Assessment of the prevalence of Subclinical Mastitis through two on-farm tests in dual-purpose livestock system of Colombian Orinoquia

ABSTRACT

The assessment of the prevalence of Subclinical Mastitis (SCM) in dairy farms is essential to evaluate the health status of mammary gland. The aim of the study was to assess the prevalence of SCM in dual-purpose livestock systems in Arauca, Colombia, through the analysis the values found on-farm diagnostic California Mastitis Test (CMT) and Electric Conductivity (EC) test. Milk samples were taken from the individual mammary quarters of 481 cows. The cow level prevalence, per affected individual mammary quarters prevalence and the mammary quarter level prevalence were determined according to the values obtained for each test using two methods of analysis. An ANOVA was performed to determine the difference between prevalences. Cohen’s kappa coefficient was used for intertest agreement, and an analysis of sensitivity and specificity. The cow level prevalence was similar between the tests (CMT = 31.4%; EC = 29.7%). The quarter-level prevalence was lower with EC (11.4% vs 14.3%) (p < 0.05). The agreement test between CMT (0) and EC ranged between 0.20 and 0.25. Despite the low agreement found between the tests, they can be useful in determining the health status of the mammary gland in the dual-purpose livestock systems of the Colombian Orinoquia.

KEYWORDS:

• Electric conductivity
• health status
• level prevalence
• mammary gland
• on-farm tests

Introduction

In Colombia, dairy production is one of the main agroeconomic activities carried out, being recognized as the fourth country with the highest milk production in Latin America, with a per capita consumption of 143 liters/inhabitant/year (PROCOLOMBIA 2018). One of the problems that affects milking cows is the manifestation of mastitis which causes high economic losses to the owners due to the decrease in the quality and quantity of milk, increase in treatment costs, costs of veterinary services and costs of culling (Tomasinsig et al. 2010; Mera Andrade et al. 2017). This disease is relevant to public health, since contaminated milk (exposed to bacteria or viruses) can transmit zoonotic diseases such as tuberculosis, brucellosis and streptococcal pharyngitis (Wolter et al. 2004; Aguilar Aldrete et al. 2014).

Subclinical Mastitis (SCM) is an inflammatory process of the mammary gland in one or more quarters usually imperceptible in the appearance of the mammary system and in the organoleptic characteristics of the milk (Florio-Luis de Pineda et al. 2015). This pathological condition translates into a reduction in milk production in the affected mammary quarter that will depend on the lactation days and the dairy production potential of the animal (Ponce et al. 2010; Florio-Luis de Pineda et al. 2015). At the farm level, most animals affected by SCM go undetected in clinical evaluation, because the infection of the mammary gland is not visible (Mendoza et al. 2017). Cows are milked normally, because this condition does not lead to visible changes in the milk or udder, but does lead to altered milk composition and reduced production (García-Sánchez et al. 2018; Ruíz-García and Sandoval-Monzón 2018). The presence of SCM is difficult to eradicate since it is caused by multiple factors including management, production system, production level, as well as nutritional and infectious processes associated with both the cow and its environment (Andresen 2001; Ruegg 2017). Poor milking practices are the common cause of spread and prevalence of mastitis on farms (Pinzón et al. 2009). The presence of mastitis indicates a risk to human health mainly due to the consumption of raw milk (Murphy et al. 2016).

Under field conditions in different countries, a high prevalence of mastitis in cows has been reported since farmers do not perform periodic tests for SCM detection (Aguilar Aldrete et al. 2014). In Asian countries, 35.25% mammary quarter level and 45% animal level prevalence of mastitis are reported (Bachaya et al. 2011). In the Caribbean islands, mastitis was reported with prevalence of 60% herd (García-Sánchez et al. 2018). Studies carried out in South American countries indicate a general prevalence of SCM of 39.9% to 42% under hand milking (Ruiz et al. 2011; Gómez-Quispe et al. 2015). The control of SCM is carried out with antibiotics that bring with them economic losses and generate problems of resistance to antibiotics, constituting a public health problem (Acosta Moreno et al. 2017).

In Colombia, SCM has been the depressing factor affecting both the quantity and quality of milk; however, most cattle breeders ignore the problem or consider it a transitory effect given the little commitment to carry out sanitary processes for the control and prevention of the disease (Mendoza et al. 2017). Studies conducted in Colombia report prevalence of SCM 11.4% to 21.6% for total mammary quarters in dual-purpose livestock systems (Calderon et al. 2011; Mendoza et al. 2017), however, there is a need to evaluate the presence of SCM in dual-purpose livestock systems that are developed with Bos taurus-Bos indicus crosses in the Colombian Orinoquia.

In Arauca, Colombia, milk production comes mainly from the dual-purpose livestock system that is developed under tropical conditions with low inputs and technological levels. The dual-purpose livestock system corresponds to multiracial groups as a product of crossbreeding of zebu cattle with European breeds. The udder health status of milking animals is unknown, due to the scarce reports available in the area. Therefore, it is important to conduct studies that provide information on the level of infection caused by bovine mastitis (Medina and Montaldo 2003; Mera Andrade et al. 2017). The performance of microbiological examinations for the SCM detection is difficult due to problems for access to farms due to the lack of passable access roads, public order problems and distance from urban centres.

At the farm level, the California Mastitis Test (CMT) has been used for decades and is the most used test in field conditions for mastitis diagnosis in cattle (Deb et al. 2013; Badiuzzaman et al. 2015; Maldonado-Arias et al. 2022). The CMT is a rapid, practical and low-cost test with reliable results (Sanford et al. 2006; Bedoya et al. 2007; Gómez-Quispe et al. 2015). The test does not provide a numerical result, but rather a categorical result, so any result above a vestigial reaction is considered suspicious, presenting validity for the SCM diagnosis (AL-Edany et al. 2012; Hernández and Bedoya 2008).

Another widely used and highly efficient method for SCM diagnosis is the measurement of Electric Conductivity (EC), used since 1942 with different types of conductivity metres (Musser et al. 1998; Romero et al. 2012; Pérez-Ruano and Tarafa-Zambrana 2017). The EC test has been used in recent decades to monitor and diagnose SCM at the farm level (Saltijeral et al. 2004; Bedoya et al. 2007; Hernández and Bedoya 2008). The EC test detects milk changes due to the increase in chlorine and sodium ions that occur in inflammatory processes, which increases the electrolytic milk component (Medina and Montaldo 2003; Hernández and Bedoya 2008; Acosta Moreno et al. 2017).

The practical applicability of the CMT and EC tests on the farm allows a rapid diagnosis of SCM without the need to perform a microbiological culture that is expensive and takes a long time(Salamanca Carreño et al. 2023). Since the mentioned tests are used in mastitis diagnosis under field conditions, the aim of the study was to assess the prevalence of SCM in dual-purpose livestock systems in Arauca, Colombian Orinoquia, through the analysis of the values found by CMT and EC tests.

Material and methods

Study site

This research was carried out in Arauca, a tropical region of Orinoquia, in eastern Colombia (Latitude: 7° 5′ 5″ N; Longitude: 70° 45′ 32.7″ W) (Figure 1). The climatic regime corresponds to a period of drought (December-March) and a period of rain (April–November). Annual rainfall is less than 1500 mm, with an ambient temperature of 35°C; a relative humidity of 65% in March, 85% in June and July (IDEAM 2000; Arauca 2022).

Figure 1. Location of the Arauca department (red colour), eastern Colombia.

The farms where the animals were sampled belong to small producers with a low technological level. The farms were chosen for convenience after meeting with the Livestock Producers Association of the region. On farms, an average of 18 milking cows (range 7–57) were found. Farm sizes ranged from 30 to 100 hectares. Milking is done by hand with the presence of the calf (Figure 2(a)) and it is done under covered stables or in open environments. Sanitary control is scarce and the destination of most of the milk is for self-consumption, home or artisan processing, reception in collection centres or direct sale to consumers (Salamanca-Carreño et al. 2018). On most farm animal feed is based on extensive grazing systems with native grasses and legumes (Figure 2(b)). In some systems, Sodium chloride (white salt) is supplied and to a lesser extent mineralized salts are provided.

Figure 2. Small cattle breeders farms with manual milking with the calf presence (a) and cows from indefinite crosses Bos taurus-Bos indicus in extensive grazing in native pastures (b).

Evaluated animals

Individual milk samples from 1924 mammary quarter belonging to 481 multiracial dairy cows (indefinite Bos indicus × Bos taurus crosses, Figure 2(b)) were evaluated. The cows were located on 28 farms, ranging in age from 3 to 10 years, from 1 to 8 calvings and with 30 to 270 lactation days.

The cows that were being milked at the time of the visit were sampled. Only milking cows with functional mammary quarters and without antibiotic treatments during the last three months were included in the study.

Cow mammary quarters were sampled during morning milking (4:00 am to 6:00 am) from June 2021 to March 2022, following the available recommendations in the literature (Wolter et al. 2004). Mammary quarters were identified, and samples were taken in that order (Figure 3(a)).

Figure 3. Mammary quarters position for taking milk samples for the EC and CMT tests. EC = Electric Conductivity; CMT = California Mastitis Test (a); EC = Electric Conductivity (b); RP = Right Posterior; LP = Left Posterior; RA = Right Anterior; LA = Left Anterior.

California Mastitis Test (CMT)

The CMT is a hand test that measures the quantity of somatic cells in milk generated by inflammatory processes (Mellenberger and Roth 2000; NMC 2017). Somatic cells migrate from blood to milk due to infection response (Philpot and Nickerson 2001). The CMT identifies the inflammatory response based on the gel viscosity. The test contains a dye (bromocresol purple) that indicates pH changes that occur in milk, as a result of inflammation (Guízar and Bedolla 2008).

To sample each mammary quarter, 2 ml of milk was deposited in each receptacle of the plastic test paddle and mixed with 2 ml of CMT reagent (Mastit read) (Figure 3(a)), following the standard procedure (Mellenberger and Roth 2000; Guízar and Bedolla 2008; NMC 2017). The mixture was homogenized for 10 s with circular movements (Mellenberger and Roth 2000) and with a 45° inclination, the test results were read. Interpretation of results was based on the standard procedure following the CMT grades: 0 =  Negative; Trace = possible infection; + =  positive grade 1(infected); ++ = positive grade 2 (infected); +++ = positive grade 3 (infected) (Mellenberger and Roth 2000; Guízar and Bedolla 2008; (NMC 2017). In the study, all results with some degree to CMT reaction were counted as SCM positive (Gómez-Quispe et al. 2015).

Electric conductivity test (EC)

The EC test for SCM detection is based on the differences in salt concentration (chlorine and sodium ions) between infected and non-infected mammary quarters of the same cow (Romero et al. 2012; Zaninelli et al. 2014; Pérez-Ruano and Tarafa-Zambrana 2017).

For the EC test, the Draminski Mastitis Detector MD4 × 4Q equipment (Draminski S.A., Owocowa, PL) was used (Draminski 1987). The equipment has a platform with four measuring containers (cups) to deposit the ‘milk squirts’ and a measurement and reading block with a special liquid crystal display, where the results of the mammary quarters are read for three seconds. The equipment reflects the differences between mammary quarters (Figure 3(b)). This test was performed before the CMT since according to the equipment user instructions, the first milk squirts from each mammary quarter are required to provide a better result (Draminski 1987).

The results were interpreted according to the manufacturer’s instructions available in the ISO 9001 equipment user manual (Draminski 1987, p. 1) readings below 250 units indicate subclinical inflammation of the mammary quarter or a high risk of becoming acute (2) a difference greater than 40–50 units between the highest and lowest quarter scores of the examined cow indicates SCM. In general terms, it can be assumed that a result between 250 and 300 units indicates an intermediate phase between SCM and a healthy mammary quarter.

Example of EC results interpretation among mammary quarters:

Table

Right Posterior (RP):	370u
Left Posterior (LP):	310u
Right Anterior (RA):	380u
Left Anterior (LA):	370u

Interpretation: the mammary quarter affected with SCM is the left posterior (LP) (380u−310u = 70u), the other mammary quarters remain normal.

Prevalence determination

Prevalence refers to the number of cases of a disease or an infectious event that occurs in a given place and time. It is an indicator of existence or ‘stock’, since it considers all present cases, whether new or old (Manterola and Otzen 2015). The cow level prevalence of SCM (with at least one affected mammary quarter), quarter level prevalence and individual mammary quarters prevalence were processed according to the following mathematical formulas (Santivañez-Ballón et al. 2013; Gómez-Quispe et al. 2015; Ramírez Sánchez 2015):

• Cow level prevalence = (Number of positive cows/Total number of sampled cows) × 100

• Quarter level prevalence = (Total number of positive mammary quarters/Total number of sampled mammary quarters) × 100

• Individual mammary quarters prevalence = (Total number of positive mammary quarters per position/Total number of mammary quarters per position) × 100

Statistical analysis

The test results at the farm level were compiled and organized in Excel format for subsequent analysis according to the following phases:

First phase. To compare the prevalence using the two tests, two variables were generated: (a) Considering negative SCM values, those negative results obtained with the CMT test (CMT 0); and (b) the variable obtained with the EC test.

Frequency tables were estimated for the variables related to the number of positive cows (animals with any SCM positivity degree on any mammary quarter declared by the tests), number of positive quarters by cow (1 to 4) and positivity by mammary quarters (RP, LP, RA, LA) obtained in each method of analysis. This information was used to estimate SCM cow level prevalence (%) and individual mammary quarters (RP, LP, RA, LA), prevalence (%). Similarly, to identify the existence of statistical differences in the prevalence by mammary quarters between the studied method, a one-way ANOVA was performed, where the fixed effects were the analysed methods (CMT (0) and EC) which counted with 4 repetitions determined by the SCM prevalence (%) estimated in each mammary quarter (RP, LP, RA, LA).To validate ANOVA assumptions, residuals were subjected to Shapiro–Wilks normality test and the homoscedasticity between treatments was evaluated using the absolute residuals through the Levene test. When statistical differences were found, Tukey’s test was used for the means differentiation (p < 0.05).

Second phase. Confusion matrices were established to assess the concordance between the methods and to identify whether the EC test classified the animals as positive or negative in a similar way to the CMT (0) test. In this case, the values obtained with the CMT test were considered as the reference values. With this information, the following parameters were estimated: (a) sensitivity (probability that animals classified as positive for SCM with the CMT test, have also been classified positive by the EC test), (b) specificity (probability that animals classified as negative for SCM with the CMT test, have also been classified as negative by the EC test), (c) false negatives (probability that animals classified as positive for SCM with the CMT test, have been classified as negative by the EC test) and (d) false positives (probability that animals classified as negative for SCM with the CMT test, have been classified as positive by the EC test). The objective of the analysis is to achieve values close to 1 in sensitivity and specificity (agreement in the classification between tests) and close to zero in classification errors (false positives and false negatives). The formulas used to estimate the parameters are shown below (Tarabla and Signorini 2020):

Table

$\mathrm{Se}={\displaystyle \frac{\mathrm{TP}}{\mathrm{TP}+\mathrm{FN}}}$	$\mathrm{Sp}={\displaystyle \frac{\mathrm{TN}}{\mathrm{TN}+\mathrm{FP}}}$
$\mathrm{FN}={\displaystyle \frac{\mathrm{FN}}{\mathrm{FN}+\mathrm{TP}}}$	$\mathrm{FP}={\displaystyle \frac{\mathrm{FP}}{\mathrm{FP}+\mathrm{TN}}}$

Note: Se = Sensitivity; Sp = Specificity; TP = True Positives; FN = False Negatives; TN = True Negatives; FP = False Positives.

With this information Cohen’s kappa coefficient was also estimated to evaluate the agreement between the test, using the following formulas (McHugh 2012):

Table

$\mathrm{Po}={\displaystyle \frac{\mathrm{TP}+\mathrm{TN}}{n}}$	$\mathrm{Pe}={\displaystyle \frac{{\displaystyle \frac{\left(\mathrm{Cm}1\times \mathrm{Rm}1\right)}{n}+{\displaystyle \frac{\left(\mathrm{Cm}2\times \mathrm{Rm}2\right)}{n}}}}{n}}$
$\mathrm{KC}={\displaystyle \frac{\mathrm{Po}-\mathrm{Pe}}{1-\mathrm{Pe}}}$

Note: Po = observed agreement; Pe = expected agreement; TP = true positive; TN = true negative; Cm 1 and 2 = marginal column 1 and 2 respectively; Rm 1 and 2 = marginal row 1 and 2 respectively.

The interpretation of the Kappa coefficient is as follows: (<0.00), slight (0.00–0.20), fair (0.21–0.40), moderate (0.41–0.60), substantial (0.61–80), almost perfect (0.81–1.00) (McHugh 2012; Iraguha et al. 2017).

All data were processed in InfoStat software (InfoStat 2020).

Results

A total of 481 cows corresponding to 1924 mammary quarters were sampled. The distribution of the cows’ number with SCM-affected mammary quarters, obtained with the CMT and EC methods within the evaluated sample is shown in Table 1. A preprints.org has previously been published (Salamanca Carreño et al. 2023).

Table 1. Results of the number of cows with affected mammary quarters in the sample evaluated according to the CMT and EC methods in crossbred cows of dual purpose livestock system, Arauca, Colombia.

Method	Total	Affected mammary Quarter					Positive
		0	1^1	2^1	3^1	4^1
CMT (0)	481	330	85	30	14	22	151
EC	481	338	87	40	13	3	143

Note: CMT (0) = California Mastitis Test considering (0) values as negative result; EC = Electric Conductivity; 0 = Negative SCM result; ¹= Cows with 1,2,3 and 4 mammary quarters affected.

Results showed that 69% (n = 330) and 70% (n = 338) of the analysed cows did not present any mammary quarter affected, with the CMT (0) and EC test respectively (Table 1). Similarly, the results showed that CMT (0) and EC methods presented the greatest agreement in classifying the SCM degree among mammary quarters. Of the 481 studied mammary quarters, the percentage difference between CMT (0) and EC methods classifying 0, 1, 2 and 3 affected quarters was low, with values between 0.2 to 2.1%. However, a larger difference was observed when classifying animals with 4 mammary quarters affected (4%).

The SCM prevalence per cow and mammary quarters is shown in Table 2. The cow level prevalence indicates whether a cow presented the disease or if at least one of her mammary quarters was positive. The cow level prevalence values were 31.4% (151/481) and 29.7% (143/481) for the CMT (0) and EC tests, respectively. The CMT (0) method showed a higher prevalence in the RP (16.0%) and RA (14.8%) individual mammary quarters, while the EC method showed it in the LA (12.1%) and RP (11.9%) individual mammary quarters.

Table 2. SCM cow level prevalence (%), individual mammary quarters prevalence by position (%) (RP, LP, RA, LA), and quarter level prevalence according to CMT and EC test in crossbred cows of dual purpose livestock system, Arauca, Colombia.

Method	Cow level prevalence		RP		LP		RA		LA		Total Quarter	Positive Quarter	Mean ± S.E Quarter level Prevalence
	N	%	n	%	n	%	n	%	n	%
CMT (0)	151	31.4	77	16.0	64	13.3	71	14.8	63	13.1	1924	275	14.3 ± 0.57 a
EC	143	29.7	57	11.9	54	11.2	49	10.2	58	12.1	1924	218	11.4 ± 0.57 b

Note: CMT (0) = California Mastitis Test considering (0) values as negative result; EC = Electric Conductivity; RP = Right Posterior; LP = Left Posterior; RA = Right Anterior; LA = Left Anterior; S. E = Standard error; N = number of positive cows; n = number of mammary quarters positives by position. Different letters indicate statistical differences (p < 0.05).

The one-way ANOVA performed on the SCM prevalence (%) by mammary quarter showed significant differences between the analysed methods (p < 0.05), although the values were close: 14.3% (275/1924) for CMT (0) and 11.4% (218/1924) for EC (Table 2) (residual were normally distributed (p-value  = 0.3144) and method variances were homoscedastic (p-value  = 0.2134).

The values of sensitivity, specificity, false negatives and false positives between the EC and CMT (0) test are shown in Figure 4.

Figure 4. Estimates of sensitivity, specificity, false negatives, and false positives between the EC and CMT (0) test. RP = Right Posterior; LP = Left Posterior; RA = Right Anterior; LA = Left Anterior; PA = Positive animals.

Values of sensitivity and specificity ranged between 0.35 to 0.45 and 0.75 to 0.90, respectively, between the methods. These results suggest that both methods were similar, classifying cows and their mammary quarters with negative SCM values; however, when classifying cows and their mammary quarters as SCM positive, the concordance between the techniques was low. This was reflected in many classification errors, with false negatives ranging among 0.55–0.65, while false positives presented lower values between 0.10 and 0.25.

The Cohen’s kappa coefficient between the CMT (0) and EC methods ranged between 0.20 to 0.25. According to these values, it is indicated that there is a slight to fair agreement (McHugh 2012; Iraguha et al. 2017; Pérez-Ruano and Tarafa-Zambrana 2017).

Discussion

Comparable SCM estimation between CMT(0) and EC

In this study, the result of cow-level prevalence with CMT (0) was slightly higher compared to EC. The mammary quarter-level prevalence presented a statistical difference (p < 0.05) and was higher for CMT (0) compared to EC. A similar and lower prevalence was observed between individual mammary quarters with the EC test compared with CMT (0). This result is not consistent with those of another study, where the CMT and EC tests were used as farm diagnostic methods, in which the test results were not significant (p > 0.05) (Reyes Sánchez and Argüello Sánchez 2015). The values of cow level prevalence and mammary quarter level prevalence are lower than those reported in other studies in Colombia (54.6%) and 31.53%) in dual-purpose livestock systems (Calderon et al. 2011; Mendoza et al. 2017).

In the CMT (0) test, a higher prevalence in the RP and RA mammary quarter was found, while the EC test findings showed the highest prevalence in the LA and RP mammary quarters. According to observed on farms by the authors of this study, the above may be due to the fact that cows during their rest period take a sternal decubitus position frequently to the right side, allowing the mammary gland to be more vulnerable to environmental microorganisms infection.

The authors of this study observed on farms that manual milking usually begins from the two right mammary quarters; or in a crossed way starting with RP and LA mammary quarters. This way in which milking is carried out can affect the presence of mastitis in individual mammary quarters. The study reports that SCM prevalence is caused by milking hygiene problems, lack of education and incentives for milkers, as well as failures in the supply of adequate materials for a good performance of their work (Trujillo et al. 2011), the factors that were observed in this study.

The SCM prevalence at the cow level found in this study is lower compared to other reports (Bachaya et al. 2011; Ruiz et al. 2011; Gómez-Quispe et al. 2015; García-Sánchez et al. 2018), possibly because most of those studies were carried out with specialized dairy breeds, while our study was carried out in crossbred breeds (Bos taurus × Bos indicus), with the presence of the calf and milking is done by hand. On the other hand, milking with the presence of the calf causes a greater amount of milk to be sucked from the mammary gland, and due to the bactericidal effect of the saliva it can prevent the microorganisms proliferation (Sánchez-Herencia and Mamani-Mango 2022). Under farm conditions, it has been pointed out that the observed SCM prevalence by cow level and mammary quarter is due to the fact that cattle breeders do not practice periodic tests for disease detection (Bachaya et al. 2011). It is also indicated that the presence of SCM in dual-purpose systems (Bos taurus × Bos indicus) is common in cows with a high degree of European purity (Castillo et al. 2007).

The CMT test presents limitations such as the variation of interpretation by the operators, factors that influence the reading, such as, the time elapsed after combining the reagent with the milk, the samples homogenization, the time elapsed for the analysis and the reagent quality (Cepero et al. 2006). Hence, the CMT is a visual test for subjective evaluation, which despite its simplicity, requires training in order to reduce the variability between operators, which is the most frequent source of variation (Ruíz-García and Sandoval-Monzón 2018). On the other hand, the evaluation of the CMT results (which is the viscosity produced by the mixture of milk and reagent), varies depending on the amount of reagent in the mixture and is affected by the pH of the milk (Ruíz-García and Sandoval-Monzón 2018). The determination of reactive degrees to CMT constitutes a valid criterion to indicate an emergency situation of SCM in a herd due to the presence of high-risk pathogens. This test allows a rapid diagnosis in the farm and without many technical requirements (Cepero et al. 2006).

In a SCM detection study in different crossbred cows, it was found that 65.2% of the sampled animals were positive using EC, so cattle breeders can easily use the technique to determine the presence of SCM in their dairy cows (Shahid et al. 2011). In another study, they conclude that the EC test showed similarity to CMT in the detection of SCM and its reliability would increase even more when used together with the other diagnostic methods (Kaşikci et al. 2012).

Past studies indicate that the EC test is efficient in detecting SCM (Mohammadsadegh et al. 2007); and it has a similar behaviour to CMT test (Kaşikci et al. 2012; Kamal et al. 2014), likewise reliable results are reported in the diagnosis of SCM, when both tests are used (Saltijeral et al. 2004). In general, similar results are reported, when using the two tests SCM detection (Petzer et al. 2013). Thus, the main advantage of the EC test is that diagnostic results can be achieved with objective, immediate measurements and a relatively lower cost, compared to other mastitis detection methods (Romero et al. 2012).

Although it is necessary to know the factors influencing the results of the EC test and the small differences between a healthy gland and an infected one dictate that the conductivity metre must be simple and with automatic calibration (Ying et al. 2004); that is, that the results the EC test may be masked by the equipment calibration (Romero et al. 2012). The EC test is useful for early identification of animals with intramammary infection, it is inexpensive and shows immediate results (Elizalde et al. 2019). In several countries, portable equipment has been used for the determination of SCM through EC, evaluating the concordance levels with the CMT test (Pérez-Ruano and Tarafa-Zambrana 2017). EC is influenced by potassium, calcium, magnesium and other ions; factors other like breed, lactation stage, milking interval and milk composition may affect milk EC (Kamal et al. 2014). EC represents a possible indicator for the detection of SCM, however, some authors mention that genetic factors may have control in milk EC (Ilie et al. 2010). Despite the fact that EC is one of the most useful methods as a predictor of mastitis, obtaining variable results raises questions (Khatun et al. 2017).

CMT and EC test concordance

The low concordance between the methods was corroborated with the low Cohen’s kappa coefficient suggesting a slight to fair agreement (McHugh 2012; Iraguha et al. 2017). These results indicated that the cows declared positive were different between the methods. That is, despite the fact that the two tests present similar SCM prevalence, some cows that were positive for one test were not for the other. The Cohen’s kappa coefficient observed in this study, differed from those observed in other studies (0.36 to 0.77) for the same tests (Pérez-Ruano and Tarafa-Zambrana 2017; Ruíz-García and Sandoval-Monzón 2018). Other studies have reported a concordance value of 0.37 for the same tests and recommend that the tests can be used for the detection of SCM (Iraguha et al. 2017). In contrast, a study in Cuba comparatively evaluated the EC and CMT tests, and found a considerable concordance level between them (Pérez-Ruano and Tarafa-Zambrana 2017). The low concordance between the tests indicates that the methods did not show the same results for the SCM diagnosis; however, they help to identify the health status of the mammary gland.

The sensitivity and specificity obtained with the EC test with respect to the CMT (0) test, indicated that both methods presented a good agreement classifying cows and their mammary quarters as SCM negative; however, in the classification of positive cases, the concordance was low. These results can be attributed to the fact that the data were not balanced between positive and negative cases (there were more negative than positive cases). In this way, the probability that EC classified the animals and mammary quarters as SCM negative was higher than the classification of positive cases. The sensitivity values obtained in this study are different from the 0.97 value reported in another study in livestock systems; while the 0.93 specificity value is similar to what was found in the present study (Bedoya et al. 2007). Other authors have reported low sensitivity of the CMT and have compared the results with other tests, concluding that it was never greater than 0.60 (Kaşikci et al. 2012; Ruíz-García and Sandoval-Monzón 2018). Some studies reported a 50% false positive with the EC method (Elizalde et al. 2019), indicating lower sensitivity and specificity values compared with the results of the present study.

Finally, we consider that the age, calvings number and lactation days of the cows may be affecting the results (Ponce et al. 2010; Florio-Luis de Pineda et al. 2015), and it may also be associated with the way the two tests work; but the purpose of the study was to evaluate the practical applicability of the CMT and EC tests on farm for a fast diagnosis without the need to perform a microbiological culture that is expensive and takes a long time. Consequently, we only analysed if both tests are consistent in terms of the ability to detect SCM in cattle. Future studies are expected to determine the factors that affect the SCM presence.

Conclusions

The CMT (0) and the EC tests are able to assess the prevalence of SCM in the evaluated systems, although, compared to CMT, the EC test showed a lower SCM prevalence. The EC test for the diagnosis of SCM in small producer farms is a feasible alternative. It is low cost and quick to perform, without the effects of visual subjectivity of CMT that can occur in operators. The CMT (0) and the EC methods showed similar results classifying the number of mammary quarters affected according to the SCM scale, however, the two tests were not consistent in detecting the same animals with or without the infection, indicating that some cows that were positive for one test were not to the other. Despite the low agreement found between the tests, they can be useful in determining the health status of the mammary gland in the dual-purpose livestock systems of the Colombian Orinoquia. The SCM diagnosis with any of the two tests should be complemented with sampling for microbiological analysis to determine the causative agent of the degree of infection in each quarter of the mammary gland.

Author contributions

Writing, Preparation of the Original Manuscript, Methodology and Formal Analysis, A. S-C.; M. V-T R.T.; Statistic Analysis: M.V-T; R.T; Review, Edition and Conceptualization, R. J-J., P.M.P-C., J. N. A-L. All Authors have read and accepted the published version of the manuscript.

Ethical approval

Bioethical Concept No. BIO163 issued by the Research Bioethics Subcommittee of the Bucaramanga Sectional Universidad Cooperativa de Colombia (Act No. 4, November 04, 2021).

Data availability statement

Data are available upon reasonable request to the second author.

Acknowledgements

The Authors thank the auxiliary students who helped in the collection of farm data. To the farmers for providing the animals for taking milk samples. All persons included in this section have consented to the acknowledgments.

Disclosure statement

No potential conflict of interest was reported by the author(s).

Additional information

Funding

This research was funded by the Research Committee—CONADI (ID2958; ID3058)—of the Universidad Cooperativa de Colombia.