Sensitivity and Specificity of Somatic Cell Count and California Mastitis Test for Identifying Intramammary Infection in Early Lactation1

J. Dairy Sci. 84:2018–2024  American Dairy Science Association, 2001.

Sensitivity and Specificity of Somatic Cell Count and California Mastitis Test for Identifying Intramammary Infection in Early Lactation1 J. M. Sargeant,* K. E. Leslie,† J. E. Shirley,‡ B. J. Pulkrabek,‡ and G. H. Lim† *Food Animal Health and Management Center and ‡Department of Animal Sciences and Industry, Kansas State University, Manhattan, KS 66506 †Department of Population Medicine, University of Guelph, Guelph, ON, CANADA, N1G 2W1

ABSTRACT

INTRODUCTION

Associations between values for the somatic cell count (SCC) or the California Mastitis Test (CMT) and intramammary infection (IMI) were studied in 131 dairy cows from three herds during the first 10 d postcalving. Intramammary infection was defined as the presence of one or two bacterial species in one or both quarter milk samples taken within 12 h of calving and at d 3 postcalving. Quarter milk samples identified IMI in 36% of glands. Values for SCC declined at a significantly faster rate over the first 10 d postcalving in noninfected quarters than in infected quarters. The usefulness of quarter milk SCC and CMT for screening was evaluated by calculating the sensitivity and specificity for various threshold values and days postcalving. A SCC threshold of 100,000 cells/ml for quarter samples evaluated on d 5 postcalving had the maximal sensitivity and specificity for detecting IMI. Evaluation of the CMT samples taken on d 3 postcalving using a threshold reaction of greater than zero had the highest sensitivity and specificity for detecting IMI. With this CMT sampling scheme, the sensitivities for detecting IMI with any pathogen, IMI with a major pathogen, and IMI with a minor pathogen were 56.7, 66.7, and 49.5, respectively. The CMT could have a useful role in dairy herd monitoring programs as a screening test to detect fresh cows with IMI caused by major pathogens. (Key words: intramammary infection, calving, somatic cell count, California mastitis test)

Mastitis is the most costly disease of dairy cattle due to economic losses from reduced milk production, treatment costs, increased labor, milk withheld following treatment, death, and premature culling (Kaneene and Hurd, 1990; Lightner et al., 1988; Miller et al., 1993). Control measures include the use of premilking udder hygiene, postmilking teat dipping, dry cow therapy with long-acting antibiotics, segregation and culling strategies for chronically infected animals, and environmental control during the dry cow and calving periods (Radostits et al., 1994). Each of these control measures is aimed at the management of specific pathogen types. For example, premilking udder hygiene and teat dipping are aimed at preventing new infections, primarily caused by contagious pathogens, during the milking period (Boddie et al., 1993; Natzke, 1981; Pankey, 1989). Dry cow therapy with long-acting antibiotics is used to cure infections present at the time of dry-off (Eberhart, 1986; Natzke, 1981; Rainard et al., 1990). Dry cow therapy is used with other management efforts to reduce the occurrence of new cases of environmental streptococcal infections during the early dry period (Smith et al., 1985). Environmental management during the transition and calving periods is targeted primarily at preventing new infections with environmental streptococcal species and coliform bacteria (e.g., Escherichia coli, Klebsiella spp.). Over half of the environmental pathogens acquired during the dry period persist to lactation (Smith et al., 1985; Todhunter et al., 1995). Therefore, the frequency of IMI present at the time of calving, as well as knowledge of the specific pathogens implicated, would provide a way to monitor the effectiveness of existing udder health programs and assess the need for new control strategies. Bacteriological culture is the standard method for identifying IMI. However, logistic and financial considerations involved in sampling all quarters at the time of calving have precluded widespread adoption of this strategy in the dairy industry. If an effective means to

Abbreviation key: CMT = California mastitis test.

Received August 2, 2000. Accepted April 23, 2001. Corresponding author: J. M. Sargeant; e-mail: sargeant@vet. ksu.edu. 1 Submission # 01-14-J from the Kansas State Agricultural Experiment Station. The authors gratefully acknowledge the technical assistance of Linda Cox (Kansas State University) and Anna Bashiri (University of Guelph).

2018

IDENTIFICATION OF INTRAMAMMARY INFECTION AT CALVING

identify fresh cows at a high risk for IMI were available and validated, it might increase the adoption of milk culture. The SCC has been widely implemented as a screening test to identify IMI in lactating cows. A threshold of 200,000 cells/ml has been shown to have a high sensitivity and specificity for identifying infections (Dohoo and Leslie, 1991; Timms and Schultz, 1987). However, thresholds for SCC during the preceding lactation or at dry-off did not provide accurate identification of IMI status at the subsequent calving (Kirk et al., 1996). In addition, SCC is usually elevated at the time of calving and then decreases at a rapid rate, particularly during the first 2 wk of lactation (Dohoo, 1993). Barkema et al. (1999) compared geometric mean SCC at the quarter level over the first six milkings postcalving and found that it decreased more rapidly in uninfected quarters than in infected quarters. In a study of colostral milk sampled at the quarter level at the time of calving, results of both SCC and the California Mastitis Test (CMT) were associated significantly with IMI status (Maunsell et al., 1999). Therefore, the potential exists to use SCC or CMT in the early postpartum period as a screening method to select quarters for bacteriological culture. Once validated, this monitoring program would provide the necessary information on pathogen prevalence and type to evaluate udder health programs, particularly during the dry period. At the same time, the selective nature of the sampling would reduce the time and expense of culturing noninfected quarters. It is hypothesized that CMT is an efficient cow-side proxy for SCC (Schalm and Noorlander, 1957) and that both tests are useful predictors of IMI in fresh cows. The aim of this study was to expand on previous work to investigate the use of SCC and CMT to determine IMI at calving caused by any mastitis pathogens, major pathogens, and minor pathogens. Specifically, the objectives were to determine threshold values of SCC and CMT at the quarter level for identifying IMI and to determine the most appropriate sampling times to maximize the ability of SCC and CMT to identify infected cows for which bacteriological culture is appropriate. MATERIALS AND METHODS Study Population The study group consisted of 131 Holstein cows in the Kansas State University dairy research herd, the Elora research herd (University of Guelph, Ontario), and the Ponsonby research herd (University of Guelph, Ontario). Cows in all three herds were housed in tiestall barns. Two of the herds used milking parlors, and the third (Ponsonby) milked with a pipeline system. All cows were milked twice daily. All multiparous cows were treated with long-acting intramammary antibiot-

2019

ics at the dry-off period immediately preceding the lactation used in this study. Sampling Strategy and Analysis of Samples Quarter foremilk samples were collected by trained technicians for SCC analyses within 12 h of calving (d 1) and at the morning milking on d 2 to 10 postcalving. Samples were analyzed for SCC using Fossomatic cell counters at Heart of America DHI (Kansas State University herd; Manhattan, KS) and at the University of Guelph Mastitis Research Laboratory (Elora and Ponsonby herds; Guelph, Ontario). The CMT were performed cow-side and according to manufacturer’s recommendations on quarter foremilk samples on d 1 to 10 postcalving. The CMT results were recorded as 0 (negative), 1 (trace or +), 2 (++), or 3 (+++). Quarter milk samples were collected aseptically on d 1 and 3 postcalving for bacteriological examination. The teat end was disinfected, and the first streams of milk were discarded before collection. All samples were frozen immediately without preservatives. Bacteriological culture was performed on thawed samples within 10 d of collection. Culturing of samples followed the procedures of the National Mastitis Council (1987), with the exception of inoculum volume. An inoculum of 0.01 ml was used for samples from the Ontario herds. Swab plating, with swabs containing approximately 0.03 ml, was used for the samples from Kansas. Samples in which three or more bacterial species were identified were considered contaminated. Outcome Evaluation A milk sample positive for IMI was defined as one with one or two pathogens isolated at either d 1 or d 3 postcalving. Bacteriological causes of IMI were categorized as major pathogens [E. coli, Klebsiella spp., environmental (non-agalactiae) streptococci, Staphylococcus aureus, Streptococcus agalactiae, and other) or minor pathogens (coagulase-negative staphylococci, Cornybacterium bovis)] (Harmon, 1994). If both major and minor pathogens were isolated from one or both quarter milk samples, the cow quarter was classified as infected with the major pathogen. If one of the two samples was contaminated or missing, the results from the remaining sample were used. Statistical Analysis Associations between SCC and IMI over the first 10 d postcalving were tested for statistical significance using mixed model linear regressions in the MIXED procedure of SAS (1999). Quarter within cow was the unit Journal of Dairy Science Vol. 84, No. 9, 2001

2020

SARGEANT ET AL.

of analysis. To approximate a normal distribution, the outcome was the natural logarithm of SCC (lnSCC). The model was as follows: lnSCCijklmno = µ + herdi + parityj + dimk + infection_statusl + infection_statusl*dimk + cow(herd)ijm + quarter(herd cow)ijklmn + eijklmno, where herd was controlled as a fixed effect, dim = DIM as a fixed effect, parity = parity as a fixed effect, infection status = the presence or absence of IMI on d 1 or d 3 postcalving, and infection status*dim = the interaction between infection status and DIM. Cow within herd and quarter within cow within herd were controlled as random effects. Separate models were run, with infection status defined in one of three ways: presence or absence of any IMI on d 1 or 3 postcalving, presence of a major mastitis pathogen on d 1 or 3 postcalving versus no growth or infection with a minor pathogen(s) on d 1 and 3 postcalving, and presence of a minor pathogen on d 1 or 3 postcalving versus no growth on d 1 and 3 postcalving. All associations were considered statistically significant at P ≤ 0.05. To determine the validity of SCC (CMT) for identifying infected quarters, sensitivity and specificity were determined at selected thresholds. Sensitivity was calculated as the proportion of culture-positive quarters that had SCC (CMT) values above the selected threshold, and specificity was the proportion of culture negative quarters that had SCC (CMT) values below the selected threshold (Martin et al., 1987). The threshold values used for validating SCC were 100, 250, 500, 750, 1000, 2000, 3000, 4000, and 5000 × 103. Samples that contained clots or were too thick for standard SCC determination were coded as 9999 × 103 so they were

included above the threshold in each scenario. For the CMT analysis, the thresholds were > 0 (i.e., any nonnegative result), > 1, and > 2. RESULTS Data were collected from July 1998 to April 1999. There were 81 cows from the Kansas State University research herd, 33 cows from the Elora herd, and 17 cows from the Ponsonby herd. The total group included 21 first-lactation cows, 48 second-lactation cows, and 62 cows in third parity and greater. Four quarters were nonfunctional, resulting in a total of 520 quarters used in the analyses. One cow died at 4 DIM, but available data for this cow were used in the analyses. Quarterlevel culture results are summarized in Table 1. Overall, 36% of quarters had an IMI; the most common pathogens isolated were coagulase-negative staphylococci. Streptococcus agalactiae was not identified in any sample. At the cow level, 29.0, 25.2, 26.0, 13.7, and 6.1% of cows had 0, 1, 2, 3, and 4 quarters infected, respectively. Approximately 30% of the quarters classified as infected with S. aureus and coagulase-negative staphylococci had the pathogen identified on both d 1 and 3. Approximately half of the infections with environmental streptococci, C. bovis, and Klebsiella spp. were positive on both days. Eschericha coli tended to be identified on either d 1 or 3, but not both. Mean SCC declined over the first 10 d postcalving (Table 2). In the regression models, the interaction between infection status and DIM was significantly associated with lnSCC when infection status was defined as any IMI, infection with a major pathogen, and infection with a minor pathogen. The direction of the association indicated that lnSCC decreased more slowly over the

Table 1. Intramammary infection status of 520 quarters of 131 cows in early lactation. % of infected quarters

Pathogen isolated1

Number of quarters

% of total quarters

No growth Major pathogens Environmental streptococci Klebsiella spp. Escherichia coli Mixed Staphylococcus aureus Other2

333

64.0

26 5 9 1 10 27

5.0 1.0 1.7 0.2 1.9 5.2

13.9 2.7 4.8 0.5 5.3 14.4

93 6 10

17.9 1.2 1.9

49.7 3.2 5.3

Minor pathogens Coagulase-negative staphylococci Corynebacterium bovis Mixed

1 A quarter was considered positive for a specific pathogen if the pathogen was present on d 1 or d 3 postcalving. 2 Includes Bacillus spp. (15), Enterobacter spp. (5), Pseudomonas spp. (4), Citrobacter spp. (2), Proteus spp. (1).

Journal of Dairy Science Vol. 84, No. 9, 2001

2021

IDENTIFICATION OF INTRAMAMMARY INFECTION AT CALVING Table 2. Geometric mean SCC (×1000 cells/ml) by infection status over the first 10 d postcalving for 520 quarters of 131 dairy cows. DIM

No growth

Infected with major pathogens

Infected with minor pathogens

1 2 3 4 5 6 7 8 9 10

568.6 295.6 131.7 99.3 60.8 49.5 42.2 36.4 28.7 24.8

1374.2 857.3 492.7 460.8 295.3 264.4 184.2 157.7 145.1 115.7

668.9 501.7 235.6 206.5 129.3 112.4 89.6 80.7 69.5 62.1

first 10 d in infected quarters. Details of the regression model for lnSCC when infection status was defined as any IMI are shown in Table 3. Table 4 shows the proportion of quarters in each CMT category by infection status over the first 10 d postcalving. Quarters infected with a major pathogen appeared more likely to have a higher CMT score than quarters infected with a minor pathogen or uninfected quarters, particularly during the early days postcalving. Within each of the infection categories, as the SCC threshold increased, the sensitivity of SCC for identifying infected quarters decreased and the specificity increased (Table 5). Over time, sensitivity decreased and specificity increased. These trends continued into the higher thresholds and over all 10 DIM (data not shown). Results indicated that the maximum sensitivity and specificity for this population of cows was seen at a threshold of >100 × 103 cells/ml for quarter samples taken on d 5 postcalving. At this time and threshold, the sensitivity and specificity were 57.4 (95% CI 50.2, 64.5) and 72.3 (95% CI 67.4, 77.4), respectively, for identifying quarters infected with any pathogen, 66.2 (95% CI 55.4, 77.0) and 66.4 (95% CI 61.9, 70.8), respectively, for identifying infection with major pathogens, and 51.4 (95% CI 42.0, 60.8) and 72.3 (95% CI 67.4,

77.2), respectively, for identifying infections with minor pathogens. Similar trends over increasing threshold values were observed for sensitivity and specificity of CMT for identifying IMI within each infection classification (d 1 to 3 postcalving shown in Table 6). A threshold of >0 on d 3 postcalving resulted in a sensitivity and specificity of 56.7 (95% CI 49.6, 63.8) and 56.2 (95% CI 50.8, 61.5), respectively, for identifying infection with any pathogen. The sensitivity and specificity for identifying major pathogens were 66.7 (95% CI 56.2, 77.1) and 54.8 (95% CI 50.1, 59.4), respectively, and the sensitivity and specificity for identifying minor pathogens were 49.5 (95% CI 40.2, 58.9) and 56.2 (95% CI 50.8, 61.5), respectively. DISCUSSION There is an increasing focus on milk quality in the dairy industry. Producing high quality milk will require effective udder health programs at the herd level. Management practices at the time of dry-off and during the dry period are essential to this effort, and the fresh cow period is an ideal time to evaluate these efforts. Screening tests, such as SCC and CMT, are used to detect mastitis during lactation. However, it has been

Table 3. Generalized linear model for Ln SCC and the presence of any IMI with quarter as the unit of analysis for 520 quarters of 131 dairy cows. Coefficient Cow (herd) Quarter (herd*cow) Residual Intercept Herd 2 Herd 3 Days in milk Parity Presence of infection Interaction between days in milk and infection status

SE

P value

6.07 0.16 0.15 −0.26 0.18 0.50

0.29 0.26 0.29 0.007 0.06 0.11

<0.001 0.52 0.59 <0.001 0.003 <0.001

0.06

0.009

<0.001

Variance components 0.749 0.625 0.771

Journal of Dairy Science Vol. 84, No. 9, 2001

2022

SARGEANT ET AL. Table 4. Proportion of quarters in each California mastitis test category by infection status over the first 10 d postcalving for 520 quarters of 131 dairy cows. Infected with major pathogens

No growth DIM 2

1 2 3 4 5 6 7 8 9 10

Infected with minor pathogens

0

11

2

3

0

11

2

3

0

11

2

3

0.61 0.54 0.56 0.57 0.63 0.68 0.70 0.70 0.76 0.78

0.23 0.32 0.33 0.35 0.28 0.27 0.21 0.25 0.20 0.17

0.13 0.12 0.09 0.07 0.09 0.05 0.09 0.05 0.04 0.05

0.03 0.03 0.02 0.02 0.003 0.01 0.003 0.003 0.003 0.003

0.40 0.34 0.31 0.35 0.45 0.43 0.44 0.47 0.55 0.56

0.23 0.27 0.28 0.29 0.19 0.22 0.26 0.28 0.18 0.21

0.16 0.14 0.21 0.11 0.19 0.22 0.19 0.20 0.19 0.16

0.21 0.25 0.20 0.25 0.18 0.14 0.11 0.04 0.08 0.07

0.62 0.55 0.50 0.59 0.61 0.64 0.67 0.59 0.68 0.67

0.17 0.24 0.32 0.24 0.21 0.22 0.18 0.27 0.17 0.19

0.15 0.14 0.13 0.12 0.15 0.11 0.14 0.14 0.14 0.14

0.06 0.07 0.05 0.06 0.03 0.03 0.01 0.01 0.02 0

1

Includes “trace” interpretations. Proportions within each infection status and day sum to 100%, excepting rounding errors.

2

shown that SCC are routinely elevated during the first week postcalving (Dohoo, 1993). Our results suggest that, despite this elevation, these screening tests can be useful to identify infected quarters and to reduce the number of quarters that need to be sampled to develop a herd pathogen profile of IMI in fresh cows. The cows sampled in this study were from three dairy research herds and thus not selected randomly. The prevalence of environmental streptococcal species was similar to previous estimates at the time of calving of 3.2 (Todhunter et al., 1995) and 6.3% (Smith et al., 1985), but the prevalence of coliform bacteria was lower than previously reported (Smith et al., 1985). Several studies have estimated the prevalence of IMI due to S. aureus in first calf heifers at calving and have shown wide variation between studies and between farms within study (Fox et al., 1995; Nickerson et al., 1995; Oliver et al., 1992; Pankey et al., 1991; Roberson et al., 1994). Bassel et al. (2000) reported an increasing prevalence of S. aureus IMI at calving with age; 3% of composite samples from first-calf heifers were positive compared with 8% in cows in parity three or higher.

Direct comparisons of IMI frequencies between studies is complicated by differing definitions of a positive sample. In the present study, quarter milk samples were taken on d 1 and 3, and a quarter was considered positive for a specific pathogen if the pathogen was isolated from either or both samples (parallel interpretation; Martin et al., 1987). The combined results were used in parallel to increase the sensitivity for detecting IMI when present, but may have led to the inclusion of transient infections as positive. A sensitivity of less than 100% at a given threshold means that infected quarters will not be selected for bacteriological culturing at that threshold. A specificity of less than 100% at a given threshold will result in false-positive quarters. These have values above the threshold and, therefore, are selected for sampling, but do not truly have an IMI. For screening to detect the possible presence of IMI, a high sensitivity is preferred, because the screening test is less likely to miss truly positive animals. Given that bacteriological culture of single samples does not have perfect sensitivity, we therefore used parallel testing even though many published studies on mastitis have

Table 5. Sensitivity and specificity (sensitivity/specificity) of SCC threshold values (×1000 cells/ml) for identifying infected quarters in early lactation. DIM 1

Any infection compared with no growth Infection with major pathogen compared with no growth or minor pathogen Infection with minor pathogen compared with no growth

2

3

>100

>250

>500

>100

>250

>500

>100

>250

>500

96.7/4.4

83.5/21.3

69.8/46.1

93.9/14.7

69.4/48.1

52.8/68.8

71.2/41.3

51.6/74.5

37.0/89.1

98.6/4.4

91.8/21.5

76.7/43.2

94.4/12.6

76.4/44.9

61.1/64.7

78.7/39.5

64.0/70.1

49.3/84.7

95.4/4.4

78.0/21.3

65.1/46.1

93.5/14.7

64.8/48.1

47.2/68.8

66.1/41.3

43.1/74.5

28.4/89.1

Journal of Dairy Science Vol. 84, No. 9, 2001

2023

IDENTIFICATION OF INTRAMAMMARY INFECTION AT CALVING

Table 6. Sensitivity and specificity (sensitivity/specificity) of California mastitis test threshold values for identifying infected quarters in early lactation. DIM 1

Any infection compared with no growth Infection with major pathogen compared with no growth or minor pathogen Infection with minor pathogen compared with no growth

2

3

>0

>1

>2

>0

>1

>2

>0

>1

>2

46.5/61.0

27.3/83.5

12.3/96.7

51.8/54.4

27.3/85.3

13.9/97.3

56.7/56.2

26.7/88.9

10.7/97.6

58.4/61.2

36.4/82.4

20.8/95.9

61.5/54.5

35.9/83.7

23.1/96.2

66.7/54.8

39.7/87.3

19.2/97.1

38.2/61.0

20.9/83.5

6.4/96.7

45.0/54.4

21.1/85.3

7.3/97.3

49.5/56.2

17.4/88.9

4.6/97.6

required two or more positive milk samples to define infection. False-positive animals can be ruled out with a second test, which in this case would be bacteriological culture. Although less serious in terms of identifying mastitis pathogens in a herd, false-positive quarters represent an economic loss from unnecessary labor and laboratory costs for the bacteriological culture of milk. Our selection of the best sampling strategy was based on the assumption that sensitivity and specificity were equally important. This assumption was used previously by Detilleux et al. (1999), when using receiveroperating characteristic methodology to illustrate the use of SCC to identify subclinical infection. If it was believed that either sensitivity or specificity were more important economically, weighting could be used to emphasize this difference in the definition of the best sampling strategy. In the SCC analyses, the sensitivity and specificity of selecting quarters for bacteriological culturing were only moderate for all IMI and for infection with minor pathogens. However, both sensitivity and specificity were considerably higher for distinguishing infections with major pathogens. Quarters infected with minor pathogens tended to have higher SCC values than noninfected quarters, but lower values than quarters infected with major pathogens. This mild inflammatory response associated with minor pathogens is consistent with previous reports (Rainard et al., 1990; Timms and Schultz, 1987). Although the literature reports that subclinical infection with S. aureus is associated with elevated SCC (Sheldrake et al., 1983; Wilson et al., 1997), numbers were insufficient in the present study to provide meaningful estimates of sensitivity and specificity of using SCC thresholds to sample cows specifically for this pathogen. The optimal sampling scheme identified for using CMT to select quarters for bacteriological culture was a threshold of greater than 0 (i.e., any nonnegative CMT) taken at 3 d postcalving. The sensitivities of CMT

for the corresponding categories of IMI were slightly lower than those found at the optimum threshold for SCC. However, CMT still had a relatively high sensitivity for differentiating quarters infected with major pathogens from quarters with no infection or infection with minor pathogens. The CMT has the advantage of being an inexpensive cow-side test that provides real time results. The lack of perfect sensitivity in these screening tests means that some individual quarters that are infected will not be identified. Thus, these testing strategies may not be ideal for making decisions about individual animals, such as identifying individual cows with S. aureus for segregated milking. However, the sensitivity and specificity of these tests will allow producers to establish the occurrence of IMI and estimate their frequency in groups of cows. Thus, the information provided by these screening tests will be useful primarily for monitoring herd-level udder-health programs, especially dry cow management practices. The unit of analysis in this study was the quarter. However, quarters within cow are not independent; infection within one quarter increases the probability that additional quarters in the same cow will be positive (Barkema et al., 1997). This lack of independence was controlled in the statistical analysis by including quarter nested with cow and herd as a random effect. However, a knowledge of the number of quarters above a threshold value also could be of interest in determining rational sampling schemes. This strategy would involve using the cow as the unit of analysis and would require a larger sample size than used in the present study to investigate and validate. Further research with large sample sizes is also needed to determine the sensitivity and specificity of these tests for identifying specific pathogens. In summary, SCC and CMT in early lactation can be used to select quarters for bacteriological culture to verify infection with major pathogens. Based on our Journal of Dairy Science Vol. 84, No. 9, 2001

2024

SARGEANT ET AL.

results, quarter samples taken 3 d postcalving that have any positive reaction to CMT should be cultured to determine whether an IMI is present and to identify the specific pathogen type. Information on IMI status of the herd at calving will aid producers in monitoring existing udder health programs and assessing the impact of new control practices. REFERENCES Barkema, H. W., H. A. Deluyker, Y. H. Schukken, and T. J. G. M. Lam. 1999. Quarter-milk somatic cell count at calving and at the first six milkings after calving. Prev. Vet. Med. 38:1–9. Barkema, H. W., Y. H. Schukken, T. J. G. M. Lam, D. T. Galligan, M. L. Beiboer, and A. Brand. 1997. Estimation of interdependence among quarters of the bovine udder with subclinical mastitis and implications for analysis. J. Dairy Sci. 80:1592–1599. Bassel, L., D. Kelton, A. Godkin, K. Lissemore, K. Leslie, N. Smart, C. Church, and P. Meadows. 2000. Prevalence of intramammary infection at calving in first lactation heifers from Ontario sentinel herds. Pages 169–170 in Proc. 39th Annu. Mtg. Natl. Mastitis Counc., Inc., Atlanta, GA. NMC, Madison, WI. Boddie, R. L., S. C. Nickerson, and R. W. Adkinson. 1993. Evaluation of teat germicides of low iodine concentrations for prevention of bovine mastitis by Staphylococcus aureus and Streptococcus agalactiae. Prev. Vet. Med. 16:111–117. Detilleux, J., J. Arendt, F. Lomba, and P. Leroy. 1999. Methods for estimating areas under receiver-operating characteristic curves: illustration with somatic-cell scores in subclinical intramammary infections. Prev. Vet. Med. 41:75–88. Dohoo, I. R. 1993. An evaluation of the validity of individual cow somatic cell counts from cows in early lactation. Prev. Vet. Med. 16:103–110. Dohoo, I. R., and K. E. Leslie. 1991. Evaluation of changes in somatic cell counts as indicators of new intramammary infections. Prev. Vet. Med. 10:225–237. Eberhart, R. J. 1986. Management of dry cows to reduce mastitis. J. Dairy Sci. 69:1721–1732. Fox, L. K., S. T. Chester, J. W. Hallberg, S. C. Nickerson, J. W. Pankey, and L. D. Weaver. 1995. Survey of intramammary infections in dairy heifers at breeding age and first parturition. J. Dairy Sci. 78:1619–1628. Harmon, R. J. 1994. Physiology of mastitis and factors affecting somatic cell counts. J. Dairy Sci. 77:2103–2112. Kaneene, J. B., and H. S. Hurd. 1990. The National Animal Health Monitoring System in Michigan. III. Cost Estimates of Selected Dairy Cattle Diseases. Prev. Vet. Med. 8:127–140. Kirk, J. H., S. L. Berry, J. P. Reynolds, J. P. Maas, and A. Ahmaki. 1996. Sensitivity and specificity analysis for somatic cell count (SCC) used to predict bacteriologically positive subclinical mastitis at calving in a dairy herd with low SCC. J. Am. Vet. Med. Assoc. 208:1054–1057. Lightner, J. K., G. Y. Miller, W. D. Hueston, and C. R. Dorn. 1988. Estimation of the costs of mastitis, using National Animal Health

Journal of Dairy Science Vol. 84, No. 9, 2001

Monitoring System and milk somatic cell count. J. Am. Vet. Med. Assoc. 192:1410–1413. Martin, S. W., A. H. Meek, and P. Willeberg. 1987. Veterinary Epidemiology: Principles and Methods. Iowa State University Press, Ames. Maunsell, F. P., D. E. Morin, P. D. Constable, W. L. Hurley, and G. C. McCoy. 1999. Use of mammary gland and colostral characteristics for prediction of colostral IgG1 concentration and intramammary infection in Holstein cows. J. Am. Vet. Med. Assoc. 214:1817–1823. Miller, G. Y., P. C. Bartlett, S. E. Lance, J. Anderson, and L. E. Heider. 1993. Costs of clinical mastitis and mastitis prevention in dairy herds. J. Am. Vet. Med. Assoc. 202:1230–1236. National Mastitis Council. 1987. Laboratory and field handbook on bovine mastitis. Hoard and Sons Co., Fort Atkinson, WI. Natzke, R. P. 1981. Elements of mastitis control. J. Dairy Sci. 64:1431–1442. Nickerson, S. C., W. E. Owens, and R. L. Boddie. 1995. Mastitis in dairy heifers: Initial studies on prevalence and control. J. Dairy Sci. 78:1607–1618. Oliver, S. P., M. J. Lewis, B. E. Gillespie, and H. H. Dowlen. 1992. Influence of prepartum antibiotic therapy on intramammary infections in primigravid heifers during early lactation. J. Dairy Sci. 75:406–414. Pankey, J. W., P. A. Drechsler, and E. E. Wildman. 1991. Mastitis prevalence in primigravid heifers at parturition. J. Dairy Sci. 74:1550–1552. Pankey, J. W. 1989. Premilking udder hygiene. J. Dairy Sci. 72:1308–1312. Radostits, O. M., K. E. Leslie, and J. Fetrow. 1994. Herd Health: Food Animal Production Medicine. 2nd ed. W.B. Saunders Co., Philadelphia, PA. Rainard, P., M. Ducelliez, and B. Poutrel. 1990. The contribution of mammary infections by coagulase-negative Staphylococci to the herd bulk milk somatic cell count. Vet. Res. Comm. 14:193–198. Roberson, J. R., L. K. Fox, D. D. Hancock, C. C. Gay, and T. E. Besser. 1994. Coagulase-positive Staphylococcus intramammary infections in primiparous dairy cows. J. Dairy Sci. 77:958–969. SAS User’s Guide: Version 7-1. 1999. SAS Inst., Inc., Cary, NC. Schalm, O. W., and D. O. Noorlander. 1957. Experiments and observations leading to development of the California mastitis test. J. Am. Vet. Med. Assoc. 130:199–204. Sheldrake, R. F., R. J. T. Hoare, and G. D. McGregor. 1983. Lactation stage, parity, and infection affecting somatic cells, electrical conductivity, and serum albumin in milk. J. Dairy Sci. 66:542–547. Smith, K. L., D. A. Todhunter, and P. S. Schoenberger. 1985. Environmental pathogens and intramammary infection during the dry period. J. Dairy Sci. 68:402–417. Timms, L. L., and L. H. Schultz. 1987. Dynamics and significance of coagulase-negative staphylococcal intramammary infections. J. Dairy Sci. 70:2648–2657. Todhunter, D. A., K. L. Smith, and J. S. Hogan. 1995. Environmental Streptococcal intramammary infections of the bovine mammary gland. J. Dairy Sci. 78:2366–2374. Wilson, D. J., R. N. Gonzalez, and H. Das. 1997. Bovine mastitis pathogens in New York and Pennsylvania: Prevalence and effects on somatic cell count and milk production. J. Dairy Sci. 1997:2592–2598.