Digital dermatitis, also known as papillomatous DD or footwarts, is a worldwide, contagious, painful dermatitis affecting the feet of cattle, typically resulting in clinical lameness, reduced milk production, and decreased reproductive efficiency.1 The precise etiology of DD is unknown, although several novel Treponema spp appear to play a major role in its pathogenesis.2–5 Common treatments include chemical disinfectants and extralabel use of antimicrobials applied under a bandage as a paste, as a topical spray, or as footbath additives.
No antimicrobials are currently approved for treatment of DD in the United States. Lincomycin and oxytetracycline are reportedly effective treatments, when applied topically as a spray or under a light bandage.6–9,a Recurrence of lesions after antimicrobial treatment reportedly ranges from 48% to 60% by 49 to 100 days after treatment.10,a Few studies have involved following treated cattle to determine recurrence rates or the mechanisms of recurrence of DD lesions after antimicrobial treatment. The objectives of the study reported here were to determine whether lincomycin was as effective as oxytetracycline in the treatment of DD lesions, whether results of gross visual assessment of treated and untreated DD lesions agreed with results of histologic evaluation, and whether epidermal flora from DD lesions was similar to the flora of healthy digital skin.
Materials and Methods
Cows and case definition—Twenty-five cows from a commercial free stall–housed, 1,200-cow Holstein dairy herd with a high (approx 45%) prevalence of DD were enrolled in the study. A cow was deemed as affected with DD when it had solitary or multiple digital skin lesions with diagnostic pathological features as described elsewhere.10 A healed lesion was defined as grossly healed skin on the originally affected foot. Footbaths and individual cow treatments were not used on this dairy during the study. The study protocol was approved by an institutional animal care and use committee, and owner consent was obtained.
Experimental procedure—Cows with DD were identified by examining the hind feet after cleaning with a water hose during milking from the side in the herringbone-style parlor. Eligible cows had apparently painful, characteristic gross lesions of DD as described elsewhere.10 Existence of pain was assessed by spraying water from the hose directly on the lesions. The following 2 days (baseline; day 1), enrolled cows were separated for treatment upon leaving the parlor. Each cow was positioned on a hydraulic tilt table. Lesions (including the interdigital space) were washed with water from a hose with a nozzle, and lesions were visually scored according to severity as follows: 0 = no visible lesion; 1 = early, ulcerative, flat lesion; 2 = intermediate, raised, ulcerative, granular-appearing lesion; 3 = proliferative, raised lesion with early papillae formation; and 4 = highly proliferative, raised lesion with advance papillae formation. The scoring system was not validated as such but was based on gross pathological observations and was used by our research group in a previous study.11 Visual lesion scores were also reevaluated by the senior author prior to statistical analysis. Photographs were taken of all lesions by use of a 100-mm macro lens at 1:4 magnification.
Prior to enrollment of cows, a random number generatorb was used to generate a list of 25 random values (3 for the control group, 11 for the oxytetracycline group, and 11 for the lincomycin group), and the values were sorted from smallest to largest. Cows were assigned to 1 of the 3 treatment groups from the sorted list of random numbers as they entered the chute for treatment on days 1 or 2 of the study. Cows were subsequently treated as follows: topical treatment with 10 g of lincomycin HCl soluble powderc (n = 11 cows), topical treatment with 10 g of oxytetracycline HCl soluble powderd (11), or no treatment (3; control group). Prior to the onset of the trial, it was agreed that only 3 cows would be allocated to the control group to minimize the number of cows not receiving potentially useful medication yet still providing a control group for comparisons.
For treatments, lincomycin or oxytetracycline was mixed with sufficient deionized water (3 to 4 mL) to create a paste and applied to 10 × 10-cm gauze, placed on the lesion, and held in place with a bandage. Control cows had their feet bandaged after biopsy specimen collection but did not receive any antimicrobial. Dairy personnel were instructed to remove all bandages 4 days after application.
Cows were reexamined on the hydraulic tilt table on days 14 and 30 after treatment application; lesions and healed lesion sites (including the interdigital space) were washed with water from a hose with a nozzle, visually scored, and photographed. On day 30, biopsy specimens were collected from the same sites as day 1 for histologic examination and bacteriologic evaluation. All untreated lesions in control cows were treated at the conclusion of the study on day 30.
Histologic evaluation—Lesions from each of the 25 cows were anesthetized locally with 2% lidocaine, and 6-mm-diameter, full-thickness punch biopsy specimens from lesions were collected and placed in neutralbuffered 10% formalin. The specimens were routinely paraffin embedded, sectioned at a thickness of 4 μm, and stained with H&E and Steiner silver to allow visualization of bacterial morphology.12
Histologic evaluation consisted of scoring the samples on the basis of degree of activity (0 = absent; 1 = focal; 3 = segmental; and 5 = continuous) of the pathological criteria for diagnosing DD, namely ulceration, invasion of the stratum spinosum and papillary dermis by spirochete-dominant bacteria, and reactive inflammation. Each process was interdependent and was therefore given equal weight to yield a total histologic score. Grading consisted of arbitrary assessment of the extent and intensity of each pathological criterion.
Lesions were classified histologically as active, incipient, or healed. Active lesions had ulceration and invasion by profuse mats of spirochetes in the stratum spinosum, papillary dermis, or both, associated with an inflammatory response. Active lesions were subclassified into high (diffuse involvement), moderate (segmental involvement), or slight (focal involvement). Incipient lesions differed from active lesions in that they had no ulceration, bacterial invasion, or inflammation; changes were confined to the stratum corneum, which was involved by parakeratotic hyperkeratosis and spirochetal colonization. Healed lesions had a continuous, fully cornified stratum corneum, no colonization or invasion by spirochetes, and slight to absent perivascular lymphoplasmacytic dermatitis. Those performing the histologic evaluations were unaware of treatment group assignment.
Bacteriologic evaluation—Full-thickness skin biopsy specimens were collected from active DD lesions from 5 cows prior to treatment and from areas adjacent to prior biopsy sites 1 month after treatment with tetracycline (n = 3 cows) or lincomycin (2). Additional biopsy specimens were collected from healthy skin on unaffected, contralateral feet from 2 cows in the tetracycline group, 2 cows in the lincomycin group, and 1 control cow before treatments were initiated. Biopsies were performed as described for histologic evaluation, except that after the procedure, specimens were vigorously rinsed with sterile saline (0.9% NaCl) solution to remove any superficially contaminating bacteria and then placed in a prereduced peptone-yeast-glucose brothe and transported on ice to the laboratory for processing. Direct microscopic evaluations and processing for bacterial cultures were performed within 12 hours after specimen collection.
For microscopic evaluation of bacterial morphotypes and overall numbers of bacteria present, the epidermal surfaces of the biopsy samples were scraped with a sterile scalpel blade and the scraping material was used to prepare Gram- and Steiner silver–stained slides.12,13 In addition, scraping material was suspended in an equal volume of sterile saline solution for examination by darkfield microscopy.
To further assess bacterial populations on the surface of skin lesions or healthy skin, bacterial cultures of biopsy scraping material were performed. Culture of aerobes, anaerobes, Campylobacter spp, and Mycoplasma spp was attempted, as well as culture of spirochetes associated with DD. Aerobic cultures consisted of plating biopsy scraping material on 5% sheep blood agar, MacConkey agar, and chocolate agar plates and incubating samples in 7.5% CO2 for 2 days. Anaerobe cultures were processed on prereduced, anaerobically sterilized Brucella blood agar and phenylethyl alcohol blood agare and incubated at 37°C for 5 days in an anaerobic chamber.f Cultures for Mycoplasma spp and Campylobacter spp were performed as described elsewhere.14,15 Direct plating and broth enrichment cultures for DD-associated spirochetes were also performed as described elsewhere.3
Aerobic bacterial isolates were identified through conventional or commercial identification methods.16,g Isolates not clearly identified by these methods were identified to genus level only or classified solely on the basis of morphological characteristics. Obligate anaerobic bacteria were identified to level II group and species.17 Mycoplasma isolates were identified by positive Dienes staining and susceptibility to digitonin.18 Serotyping attempts for Mycoplasma isolates were performed by use of an immunoperoxidase method19 with antisera prepared against common bovine mycoplasmas that included Mycoplasma bovis (strain 201), Mycoplasma bovigenitalium (ATCC 19852), Mycoplasma bovirhinis (ATCC 27748), Mycoplasma arginini (ATCC 23838), Mycoplasma canadense (ATCC 29418), Mycoplasma alkalescens (ATCC 29103), and Mycoplasma californicum (ATCC 33461). Spirochetes were identified as described elsewhere.3
Statistical analysis—All observations were categorical in nature. Possible scores were integer values from 0 to 4, indicating increasing severity of disease. Because categorical scores are not normally distributed and scores were assessed 3 times for each cow, a GEEh approach to data analysis was used. A multinomial link function was used to consider various combinations of the fixed effects of treatment (control, lincomycin, and oxytetracycline), limb chosen for observation (right or left), period of observation (days 0 to 8, days 9 to 20, and > day 20), and cow. Accordingly, the linear predictor for this analysis had the general form for main effects as follows:
in which ηijkl is the expected value of the mean of the multinomial link function for the lth (l = 1,2,…,25) cow in the kth (k = 1,2,3) treatment observed on the ith (i = 1,2) limb at the jth (j = 1,2,3) time. This model could be expanded to include interactions between pairs of factors (limb, period, or treatment), but for simplicity of presentation, such terms were not included in the preceding formula. The term for cow nested within the limb by treatment interaction, cow (limb•treatment)jkl, was included to accommodate the repeated observations on the same animal. To further assess the difference in gross visual response to treatment at day 14, z statistics were calculated for treated and control cows.
To evaluate the accuracy of the visual scoring system, the presence or absence of disease in visually scored cows was compared with the results of histologic evaluations. In this process, the histologic score was presumed to represent the true presence or absence of disease (ie, the criterion standard), and sensitivity, specificity, and positive and negative predictive values of visual scores were calculated.20,21
Results
Baseline comparisons—At enrollment and before treatment, 14 of 25 (56%) cows (3 control, 5 oxytetracycline-treated, and 6 lincomycin-treated) had a gross visual lesion score of 2, 10 (40%) cows (6 oxytetracycline-treated and 4 lincomycin-treated) had a lesion score of 3, and 1 (4%) cow (lincomycin-treated) had a lesion score of 4. All lesions were assessed as painful as well as visually and histologically active (evidence of ulceration and bacterial invasion) at enrollment. There was no significant (P > 0.09) difference in gross visual lesion scores among the 3 treatment groups at enrollment.
Gross observations—At 1 month after treatment, lesions in 15 of the 22 (68%) antimicrobial-treated cows appeared healed, 6 (27%) cows had a lower lesion score than at baseline, and 1 (5%) had no change in the appearance of the lesion. Of the 3 untreated control cows, 2 had lesion scores that were the same or higher than at baseline, and 1 cow appeared healed at the original site above the heel but had a developing lesion at the more typical site on the interdigital ridge of the same foot.
Predictions of a lesion score of 0 (healed) given the parameter estimates of the GEE model were graphically displayed (Figure 1). Given that the effect of limb was not significant, this characteristic was not included in the model used for these predictions. Similarly, the analysis distinguished only between control and treated cows; that is, the predictions were made from observations for both antimicrobial treatments (lincomycin and oxytetracycline) combined because of the lack of a significant (P > 0.80) difference between lesion responses to lincomycin and oxytetracycline. The prediction model included the variables treatment and period as well as the random effect of cow to account for the repeated nature of lesion scoring across evaluation periods. Probability estimates with overlapping 95% CIs were considered not significantly different.
Probabilities and 95% CIs of cows with DD having a visual lesion score of 0 before treatment (day 1) and 14 and 30 days after topical treatment with (gray bars; n = 22) or without (white bars; 3) antimicrobials (10 g of lincomycin or oxytetracycline in paste form). Values were calculated by use of a GEE model.
Citation: Journal of the American Veterinary Medical Association 237, 5; 10.2460/javma.237.5.555
Probabilities and 95% CIs of cows with DD having a visual lesion score of 0 before treatment (day 1) and 14 and 30 days after topical treatment with (gray bars; n = 22) or without (white bars; 3) antimicrobials (10 g of lincomycin or oxytetracycline in paste form). Values were calculated by use of a GEE model.
Citation: Journal of the American Veterinary Medical Association 237, 5; 10.2460/javma.237.5.555
Probabilities and 95% CIs of cows with DD having a visual lesion score of 0 before treatment (day 1) and 14 and 30 days after topical treatment with (gray bars; n = 22) or without (white bars; 3) antimicrobials (10 g of lincomycin or oxytetracycline in paste form). Values were calculated by use of a GEE model.
Citation: Journal of the American Veterinary Medical Association 237, 5; 10.2460/javma.237.5.555
To further test the hypothesis that treated or control cows were different in their probability of a gross visual lesion score of 0 (healed) at day 14, z statistics were calculated. The parameter for the probability function of a control cow at day 14 being free of disease was 0.584 (SE = 0.98), which was not significantly different from 0 (P = 0.55). When the same parameter was calculated for a cow in the combined treatment group, the value was 2.17 (SE = 0.561), which was significantly different from 0 (P < 0.001). In other words, antimicrobial treatment increased the probability of being healed at day 14, whereas the lack of treatment did not. In addition, antimicrobial-treated cows at day 30 had a significantly higher probability of a 0 lesion score (healed), compared with the control cows.
Sensitivity, the proportion of truly DD-affected cows that were correctly identified as affected by the visual scoring system, was 0.83 (33/40; 95% CI, 0.67 to 0.92). Samples were pooled across evaluation periods to increase the number of available counts in each of the 4 cells. Specificity, the proportion of truly unaffected cows that were correctly identified as unaffected, was 0.89 (8/9; 95% CI, 0.52 to 0.99). The positive predictive value of visual diagnosis was 0.97 (33/34; 95% CI, 0.84 to 0.99), whereas the negative predictive value of visual diagnosis was 0.53 (8/15; 95% CI, 0.53 to 0.79). Probability estimates with overlapping CIs were considered not significantly different.
Histologic evaluation—When the total histopathologic score was compared among treatment groups, there was no significant (P = 0.85) difference in treatment response rates between lincomycin and tetracycline; lesions treated with antimicrobials had a significant (P < 0.005) improvement in histopathologic score, compared with untreated lesions. Gross lesion score had no significant (P = 0.08) effect on histopathologic score. There was a significant (P = 0.049) decrease in histopathologic score with increasing study day.
Bacteriologic evaluation—Results of direct microscopic evaluations for all 5 pretreatment lesions were similar. Collectively, there was a predominance of gramnegative filaments and faintly staining, long, spiral-shaped rods with occasional rare gram-positive cocci and gram-positive and gram-negative rods. Silver staining revealed large numbers of long, spiral-shaped organisms and moderate to large numbers of long, straight rods, some of which appeared pointed at the ends. Rare short, plump rods were occasionally observed. Darkfield examination of suspensions of the scraping material from untreated lesions revealed numerous long, often motile, spiral-shaped organisms, some of varying thicknesses and numerous long, straight rods. Occasional plump short rods were observed.
Gram staining of similarly prepared material from the previously active lesion sites 1 month after treatment revealed small to rare numbers of gram-positive rods in 2 samples, with rare gram-negative rods also present in 1 of those samples. Rare gram-positive cocci were observed in 1 sample. No organisms were detected in the remaining samples. Silver staining and darkfield examination revealed similar bacteriology morphotypes as were observed in the pretreatment lesion biopsy specimens except that the number of gram-negative filaments and long, spiral-shaped rods was markedly reduced. The results from the 5 control skin samples were all similar. No organisms were detected in any of the smears prepared from these biopsy specimens when examined with Gram stain. Only rare short rods or cocci were seen in the control skin biopsy specimens examined with silver stain or by darkfield microscopy.
Aerobic bacterial cultures of active lesions before treatment yielded predominately gram-positive organisms that consisted of various gram-positive pleomorphic rods collectively identified as diphtheroids. Also identified were occasional Streptococcus spp, Staphylococcus spp, and various rare, nonfermentative and fermentative gram-negative rods, including Escherichia coli. In anaerobic cultures, pigmented gram-negative rods, including Porphyromonas levii, were consistently present. Fusobacterium necrophorum and members of the genera Prevotella, Porphyromonas, and Peptostreptococcus were detected, often in small to rare numbers. An obligately anaerobic gram-negative filamentous rod not identified by methods used was regularly encountered. Mycoplasma spp were isolated from all active lesion samples but did not react with any of the available antisera. No Campylobacter spp were isolated from any samples. Spirochetes consistent with those detected in papillomatous DD lesions3 were recovered from all lesion site samples.
A similar bacterial population was isolated from lesion sites after treatment on day 30, although in much smaller numbers. Spirochetes were recovered from 2 of the samples, and Mycoplasma spp were recovered from 4. No Campylobacter spp were recovered.
Cultures obtained from control-skin material did not yield any detectable Campylobacter spp, Mycoplasma spp, or spirochetes. Aerobic cultures of controlskin material consisted predominately of rare to small numbers of diphtheroid bacteria of various colony morphologies. Anaerobic cultures yielded no growth from 2 control-skin samples, and rare Prevotella spp, Porphyromonas spp, and anaerobic gram-positive rods were detected in the remaining samples.
Discussion
In the study reported here, the effectiveness of 2 antimicrobials, lincomycin and oxytetracycline, for the topical treatment of DD in dairy cattle was assessed. No significant differences were evident between lincomycin and oxytetracycline treatment groups on lesion severity score or histopathologic score. By 1 month after treatment, 15 of the 22 (68%) cows treated with antimicrobials appeared healed by gross visual observation. However, 7 of these 15 cows had histologic evidence of colonization by spirochetes. It could not be determined whether these findings represented incomplete healing or recurrence. Whereas visual lesion scores had not significantly decreased from days 1 (before treatment) to 14 (after treatment), both visual lesion scores and histopathologic scores had significantly decreased by 1 month after treatment, regardless of antimicrobial used. Both lincomycin and oxytetracycline treatment resulted in a significant increase in gross visual lesion healing relative to no treatment by 1 month after treatment, but neither antimicrobial treatment had a significant effect on visual lesion healing by 14 days after treatment, even though there appeared to be a clinical difference on day 14. When z statistics were calculated for the day 14 scores, they indicated that the treated cows had a greater than 0 probability of being healed and the control cows did not.
The lack of a significant difference in visual lesion scores between treated and control groups at day 14 according to the GEE model was likely attributable to the small number of control cows at enrollment and the fact that one of the control cows had a lesion that appeared to be healing at the original lesion site. This was the only cow of the 25 evaluated that had a DD lesion distal to the dewclaw on the pastern rather than on the typical site at the interdigital commissural ridge. The lesion site for this control cow also appeared to be healing on day 30 after treatment, and the biopsy specimen was uninterpretable by the pathologist. The cow did, however, have a small, active lesion on the typical site at the interdigital commissural ridge at day 30. We can only speculate why the lesion on this control cow appeared to be healing, but there has been at least 1 study22 that revealed different healing rates for lesions at different locations after treatment. It may be that this location on the proximal phalanx allowed the lesion to dry out and begin healing at the original site after 1 month. Our decision to enroll only a small number of cows in the control group was based on our observations in another study,10 in which we found that spontaneous healing of DD lesions is rare and that the disease is painful. Control cows had no significant decrease in histopathologic lesion scores by day 30.
A subset of the lesions from the treatment study was assessed for bacterial flora prior to and after antimicrobial treatment as a measure of treatment effect of bacterial population. Healthy skin samples were evaluated to serve as a baseline for comparison. Direct microscopic examinations revealed diverse bacterial populations in the active lesions, with certain morphotypes predominant and consistently found among all lesions. Treatment appeared to substantially reduce numbers of all bacterial morphotypes, but treatment had either not eliminated them entirely or they had reemerged but in lesser numbers by the time the day 30 biopsy specimens were collected. The same bacteria recovered from active or healing lesion material were not recovered from control-skin material from the typical site on the interdigital commissural ridge of the rear foot.
Bacterial cultures yielded organisms consistent with the findings of direct microscopic evaluation. Healthy control skin harbored only few numbers of aerobic, predominately gram-positive floras. Uncommonly, anaerobic bacteria were detected but not consistently and no spirochetes, Campylobacter spp, or Mycoplasma spp were isolated. In the active lesions, anaerobic bacteria including P levii, an unidentified obligately anaerobic gram-negative filamentous rod, and spirochetes were the most consistent groups detected along with unidentified Mycoplasma spp. Without a specific culture for Mycoplasma organisms, it is not likely their regular presence would have been recognized since they would not typically have been detected by other culture methods or by direct microscopic examination. In some instances, however, Mycoplasma colonies were detected on the spirochete isolation media. Bacterial culture of lesion sites after treatment (day 30) revealed a similar bacterial composition as that in active lesions, although with substantially lower numbers of each individual morphotype as well as overall bacterial numbers.
The direct smear and bacterial culture results were consistent with what was observed histologically. The bacterial populations that were predominant in active lesions were still present but at greatly reduced numbers or had reestablished colonization or infection by 1 month after treatment. It is likely that the incipient lesions detected histologically in the 1-month posttreatment biopsy specimens still provided surface areas for these bacteria to attach that were not provided in healthy, intact skin. Although the pathogenesis of DD is still unclear, it is obvious that both antimicrobial treatments markedly reduced the numbers of bacteria present. In this study, the treatment regimen involving 2 different antimicrobials either did not result in complete lesion healing in some of the treated cows or resulted in lesions reoccurring at the same site by 1 month posttreatment. Incomplete healing or lesion recurrence appeared to allow for sites for attachment of the bacterial types associated with active lesions. Further studies are required to determine the mechanism of DD lesion recurrence.
If treated DD lesions do not completely heal but rather enter a period of apparent latent infection, then more rigorous or extended treatment protocols may be required to control the infectious process to achieve complete healing. A second possibility is that the underlying infectious activity was due to a new infection that had not yet completely eroded the epidermis, in which situation additional research is required to determine the reason certain cows are more prone to reinfection. A third possibility is that the incompletely healed skin is more prone to reinfection than healthy healed skin, which still leads to the question of why some cows healed and others did not. This discrepancy between gross and histologic findings may also create a false low prevalence estimate of DD in dairies that rely only on gross visual examination to identify infection. This could contribute to failure of DD control programs in dairy herds because of persisting infection and treatment failures that are perceived as treatment cures on the basis of visual examination alone. It will be important to determine the true DD treatment failure and recurrence rates to understand their impact on the epidemiology of DD in dairy cattle.
ABBREVIATIONS
| CI | Confidence interval |
| DD | Digital dermatitis |
| GEE | Generalized estimating equation |
Nuccitelli B, Berry SL. The long term recurrence of digital dermatitis after treatment with lincomycin HCl on a commercial dairy (abstr), in Proceedings. 40th Annu Conv Am Assoc Bovine Pract 2007;289.
Excel 1997, Microsoft Corp, Redmond, Wash.
Lincomix Soluble Powder, Pharmacia & Upjohn Co, Kalamazoo, Mich.
Terramycin-343 Soluble Powder, Pfizer Animal Health, New York, NY.
Anaerobe Systems, Morgan Hill, Calif.
Bactron IV, Bactron Anaerobic/Environmental Chamber, Shel Lab, Sheldon Manufacturing Inc, Cornelius, Ore.
API products, bioMérieux, Hazelwood, Mo.
PROC GENMOD, SAS system for Windows, version 8.02, SAS Institute Inc, Cary, NC.
References
1. Read DH, Walker RL. Papillomatous digital dermatitis and associated lesions of dairy cattle in California: pathologic findings, in Proceedings. 8th Int Symp Disord Rumin Lameness Int Conf Bovine Lameness 1994; 156–158.
2. Choi BK, Nattermann H, Grund S, et al. Spirochetes from digital dermatitis lesions in cattle are closely related to treponemes associated with human periodontitis. Int J Syst Bacteriol 1997; 47:175-181.
3. Walker RL, Read DH, Loretz KJ, et al. Spirochetes isolated from dairy cattle with papillomatous digital dermatitis and interdigital dermatitis. Vet Microbiol 1995; 47:343-355.
4. Evans NJ, Brown JM, Demirkan I, et al. Three unique groups of spirochetes isolated from digital dermatitis lesions in UK cattle. Vet Microbiol 2008; 130:141-150.
5. Stamm LV, Bergen HL, Walker RL. Molecular typing of papillomatous digital dermatitis-associated Treponema isolates based on analysis of 16S-23S ribosomal DNA intergenic spacer regions. J Clin Microbiol 2002; 40:3463-3469.
6. Graham PD. A survey of digital dermatitis treatment regimes used by veterinarians in England and Wales, in Proceedings. 8th Int Symp Disord Rumin Lameness Int Conf Bovine Lameness 1994; 205–206.
7. Moore DA, Berry SL. Comparison of a topical non-antibiotic cream to lincomycin and no treatment for digital dermatitis, in Proceedings. 12th Int Symp Lameness Rumin 2002;369.
8. Shearer JK, Elliott JB. Papillomatous digital dermatitis: treatment and control strategies—part I. Compend Contin Educ Pract Vet 1998; 20:S158–S173.
9. Shearer JK, Hernandez J, Elliott JB. Papillomatous digital dermatitis: treatment and control strategies—part II. Compend Contin Educ Pract Vet 1998; 20:S213–223.
10. Read DH, Walker RL. Papillomatous digital dermatitis (footwarts) in California dairy cattle: clinical and gross pathologic findings. J Vet Diagn Invest 1998; 10:67-76.
11. Moore DA, Berry SL, Truscott ML, et al. Efficacy of a nonantimicrobial cream administered topically for treatment of digital dermatitis in dairy cattle. J Am Vet Med Assoc 2001; 219:1435-1438.
12. Chapin KC, Murray PR, Baron EJ, et al. Stains. In: Manual of clinical microbiology. Washington, DC: ASM Press, 1999; 1674–1678.
13. Carson FL. Microwave Steiner and Steiner procedure for spirochetes, Helicobacter and Legionella organisms. In: Histotechnology: a self-instructional text. Chicago: ASCP Press, 1997; 201–203.
14. Kinde H, Daft BM, Walker RL, et al. Mycoplasma bovis associated with decubital abscesses in Holstein calves. J Vet Diagn Invest 1993; 5:194-197.
15. Campero CM, Anderson ML, Walker RL, et al. Immunohistochemical identification of Campylobacter fetus in natural cases of bovine and ovine abortions. J Vet Med B Infect Dis Vet Public Health 2005; 52:138-141.
16. Quinn PJ, Carter ME, Markey BK, et al. Clinical veterinary microbiology. London: Wolfe Publishing, 1994.
17. Summanen P, Baron EJ, Citron D, et al. Wadsworth anaerobic bacteriology manual. Belmont, Calif: Star Publishing Co, 1993.
18. Razin S, Tully JG. Methods in mycoplasmology. Volume 1—Mycoplasma Characterization. New York: Academic Press Inc, 1983.
19. Imada Y, Uchida I, Hashimoto K. Rapid identification of mycoplasmas by indirect immunoperoxidase test using small square filter paper. J Clin Microbiol 1987; 25:17-21.
20. Altman DG, Bland JM. Diagnostic tests 1: sensitivity and specificity. BMJ 1994; 308:1552.
21. Martin SW, Meek AH, Willeberg P. Veterinary epidemiology: principles and methods. Ames, Iowa: Iowa State University Press, 1987.
22. Hernandez J, Shearer JK. Efficacy of oxytetracycline for treatment of papillomatous digital dermatitis lesions on various anatomic locations in dairy cows. J Am Vet Med Assoc 2000; 216:1288-1290.
