| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
|
|
|
|||||||
|
|||||||||
* Department of Biological Sciences, University of Warwick, Coventry, CV4 7AL, UK.
Department of Population Medicine and Diagnostic Sciences, College of Veterinary Medicine,Cornell University, Ithaca, New York, 14853, USA.
Department of Clinical Veterinary Science, University of Bristol, Langford House, Langford, Bristol, BS40 5DT, UK.
Corresponding author:
M. J. Green; e-mail:
martin{at}greenl.swinternet.co.uk.
 |
ABSTRACT |
|---|
 TOP ABSTRACT INTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Key Words: dry period • intramammary infection • mastitis
Abbreviation key: 2-C = the time period from 2 wk before calving to 1 wk after calving
|
INTRODUCTION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
The importance of the dry (nonlactating) period in the dynamics of IMI in dairy cattle has been studied over many decades (Neave et al., 1950; Oliver et al., 1962; Oliver and Michell, 1983; Smith et al., 1985; Oliver, 1988; Todhunter et al., 1991). Quarters that are infected with bacteria during the dry period produce less milk in the next lactation (Smith et al., 1968) and, therefore, cause financial loss.
Previous field studies have examined IMI in the dry period and during lactation (Eberhart and Buckalew, 1977; Smith et al., 1985; Oliver, 1988; Todhunter et al., 1991), but from these results it is difficult to quantitatively assess the impact of dry-period IMI on clinical mastitis. While not all new dry-period IMI persist through to lactation (Oliver, 1988), there is evidence that some IMI present in the dry period will result in clinical disease in lactation (Neave et al., 1950; Bramley, 1976; McDonald and Anderson, 1981). This has been recently confirmed in the UK with DNA fingerprinting of enterobacterial strains (Bradley and Green, 2001).
A greater understanding of the associations between dry-period IMI and clinical mastitis in the next lactation may assist with the prevention of clinical mastitis. Such associations can only be estimated accurately if the clustered nature of the data is accounted for; quarters being clustered within cows and cows within herds (Goldstein, 1995). Without allowing for clustering, erroneous conclusions may be drawn (McDermott and Schukken, 1994).
The purpose of this study was to examine the relationship of quarter IMI at drying off and during the dry period, with clinical mastitis in the same quarter, in the next lactation.
|
MATERIALS AND METHODS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Six herds were selected on the basis of location (Somerset), a 3-mo geometric mean bulk milk SCC under 250,000/ml, nonseasonal calving pattern, and likelihood of owner compliance. Cows that were dried off between 10 February 1997 and 26 January 1998 and calved between 22 March 1997 and 4 April 1998 were recruited to the study. Sample collection ended on 15 July 1998.
Two lacteal secretion samples were collected at each sample time. Samples were taken from all four quarters at drying off and 0 to 7 d after calving. During the dry period, samples were taken from the ipsilateral quarters (left fore and left hind, odd numbered cows or right fore and right hind, even numbered cows), once 0 to 7 d and once 8 to 14 d before the anticipated calving date. Two quarters were left unsampled throughout the dry period as controls. Any cow not calving by her expected date was sampled every week until parturition. During the subsequent lactation, milk samples were collected from all quarters with clinical mastitis identified by the herdspersons who had been previously trained in identification and sample collection. These samples were frozen and submitted once each week to the laboratory for bacteriological analysis.
Teats were thoroughly disinfected, dried, and cleaned twice with surgical spirit before they were sampled, and disinfected after sampling. Milk samples were stored in a cool box immediately after collection and maintained at or below 4°C. Bacteriology, usually performed within 24 h, was carried out at the Veterinary Laboratory Agency (Langford, UK). Ten ml of secretion was inoculated onto sheep blood agar and Edward’s agar; 100 ml of secretion was inoculated onto MacConkey agar to enhance the detection of Enterobacteriaceae. Plates were incubated at 37°C and read at 24 and 48 h. Organisms were identified and quantified using standard laboratory techniques (Quinn et al., 1994). Escherichia coli was identified by colony morphology, oxidase, and indole tests; other Enterobacteriaceae were identified using a microtube identification system (RapiD 20 E, bioMérieux, UK).
Dry cow therapy was administered to all cows immediately after collection of the drying off samples. Dry cow products containing cloxacillin (Orbenin Extra, Pfizer, Sandwich, UK), cephalonium (Cepravin, Schering-Plough, Harefield, UK), or procaine penicillin G (Mylipen, Schering-Plough) were used. Only one of these products was used on any one farm.
Definition of Terms for Analysis
Days at risk.
Quarters were considered at risk from the date of drying off, until 300 d into the following lactation, unless the cow was censored earlier because the study ended or she was dried off or culled.
Seasonality.
Seasons were defined as Winter (Dec., Jan., Feb.); Spring (Mar., Apr., May); Summer (Jun., Jul., Aug.); and Autumn (Sep., Oct., Nov.).
Dry period.
The time between the intramammary infusion of dry cow antibiotic and calving.
2-C period.
The time period from 2 wk before calving to 1 wk after calving during which three samples were taken.
Intramammary infection: Screening Samples.
Isolation of at least one colony of one to three species of bacteria was considered to be an IMI. The presence of more than three bacterial species was ‘contaminated’ and reculture of the second sample was performed. Infection status was then based on isolation of organisms in the second sample. If the second sample was contaminated, the results were not used in data analysis.
Causes of mastitis.
Major pathogen—defined as the mastitis causing organisms; Streptococcus agalactiae, Streptococcus uberis, Streptococcus dysgalactiae, coagulase-positive staphylococci, Escherichia coli, Klebsiella spp., Enterobacter spp., Citrobacter spp., Serratia spp., and Arcanobacterium pyogenes.
Minor pathogen – defined as Corynebacterium spp. and coagulase-negative staphylococci.
Causes of mastitis were defined as follows.
A quarter with clinical mastitis.
Analysis was performed at the quarter level. A quarter with one or more cases of clinical mastitis during the study period was a positive (mastitic) quarter. A cow could have up to four quarters affected. The bacterial species causing clinical mastitis in a quarter was defined as that associated with the first case during lactation.
A nonmastitic quarter.
A quarter with no clinical mastitis through the study period.
Data Recording and Analysis
Herd, cow, and quarter identity, dates and culture results were entered into a database (Microsoft Access, Microsoft Corp., Redmond, WA, US). Univariate analysis was performed using Microsoft Excel (Microsoft Corp.), Minitab 10.51 (Minitab Inc., State College, PA, US) and EGRET 2.0.3 (Cytel Software Corp., San Diego, CA, US). The associations between bacterial species isolated in screening samples and the incidence of clinical mastitis were assessed at quarter level. This was done initially using Chi square tests (Petrie and Watson, 1999) and then using binomial logistic regression (see below). Single time-point and multiple isolations of the same bacteria through the sample period in each quarter were examined to investigate associations with clinical mastitis. The following patterns of infection were considered independently and together, in relation to clinical mastitis:
Variables that could influence the relationship between IMI and clinical mastitis (potential confounders) were assessed initially with univariate analysis carried out at quarter level without accounting for clustering of quarters within cows. The effect of farm, calving season, parity, days at risk, length of dry period, and the number of dry period screening samples, on the incidence of clinical mastitis, were investigated using Chi square tests for categorical data and Kruskal Wallis tests for nonparametric continuous data (Petrie and Watson, 1999).
Logistic binomial regression with random effects for distinguishable data (EGRET 2.0.3, Cytel Software Corp.) was used to model the occurrence of clinical mastitis and patterns of bacterial IMI as previously defined. Logistic models were fitted with the outcome (response) variables being:
Since there were insufficient cases of mastitis caused by coagulase-positive staphylococci, S. dysgalactiae and S. uberis to allow pathogen-specific modeling, Chi-square analysis was used to examine the influence of IMI in the sampling period on clinical mastitis.
A random term for cow was included in the models to account for the effect of clustering of quarters within cow. Farm level variation was accounted for by including ‘herd’ as a fixed effect where this improved the model (see below). Therefore, the model for the response probability 
ij for the jth covariate pattern within the ith cluster was:
 |
where 
is the intercept, ß1 is the coefficient of covariate X1 and there are p covariates, and 
ui is a random term in which is a positive scalar coefficient (overdispersion parameter). A value of = 0 means no overdispersion, and > 0 indicates overdispersion. ui is a standardized binomial random variable.
Explanatory variables were selected from the univariate analysis if there was a trend to increase or decrease the risk of clinical mastitis (P < 0.25) and modeled using a forward stepwise procedure (Hosmer and Lemeshow, 1989, Kleinbaum, 1994). Covariates were left in the model when there was a significant reduction in deviance computed from the likelihood ratio statistic. Potential confounders and a random term for cow were included in the model when there was a biological improvement in the model; a change in the odds ratios or confidence intervals for the covariate coefficients (Hosmer and Lemeshow, 1989; Kleinbaum, 1994). The significance probability was set at 0.05 for a two-tailed test. Interactions between significant covariates remained in the model if there was a significant improvement as judged by the likelihood ratio statistic. When the prevalence of infection for an outcome was zero, which precluded calculation of an odds ratio using normal logistic regression, combinations of time points were used, rather than exact analytical methods, or the variable was removed if it was judged to be of no biological importance. Goodness of fit of the models was assessed from graphs of residuals.
Kaplan-Meier curves were constructed of survival time to clinical mastitis (Parmer and Machin, 1995) in quarters with and without IMI during the dry period. The survival function was calculated as:
 |
where nt = the number of quarters at risk (not yet having mastitis) at the start of time interval t, dt = the number of quarters that got clinical mastitis during time interval t, and S(t) = the probability of a quarter not getting clinical mastitis by time t. Curves were evaluated for statistical differences using the hazard ratio calculated from a Cox Proportional Hazards Model (Parmer and Machin, 1995). As before, the significance probability was set at 0.05 for the likelihood ratio statistic. To check that the assumption of proportionality of hazards between groups was correct, a visual assessment was performed of log-transformed data (Parmer and Machin, 1995).
|
RESULTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
.
|
. In 32 (38.1%) of the 84 cases in sampled quarters, the species of bacteria isolated at the time of clinical mastitis were also isolated in at least one of the screening samples. The prevalence of bacterial IMI in the 954 quarters sampled during the dry period is presented in Table 3.
|
|
Considering all herds together, quarters from cows of greater than third parity were at increased risk of clinical mastitis compared with those of parity 3 or less (P < 0.01). There was a trend for quarters from cows calving in the summer to have more clinical mastitis than those in spring (P = 0.17). Quarters from cows in herds 4 and 5 were at a significantly reduced risk of clinical mastitis compared with herd 1 (P < 0.01).
Quarters with clinical mastitis had significantly fewer days at risk during the study period (mean = 210.9, median = 229.0 d, interquartile range = 144.3 to 300.0 d) than nonmastitis quarters (mean = 232.8 median = 261.0, interquartile range = 169.0 to 300.0 d), (P = 0.044) because clinical mastitis occasionally resulted in premature culling. Since quarters with a case of clinical mastitis had fewer days at risk than other quarters, an increased days at risk was not a cause of an increased probability of clinical mastitis. No effect of the length of the dry period was found on clinical mastitis incidence.
Therefore, the potential confounders considered in the models were the number of dry period samples, herd, parity, and calving season.
Univariate Analysis of Clinical Mastitis and Bacteria Isolated in Screening Samples
Bacteria present at drying off.
When the following bacteria were isolated at drying off, there was a trend for an increased risk of clinical mastitis (P < 0.25) S. dysgalactiae, S. uberis, S. faecalis, coagulase-positive staphylococci, E. coli, Enterobacter spp., Serratia spp., and Corynebacterium spp.
Bacteria present from 2 wk before calving to 1 wk after calving.
A trend for increased risk of clinical mastitis (P < 0.25) was associated with isolation of S. uberis, S. faecalis, coagulase-positive staphylococci, Serratia spp., and Klebsiella spp., 8 to 14 d before calving and S. dysgalactiae, S. uberis, S. faecalis, coagulase-positive staphylococci, E. coli, Serratia spp., and coagulase negative staphylococci, 0 to 7 d before calving. Isolation of the following organisms in at least two out of three samples during the 2-C period was also associated with a trend for an increased risk of clinical mastitis (P < 0.25): S. dysgalactiae, S. uberis, S. faecalis, coagulase-positive staphylococci, E. coli, Serratia spp., Klebsiella spp., and coagulase-negative staphylococci. The presence of Corynebacterium spp. 0 to 14 d before calving was associated with a decreased likelihood of clinical mastitis (P < 0.25). When the following bacteria were isolated in the postcalving sample, there was a trend for an increased likelihood of clinical mastitis: S. uberis, coagulase-positive staphylococci, E. coli, and Klebsiella spp. (P < 0.25). When Corynebacterium spp. were isolated in the postcalving sample, there was a trend for less clinical mastitis (P < 0.25).
Modeling IMI and Clinical Mastitis
The final clinical mastitis models are presented in Tables 4 and 5.
|
|
The E. coli model indicated that the risk of clinical E. coli mastitis was increased with isolation of E. coli and coagulase-positive staphylococci at drying off and isolation of S. faecalis and E. coli at least twice in the 2-C period in quarters not infected with them at drying off (Table 5
). Quarters in cows from herds 4 and 5 were at significantly reduced risk of clinical E. coli mastitis compared to herd 1 but there was no significant effect of parity, calving season, or the number of dry period samples taken.
The random effect of ‘cow’ improved both models and was, therefore, included in the models.
The risk of clinical mastitis caused by S. uberis, S. dysgalactiae, or coagulase-positive staphylococci was significantly increased if the organism causing mastitis was isolated in at least two of the 2-C samples (Table 6).
|
) indicated that quarters with mastitis were more likely to have higher fitted values (i.e., predicted to be a case by the model) than nonmastitic quarters; the model was able to differentiate mastitic and nonmastitic quarters reasonably well.
|
. Sixty percent of quarters in group A got clinical mastitis within 14 d of calving compared with 20% in group B. Over 80% of mastitis in group A had occurred by d 120 of lactation, whereas only 50% had occurred by this time in group B. The probability of a case of clinical mastitis was fairly constant throughout lactation for group B quarters. The two curves were significantly different, with group A being at an increased hazard of mastitis (Hazard Ratio = 1.86, 95% confidence intervals = 1.17 – 2.95, P = 0.01).
|
. Quarters infected at drying off were more likely to get clinical mastitis at any time during the next lactation compared with uninfected quarters (Hazard Ratio = 1.85, 95% confidence intervals =1.20 – 2.84, P < 0.01). Quarters in which Corynebacterium spp. IMI were found during the 2-C period, with no previous major pathogen IMI during the 2-C period, were less likely to get clinical mastitis during the next lactation compared with quarters not infected with Corynebacterium spp. in the 2-C period (Hazard Ratio = 0.21, 95% confidence intervals = 0.03 – 1.49, P = 0.03).
|
|
DISCUSSION |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Over 60% of clinical mastitis in quarters in which the same pathogen was identified during the dry period, occurred within 2 wk of calving and 90% within 150 d of calving. This is in contrast to the approximately constant rate of clinical mastitis during lactation that occurred in quarters from which the same pathogens were not cultured during the dry period. The pattern and rate of clinical mastitis over lactation on a dairy farm can, therefore, give an indication of the impact of dry period IMI on clinical mastitis and moreover, the areas to target preventive measures.
The presence of S. dysgalactiae, S. faecalis, E. coli, Enterobacter spp., and Corynebacterium spp. at drying off increased the probability of subsequent lactational clinical mastitis. The influence of IMI in the 2-C period on subsequent clinical mastitis was statistically separate from the influence of IMI at drying off since both sample times remained in the models. The data do not support the hypothesis that these organisms persisted from drying off through the dry period since persistence was tested in the models and did not influence the risk of clinical mastitis. Similarly, these findings are unlikely to be a result of contamination of milk samples since the sampling technique was fastidious, sampling during the dry period was always performed by the principle investigators, and there was also no reason that contamination should occur more often at drying off in quarters that got clinical mastitis in the next lactation. One plausible explanation for the increased risk of clinical mastitis in quarters with IMI at drying off is that certain quarters are simply susceptible to IMI. An alternative explanation is that the previous bacterial infection increases the risk of subsequent infection. Quarter susceptibility may occur because of anatomical features such as short wide teat canals (Grega and Szarek, 1985; Grindal et al., 1991) or immunological characteristics, such as poor white cell function (Hill, 1981) or low white cell number (Peeler, 2001). Alternatively, previous infection may cause damage or alteration to the mammary environment in some way and reduce innate defense mechanisms.
An interesting finding in this study was that the timing of IMI with Corynebacterium spp. was important in determining its association with clinical mastitis. If infection was present at drying off, then the quarter was at greater risk of clinical mastitis, and the Kaplan-Meier curves show that these cases occurred evenly throughout lactation. It is possible that IMI with Corynebacterium spp. occurred because of poor teat hygiene procedures, and this continued into the next lactation and predisposed the quarters to clinical mastitis. Alternatively, it is possible that Corynebacterium spp. IMI at drying off was an indicator of quarters that had poor defense mechanisms such as poor teat closure. This would mean these quarters were generally prone to IMI and would be more likely to get clinical mastitis in the next lactation. A previous field study has also found that quarters with Corynebacterium bovis IMI were at an increased risk of subsequent IMI (Hogan et al., 1988).
In this study, however, if Corynebacterium spp. were isolated in the 2-C period, in a quarter not already infected with a major pathogen, then the quarter was at a reduced risk of clinical mastitis. The data, therefore, suggest that the timing of Corynebacterium spp. infection may be important in determining whether a quarter is at greater or reduced risk of clinical mastitis. This may explain why some authors have found protective effects (Lam et al., 1997; Rainard and Poutrel, 1988), while others have found an increase in susceptibility (Hogan et al., 1988). It would be useful to understand the exact conditions that dictate whether infection with Corynebacterium spp. indicates either a more susceptible quarter or a quarter at reduced risk of clinical mastitis. These conditions may include the species and strains of the Corynebacterium involved, the timing of infection, and whether the Corynebacterium spp. IMI is present at the actual time of challenge by a major pathogen.
The clinical value of protection afforded by Corynebacterium spp. in this study was limited. Only 42 out of 954 (4.4%) quarters were infected with Corynebacterium spp. in the 2-C period, indicating that the vast majority of quarters did not benefit from protection. Since the prevalence of Corynebacterium spp. has been found to decrease in the dry period, even when no antibiotic dry cow therapy is used (Harmon et al., 1986; Oliver and Juneja, 1990), any protective effect is likely to be of value in the field only if methods to safely increase its prevalence are found.
The current study could not determine whether there was a varying level of protection from Corynebacterium spp. IMI to different major pathogens. There were insufficient Corynebacterium spp. infections (n = 42) in the 2-C period to perform this analysis. It remains possible, therefore, that protection may occur for some major pathogens and not others, as has been reported experimentally (Linde et al., 1980; Brooks and Barnum, 1984; Pankey et al., 1985; Doane et al., 1987).
S. faecalis IMI in the 2-C period was also a risk factor for clinical mastitis (Model 1). Again, this may be an indicator of increased quarter susceptibility to IMI, or it may be a result of synergistic action between S. faecalis and other organisms, particularly E. coli, for which it increased the risk of clinical mastitis (Model 2). Previous reports have suggested that there is more likely to be competition and exclusion, however, rather than synergy between S. faecalis and E. coli (Dalhoff, 1982). Further research into the interactions between S. faecalis and other pathogenic bacteria in the mammary gland may be fruitful, but these data suggest that S. faecalis may not necessarily be indicative of a contaminated milk sample, as would commonly be reported from UK laboratories.
Data from this study highlight the difficulties of defining IMI. Previous longitudinal studies have defined an IMI as present when a pathogen is cultured on two out of three occasions from consecutive samples 1 to 30 d apart (Smith et al., 1985; Hogan et al., 1988; Lam et al., 1997). This definition is likely to represent long-term infections and will miss those that are transient. Various combinations of IMI were considered in this study; single infections, at least one infection in three consecutive 2-C period samples, and at least two infections in three consecutive 2-C period samples. While two infections out of three samples was the best indicator of clinical mastitis in the models, it was found that this definition of an IMI gave a greater specificity and predictive value for clinical mastitis but a poorer sensitivity than including single infections. This is likely to be because while some single bacterial isolates can lead to clinical mastitis, many will be removed by cow defenses without the appearance of clinical signs.
The use of a single milk sample at drying off to diagnose IMI, rather than two or three, will have reduced the sensitivity of IMI diagnosis. It is considered that false-negative samples are most likely to occur with coliforms and coagulase-positive staphylococci (National Mastitis Council, 1999), although it has been estimated that a single quarter milk sample from a lactating gland has 75 to 90% sensitivity and >97% specificity for diagnosis of coagulase-positive staphylococci IMI (Sears et al., 1990; Buelow et al., 1996). Under diagnosis of true IMI at drying off in this study will have caused an underestimate of the importance of IMI at this time and an overestimate of IMI considered to be ‘new’ in the 2-C period. False-negative coliform cultures were reduced by using 0.1 ml secretion instead of 0.01 ml (National Mastitis Council, 1999). While recognizing the constraints of bacteriological culture, both total IMI in the 2-C period and apparent new IMI were modeled and it was found that only the new IMI remained in the models. Although not all of these may be truly new IMI, this supports the view that new IMI in the dry period are important (Smith et al., 1985; Eberhart, 1986; Todhunter et al., 1991).
The graphs of fitted values (prediction by the model for each quarter to get mastitis or not) showed a reasonably good differentiation between mastitic and normal quarters (Figure 1). The relatively high fitted values (>0.2) for mastitic quarters related to quarters with dry period IMI and the lower values to quarters with mastitis not associated with dry period IMI. The model could not predict all quarters that got mastitis, and this was expected because not all clinical mastitis was associated with dry period IMI. The number of quarters (32 out of 84), however, in which clinical mastitis was associated with IMI by the same organism during the sampling period, was surprisingly high. The model could differentiate quarters that did not get mastitis as shown by low fitted values, and this indicated that few of these non-mastitis quarters had dry period IMI.
Confounders, such as parity and herd, were included in the logistic regression analysis to control for variation that may affect the relationship between dry period IMI and clinical mastitis. This meant that estimates of the coefficients of the explanatory variables were adjusted and of greater accuracy (Hosmer and Lemeshow, 1989; Kleinbaum, 1994). Herd and parity were the only confounders identified in the final models, and it is noteworthy that quarters from cows of parity greater than 3 were at approximately twice the risk of clinical mastitis compared to those of parity 2 and 3. Cows of parity 1 did not participate in the study because they did not have a dry period. There was a small increase in the proportion of quarters that got clinical mastitis when sampled during the dry period compared to those not sampled. This was mainly because of an increased risk in quarters sampled four times during the dry period. The number of dry period samples taken was, therefore, used as a covariate for modeling clinical mastitis, but it did not remain in any model, indicating that sampling did not significantly influence the risk of clinical mastitis.
Improvement of the models with the inclusion of a random term for cow indicated that quarters within a cow were not behaving independently. That is, two quarters of the same cow were more likely to show similar patterns of IMI and clinical mastitis than two quarters of different cows. Correcting for this potential bias in the data improved the accuracy of the confidence intervals in the model and shows that this is an important consideration when handling quarter level mastitis data from dairy cows.
|
CONCLUSIONS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
|
ACKNOWLEDGEMENTS |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
Received for publication July 12, 2001. Accepted for publication April 9, 2002.
|
REFERENCES |
|---|
|
TOPABSTRACTINTRODUCTIONMATERIALS AND METHODSRESULTSDISCUSSIONCONCLUSIONSACKNOWLEDGEMENTSREFERENCES |
|---|
This article has been cited by other articles:
|
|
 |
|
  K. I. Parker, C. W. R. Compton, F. M. Anniss, C. Heuer, and S. McDougall Quarter-Level Analysis of Subclinical and Clinical Mastitis in Primiparous Heifers Following the Use of a Teat Sealant or an Injectable Antibiotic, or Both, Precalving J Dairy Sci, January 1, 2008; 91(1): 169 - 181. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. J. Green, A. J. Bradley, G. F. Medley, and W. J. Browne Cow, Farm, and Management Factors During the Dry Period that Determine the Rate of Clinical Mastitis After Calving J Dairy Sci, August 1, 2007; 90(8): 3764 - 3776. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
K. I. Parker, C. Compton, F. M. Anniss, A. Weir, C. Heuer, and S. McDougall Subclinical and Clinical Mastitis in Heifers Following the Use of a Teat Sealant Precalving J Dairy Sci, January 1, 2007; 90(1): 207 - 218. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
M. Chassagne, J. Barnouin, and M. Le Guenic Expert Assessment Study of Milking and Hygiene Practices Characterizing Very Low Somatic Cell Score Herds in France J Dairy Sci, May 1, 2005; 88(5): 1909 - 1916. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
E. L. Annen, R. J. Collier, M. A. McGuire, J. L. Vicini, J. M. Ballam, and M. J. Lormore Effect of Modified Dry Period Lengths and Bovine Somatotropin on Yield and Composition of Milk from Dairy Cows J Dairy Sci, November 1, 2004; 87(11): 3746 - 3761. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
S. Godden, P. Rapnicki, S. Stewart, J. Fetrow, A. Johnson, R. Bey, and R. Farnsworth Effectiveness of an Internal Teat Seal in the Prevention of New Intramammary Infections During the Dry and Early-Lactation Periods in Dairy Cows when used with a Dry Cow Intramammary Antibiotic J Dairy Sci, December 1, 2003; 86(12): 3899 - 3911. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| HOME | HELP | FEEDBACK | SUBSCRIPTIONS | ARCHIVE | SEARCH | TABLE OF CONTENTS |
