1 (0.3)
1 (0.3) 1 (0.3)
“ Ambient chickens were stored at 20 to 30°C, chilled chickens were stored at 4 to 10°C, and frozen chickens were stored at - 1 0 to -2 0 °C . Differences among percentage of Salmonella -positive samples by market type or storage temperature were not significant IP > 0.05). ‘ Differences among percentage of Salmonella -positive samples by market type or storage temperature were significant (P < 0.05). These differences were based on a likelihood ratio chi-square test of the differences in risk by market type and storage temperature for the top five most prevalent serotypes.
antimicrobial agents; 44, 33, 30, 22, and 18% of Thompson, Enteritidis, Typhimurium, Agona, and Indiana isolates, re spectively, were resistant to 6 to 10 agents; and 74, 38, 13, 7, and 6% of Indiana, Thompson, Agona, Typhimurium, and Enteritidis isolates, respectively, were resistant to 11 to 15 agents. High percentages (91 and 82%) of isolates of two serotypes (Indiana and Thompson, respectively) were resistant to 6 to 15 antimicrobial agents. DISCUSSION In our previous study (35), 52.2% of 1,152 retail chickens were Salmonella positive, whereas in the present study only 43.3% of 300 retail chickens were Salmonella positive. The overall mean Salmonella counts were 1.7 log MPN per carcass (95% Cl, 1.6 to 1.8 log MPN per carcass). This mean is similar that reported on postchilled chickens at processing plants in the United States (1.7 log MPN per carcass) (28). When comparing Salmonella count data from China (Fig. 1) and the United States (28), Salmonella counts of 1.0 to 2.0 log MPN per carcass were found in 34% of carcasses in the present study compared with 3.7% of U.S. carcasses (28); counts of 2.1 to 3.0 log MPN per carcass were found in 7.3% of our samples compared with 1.2% of U.S. carcasses; and counts of 3.1 to
3.8 log MPN per carcass were found in 2.3% of our samples compared with 0.27% of U.S. samples. Compared with other countries, the mean MPN for Salmonella found on the chickens in China was slightly higher than that on fresh retail chicken in The Netherlands where 89% of fresh (nonfrozen) chickens had less than 1 log MPN per carcass (10) and in Cambodia where Salmonella counts were 3 to 4 log CFU/g in 22.4% of the poultry samples (n = 152) (21). Differences in type of chicken meat sample, sampling times, sample number, slaughterhouse and retail store sani tation, and Salmonella quantification and isolation methods should be taken into consideration when comparing findings among studies. Salmonella Enteritidis was the most frequently identi fied serotype in China overall and in Beijing City and Shaanxi, Henan, Guangxi, and Fujian provinces. This serotype was found on chickens from various market types (wet, large, and small) and on chickens stored at various temperatures (ambient, chilled, and frozen). Salmonella Enteritidis was also identified as the predominant serotype in poultry products in other studies (1, 2, 16, 19, 34) and as one of the most common serotypes that cause human salmonellosis in many countries (7, 17, 18, 26). According to a laboratory-based survey of nontyphoidal Salmonella
900
J. Food Prot., Vol. 77, No. 6
YANG ET AL.
TABLE 4. Anitimicrobial resistance of Salmonella isolates from raw poultry collected at three types of retail markets under three storage conditions in Chinaa No. (%) of resistant isolates Market type
Sulftsoxazole Tetracycline Ampicillin Nalidixic acid Trimethoprimsulfamethoxazole Amoxicillin-clavulanic acid Chloramphenicol Kanamycin Streptomycin Gentamicin Ceftiofur Amikacin Ceftriaxone Ciprofloxacin Cefoxitin
Wet
(n = 779)
Large (n = 248)
Small (n = 260)
(n = 271)
577 554 481 479
185 173 154 154
191 196 169 158
201 185 158 167
Total Antimicrobial agent
Storage temp
P value"
Ambient (.n = 235)
Frozen (n = 126)
Chilled (n = 418)
P value
96 86 82 88
307 305 262 254
(74.2) (68.3) (58.3) (61.6)
0.982 0.489 0.693 0.937
174 163 137 137
(73.4) (73.0) (62.7) (60.8)
0.822 0.841 0.886 0.868
441 (56.6) 139 (56.0) 162 (62.3) 140 (51.7)
0.113
125 (53.2) 107 (84.9) 209 (50.0)
0.492
116 (46.8) 126 (48.5) 117 (43.2) 100 (40.3) 127 (48.8) 102 (37.6) 107 (43.1) 111 (42.7) 88 (32.5) 101 (40.7) 98 (37.7) 85 (31.4) 90 (36.3) 97 (37.3) 86 (31.7) 69 (27.8) 71 (27.3) 59 (21.8) 69 (27.8) 67 (25.8) 49 (18.1) 65 (26.2) 58 (22.3) 59 (21.8) 62 (25.0) 55 (21.2) 58 (21.4) 49 (19.8) 52 (20.0) 48 (17.7)
0.638 0.054 0.049 0.360 0.447 0.120 0.181 0.472 0.495 0.670
106 (45.1) 94 (40.0) 72 (30.6) 72 (30.6) 71 (30.2) 50 (21.3) 46 (19.6) 50 (21.3) 51 (21.7) 42 (17.9)
(74.1) (71.1) (61.7) (61.5)
359 (46.1) 329 (42.2) 306 (39.3) 284 (36.5) 273 (35.0) 199 (25.5) 185 (23.7) 182 (23.4) 175 (22.5) 149 (19.1)
(74.6) (69.8) (62.1) (62.1)
(73.5) (75.4) (65.0) (60.8)
(74.0) (69.4) (58.3) (58.3)
69 82 65 55 51 48 37 43 35 31
(76.2) (68.3) (65.1) (69.8)
(54.8) (65.1) (51.6) (43.7) (40.5) (38.1) (29.4) (34.1) (27.8) (24.6)
184 153 169 157 151 101 102 89 89 76
(44.0) (36.6) (40.4) (37.6) (36.1) (24.2) (24.4) (21.3) (21.3) (18.2)
0.460 0.524 0.020 0.298 0.227 0.008 0.483 0.052 0.665 0.249
a p values are based on a likelihood ratio chi-square test of the differences in risk by market type and storage conditions.
infections in humans in China conducted by the Chinese Center for Disease Control and Prevention, Salmonella Enteritidis was the most frequently identified serotype (31%, n = 662 isolates) (24). Data from other regions have revealed that Salmonella Enteritidis is the most common serotype among human salmonellosis cases in Africa, Asia,
Europe, Latin America, and the Caribbean and is the second common serotype in North America (13). Other Salmonella serotypes, including Indiana, Typhimurium, Agona, Thompson, and Derby, were repeatedly recovered from chicken meat in the present study. Our findings are in agreement with those described previously, indicating that
TABLE 5. Antimicrobial resistance by most common serotype of Salmonella isolates from raw chicken collected in China No. (%) of resistant Salmonella isolates
Antimicrobial agent
Tetracycline Sulftsoxazole Nalidixic acid Ampicillin Trimethoprimsulfamethoxazole Amoxicillinclavulanic acid Chloramphenicol Kanamycin Gentamicin Streptomycin Ceftiofur Amikacin Ciprofloxacin Ceftriaxone Cefoxitin
Typhimurium (n = 102)
Agona (« = 46)
Thompson (n = 55)
30 37 23 22
70 66 39 66
Enteritidis
Indiana
(n = 146)
(n = 114)
178 189 197 154
(62.7) (66.6) (69.4) (54.2)
104 (91.2) 102 (90.3) 106 (93.0) 105 (92.1)
90 95 79 68
119 (41.9)
105 (92.9)
65 (48.5)
23 (50.0)
111 (39.1) 84 (29.6) 82 (28.9) 64 (22.5) 76 (26.8) 36 (12.7) 42 (14.8) 23 (8.1) 44 (17.9) 44 (15.5)
91 (79.8) 97 (85.1) 93 (81.6) 93 (81.6) 83 (73.5) 79 (69.9) 61 (53.5) 93 (81.6) 73 (64.0) 42 (36.8)
44 (32.8) 40 (29.9) 42 (31.3) 38 (28.4) 57 (42.5) 19 (14.2) 12 (9.0) 22 (16.4) 12 (10.4) 9 (6.7)
16 (34.8) 14 (30.4) 11 (23.9) 6 (13.0) 14 (30.4) 7 (15.2) 4 (8.7) 10 (21.7) 6 (15.4) 8 (17.4)
(67.2) (70.9) (59.0) (50.8)
(65.2) (80.4) (50.0) (47.8)
Other serotypes Cn = 316)"
P value*
(71.3) (76.5) (58.3) (57.4)
