Prevalence and identification of Listeria species
Out of 482 samples tested, 37 (7.67%; 95%: CI: 5.46, 10.42) were found to be positive for the Listeria species. The prevalence of Listeria species in Ambo town (10.98%) was the highest when compared to Holeta (6.89%) and Bako 3(3.90%) towns (Table 1).
Table 1 The overall prevalence of Listeria species isolated from dairy cow milk and milk products in study towns
Study areas
|
Sample type
|
Listeria species
|
|
|
|
Number tested
|
Prevalence (%)
|
95% CI
|
Ambo
|
Cow milk
|
127
|
14 (11.02)
|
1.59-4.82
|
|
Curdle milk
|
10
|
1 (10.00)
|
0.00-1.15
|
|
Bulk tank milk
|
17
|
3 (17.65)
|
0.12-1.80
|
|
Cottage cheese
|
10
|
0 (0.00)
|
0.00-1.15
|
|
Subtotal
|
164
|
18 (10.98)
|
2.22-5.83
|
Holeta
|
Cow milk
|
190
|
12 (6.32)
|
1.29-4.30
|
|
Curdle milk
|
15
|
1(6.67)
|
0.00-1.15
|
|
Bulk tank milk
|
7
|
1 (14.29)
|
0.00-1.15
|
|
Cottage cheese
|
20
|
2 (10.00)
|
0.00-1.49
|
|
Subtotal
|
232
|
16 (6.90)
|
1.90-5.33
|
Bako
|
Cow milk
|
67
|
1 (1.49)
|
0.00-1.15
|
|
Curdle milk
|
5
|
0 (0.00)
|
|
|
Bulk tank milk
|
9
|
1 (11.11)
|
0.00-1.15
|
|
Cottage cheese
|
5
|
1 (20.00)
|
0.00 – 1.15
|
|
Subtotal
|
86
|
3 (3.49)
|
0.12- 1.80
|
Overall
|
|
482
|
37 (7.68)
|
5.46 -10.42
|
CI, Confidence Interval
The highest prevalence of Listeria species (15.15%; 95% CI: 5.11-31.90) and L. monocytogenes (9.09; 95% CI: 1.92-24.33) was detected in bulk tank milk. Similarly, the lowest prevalence of Listeria species (6.67%; 95% CI: 0.82-22.07) and L. monocytogenes (0.00; 95% CI: 0.00-1.15) was found in curdling milk (Table 2). The overall prevalence of Listeria species from milk and milk products was 7.68% (37/482) and from which, the highest prevalence was recorded for Listeria grayi (2.49%) and the lowest for L. weshimeri (0.62%). The prevalence of L. monocytogenes, L. ivanovii, and L. seeligeri was 2.28%, 1.24%, and 1.04% respectively.
Table 2 Comparison of prevalence of Listeria species from different samples
Sample type
|
No.
tested
|
Listeria species
|
L. monocytogenes
|
No. Pos.
|
% Prevalence (95% CI)
|
No. Pos.
|
% Prevalence (95% CI)
|
Bulk tank milk
|
33
|
5
|
15.15 (5.11-31.90)
|
3
|
9.09 (1.92-24.33)
|
Cottage Cheese
|
35
|
3
|
8.57 (1.80-23.06)
|
1
|
2.85 (0.00-1.15)
|
Cow level raw milk
|
384
|
27
|
7.03 (4.68-10.07)
|
7
|
1.82 (0.58-2.96)
|
Curdle milk
|
30
|
2
|
6.67 (0.82-22.07)
|
0
|
0.00 (0.00-1.15)
|
Total
|
482
|
37
|
9.64 (6.88-13.04)
|
11
|
2.28 (1.14-4.04)
|
CI, Confidence Interval; No., number; pos, sample positive for L. monocytogenes
Risk factor analyses
The farm-level prevalence of Listeria species was significantly high (p<0.05) in Holeta (63.64%) than Bako (18.75%) town, in large herd size (85.71%) than smallholder (12.50%), in large farms (58.33%) than small farms (12.50%), and in intensively (50.00%) than extensively (8.33%) managed cows (Table 3). The following independent variables have univariable p-values less than 0.25 hence are potential variables for inclusion into the multivariable model: town, management, farm hygiene, education level, herd size, number of lactating cows, farm size, time of udder washing, materials used for udder drying, and source of water for udder and hand washing. Among these variables, the multicollinearity matrix showed that the following are collinear. Management vs town (r=0.69), education vs management (r=0.51), number of lactating cow’s vs farm hygiene (r=0.58), number of lactating cow’s vs herd size (r=0.91), farm size vs herd size (r=0.90), and farm size vs number of lactating cows (r=0.83). Among the collinear variables management, farm hygiene, and herd size were selected for inclusion into the multivariable model due to biological plausibility. Moreover, the time of udder washing, materials used for udder drying, and source of water for udder and utensil washing were also included in the multivariable model (Table 3). Finally, after running the full model, herd size was removed from the model due to confounding identified through the change in OR >30% between univariable and multivariable models.[28] The Hosmer-Lemeshow goodness of fit test revealed that the model predicts or fitted the data well (HLX2=5.06; p=0.7511, sensitivity=35.3%, specificity =83.9%, positive predicting value 54.6%, negative predicting value =70.3%, ROC=0.7581). However, none of the variables were independent and significant predictors of Listeria species isolation rate.
Table 3 Results of logistic regression analysis of farm-level potential risk factors for isolation of Listeria species
Risk factors
|
Categories
|
No.
tested
|
No. positive
(%)
|
Univariable
|
Multivariable
|
OR (95% CI)
|
P-value
|
OR (95% CI)
|
P-value
|
Town
|
Bako
|
16
|
3 (18.75)
|
1.0
|
-
|
|
|
Ambo
|
21
|
7 (33.33)
|
2.17 (0.46-10.20)
|
0.328
|
|
|
Holeta
|
11
|
7 (63.64)
|
7.58 (1.31-43.92)
|
0.024
|
|
|
Farm hygiene
|
Poor
|
23
|
6 (26.09)
|
1.0
|
|
1.0
|
-
|
Good
|
13
|
5 (38.46)
|
1.77 (0.41-7.58)
|
0.441
|
1.69 (0.27-10.46)
|
0.570
|
Moderate
|
12
|
6 (50.0)
|
2.83 (0.65-12.26)
|
0.164
|
3.53 (0.51-24.59)
|
0.203
|
Level of education of workers
|
College and above
|
14
|
4 (28.27)
|
1.0
|
|
|
|
Secondary
|
13
|
4 (30.77)
|
1.11 (0.21-5.80)
|
0.901
|
|
|
Elementary
|
15
|
5 (33.33)
|
1.25 (0.26-6.07)
|
0.782
|
|
|
Illiterate
|
6
|
4 (66.67)
|
5.0 (0.64-39.06)
|
0.125
|
|
|
Herd size
|
Smallholder (≤10)
|
24
|
3 (12.50)
|
1.0
|
-
|
|
|
Medium (10 – 50)
|
17
|
8 (47.06)
|
6.22 (1.33-29.01)
|
0.020
|
|
|
Large (≥ 50)
|
7
|
6 (85.71)
|
42 (3.67-481.03)
|
0.003
|
|
|
Farm size
|
Small
|
24
|
3 (12.50)
|
1.0
|
-
|
|
|
Large
|
24
|
14 (58.33)
|
9.8 (2.28-42.06)
|
0.002
|
|
|
No. of lactating cows
|
≤5
|
21
|
2 (9.52)
|
1.0
|
-
|
|
|
6 – 13
|
18
|
8 (44.44)
|
7.6 (1.35-42.80)
|
0.021
|
|
|
≥ 14
|
9
|
7 (77.78)
|
33.25 (3.90-283.45)
|
0.001
|
|
|
Management system
|
Extensive
|
12
|
1(8.33)
|
1.0
|
|
1.0
|
-
|
Semi-intensive
|
14
|
5(35.71)
|
6.11 (0.60-62.23)
|
0.126
|
2.93 (0.18-48.45)
|
0.452
|
Intensive
|
22
|
11(50.00)
|
11(1.21-100.39)
|
0.034
|
6.38 (0.54-75.95)
|
0.143
|
Training
on food safety
|
Yes
|
23
|
8(34.78)
|
1.0
|
|
|
|
No
|
25
|
9(36.00)
|
1.05 (0.32-3.45)
|
0.930
|
|
|
Washing cows
Udder before milking
|
No
|
7
|
2(28.58)
|
1.0
|
-
|
|
|
Yes
|
41
|
15(36.59)
|
1.44 (0.25-8.37)
|
0.683
|
|
|
Materials used for udder drying
|
Collective towel
|
13
|
2 (15.38)
|
1.0
|
-
|
1.0
|
-
|
Without towel/just with hand
|
10
|
4 (40.00)
|
3.67 (0.51-26.22)
|
0.196
|
3.72 (0.26-53.45)
|
0.333
|
Individual towel
|
25
|
11(44.00)
|
4.32 (0.79-23.68)
|
0.092
|
2.52 (0.31-20.52)
|
0.388
|
Source of water for
Udder and utensil washing
|
River
|
12
|
2(16.67)
|
1.0
|
-
|
1.0
|
-
|
Pipe
|
36
|
15(41.67)
|
3.57 (0.68-18.72)
|
0.132
|
1.53 (0.21-11.11)
|
0.672
|
Washing hands before milking
|
Yes
|
30
|
10(33.33)
|
1.0
|
-
|
|
|
No
|
18
|
7(38.89)
|
1.27 (0.38-4.29)
|
0.697
|
|
|
timetimeTTime Time of udder washing
|
Before milking
|
25
|
7 (28.00)
|
1.0
|
-
|
1.0
|
-
|
No washing
|
6
|
2 (33.33)
|
1.29 (0.19-8.67)
|
0.796
|
1.54 (0.16-14.73)
|
0.710
|
Before and after milking
|
17
|
8 (47.06)
|
2.29 (0.63-8.32)
|
0.210
|
2.29 (0.46-11.53)
|
0.314
|
Univariable logistic regression analysis of animal level risk factors showed that the risk of contamination of raw milk by Listeria species was 8.12 times higher in Ambo as compared to Bako town (p=0.045). Similarly, the likelihood of contamination of milk by Listeria species was 2.82 times higher in intensively managed cows as compared to extensively managed cows (p=0.031). Independent variables like the season, breed, age, parity, washing hands before milking, washing udder before milking, materials used for udder drying, and source of water for udder/hand washing were not significantly associated with the isolation of Listeria species from raw cow milk (p>0.05). None of the independent variables studied were collinear with each other (/r<0.5/).
Independent variables with univariable p<0.25 considered for the multivariable model include town, season, breed, management system, materials used for udder drying, and source of water used for udder/hand washing. The final multivariable logistic regression model revealed that the management system is an independent predictor of Listeria species isolation from milk (Table 4).
Table 4 Logistic regression analyses of potential risk factors for Listeria species isolation rate in study towns at cow level
Risk factors
|
Category
|
Univariable
|
Multivariable
|
OR (95%CI)
|
P-value
|
OR (95%CI)
|
P-value
|
Town
|
Bako
|
1.0
|
-
|
1.0
|
-
|
Holeta
|
4.45 (0.57-34.89)
|
0.155
|
3.42 (0.39-29.97
|
0.267
|
Ambo
|
8.18 (1.05-63.60)
|
0.045
|
5.87 (0.71-48.32)
|
0.100
|
Breed
|
Local
|
1.0
|
-
|
1.0
|
-
|
Cross
|
2.57 (0.58-11.23)
|
0.209
|
1.73 (0.33-9.01)
|
0.513
|
Jersey
|
2.61 (0.42-16.41)
|
0.305
|
2.64 (0.39-17.99)
|
0.323
|
Age in years
|
2 - 4
|
1.0
|
-
|
|
|
5 - 7
|
1.46 (0.55 – 3.87)
|
0.448
|
|
|
8 - 15
|
1.69 (0.52 – 5.45)
|
0.382
|
|
|
Seasons
|
Wet
|
1.0
|
-
|
1.0
|
-
|
Dry
|
1.68 (0.72-3.94)
|
0.232
|
1.04 (0.30-3.63)
|
0.951
|
Management system
|
Extensive
|
-
|
-
|
1.0
|
|
Semi-intensive
|
1.03 (0.34-3.19)
|
0.943
|
1.38 (0.37 – 5.20)
|
0.635
|
Intensive
|
2.82 (1.10-7.24)
|
0.031
|
3.38 (1.02 – 11.18)
|
0.046
|
Parity
|
Second
|
1.0
|
-
|
|
|
Fourth
|
1.09 (0.32 -3.69)
|
0.895
|
|
|
First
|
1.38 (0.49 – 3.84)
|
0.544
|
|
|
Third
|
1.65 (0.48 – 5.67)
|
0.427
|
|
|
Washing hands before milking
|
Yes
|
1.0
|
-
|
|
|
No
|
1.37(0.63-3.00)
|
0.429
|
|
|
Washing udder before milking
|
Yes
|
1.0
|
-
|
|
|
No
|
1.25 (0.45 -3.44)
|
0.668
|
|
|
Materials used for udder drying
|
Collective/shared/ towel
|
1.0
|
-
|
1.0
|
-
|
Individual towel
|
1.36 (0.57 – 3.23)
|
0.493
|
1.47 (0.52 – 4.11)
|
|
Just with hand/no towel
|
2.45 (0.81 – 7.40)
|
0.112
|
2.35 (0.63 – 8.71)
|
0.203
|
Source of water used for udder/utensil washing
|
River
|
1.0
|
-
|
1.0
|
-
|
Tap
|
2.09 (0.61 – 7.14)
|
0.238
|
1.70 (0.45 – 6.41)
|
0.436
|
Chi-square analysis of the association between potential risk factors and prevalence of Listeria species in bulk tank milk, curdled milk, and cottage cheese revealed that none of the factors considered were significantly associated (p>0.05) (Table 5).
Table 5 Results of Chi-square analysis of the association between the prevalence of Listeria species in bulk tank milk, curdled milk, and cottage cheese and potential risk factors for contamination
Variable
|
Categories
|
No. tested
|
No. positive
|
Prevalence (%)
|
Chi-square
|
P-value
|
Town
|
Holeta
|
42
|
4
|
9.52
|
0.0382
|
0.981
|
Bako
|
19
|
2
|
10.53
|
|
|
Ambo
|
37
|
4
|
10.81
|
|
|
Sample type
|
Curdle milk
|
30
|
2
|
6.67
|
1.3931
|
0.498
|
Cottage cheese
|
35
|
3
|
8.57
|
|
|
Bulk tank milk
|
33
|
5
|
15.15
|
|
|
Gender of handler
|
Female
|
66
|
6
|
9.09
|
0.2733
|
0.601
|
Male
|
32
|
4
|
12.50
|
|
|
Level of Education
|
Primary
|
23
|
1
|
4.35
|
1.2869
|
0.732
|
Illiterate
|
44
|
5
|
11.36
|
|
|
Secondary
|
25
|
3
|
12.00
|
|
|
College and above
|
6
|
1
|
16.67
|
|
|
Food safety information
|
No
|
32
|
3
|
9.38
|
0.0356
|
0.850
|
Yes
|
66
|
7
|
10.61
|
|
|
Received training on food safety
|
No
|
70
|
5
|
7.14
|
2.5057
|
0.113
|
Yes
|
28
|
5
|
17.86
|
|
|
Type of container used for handling
|
Stainless steel
|
31
|
3
|
9.68
|
0.0137
|
0.907
|
Plastic
|
67
|
7
|
10.45
|
|
|
Antimicrobial Susceptibility
In this study, twelve antimicrobial discs were tested against a total of eleven isolates of Listeria monocytogenes for antimicrobial susceptibility test. The isolates showed high resistance to both oxacillin (100%), amoxicillin (90.91), and ampicillin (72.73%) whereas the isolates showed 100% susceptibility against gentamycin and norfloxacin. Of the total 11isolates subjected for antimicrobial susceptibility test, 11(100%) exhibited resistance for oxacillin, 10 (90.91%) for amoxicillin, 9 (81.82%) for vancomycin. Also, 8(72.73%) and 7 (63.64%) of the isolates were susceptible to chloramphenicol and nitrofurantoin (Table 6).
Table 6 Results of antimicrobial susceptibility testing of L. monocytogenes isolates (N=11) from all samples in study towns
Antimicrobial
classes
|
Antimicrobials
|
Resistance
|
Intermediate
|
Susceptible
|
|
Number
|
%
|
Number
|
%
|
Number
|
%
|
Glycopeptide
|
Vancomycin
|
9
|
81.82
|
1
|
9.09
|
1
|
9.09
|
Aminoglycosides
|
Gentamycin
|
0
|
0
|
0
|
0
|
11
|
100
|
Cephem
|
Cefotaxime
|
5
|
45.45
|
3
|
27.27
|
3
|
27.27
|
β-lactams
|
Oxacillin
|
11
|
100
|
0
|
0
|
0
|
0
|
Ampicillin
|
8
|
72.73
|
0
|
0
|
3
|
27.27
|
Amoxicillin
|
10
|
90.91
|
0
|
0
|
1
|
9.09
|
Quinolones
|
Nalidixic acid
|
6
|
54.55
|
2
|
18.18
|
3
|
27.27
|
Norfloxacin
|
0
|
0
|
0
|
0
|
11
|
100
|
Nitrofuran
|
Nitrofurantoin
|
1
|
9.09
|
3
|
27.27
|
7
|
63.64
|
Macrolide
|
Azithromycin
|
3
|
27.27
|
4
|
36.36
|
4
|
36.36
|
Phenicol
|
Chloramphenicol
|
3
|
27.27
|
0
|
0
|
8
|
72.73
|
Tetracycline
|
Tetracycline
|
4
|
36.36
|
0
|
0
|
7
|
63.64
|
Multi-drug resistance
The majority of the L. monocytogenes isolates 9 (81.82%) showed MDR. Out of the 9 MDR L. monocytogenes isolates, two isolates showed resistance to 6 classes of drugs (Table 7).
Table 7 Multidrug resistance patterns of L. monocytogenes isolates (N=11)
Antimicrobial resistance pattern
(No of resistant isolates)
|
MDR
pattern
|
Percent of MDR isolates
|
OXC-NAL (1)
|
2
|
9.1
|
VAN-OXC-AMP-AMX (1)
|
2
|
9.1
|
VAN-OXC-AMX-TET (1)
|
3
|
9.1
|
VAN-OXC-AZM-CHL-AMX (1)
|
4
|
9.1
|
VAN-CTX-OXC-AMP-AMX (1)
|
3
|
9.1
|
OXC-AMP-NAL-AZM-AMX (1)
|
3
|
9.1
|
VAN-CTX-OXC-AMP-AZM-AMX (1)
|
4
|
9.1
|
VAN-OXC-AMP-NAL-AMX-TET (1)
|
4
|
9.1
|
VAN-CTX-OXC-AMP-NAL-AMX-TET (1)
|
5
|
9.1
|
VAN-CTX-OXC-AMP-NAL-CHL-AMX-TET (1)
|
6
|
9.1
|
VAN-CTX-OXC-AMP-NAL-NIT-CHL-AMX (1)
|
6
|
9.1
|
OXC, Oxacillin; NAL, Nalidixic acid; VAN, Vancomycin; AMP, Ampicillin; AMX, Amoxicillin; TET, Tetracycline, AZM, Azithromycin; CHL, Chloramphenicol; CTX, Cefotaxime, NIT, Nitrofurantoin; MDR, multidrug resistance
Out of the L. monocytogenes isolated from raw milk collected from individual cows (N=7), 5 (71.43%) isolates showed MDR and all isolates from bulk tank milk and cottage cheese were MDR (Table 8).
Table 8 Comparison of antimicrobial resistance profile of L. monocytogenes isolated from different sources
Sample Sources
|
No. of L.monocytogenes isolates
|
Sensitive to all
drugs
|
Intermediate
susceptibility
|
Resistance to Single class of drug
|
resistance to two classes of drugs
|
Multiple drug resistance (≥3 classes of drugs)
|
Cow
|
7
|
0(0.00)
|
4(57.14)
|
0(0.00)
|
2(28.57)
|
5 (71.43)
|
BTM
|
3
|
0(0.00)
|
2(66.67)
|
0(0.00)
|
0(0.00)
|
3 (100)
|
CC
|
1
|
0 (0.00)
|
0 (0.00)
|
0 (0.00)
|
0 (0.00)
|
1(100)
|
Total
|
11
|
0 (0.00)
|
6 (54.55)
|
0 (0.00)
|
2 (18.18)
|
9 (81.82)
|
BTM, Bulk tank milk; CC, cottage cheese; No., number