If you don't remember your password, you can reset it by entering your email address and clicking the Reset Password button. You will then receive an email that contains a secure link for resetting your password
If the address matches a valid account an email will be sent to __email__ with instructions for resetting your password
Department of Production Animal Health, Faculty of Veterinary Medicine, University of Calgary, Calgary, Alberta, Canada T2N 4N1Alberta Milk, Edmonton, Alberta, Canada T6X 1H1
Digital dermatitis (DD) is the most prevalent foot lesion affecting dairy herds worldwide. Its implications include production losses and decreased animal welfare. Footbathing is the most common herd-level prevention strategy for DD. Because many common footbath products have negative environmental and health consequences, replacement products expected to have improved safety but equal efficacy are being developed. Therefore, the aim of this study was to evaluate the efficacy of a new quaternary ammonium-based commercial footbath product (QAC) for reducing the prevalence of active DD lesions compared with an industry standard (copper sulfate; CuSO4) and typical on-farm footbath practices. A controlled intervention trial was conducted on 19 Alberta dairy farms over 12 wk, with 9 farms allocated to the QAC group (1% QAC daily, 5 d/wk), 5 to the CuSO4 group (5% CuSO4 daily, 5 d/wk), and 5 to a noninterference group (maintained typical footbath practices). A total of 22,285 observations on 3,465 lactating cows were assessed for DD lesions and leg cleanliness in the milking parlor. Five farms discontinued use of the QAC product for various reasons. Noninferiority analysis was used to assess QAC ability to decrease the proportion of cows with 1 or more active DD lesions compared with CuSO4 after 6 wk. Multilevel logistic regression models for repeated measures were used to evaluate efficacy of QAC compared with CuSO4 and noninterference farms in reducing the prevalence of active DD lesions at the foot level over 12 wk. The noninferiority analysis determined that the proportion of cows with 1 or more active DD lesion decreased 2.19 (95% CI: 1.39–3.46) times less after 6 wk of study on the QAC farms compared with CuSO4 farms, making QAC inferior to CuSO4. The multilevel logistic regression models determined that the proportion of active DD lesions increased in the QAC herds, whereas this proportion decreased in the CuSO4 and noninterference herds over 12 wk. Additionally, cows in mid- and late-lactation had a higher odds of having active DD compared with fresh cows. Older cows (parity 3 and ≥4) had a decreased odds of active DD compared with first-parity cows. At the farm level, a higher baseline active DD prevalence resulted in increased odds of active DD; however, this did not modify the effect of treatment or week of study. We concluded that QAC was inferior to CuSO4 and typical on-farm footbath practices, and further development of novel footbath products is required to develop an ideal alternative.
Digital dermatitis (DD) is an infectious disease causing erosive, ulcerative, often painful lesions on the heel bulbs of cattle. The consequences of DD in dairy cattle include lameness, hoof conformation changes, losses in milk production, and decreased cow comfort, longevity, and fertility (
) and was the most common (40%) of all foot lesions recorded by hoof trimmers in Alberta, British Columbia, and Ontario dairy herds over a 3-yr period (
) that play a role in disease manifestation and progression. Farm-level risk factors for high prevalence of DD include poor environmental and animal hygiene (
Control of DD typically focuses on improving hygiene to limit bacterial spread and using footbaths with various antibacterial or disinfecting products to prevent DD lesions and control infectious foot diseases (
). Compared with a negative control (no footbath), the use of CuSO4 footbaths resulted in more cows with no lesions (45 vs. 30%) and fewer cows with active (M1 + M2) lesions (21 vs. 70%;
) when used at a 5% solution, 4 consecutives milkings weekly for 5 wk. However, both CuSO4 and formalin have undesirable characteristics. Formalin is carcinogenic (
). This has prompted development of alternative footbath products with equivalent effectiveness to CuSO4 or formalin but without the negative repercussions for the environment and human health.
The experimental product (QAC, DeLaval Manufacturing, Kansas City, MO) is a formulation containing quaternary ammonium compounds, specifically didecyldimethylammonium chloride and alkyl (C12–C16) dimethylbenzylammonium chloride, acting as broad-spectrum bactericidal and fungicidal disinfectants for the prevention of DD. When evaluated in vitro on Treponema culture without the presence of manure after 30 s and 10 min exposure, MIC and minimum bactericidal concentration (MBC) measurements suggest that QAC was equal to CuSO4 (equal MIC and MBC) but less effective (lower MIC and MBC) than formalin (
). In the presence of 10% manure, QAC was more effective (higher MIC and MBC) than CuSO4 and formalin; however, at 20% manure, the MIC and MBC of QAC were lower than CuSO4 and formalin. Further testing on dairy farms was required, as in vitro models cannot replicate the complexity of situations in the field. Therefore, the aim of this study was to evaluate the efficacy of QAC for decreasing the prevalence of active DD lesions. This was completed using 2 methods: (1) comparison to CuSO4 (active control) using a noninferiority hypothesis that QAC is at least as efficacious as CuSO4, and (2) comparison of the prevalence of active lesions in QAC to those in CuSO4 and a noninterference group (no change in footbath protocol), therefore accounting for farm- and animal-level factors.
MATERIALS AND METHODS
Study Design
The study was designed as a randomized noninferiority intervention study with farms enrolled in 2 different footbath intervention groups. However, the parameters for a noninferiority study should ideally be determined using a meta-analysis or large studies, and results were only available from a single trial, with a small sample size and a relatively short duration of 6 wk (
). Therefore, in addition to the 6-wk noninferiority study, a further evaluation over a 12-wk period was carried out to determine the efficacy of QAC compared with both CuSO4 and a noninterference group for reducing the prevalence of active DD lesions. A noninterference group of farms was added to compare QAC to typical farm footbath protocols.
The objective of noninferiority trials is to demonstrate that an experimental treatment is at least as effective as an established treatment within a pre-specified margin of noninferiority (∂;
). The margin of noninferiority is determined as the smallest reliable effect size based on prior experience with an active control compared with no treatment and further determining the clinically relevant margin of indifference (Δ). This is the margin used in the null hypothesis that a 1% QAC protocol is inferior compared with a 5% CuSO4 protocol in terms of the proportion of cows with active DD lesions over 6 wk expressed as a risk ratio (RR). The ∂ was determined by the smallest reliable effect size of CuSO4 based on the one reported superiority study (
). This effect size corresponded to the lower boundary of the 2-sided 95% confidence interval of the risk reduction of CuSO4 relative to no treatment (a risk reduction of 25.1%), resulting in a ∂ = 1.335 [RR calculated by dividing the CuSO4 effect versus no treatment (0.749) into 1 (1 ÷ 0.749 = 1.335)]. Thereafter, Δ was calculated to maintain at least 50% of the smallest reliable effect of CuSO4 (50% retention of 25.1% risk reduction;
), calculated by taking 50% of the natural logarithm of RR and converting it back to a RR (e[ln(1.355)×0.5]). Thus, Δ was defined as Δ = 1.155. Noninferiority of QAC compared with CuSO4 was concluded if the upper bound of the 2-sided 95% CI of the RR between QAC and CuSO4 was smaller than Δ.
Farm Selection and Enrollment
Sample size was calculated based on the effect of the reference treatment (CuSO4) from the trial comparing CuSO4 to a negative control (
) and an assumption of a 20% difference between the standard treatment (CuSO4) and the experimental treatment (QAC) with a noninferiority limit of 50%, resulting in a minimum of 58 active DD lesions per group. Assuming a prevalence of active DD lesions of 15% and an average herd size of 140 cows, 5 farms would be required per treatment group allowing for cows being dried off or culled.
Fifteen Alberta dairy herds were recruited by phone and at the 2013 Western Canadian Dairy Seminar (Red Deer, AB, Canada), based on the following criteria to correctly reflect the target population of typical Alberta dairy farms: willingness to participate, adequate number of footbaths, free stall housing, use of a milking parlor, ≥60 lactating dairy cows, >90% Holstein-Friesian cows, and a maximal distance from the University of Calgary (Calgary, AB, Canada) of 200 km. Individual cow data on DIM, parity, and test-day milk production were obtained from CanWest DHI (Guelph, ON, Canada); however, 2 farms did not participate in the CanWest DHI program. However, data collected on these 2 non-DHI farms were similar to data from farms participating in DHI. Information on each farm's footbath protocol before the intervention was also collected. Data were collected between April 23, 2013, and October 28, 2013. All methods were approved by the Animal Care Committee (AC13–0055) of the University of Calgary before contacting participating farmers.
Participating farms were listed in order of enrollment and the first farm was randomly assigned to a group, and subsequent farms were sequentially assigned to either a CuSO4 footbath treatment (denoted C1–C5) or the QAC footbath treatment (denoted Q1–Q5). Additionally, another 5 farms were enrolled and designated as noninterference farms (denoted NI1–NI5) that maintained their current DD management protocol for the length of the study. Four farms designated to the QAC treatment discontinued use of the product after wk 6 of the study. To replace these 4 farms, 4 additional farms were enrolled from the pool of volunteers recruited at the beginning of the study and designated to the experimental treatment (denoted Q6–Q9) after 4 to 6 wk. One of these 4 additional farms discontinued use of the product after wk 3 of the study; this farm was not replaced. For animal welfare and ethical reasons, farmers were allowed to continue their regular hoof trimming and individual cow DD treatments.
Footbath Protocols
Baseline DD prevalence was determined at the first visit (wk 0). The Monday following this visit, the farms started using their assigned footbath protocol. Farms assigned to CuSO4 (C1–C5) or QAC (Q1–Q9) used a freshly prepared footbath once a day after milking for 5 consecutive days (Monday to Friday), for 12 wk. The number of cow passages through the footbath did not exceed 150. Noninterference farms (NI1–NI5) continued their routine footbath protocols (AppendixTable A1). Footbath concentrations were 1% for QAC, as recommended by the manufacturer, and 5% for CuSO4 (
. In summary, M0 is defined as normal digital skin with no evidence of dermatitis; M1 if a small (<2 cm in diameter) circumscribed red to gray epithelial defect is present; M2 if an ulcerative active lesion ≥2 cm in diameter with a red-gray surface; M3 (healing stage) after M2 lesion surface becomes firm and scar-like; M4 (chronic stage) if the lesion surface is raised with brown or black tissue, hyperkeratotic, scaly, or proliferative. The M4.1 lesions, small red circumscribed lesions occurring within the boundaries of an existing M4 lesion (
), were not scored as such, and therefore lesions of this description would exist within the M1 and active categories. Assessment of DD in the CuSO4 and QAC groups was conducted at 3-wk intervals for 12 wk (for a total of 5 evaluations), whereas in noninterference groups feet were scored at 6-wk intervals for 12 wk (a total of 3 evaluations). A single observer (CJ) was trained in the detection of DD lesions through the use of images and definitions, and this observer scored all feet in the study to remove inter-observer discrepancies.
Leg Cleanliness
Cows were scored for leg cleanliness at each parlor inspection using the scoring system developed by
. In short, cleanliness on the lateral lower hind legs, from the coronary band to the middle of the tarsal joint, was scored from 1 to 4 according to varying contamination: 1 = fresh manure for <50%; 2 = fresh manure for >50%; 3 = dried caked and fresh manure for >50%; and 4 = entire area with dried caked manure.
Statistical Analyses
Data were entered into Microsoft Excel (Microsoft Corp., Redmond, WA) and statistical analyses were performed using Stata 13.1 (StataCorp, 2013, College Station, TX). For all analyses, a P-value of <0.05 was considered significant. Descriptive analyses were performed using the M-stage scoring system and also using a simplified scoring system: no DD lesions (M0), active DD lesions (M1 and M2 merged), and chronic DD lesions (M3 and M4 merged). Merging of lesion stages into active and chronic lesions was because of the common misclassification of M3 and M4 stages (
). The difference in the proportion of lesions at baseline (wk 0), middle (wk 6), and end of the study (wk 12) was assessed using McNemar's test. Comparisons of lesion prevalence between treatments was assessed using chi-squared and 2-sample tests of proportions. Cow-level characteristics included leg cleanliness [collapsed into clean (score 1 and 2) and dirty (score 3 and 4) due to the low frequency of scores 1 and 4], daily milk production, DIM, and parity. Baseline active DD prevalence was considered a herd-level variable. Comparisons of herd-level and cow-level characteristics were done using chi-squared and Student's t-tests for categorical and continuous variables, respectively.
Efficacy of QAC for the noninferiority hypothesis was assessed in a multilevel generalized linear model using a modified Poisson approach to estimate relative risk with robust error variances (
) with the prevalence of cows with active DD as the outcome, accounting for clustering at the farm level with treatment as a fixed effect. Treatment effect over time was included in the model and milk production was included as a random effect as mean milk production varied across the 3 groups of farms (Table 1).
Table 1Characteristics (mean ± SD) and baseline active (M1 + M2)
M1 = a small (<2 cm in diameter) circumscribed red to gray epithelial defect; M2 = an ulcerative active lesion ≥2 cm in diameter with a red-gray surface.
digital dermatitis (DD) prevalence in percentage for 19 participating freestall dairy farms in Alberta, Canada
Within a row, means without a common superscript differ (P < 0.05).
a–c Within a row, means without a common superscript differ (P < 0.05).
1 M1 = a small (<2 cm in diameter) circumscribed red to gray epithelial defect; M2 = an ulcerative active lesion ≥2 cm in diameter with a red-gray surface.
Further analysis on the efficacy of CuSO4 and noninterference farms compared with QAC in reducing active DD lesions over 12 wk was assessed using a multilevel logistic regression model with repeated measures. Models including all drop-out farms were compared with models excluding drop-out farms, to assess the effect of drop-out farms on the treatment effect. Active DD lesions over 12 wk were the outcomes of interest. The statistical unit was the foot within cow. Univariable analyses were performed to assess associations and outcome covariates. Treatment group was forced into the final regression model as a fixed effect, whereas farm, cow, and foot were included as random effects. Baseline active DD prevalence was included in the analysis as a herd-level variable. Cow- and herd-level variables significant at P ≤ 0.05 were retained in the final model. Additionally, if confounding was present (i.e., removal of any variable resulted in a 30% change in the estimate of any other significant predictor), that variable was also retained in the final model. Baseline active DD prevalence and daily milk production were included as continuous variables, and DIM was categorized as follows: fresh: 0 to 44 DIM, early: 45 to 99 DIM, mid: 100 to 199 DIM, and late lactation: ≥200 DIM, whereas parity was categorized as 1, 2, 3, and ≥4. Predictor variables considered included treatment group, baseline active DD prevalence (centered on the lowest prevalence), week of trial (centered on wk 0), parity, DIM, and leg cleanliness. Two-way biologically relevant interactions (e.g., parity and DIM, parity and leg cleanliness) were tested among the significant predictors in the main effects model, but none was retained (P > 0.05 in all cases).
RESULTS
Descriptive Results
Average herd size was 143 lactating cows (ranging from 71–274), whereas the QAC, CuSO4, and noninterference farms on average had 148, 173, and 103 lactating cows, respectively (Table 1).
All 5 CuS04 farms completed the 12-wk study. Of the 9 enrolled QAC farms, 4 (Q5, Q6, Q7, and Q8) completed the 12-wk study. These 4 farms discontinued use of QAC product because of a real or perceived increase in incidence and prevalence of DD (AppendixTable A2) and did not complete the 12-wk intervention. Farm Q9 discontinued use of the product after wk 3, did not complete the 12-wk intervention, and was therefore excluded from analyses. Farm Q9 also discontinued use of the QAC product because of a perceived increased in incidence and prevalence of DD.
A total of 22,285 observations on 3,465 lactating cows were collected throughout the study period. Parity and DIM were not different between QAC, CuSO4, and noninterference farms; however, mean daily milk production was higher in the CuSO4 farms than in the QAC and noninterference farms (Table 1). At baseline (wk 0), 60% of feet were affected by any stage of DD (ranging among herds from 42 to 85%), whereas the prevalence of feet with active lesions (M1 or M2) was 4.9% (range 0.7–24%; Figure 1, AppendixTable A1). The baseline prevalence of feet with active lesions differed between groups with 5.2% in the QAC group, 2.4% in the CuSO4 group, and 8.7% in the noninterference group (Figure 1).
Figure 1Foot-level prevalence of active (M1 + M2) and chronic (M3 + M4) digital dermatitis lesions in each of the 3 treatment groups [quaternary ammonium compound product (QAC; DeLaval Manufacturing, Kansas City, MO), copper sulfate, and noninterference] over the course of the study. a–cWithin a treatment group, means without a common letter differ (P < 0.05). M1 = a small (<2 cm in diameter) circumscribed red to gray epithelial defect; M2 = an ulcerative active lesion ≥2 cm in diameter with a red-gray surface; M3 = (healing stage) after M2 lesion surface becomes firm and scar-like; M4 = (chronic stage) the lesion surface is raised with brown or black tissue, hyperkeratotic, scaly, or proliferative.
In the CuSO4 and noninterference group, prevalence of feet with no DD lesion (M0) increased throughout the 12 wk of the study (Figure 1). In the QAC group, prevalence of feet with no DD lesion increased from 0 to 6 wk (38 to 44%); however, the prevalence of feet without DD lesions was not different in wk 12 from wk 0 or 6. Prevalence of feet with active lesions did not change throughout the study period in the CuSO4 group. The noninterference group experienced a decrease in active lesions from 0 to 6 wk (8 to 5%) and then a subsequent increase to higher than baseline levels in 12 wk (11%). In the QAC group, the prevalence of active lesions increased from 0 to 6 wk (5 to 8%) and then was not different to 0 and 6 wk in 12 wk (9%). Prevalence of feet with chronic lesions decreased over 12 wk in the CuSO4 and noninterference group by 19 and 17%, respectively, whereas in the QAC group, prevalence of chronic lesions decreased by 8% from 0 to 6 wk and then remained constant to 12 wk.
Noninferiority Analysis
The proportion of cows with one or more active DD lesion decreased 2.19 (95% CI: 1.39–3.46; P = 0.001) times less after 6 wk of study on the QAC farms compared with CuSO4 farms (Table 2). The upper bound of this RR (3.46) was greater than the previously described noninferiority margin (Δ = 1.16). Therefore, we failed to reject the null hypothesis and thus cannot conclude noninferiority. In fact, no evidence indicated that QAC had any effect, as the RR was greater than both ∂ and Δ [2.19 > 1.16 (Δ) and 1.36 (∂)]. Additionally, because the lower bound of the 95% confidence interval of the effect of QAC (1.39) was also greater than the noninferiority margin (1.16), QAC was inferior to CuSO4 at decreasing the proportion of cows with one or more active DD lesion over 6 wk of footbathing.
Table 2Final multilevel generalized linear model using a modified Poisson approach to estimate relative risk with robust error variances with the prevalence of cows with active digital dermatitis (DD; M1 + M2) as the outcome after 6-wk intervention with 2 treatment groups on 11
Two farms excluded from analysis [quaternary ammonium compound product (QAC; DeLaval Manufacturing, Kansas City, MO) group] because they were not DHI participants.
M1 = a small (<2 cm in diameter) circumscribed red to gray epithelial defect; M2 = an ulcerative active lesion ≥2 cm in diameter with a red-gray surface.
1 Two farms excluded from analysis [quaternary ammonium compound product (QAC; DeLaval Manufacturing, Kansas City, MO) group] because they were not DHI participants.
2 M1 = a small (<2 cm in diameter) circumscribed red to gray epithelial defect; M2 = an ulcerative active lesion ≥2 cm in diameter with a red-gray surface.
In the 12 wk of the study, the proportion of active DD lesions increased in the QAC herds, whereas this proportion decreased in the CuSO4 and noninterference herds (Table 3, Figure 2). Additionally, mid- and late-lactation cows had a higher odds of having active DD compared with fresh cows. Older cows (parity 3 and ≥4) had a decreased odds of active DD compared with first-parity cows. Additionally, farms with increased baseline active DD prevalence, the odds of active DD increased; however, this did not modify the effect of treatment or week of study. Excluding drop-out farms did not alter the results of the model; therefore, all farms were retained in the final model.
Table 3Final repeated measures multilevel logistic regression model for active digital dermatitis (DD) lesions detected in the milking parlor after 12-wk intervention with 3 treatment groups on 16
Two farms were excluded from analysis [quaternary ammonium compound product (QAC; DeLaval Manufacturing, Kansas City, MO) group] because they were not DHI participants.
1 Two farms were excluded from analysis [quaternary ammonium compound product (QAC; DeLaval Manufacturing, Kansas City, MO) group] because they were not DHI participants.
2 1% QAC footbath daily 5 d/wk.
3 CuSO4 = 5% CuSO4 footbath daily 5 d/wk.
4 Noninterference = no change in footbath protocol.
Figure 2Predicted foot-level prevalence of active (M1 + M2) digital dermatitis (DD) lesions in each of the 3 treatment groups [quaternary ammonium compound product (QAC; DeLaval Manufacturing, Kansas City, MO), copper sulfate, and noninterference] over the course of the study. M1 = a small (<2 cm in diameter) circumscribed red to gray epithelial defect; M2 = an ulcerative active lesion ≥2 cm in diameter with a red-gray surface.
Results of the noninferiority analysis were that 1% QAC was inferior to 5% CuSO4 after 6 wk of the study at reducing the number of cows with active DD lesions. This analysis required the results of
for comparison parameters, however, these studies were not identical, as there were variations in baseline prevalence, footbath use, and cow-level characteristics that we attempted to take into account. The marked difference in baseline active DD prevalence between farms enrolled in this study (5 and 9% for CuSO4 and QAC, respectively) and the cows included in the
study (39 and 56% for the CuSO4 and control groups, respectively) may have limited the effect of CuSO4 in this study, as the prevalence of active DD lesions in the CuSO4 group was already relatively low. After 6 wk, a 20% decrease of active DD lesions was observed in the CuSO4 group compared with a 37.5% decrease in the
study. Conversely, a 44% increase in active DD lesions occurred in the QAC group, whereas a 79.5% increase in active DD lesions was reported in the control group of
study was likely due to footbath protocols used before the study. Seventeen of the 19 farms enrolled in our study used a routine footbath protocol before the start of the study, which was likely responsible for the relatively low DD prevalence. However, no mention was made of the baseline adoption of footbathing by
study, calculation of Δ using a RR (as opposed to a risk difference, which is the risk of an event in the experimental group minus the risk of the event in the control group) was done to improve validity (
and the present study were that average DIM, parity, and milk production were higher in the present study. However, only milk production differed between the QAC and CuSO4 farms, but did not modify the results of statistical analyses. For future noninferiority trials, it would be advisable to include disease prevalence at the baseline in the enrollment criteria to ensure the effect of the reference control.
To our knowledge, this was the first attempt to compare footbath protocols using noninferiority. The advantage of noninferiority trials is that they are intended to show whether a new treatment is as efficacious as an established treatment, and typically the new treatment has other advantages. This is congruent with the desire to develop footbath products with similar efficacy to CuSO4 without the environmental consequences for a disease in which having no treatment controls may be unethical. However, the paucity of evidence on the effectiveness of footbath protocols compared with no treatment makes this type of analysis challenging.
Further analysis of the results indicated that the CuSO4 protocol reduced the prevalence of active DD lesions compared with the QAC protocol. The prevalence of active DD lesions in the noninterference groups also decreased compared with the QAC group, suggesting QAC was less effective than footbathing practices in the 5 noninterference herds. This could be due to farmers' increased awareness of DD knowing that DD lesions were being routinely identified and recorded, and potentially improving routine control practices on farms. However, this increased awareness would likely be similar across treatment groups, and therefore the ineffectiveness of QAC occurred despite it. The wide variation of footbath protocols in the noninterference group was expected (
). Because we deemed it unethical to include a negative control group while a product with proven efficacy, CuSO4, is available to farmers, we decided to include a noninterference group to compare results with typical on-farm practices. Indeed, protocols of the noninterference group were representative of those identified on 81 Alberta farms by
in terms of product, frequency, concentration, and dimensions. Unfortunately, this noninterference group was added after commencement of the trial, and these 5 farms were not included in the randomization procedure. Notwithstanding, the farms were similar in terms of herd characteristics (parity, DIM, and milk yield) and therefore results were likely unaffected.
Prevalence of active DD lesions was also influenced by parity and DIM, with higher parity cows having decreased odds of having active lesions and mid- and late-lactation cows having increased odds of active lesions, consistent with findings in other studies (
). Additionally, farms with higher active DD prevalence at baseline were more likely to have higher active DD prevalence throughout the course of the study, but this did not influence treatment or week of study. Despite baseline prevalence being less than the 15% assumed in the sample size calculation, there was sufficient power to detect a difference between the QAC and CuSO4 groups.
A 3-wk scoring interval was chosen in the CuSO4 and QAC groups to best capture the effect of the footbath protocols being tested while visiting such a large number of farms. The noninterference group acted as a control to compare the more intensive footbath strategies to what typically occurs on farm, and thus only 6-wk intervals were used in this group. Both the 3- and 6-wk intervals provide an opportunity of missing transition stages between inspections (
), but this detail was not required for the noninterference group and should be limited between the 3-wk intervals. Unfortunately, the 3-wk scoring intervals allowed for the possibility of topical treatment of DD lesions to occur between visits that may not have been recorded. Feet with wraps (from treatment by farmers or hoof trimmers) present during observations were scored as such, and that observation was considered missing from the analysis. During the study, feet observed with wraps only occurred for 0.8% (186/22,952) of the observations and the results of the study would not differ due to these treatments. Additionally, the routine treatment and hoof trimming schedule was maintained for all farms over the course of the study, and therefore the efficacy of the footbath protocols was assessed as a component of the farms' current DD management. Furthermore, milking parlor DD inspections, although highly accurate and reliable at determining DD presence, have variable accuracy among M-stages, and DD lesions that occur in the interdigital space are not detected (
). This study endeavored to address DD prevalence as seen by the farmer, which would typically occur in the milking parlor. In another study conducted in Alberta, Canada, 18% of DD lesions occurred in the interdigital space, and only 6% occurred on the front feet (
). Therefore, it is promising that the effect of footbath protocols on DD presence and active DD lesion prevalence can be captured with routine, noninvasive parlor inspection.
This study has been one of few footbath protocol evaluations conducted on multiple farms. Most research on footbath protocols has been conducted on single farms (
Effectiveness of different regimens of a collective topical treatment using a solution of copper and zinc chelates in the cure of digital dermatitis in dairy farms under field conditions.
) has demonstrated that including multiple farms with varying characteristics can improve outcome precision.
The high number of farms that opted to stop participating raises concerns regarding the farmer-perceived effectiveness of QAC. Although not all drop-outs were due to outbreaks (AppendixTable A2), the proportion of active lesions in the QAC group did increase considerably in 6 wk (Figure 1), which warranted the decision to discontinue use of the product. Other than increased proportion of lesions, skepticism regarding the product by farmers was a major challenge associated with the evaluation of QAC. Although it is very difficult to achieve under practical circumstances, blinding farmers and researchers to treatments should be a priority for future footbath product evaluations. This challenge also highlighted that from a farmer's perspective, results may be expected within the first 6 wk. Additionally, the use of a noninterference group allowed a comparison to typical footbath protocols and the variation that may occur with these characteristics. This noninterference group acts as a control in a situation where a negative control would have been unethical. Although QAC increased active DD lesions, the comparison to the noninterference group serves to inform that QAC is neither equivalent to CuSO4 or what farmers are typically doing.
Additional factors that may need to be considered for footbath protocols and study comparisons are footbath dimensions, temperature, and water pH. Literature recommendations are a length of 3 m and a depth of 28 cm to ensure optimal contact with the foot (
) are being emphasized as potential factors affecting footbath efficacy, although limited evidence is available regarding optimal levels of these factors. During this study, average weekly environmental temperatures ranged from −1 to 17°C, but pH was not recorded. Further research should assess the effects of these factors on footbath efficacy. Many risk factors differ among farms (animal and environmental hygiene, purchase of cattle, herd size, and cow-level factors) and the combinations of these will affect DD prevalence and need to be considered when developing DD control strategies.
Unfortunately, the issue of environmental residues associated with CuSO4 remains, and future studies involving this and other products are needed to determine the most effective protocol to decrease prevalence of DD lesions. Although field studies with multiple farms provide a comprehensive assessment of new footbath protocols, methods to attempt to reduce the effect of farmer skepticism of new products should be identified.
CONCLUSIONS
Noninferiority of QAC (1% solution, once a day for 5 d) compared with CuSO4 (5% solution, once a day for 5 d) for reducing the prevalence of active DD lesions could not be demonstrated, and QAC was inferior to CuSO4 in decreasing the proportion of cows with active DD lesions. On farms using CuSO4 or noninterference (no change in footbath protocol), farms experienced a decrease in active DD lesions compared with QAC. At the manufacturer-recommended protocol, QAC should not be considered as a replacement for CuSO4 footbaths.
ACKNOWLEDGMENTS
The authors thank participating farmers for their willingness, time, and cooperation on this project and Miranda Marchant (University of British Columbia, Vancouver, British Columbia, Canada) for her help in data collection. We thank Laura Solano (University of Calgary, Calgary, Alberta, Canada) for her knowledge and support during the field trial and her mentorship overall. Additionally, we thank Diego Nobrega (University of Calgary, Calgary, Alberta, Canada) for his support in t statistical analyses. We are also grateful to Elizabeth French and Mario Lopez Benavides (both of DeLaval Manufacturing, Kansas City, MO) for contributions to study design, sponsorship, and funding.
APPENDIX
Table A1Characteristics of footbathing practices for lactating cows before intervention [with copper sulfate (C1–C5), quaternary ammonium compound product (QAC; Q1–Q9), or nonintervention (NI1–NI5)] on 19 participating freestall dairy farms in Alberta, Canada
M1 = a small (<2 cm in diameter) circumscribed red to gray epithelial defect; M2 = an ulcerative active lesion ≥2 cm in diameter with a red-gray surface. DD = digital dermatitis.
Farm Q9 discontinued before wk 3 and was excluded from the analysis.
6.0
1 footbath
204 × 108 × 10
5% CuSO4
2 milkings/wk
150
NI1
1.4
1 footbath, alternate between exit lanes
230 × 30 × 19
5% CuSO4, formalin (unknown)
3 wk CuSO4, 1 wk formalin
290
4 consecutive milkings, 2 d/wk 1 exit lane
2 d/wk other exit lane
NI2
0.7
1 footbath, split design
232 × 36 × 18
7% CuSO4, 5% HealMax
CuSO4: 3 consecutive milkings/wk
230
HealMax: 2 consecutive milkings/wk
NI3
1.4
1 footbath
269 × 69 × 18
1% CuSO4, 4% formalin
CuSO4: 3 milkings/wk
140
Formalin: 4 milkings/wk
NI4
23.6
1 footbath
79 × 305 × 9
5% acidified CuSO4, 5% acidified ZnSO4
CuSO4: 2 milking/wk, every 2 wk
120
ZnSO4: 2 milkings/wk, every 2 wk
NI5
9.4
No footbath protocol
NA
NA
NA
NA
1 M1 = a small (<2 cm in diameter) circumscribed red to gray epithelial defect; M2 = an ulcerative active lesion ≥2 cm in diameter with a red-gray surface. DD = digital dermatitis.
2 Victory (Westfalia-Surge, GEA North America, Drummondville, Quebec, Canada); HealMax (AgroChem Inc. Saratoga Springs, NY).
3 Farms Q1 to Q4 discontinued use of QAC (DeLaval Manufacturing, Kansas City, MO) protocol after wk 6 and were included until wk 6.
4 NA = not available.
5 Farm Q9 discontinued before wk 3 and was excluded from the analysis.
M1 = a small (<2 cm in diameter) circumscribed red to gray epithelial defect; M2 = an ulcerative active lesion ≥2 cm in diameter with a red-gray surface; M3 = (healing stage) after M2 lesion surface becomes firm and scar-like; M4 = (chronic stage) the lesion surface is raised with brown or black tissue, hyperkeratotic, scaly, or proliferative.
Farms denoted C1 to C5 used 5% copper sulfate (CuSO4) footbath daily 5 d/wk; farms denoted Q1 to Q8 used 1% quaternary ammonium compound (QAC; DeLaval Manufacturing, Kansas City, MO) footbath daily 5 d/wk; farms denoted NI1 to NI5 maintained their regular footbath protocols. The CuSO4 and QAC farms were visited every 3 wk and NI farms were visited every 6 wk. Farms Q1 to Q4 discontinued the QAC protocol after 6 wk.
Item
Wk 0
Wk 3
Wk 6
Wk 9
Wk 12
Farm C1
No DD
70 (36.8)
100 (47.9)
106 (52.7)
108 (53.7)
146 (69.5)
DD active
10 (5.3)
4 (1.9)
2 (1.0)
3 (1.5)
8 (3.8)
DD chronic
110 (57.9)
105 (50.2)
93 (46.3)
90 (44.8)
56 (26.7)
Total
190
209
201
201
210
Farm C2
No DD
314 (57.3)
288 (51.7)
335 (56.1)
379 (64.5)
406 (70.1)
DD active
5 (0.9)
3 (0.5)
2 (0.3)
2 (0.3)
2 (0.4)
DD chronic
229 (41.8)
266 (47.8)
260 (43.6)
207 (35.2)
171 (29.5)
Total
548
557
597
588
579
Farm C3
No DD
132 (25.9)
148 (29.8)
169 (33.3)
170 (32.8)
190 (40.7)
DD active
19 (3.7)
26 (5.2)
21 (4.1)
18 (3.5)
22 (4.7)
DD chronic
359 (70.4)
323 (65.0)
318 (62.6)
331 (63.8)
255 (54.6)
Total
510
497
508
519
467
Farm C4
No DD
123 (38.7)
170 (51.5)
191 (59.3)
212 (62.7)
222 (67.5)
DD active
4 (1.3)
8 (2.4)
11 (3.4)
2 (0.6)
5 (1.5)
DD chronic
191 (60.1)
152 (46.1)
120 (37.3)
124 (36.7)
102 (31.0)
Total
318
330
322
338
329
Farm C5
No DD
60 (36.5)
72 (45.0)
67 (42.4)
87 (55.1)
90 (54.9)
DD active
3 (1.8)
0 (0)
1 (0.6)
3 (1.9)
2 (1.2)
DD chronic
101 (61.6)
88 (55.0)
90 (57.0)
68 (43.0)
72 (43.9)
Total
164
160
158
158
164
Farm Q1
No DD
52 (23.6)
61 (35.1)
89 (41.4)
—
—
DD active
6 (2.7)
6 (3.5)
14 (6.5)
—
—
DD chronic
162 (73.6)
107 (61.5)
112 (52.1)
—
—
Total
220
174
215
—
—
Farm Q2
No DD
97 (33.9)
58 (20.8)
76 (27.4)
—
—
DD active
24 (8.4)
37 (13.3)
63 (22.7)
—
—
DD chronic
165 (57.7)
184 (66.0)
138 (49.8)
—
—
Total
286
279
277
—
—
Farm Q3
No DD
280 (52.3)
381 (65.6)
365 (61.2)
—
—
DD active
24 (4.5)
26 (4.5)
14 (2.4)
—
—
DD chronic
231 (43.2)
174 (30.0)
217 (36.4)
—
—
Total
535
581
596
—
—
Farm Q4
No DD
94 (28.9)
95 (29.7)
116 (34.5)
—
—
DD active
5 (1.5)
17 (5.3)
14 (4.2)
—
—
DD chronic
226 (69.5)
208 (65.0)
206 (61.3)
—
—
Total
325
320
336
—
—
Farm Q5
No DD
36 (15)
37 (15.2)
43 (16.9)
43 (16.0)
64 (24.1)
DD active
19 (7.9)
8 (3.3)
11 (4.3)
9 (3.4)
14 (5.3)
DD chronic
185 (77.1)
198 (81.5)
200 (78.7)
216 (80.6)
188 (70.7)
Total
240
243
254
268
266
Farm Q6
No DD
155 (42.9)
144 (42.4)
174 (44.3)
164 (45.2)
142 (36.5)
DD active
24 (6.7)
12 (3.5)
27 (6.9)
26 (7.2)
40 (10.3)
DD chronic
182 (50.4)
184 (54.1)
192 (48.9)
173 (47.7)
207 (53.2)
Total
361
340
393
363
389
Farm Q7
No DD
92 (46.0)
98 (51.6)
104 (55.0)
90 (46.9)
104 (55.3)
DD active
9 (4.5)
20 (10.5)
18 (9.5)
15 (7.8)
16 (8.5)
DD chronic
99 (49.5)
72 (37.9)
67 (35.5)
87 (45.3)
68 (36.2)
Total
200
190
189
192
188
Farm Q8
No DD
92 (51.9)
95 (53.1)
103 (58.5)
102 (52.6)
81 (47.4)
DD active
11 (6.2)
23 (12.9)
22 (12.5)
15 (7.7)
21 (12.3)
DD chronic
74 (41.2)
61 (34.1)
51 (29.0)
77 (39.7)
69 (40.4)
Total
177
179
176
194
171
Farm NI1
No DD
71 (51.1)
—
90 (70.9)
—
99 (68.8)
DD active
2 (1.4)
—
3 (2.4)
—
4 (2.8)
DD chronic
66 (47.5)
—
34 (26.8)
—
41 (28.5)
Total
139
—
127
—
144
Farm NI2
No DD
81 (54.4)
—
86 (58.9)
—
111 (72.1)
DD active
1 (0.7)
—
0 (0)
—
4 (2.6)
DD chronic
67 (45.0)
—
60 (41.1)
—
39 (25.3)
Total
149
—
146
—
154
Farm NI3
No DD
147 (52.7)
—
233 (81.8)
—
256 (82.9)
DD active
4 (1.4)
—
2 (0.7)
—
1 (0.3)
DD chronic
128 (45.9)
—
50 (17.5)
—
52 (16.8)
Total
279
—
285
—
309
Farm NI4
No DD
66 (28.1)
—
103 (40.7)
—
89 (31.3)
DD active
57 (24.3)
—
28 (11.1)
—
79 (27.8)
DD chronic
112 (47.7)
—
122 (48.2)
—
116 (40.9)
Total
235
—
253
—
284
Farm NI5
No DD
80 (35.1)
—
111 (45.7)
—
109 (42.4)
DD active
22 (9.7)
—
21 (8.6)
—
41 (16.0)
DD chronic
126 (55.3)
—
111 (45.7)
—
107 (41.6)
Total
228
—
243
—
257
1 M1 = a small (<2 cm in diameter) circumscribed red to gray epithelial defect; M2 = an ulcerative active lesion ≥2 cm in diameter with a red-gray surface; M3 = (healing stage) after M2 lesion surface becomes firm and scar-like; M4 = (chronic stage) the lesion surface is raised with brown or black tissue, hyperkeratotic, scaly, or proliferative.
2 Farms denoted C1 to C5 used 5% copper sulfate (CuSO4) footbath daily 5 d/wk; farms denoted Q1 to Q8 used 1% quaternary ammonium compound (QAC; DeLaval Manufacturing, Kansas City, MO) footbath daily 5 d/wk; farms denoted NI1 to NI5 maintained their regular footbath protocols. The CuSO4 and QAC farms were visited every 3 wk and NI farms were visited every 6 wk. Farms Q1 to Q4 discontinued the QAC protocol after 6 wk.
Copper sulfate for footbaths—Issues and alternatives.
in: Proceedings of the 2007 Tri-State Dairy Nutrition Conference. 2007 (Fort Wayne, Indiana, USA, 24–25 April 2007. M. L. Eastridge, ed. Ohio State University, Columbus)
Guidance for industry non-interferiority clinical trials. 2010 (U.S. Department of Health and Human Services, Center for Drug Evaluation and Research, Center for Biologics Evaluation and Research. Accessed Jul. 15, 2016.)
Effectiveness of different regimens of a collective topical treatment using a solution of copper and zinc chelates in the cure of digital dermatitis in dairy farms under field conditions.
30377-6&id=gr1.jpg){kind=link}
30377-6&id=gr2.jpg){kind=link}