Lottery system was used to select the one in Modjo city from six (6) export abattoirs in the country for this study. The study was conducted from November 2012 to April 2013 at the export abattoir in Modjo town, Ethiopia. Modjo is the center of Lume District, eastern Showa administrative zone of Oromia Regional State. It is located 73 km from Addis Ababa, with altitude of 1777 meters above sea level. The average minimum and maximum temperature are 18oc and 28oc respectively .
At this abattoir, 500-1500 goats are slaughtered every day along with 200-600 sheep twice per week. The export abattoir where the study was conducted is well equipped with modern facilities and it is certified of International Organization for Standardization (ISO) 22000. As an export standard abattoir, there is implementation of HACCP practices for maintaining hygienic standards of the abattoir. Sheep and goat are slaughtered separately and the same personnel are involved in slaughtering both groups of animals. All slaughtering operations are performed on overhead rails. Animals bled using `Halal` methods. The skins are washed by pressurized water, cleaned and trimmed and finally kept in chilling room. In the abattoir, there are clean areas for bleeding, dressing, evisceration and meat inspection. Animals slaughtered in the abattoir are exported to Arab countries, in Middle East.
The study was conducted on apparently healthy male sheep and goats slaughtered in the export abattoir during the study period. Animals are originated from different parts of the country mainly from Geanear (Bale), Somali, Awash-Metehara, Jima, Ambo, Borena, Arbaminch and Bati (Wollo). Most of them were transported to the abattoir by open aired vehicles, and this study considers these animals starting from the lairage (Figure 1).
Figure 1 Catchment areas for sheep and goats slaughtered at the abattoir
A survey was conducted to determine the prevalence of E. coli O157:H7 on skin, feces, intestinal mucosal swab and carcasses of slaughtered animals and abattoir environment (carcass in contacts) particularly knives, water, hook and worker’s hand. Swab samples were collected from November 2012 to April 2013.
Systematic random sampling was used to select the sampled animals. Fecal samples and skin, carcass and intestinal mucosal swab samples were collected from each selected animal following their rail (tag) along the line of operation. Swab from abattoir environment, which are in contact with the carcass, were sampled once on each sampling day. Knives, hook, workers hands and the tap water were considered to be carcass in contacts. Water, which was used to wash the carcass, was sampled directly from the tap.
All samples were transported in icebox to Microbiology laboratory, College of Veterinary Medicine and Agriculture, Addis Ababa University (CVMA, AAU) and stored at 4°C until processed. All samples were processed in 12 hours interval. Culturing, isolation, identification and PCR were performed to determine the presence of E. coli O157:H7 in each sample and most effective drugs against the bacteria were selected after antibiotic sensitivity testing.
Sample Size Determination
The number of study animals was determined based on the expected prevalence of E. coli O157:H7 and the desired absolute precision according to the formula stated on Thrusfield ;
n= 1.962 Pexp(1-Pexp) / d2
Where:- n=required sample size; Pexp =Expected prevalence; d = desired absolute precision
Based on a previous study done in Modjo and Debre-zeit export and municipal abattoirs, the prevalence rates of E. coli O157:H7 in goat and sheep were 2% and 2.5%, respectively . Using these two expected prevalence, 95% confidence interval and 5% absolute precision; the number of sampled goats and sheep were estimated to be 31 and 38, respectively.
Skin swab samples were taken according to McEvoy et al. , by using 2x3 cm sterile cotton tipped swabs soaked in approximately 10ml of buffered peptone water (Oxoid Ltd., Hampshire, England). Skins were swabbed from the neck of animals over the line of bleeding before slaughtering near the bleeding area at an area of approximately 10x10 cm. Skin of the ventral midline part of the animal was also swabbed similarly at the mid line to determine its contact to the carcass during flaying (Figure 2). The shaft of the swab was then broken by pressing it against the inner wall of the test tube and disposed. Finally this was repeated but with a dry sterile cotton.
Approximately 25g of fecal samples were taken after complete evisceration directly by opening the rectum according to the method described by Elder et al. . The whole abdominal digestive organs were separated from the slaughtering line in a plastic bucket and the rectum was opened using a sterile surgical blade (Figure 2); the fecal sample was then put in to sterile universal bottle. Whereas for intestinal mucosal swab sampling, the distal colon was ligated and opened using a sterile surgical blade proximal to the rectum and the lumen was swabbed by using sterile swab. The swab was then introduced in to approximately 10 ml buffered peptone water in a sterile test tube.
The external parts of carcasses were swabbed from rump, midline and brisket area, just before chilling according to the methods described by McEvoy et al. . Sterile cotton swabs soaked in approx. 10 ml of buffered peptone water was used to rub against the carcass first horizontally then vertically. A second dry sterile cotton swab was also rubbed at exactly the same area. For the internal part of the carcass, where contacts with other carcass is not possible, the thoracic and the pelvic parts of both sides, through the evisceration opening, were swabbed using the same procedure as above (Figure 2). Disposable sterile gloves were used for each carcass and changed.
Similarly, abattoir utensils and other carcass in contacts such as knife, hook, water and workers hands were also swabbed as a sample using sterile cotton swabs soaked in approx. 10 ml of buffered peptone water. A pooled sample was taken from knives used for evisceration and carcass trimming on each sampling day. Only hooks used to hang the sampled carcass were also swabbed each sampling day as a pooled sample. Whereas abattoir workers whose hands have direct access to the washed carcass were swabbed for sampling, at their palm surface and fingers. Hand washing is practiced almost regularly. They wash their hands between each work which was practiced according to the abattoirs particular HACCP procedures. Moreover, 25ml of water sample was collected directly from the tap which is used for washing of the carcass.
For each and every sampling a sterile latex glove was used to avoid cross contamination and every procedure was done as aseptic as possible. All samples were transported in ice box to the laboratory.
Bacteriological sample processing
Fecal samples were measured for accuracy and placed in to sterile stomacher bag and a 1:9 ratio of a modified tryptone soya broth (Oxoid Ltd., Hampshire, England) containing 20 mg|l novobiocin (Sigma, Steinheim, Germany) (mTSB+n) was added in it. It was then put in to stomacher (Seward Stomacher 400, Seward, London, UK) for agitation at low speed for 30 seconds.
In to all other swab samples 90 ml of mTSB+n was added and homogenized using vortex mixer, and also for the 25 ml water sample 225ml mTSB+n was added just to keep the 1:9 ratio.
Isolation and identification of E. coli O157:H7
Microbiological samples for the isolation and identification of this bacterium were processed as described as follows.
Modified tryptone soya broth containing 20 mg/l novobiocin (Oxoid, Ltd, Hampshire, England) was used at 1:9 ratios as mentioned above, for selective enrichment of all the samples. Then, all the samples types were incubated at 41.5 oC for 24 h.
Isolation by immuno magnetic separation (IMS) and culturing of the isolates
After 24 h of incubation all enriched broth culture were subjected to IMS using Dynabeads anti-E. coli O157 (Dynal Biotech AS, ThermoFisher Scientific, Oslo, Norway) as follows. Both enriched broth culture and the paramagnetic beads were vortexed to be homogenized and 1ml of the enriched culture was put in to a sterile screw cupped eppendrof tube. A 20µl of resuspended paramagnetic beads (Dynal Biotech AS, ThermoFisher Scientific, Oslo, Norway) was then transferred in to the same eppendrof tube. Then each tube was briefly vortexed on the dynal mixer (Dynal MX4sample mixer) (Dynal Biotech AS, ThermoFisher Scientific, Oslo, Norway) at 20 rpm for 30 minutes at room temperature, for the bacteria to attach to antibody surface on the beads. The tubes were then put in to the manual magnetic particle concentrator (MPC-S) (Dynal Biotech AS, ThermoFisher Scientific, Oslo, Norway) with the magnetic strip in place, inverted 3 to 4 times and left to settle for about 5 minutes. It was then gently rotated for the magnetic beads to concentrate at the back of the tube. The cap of the tube carefully opened and the supernatant was discarded by carefully aspirating it with sterile fine tipped pipette, without touching the back wall of the tube. Then magnetic strip was removed and 1ml of phosphate buffered saline containing 0.05% tween 20 (PBST, Sigma chemicals Co, Saint Louis, USA) was added to each tube using another disposable fine tipped pipette. It was then inverted 3 times after the tubes were clothed, the magnetic strip replaced and the above step repeated at least twice. To prevent cross contamination the PBST was put in different small containers and for each sample and each material transferring new pipette tips were used. Finally the supernatant was aspirated; the magnetic strip was removed and about 100µl of PBST was added in each tube and mixed gently.
Around 50 µl of IMS bead and bacteria complex were streaked onto Sorbitol MacConkey agar (Difco, Becton Dickinson, Claix, France) containing 0.05 mg/l cefixime and 2.5 mg /l potassium tellurite (Dynal Biotech ASA, Oslo, Norway) (CT-SMAC). Culturing was carefully to obtain pure colonies and plates were incubated at 37oC for 20–24 h. The CT-SMAC agar plates were examined for the presence of non-sorbitol fermenting colonies [23, 24, 25, 26].
The non-sorbitol fermenting colonies on CT-SMAC appear as slightly transparent, almost colorless with a weak pale brownish appearance with a diameter of one mm [24, 23, 26]. Such colonies are sub cultured on CT-SMAC for further a confirmatory test.
Confirmatory test by latex agglutination
Latex agglutination was performed for confirmation of E. coli O157:H7 using latex kit (ThermoFisher Scientific, Oslo, Norway). The latex kit consists of four components: latex test reagent, latex control reagent, the positive controls and negative controls. The test reagent is latex particles sensitized with specific rabbit antibody against O157 antigen and the control reagent consists of latex particles sensitized with rabbit globulin. The positive and negative controls are suspension of inactivated E. coli O157:H7 cells and inactivated non-specific E. coli cells respectively.
The test was performed according to the manufacturer instructions (Oxoid Ltd, Hampshire, England). But first the latex kit was checked for its performance by using the control suspensions in the kit, the test was continued after the positive control reacts with the test latex showing positive result. A drop of test latex and 0.085% sterile saline water were dispensed in to the reaction card separately. A few presumptive colonies (an average of 2 colonies) of E. coli O157 were taken and emulsified in to the saline water on the latex card, then slowly mixed with the test latex and checked for agglutination within two minutes. Isolates showing visible agglutination by reacting with the test latex solution are again sub cultured for virulent gene identification.
Figure 2 Procedures followed during sample collection and processing
Determination of virulence genes by polymerase chain reaction
Multiplex PCR was conducted to assess the presence of virulence genes (stx1, stx2 and eaeA) methods described in Mora et al.  and Inat and Siriken . DNA extraction of isolates was conducted using boiling method. Briefly, each suspect colony was inoculated on CT-SMAC and incubated for 24 h at 37oC to get fresh colony. Few colonies were then selected and suspended separately in 100 μl of sterile distilled water in eppendorf tubes, after which they were boiled at 92.5oC for 17 min in a water bath. The tubes were then centrifuged at 13000rpm for 10 min, after which the supernatant containing the template DNA was transferred into Dnase/Rnase-free eppendorf tubes. The extracted DNA samples were stored at -20oC until use.
Detection of the stx1, stx2 and eaeA genes was performed according to the protocol indicated in Inat and Siriken  with slight modification. Thus, 2 μl of extracted DNA was used as a template in a reaction mixture with a final volume of 25 μl that contained 10 mM of each dNTP, 25 nM stx1 primer, 25 nM stx2 primers, 25 nM eaeA primer (Table 1), 1U of Taq DNA polymerase (Qiagen, Hilden, Germany) in 1× PCR buffer and 2 mM of MgCl2. Amplification of DNA was conducted using initial denaturation at 95oC for 3 min, 35 cycles of denaturation at 95oC for 20 sec, annealing at 58oC for 40 sec, extension at 72oC for 1 min, and final extension at 72ºC for 8 min.
For gel electrophoresis, the 10-μl amplicon mixture was loaded onto a 1.5% agarose gel. Electrophoresis was conducted at 125 V for 1 h. A 100 up to 1,000 bp molecular weight marker was used to identify the amplified products as a ladder, which was visualized by UV illumination.
Table 1 Primer sequences used in multiplex PCR for amplification of stx1, stx2 and eaeA genes
Antimicrobial susceptibility pattern
Antimicrobial susceptibility testing was performed following the standard agar disk diffusion method according to CSLI  using commercial antimicrobial disks (Oxoid Ltd, Hants, UK). The selected antimicrobials their symbols and inhibition zone size interpretations are listed in table 2.
Pure colonies, incubated for 6 hours in Tryptone Soya Broth (Oxoid Ltd, Hants, UK) were processed to a turbidity of 0.5 McFarland standards (approximately 3x108 CFU per ml) in a sterile saline solution. Then, they were inoculated on Muller-Hinton agar plates (Becton Dickinson company, Cockeysville USA) using sterile cotton swab, making sure that all the surface of the media is immersed with the bacterial suspension. Antibiotic discs (Oxoid Ltd, Hants, UK) were then dispensed and plates were incubated for 24 hours at 370c. Diameters of the zone of inhibition were measured and the results were classified as resistant, intermediate and susceptible according to CLIS . E. coli ATCC 25922 type strains were used as a positive control.
Table 2 Antimicrobials used, their symbols and inhibition zone size interpretation for Gram-negative enteric bacteria.
Data collection, management and analysis
The establishment of computer database and the necessary manipulations such as variable coding was performed using MS Excel (Microsoft® Excel® 2010, Microsoft Corporation; Santa Rosa, California, USA). The database was transferred to SPSS version 11.5 (IBM Corporation, New York, USA)  for analysis. Descriptive statistics such as proportions, standard deviations and 95 % confidence intervals were performed. Over all and sample-specific prevalence were determined by dividing the number of positive samples to the total number of samples examined. Difference among and between proportions of the groups with certain determinant factors was determined by chi-square (X2) test. ORs were calculated using univariable logistic regression to determine the degree of associations of carcass contamination with fecal, skin and carcass in contact surfaces’ status. A p-value <0.05 was considered indicative of a statistical significance difference.
Furthermore, Kappa statistics was performed to see whether there is agreement between fecal sample and intestinal mucosal as well as carcass inside and carcass outside swab E. coli O157: H7 status. Interpretation of the Kappa test was based on Rules-of-thumb for kappa: values less than 0.40 indicate low association; values between 0.40 and 0.75 indicate medium association; and values greater than 0.75 indicate high association between the two raters.