Effect of footbath protocols for prevention and treatment of digital dermatitis in dairy cattle: A systematic review and network meta-analysis

Preventive Veterinary Medicine 164 (2019) 56–71

Contents lists available at ScienceDirect

Preventive Veterinary Medicine journal homepage: www.elsevier.com/locate/prevetmed

Effect of footbath protocols for prevention and treatment of digital dermatitis in dairy cattle: A systematic review and network meta-analysis

T

C. Jacobsa, , C. Beningera, G.S. Hazlewoodb,c, K. Orsela, H.W. Barkemaa,b ⁎

a

Department of Production Animal Health, Faculty of Veterinary Medicine, University of Calgary, Calgary, AB, Canada Departments of Medicine and Community Health Sciences, University of Calgary, Calgary, AB, Canada c McCaigInstitute for Bone and Joint Health, University of Calgary, Calgary, AB, Canada b

ARTICLE INFO

ABSTRACT

Keywords: Digital dermatitis Footbath Network meta-analysis Systematic review

Digital dermatitis (DD) is an infectious bacterial disease affecting cattle feet. Footbaths are a common herd-level control method for DD; however, variations in product, concentration, and frequency of use complicate comparisons between protocols. The objective of this systematic review was to evaluate all walk-through footbath protocols reported that determined efficacy for prevention and treatment of DD lesions in dairy cattle. An extensive literature search was conducted, including electronic databases and gray literature updated until March 2018. Studies identified included all liquid walk-through footbath protocols that were compared to other footbath protocols or no footbath. Only studies with treatment or prevention of DD lesions as an outcome were included. Literature search and subsequent screening identified 14 publications with 24 treatment comparisons and 24 prevention comparisons. Studies included mostly had low and/or unclear risks of bias. Descriptive analyses were performed according to prevention and treatment outcomes, with case and success definitions summarized as odds ratios (OR). A subsequent network meta-analysis was conducted of 11 studies, comparing 17 protocol comparisons for the prevention outcome and 10 studies comparing 19 protocol comparisons for the treatment outcome, using semi-informative priors in a Bayesian statistical framework. Results of a random effects Bayesian network meta-analysis indicated only 5% copper sulfate used at least 4 times/wk was superior to both no footbath (OR: 5.26; 95% CrI: 1.27–28.8) and a water placebo (OR: 9.47; 95% CrI: 1.03–85.8) in treatment of DD. No other protocol was associated with a reduction in DD, and there were no differences in pairwise comparisons between any active treatments. Unfortunately, for both outcomes (treatment and prevention), small sample sizes (adjusted for clustering) limited the power to detect substantial differences between protocol effects. Thus, despite widespread use of footbaths, limited strength of evidence for use remains and standardized protocols with large sample sizes are needed to further investigate effectiveness of footbath protocols for control of DD. This research did not receive any specific grant from funding agencies in the public, commercial, or notfor-profit sectors.

1. Introduction Digital dermatitis (DD), an infectious bacterial disease affecting cattle feet, is the most common foot lesion affecting dairy cattle in Alberta, Canada (Solano et al., 2016) and most other regions of the world (Orsel et al., 2018). Digital dermatitis affects up to 92% of herds in Europe and North America, with within-herd prevalence ranging from 0 to 74% (Somers et al., 2005; Holzhauer et al., 2006b; Cramer et al., 2008; Solano et al., 2016). Digital dermatitis is often painful, causes lameness and is associated with decreased milk production, poor fertility, hoof conformation changes and impaired animal welfare (Holzhauer et al., 2008; Gomez et al., 2015).

⁎

The etiology of DD is complex, with multiple bacterial species present during stages of disease progression; however, Treponema spp. have been highlighted as the probable causative agent (Krull et al., 2014). This anaerobic spirochete bacteria is present within lesions (Blowey et al., 1994), suggesting their presence to be invasive (Edwards et al., 2003). Possible routes of transmission include contact with manure slurry, direct skin-to-skin contact and transmission from contaminated hoof trimming equipment (Evans et al., 2012; Blowey et al., 2013). Therefore, a common risk factor for DD is contact with cows’ feet and manure slurry in damp conditions, in both free stall and corralbased systems (Rodríguez Lainz et al., 1996; Read and Walker, 1998; Laven, 1999). Thus, DD control typically includes maintaining a clean,

Corresponding author at: University of Calgary, Faculty of Veterinary Medicine, 3330 Hospital Drive NW, Calgary, AB, T2N 4N1, Canada. E-mail address: [email protected] (C. Jacobs).

https://doi.org/10.1016/j.prevetmed.2019.01.011 Received 2 October 2018; Received in revised form 25 January 2019; Accepted 25 January 2019 0167-5877/ © 2019 Elsevier B.V. All rights reserved.

Preventive Veterinary Medicine 164 (2019) 56–71

C. Jacobs et al.

dry environment. A French study demonstrated a relationship between herds with higher (dirtier) leg cleanliness scores having a higher prevalence of DD (Relun et al., 2013). Furthermore, Solano et al. (2017) reported that at the cow level, poor leg cleanliness was associated with a higher prevalence of active DD lesions. Currently, footbathing is the most common herd-level control method for DD; however, numerous products are used in varying concentrations and frequencies. A recent study in Canada reported that of 141 freestall farms, 87% used footbaths, with a range of 1 to 4 products used per farm (Solano et al., 2015). Of these, the most common products used were copper sulfate (CuSO4) and formalin. Unfortunately, CuSO4 has negative environmental consequences due to soil accumulation (Flemming and Trevors, 1989; Hoff et al., 1998; Epperson and Midla, 2007) and formalin is carcinogenic (Doane and Sarenbo, 2014), prompting the need for alternatives. Reported effectiveness of footbath protocols is varied, with most support coming from clinical practice and anecdotal evidence. Comparing footbath protocols available in the literature is difficult; varying products, various concentrations and frequencies of use, combined with using a variety of comparators make assessing and comparing footbath protocols challenging. Additionally, the multi-bacterial etiology of DD combined with farm and management factors, previous diagnoses of DD, individual cow treatment and method of clinical diagnosis further complicates comparisons. A systematic review determined that effectiveness of CuSO4 footbaths was not supported by evidence (Thomsen, 2015). An additional systematic review and meta-analysis revealed that the use of collective therapy (foot bathing and herd-level topical treatments) for treatment and prevention of DD remains uncertain and must be further investigated (Ariza et al., 2017). However, Ariza et al. (2017) combined all collective therapies (regardless of frequency, concentration, and product used) available in the literature compared to a negative control (excluding all studies without a negative control). The aim of this systematic review was, therefore, to evaluate the evidence of footbath protocols with varying doses, products, and frequencies for the treatment and prevention of DD and use a network meta-analysis to compare the efficacy of footbath protocols reported in the literature.

2.2. Literature search and study selection A systematic review and network meta-analysis was performed according to the Preferred Reporting Items for Systematic Reviews and Meta-analyses (PRISMA) extension statement for incorporating network meta-analyses of health care interventions (Hutton et al., 2015). No review protocol was submitted prior to systematic review being performed. CAB Abstracts via EBSCO (no date restriction), Web of Science (all databases; no date restriction), and Scopus (Elsevier; no date restriction) were systematically searched (Appendix Table A1) for relevant randomized controlled trials (RCTs). Relevant text words and subject headings were used for footbath protocols related to controlling DD. The search was not limited by language and non-English abstracts/ articles were translated to assess eligibility. Conference proceedings and grey literature identified in searches were included as eligible articles. Two reviewers (CJ and CB) worked independently, using a screening tool to verify identified abstracts for eligibility for the full text review. A third reviewer (HB) was consulted if discrepancies existed between reviewers. Additionally, one reviewer (CJ) manually searched reference lists of all included articles, systematic reviews and metaanalyses for additional eligible articles. RCTs of walk-through footbath protocols used on dairy cattle for prevention or control of DD were included according to the previously stated eligibility criteria. The focus was on RCTs and cluster RCTs (RCTs in which groups of cows were allocated to a treatment) as footbath protocol interventions are often tested at the herd level rather than the individual animal. Studies were restricted to trials that measured treatment or prevention of DD lesions. 2.3. Data collection and quality assessment Information from articles selected for inclusion in the systematic review and network meta-analysis were extracted and transferred into an Microsoft Excel spreadsheet by one reviewer (CJ). Data abstracted included study setting (country, cluster design), footbath protocol details (product, frequency, dimensions of footbath), comparator(s), participant demographics (number of herds, number of cattle, mean parity, mean DIM), and outcome events as previously described. For each outcome event in each trial arm, total number of cows enrolled and total number of cows to experience the outcome were abstracted. If only percentages were reported, number of cows with the outcome was calculated and rounded to the nearest whole number. Outcome events for both outcomes at the end of trials were extracted. Risk of bias of all eligible articles was independently assessed using SYRCLE’s risk of bias tool for animal studies (Hooijmans et al., 2014). Each study was assigned a risk of bias score from low to high, based on relevant risk of bias types, including selection bias, performance bias, detection bias, attrition bias, reporting bias and other potential threats to validity as outlined in the tool. Selection bias assessed sequence generation, baseline characteristics and allocation concealment to assess the efforts used to randomly assign subjects or to balance baseline risks among intervention groups. Performance bias assessed measures taken to conceal allocation of subjects (blinding) as well as to house animals randomly within a room (however, the latter domain of the performance bias type is only applicable to lab animal studies and did not apply to the current study framework). Detection bias assessed methods and objectivity of the outcome assessment (lesion scoring system). Attrition bias considered methods by which incomplete outcome data were handled. Finally, other sources of bias considered possibly carryover effects and any other identified biases. Sources and risk of bias were used to inform data analysis (in the case of high risk of bias) as well as to inform interpretation of the results.

2. Materials and methods 2.1. Eligibility criteria The review criteria followed the evidence-based veterinary medicine concept of PICO terms: population (P), intervention (I), comparator (C) and outcomes (O) (Richardson et al., 1995). All controlled clinical trials that assessed walk-through footbath protocols for the prevention or treatment of DD in dairy cattle (heifers, lactating or dry cows) were included. These trials were required to have a comparator group (ranging from no treatment to another footbath protocol). Studies assessing antibiotic footbath treatments were excluded, based on the desire to discontinue use of antibiotics and that these practices are not supported by European Union policies (Laven and Logue, 2006). Additionally, footbath protocols needed to be relevant for on-farm use and utilize a walk-through footbath; therefore, only protocols designed to be used on groups of animals were included. Thus, studies were excluded if they investigated other forms of collective therapy, e.g. spraying feet in the milking parlour, foam footbath products, or footbath protocols that required cattle to stand in the product for prolonged intervals, as opposed to passing through. Comparators included other footbath protocols or no footbath. Primary outcomes were prevention and treatment of DD lesions. Prevention was measured as the incidence of new clinical DD lesions within the follow-up period in previously healthy cattle, whereas treatment was measured as the proportion of DD lesions that clinically cured or the proportion of existing clinical DD lesions that were reduced in severity within the follow-up period. Both outcomes had to be assessed by visual diagnosis using a lesion scoring system.

2.4. Data synthesis and network meta-analysis Prevention and treatment outcomes were the outcomes used in the 57

Preventive Veterinary Medicine 164 (2019) 56–71

C. Jacobs et al.

network meta-analysis. For these binary outcomes, individual study outcomes were expressed as odds ratios (OR) with 95% confidence intervals (CI) calculated from event numbers extracted from each trial (Tables 2 and 3). If event numbers were not provided, the reported OR was extracted. Odds ratios were calculated with proportions of events and non-events as the numerator and denominator, respectively. For the prevention outcome, the event considered was incidence of new DD lesions. For the treatment outcome, the event considered was cure or improvement of existing DD lesions. To account for relative variability within and between randomized clusters, appropriate adjustments were made using the intraclass correlation coefficient (Higgins, 2011) at a value of 0.30 obtained from previous studies (Holzhauer et al., 2006a; Cramer et al., 2008) prior to pooling. For protocols using CuSO4, study effects were combined based on the concentration and frequency of use into 4 categories: 1) ≥ 5% CuSO4, ≥ 4 times/wk; 2) ≥ 5% CuSO4, < 4 times/wk; 3) < 5% CuSO4, ≥ 4 times/wk; and 4) < 5% CuSO4, < 4 times/wk; these were based on recommendations of 5% CuSO4 4 times/ wk (Speijers et al., 2010; Relun et al., 2012; Speijers et al., 2012; Solano et al., 2017). A random-effects Bayesian network meta-analysis was conducted to inform pairwise comparative efficacy of footbath protocols for prevention and treatment of DD. These Bayesian analyses used the published informative prior of Turner et al. (2012) for the between study variability with a subjective outcome in trials comparing treatment versus treatment. This approach is helpful in sparse evidence networks to improve precision around treatment effects. Sensitivity analyses were conducted using an uninformative prior in a random-effect and a fixedeffect model. All chains were run with 10,000 burn-in iterations followed by 10,000 monitoring iterations. Convergence was assessed by running 3 chains, inspecting sampling history plots, and calculating Gelman-Rubin-Brooke statistics (Gelman and Rubin, 1992; Brooks and Gelman, 1998). All Bayesian analyses were conducted using R statistical software 2.15.2 with the rjags package version 3.3.0 (www.rproject.org).

Fig. 1. Flow diagram of studies considered for inclusion.

or acidified, ionized copper (Laven and Hunt, 2002; Manske et al., 2002; Bergsten et al., 2007; Goossens and Hemling, 2007; Speijers et al., 2010; Teixeira et al., 2010; Speijers et al., 2012; Döpfer et al., 2013; Smith et al., 2014), and 4 included footbath protocols using formalin/formaldehyde (Laven and Hunt, 2002; Teixeira et al., 2010; Cramer and Benavides, 2012; Holzhauer et al., 2012). Two studies were unclear in reporting of effectiveness of the footbath protocol in terms of treatment or prevention; therefore, it was not possible to calculate ORs (Bergsten et al., 2007; Döpfer et al., 2013). Two studies did not connect in the network for prevention or treatment and were not included in the network meta-analysis (Cramer and Benavides, 2012; Holzhauer et al., 2012). Eleven studies contained appropriate information to use in the network meta-analysis and represented a total of 14 unique footbath products which included 2 experimental products (Cramer and Benavides, 2012), Thymox (Döpfer et al., 2013), DoubleAction (Goossens and Hemling, 2007), Digiderm+ (Holzhauer et al., 2012), Klingon Blue (Klaas et al., 2008), Hyperox (paracetic acid; (Laven and Hunt, 2002), Hoof-Fit Bath (Relun et al., 2012), Provita Hoofsure Endurance (Smith et al., 2014), sodium hypochlorite (Speijers et al., 2010), Dragonhyde (Teixeira et al., 2010), Virocid (glutaraldehyde; (Thomsen et al., 2008), Hoofcare DA (quaternary ammonium compounds; (Thomsen et al., 2008), and Kickstart 2 (organic acids; (Thomsen et al., 2008). All 11 studies included in the network meta-analysis evaluated prevention of DD lesions assessing 17 comparisons (Fig. 2a) and 10 studies evaluated treatment of DD lesions assessing 19 comparisons (Fig. 2b). Of the 9 studies including CuSO4 or acidified ionized copper, 4 contributed to the ≥ 5% CuSO4, ≥ 4 times/wk category (Goossens and Hemling, 2007; Speijers et al., 2010, 2012; Smith et al., 2014), 4 contributed to the ≥ 5% CuSO4, < 4 times/wk category (Klaas et al., 2008; Speijers et al., 2010; Teixeira et al., 2010; Speijers et al., 2012), 2 contributed to the < 5% CuSO4, ≥ 4 times/wk category (Manske et al., 2002; Speijers et al., 2010), and 1 contributed to the < 5% CuSO4, < 4

3. Results 3.1. Literature search and trial characteristics The literature search yielded 172 articles, 72 from CAB abstracts, 33 from Scopus and 67 from Web of Science (Fig. 1). Following title and abstract and full-text review, 14 studies were included in the qualitative analysis and 11 of these were included in the quantitative network meta-analyses. 3.2. Summary of network Table 1 summarizes characteristics of the 14 studies included in the qualitative analysis. The majority of studies were conducted in Europe, with only 3 studies conducted in North America. Initial prevalence of DD ranged from 0 to 100%. Prevalence of 0% occurred by selecting cows free of DD at the start of the study and following this subgroup throughout to determine effect of foot bathing on preventing DD lesions. Alternatively, selecting cows that were affected with DD resulted in a subgroup with a prevalence of 100% to assess effect of foot bathing on treatment of DD. Duration of follow-up at assessment of DD ranged from 4 to 24 wk for the prevention outcome and 3 to 24 wk for the treatment outcome. Five studies included individual topical treatment of ulcerative lesion stages as part of their protocol (Klaas et al., 2008; Speijers et al., 2010; Holzhauer et al., 2012; Relun et al., 2012; Speijers et al., 2012). Of the 14 studies included in the qualitative analysis (Tables 2 and 3),5 compared a footbath protocol to no footbath (Bergsten et al., 2007; Klaas et al., 2008; Thomsen et al., 2008; Speijers et al., 2010; Relun et al., 2012), 2 compared a footbath protocol to water (Manske et al., 2002; Bergsten et al., 2007), 9 included footbath protocols using CuSO4 58

59

1

CRT

CrT

NRCS

1

2

NRCS NRCS

1 1

NRCS NRCS NRCS NRCS NRCS CrT

CrT

1

1 2 3 1 2 1

NRCS

1

NRCS

NRCS

1

1

NRCS

1

8

4

5 8 10 14 14 4

9

24

3 24

4

16

24

12

8

16

16

Duration of study (wk)

4

1

1 1 1 1 1 1

3

52

1 1

1

1

1

4

1

1

1

Herds

400

356

118 117 95 70 64 406

286

4,677

169 44

114

115

66

650

265

167

112

Cows

Enrolled (no.)

Red or Holstein Red or Holstein

Holstein

Holstein

2

Foot

Cow

Cow Cow Cow Cow Cow Cow

Foot

NS3 Holstein Holstein Holstein Holstein Holstein Holstein

Foot

Hind feet Cow

All feet

Hind feet

Hind feet

Foot

Foot

Foot

Foot

Outcome unit

Holstein/ Normande

Holstein Swedish Holstein

Holstein

Holstein

Holstein

NS3

Swedish Swedish Swedish Swedish NS3

Breed

CRT = crossover trial; CRT = cluster randomized trial; NRCS = non-randomized control study. DIM at enrollment. 3 Not specified.

1

Thomsen et al. (2008); Denmark

Speijers et al. (2012); Northern Ireland Teixeira et al. (2010); USA

Smith et al. (2014); United States Speijers et al. (2010); Northern Ireland

NRCS

1

NRCS

2

Cramer and Benavides (2012); Canada Döpfer et al. (2013); Not specified Goossens and Hemling (2007); Belgium Holzhauer et al. (2012); the Netherlands Klaas et al. (2008); Denmark Laven and Hunt (2002); UK Manske et al. (2002); Sweden Relun et al. (2012/2013); France

NRCS

1

Bergsten et al. (2007); Sweden

Type of study1

Experiment

Study; country

21.8 – 22.7

24

59 74 53 95 0 20

57

6-13

100 100

16-18 (active DD)

35

41 – 49

NS3

28

17

17

Initial DD prevalence (%)

Freestalls

NS3

10 herds no access; 2 herds partial access (2 wk)

No

No No No No No No

NS3

Grazing season (6 farms no access)

Not specified Total access

No

Partially (8 wk)

Partial (12 wk)

NS3

Freestalls

Freestalls

Freestalls Freestalls Freestalls Freestalls Freestalls Freestalls

NS3

Freestalls

NS3 Freestalls

Freestalls

Freestalls

Freestalls

NS3

Freestalls

Freestalls

NS3

No

Housing system

Pasture access

Herringbone and rotary

Parallel

Rotary parlour Rotary parlour Rotary parlour Rotary parlour Rotary parlour Parallel

NS3

Herringbone

Mean = 3 Mean = 3 Mean = 3 Mean < 3 Mean < 2.5 1 (56%) > 1 (44%) 1 (46%) > 1 (54%) NS3

1 (35%) 2 (27%) 3 (17%) > 4 (20%) NS3

NS3 NS3

NS3

NS3 Automatic milking system NS3 NS3

NS3

NS3

NS3

Mean > 100 Mean > 100 Mean < 90 Mean < 60 Mean < 65 Dry-off to -25 DIM Dry-off to 25 DIM NS3

< 90 (35%) 90–150 (23%) > 150 (43%) NS3

NS3 NS3

NS3

NS3

NS3

NS3

NS3 NS3

NS3

NS3

NS3

NS3

Days in milk (DIM)2

NS3

NS3

NS3

Parity

Rotary parlour

Rotary parlour

Milking system

Table 1 Descriptive summary of study design and characteristics of 14 publications included in a systematic review assessing effectiveness of footbath protocols for prevention and treatment of digital dermatitis (DD).

C. Jacobs et al.

Preventive Veterinary Medicine 164 (2019) 56–71

Case definition

Lesions were scored from 0 (no lesion) to 2 (severe lesion)

Lesions scored using a graded scoring system (no lesion (O), hyperkeratosis (H), ulcerative (U), granulomatous (G) and papillomatous (P))

Lesions scored according to M-stage system

Presence/absence; Not specified scoring method

Lesions scored according to M-stage scoring system, M2 lesions were the outcome of interest

New cases defined as active lesions (scores 1 to 3) developing from feet without active lesions (scores 0, 4, 5) (Manske et al., 2002)

New cases defined as active lesions (scores 1 to 3) developing from feet without active lesions (scores 0, 4, 5)

Active lesions defined as early or acute DD (M1 or M2)

Presence of lesions defined as feet with active (M1 and M2) lesions

Study

Bergsten et al. (2006)

Cramer and Benavides (2012)

Döpfer et al. (2013)

Goossens and Hemling (2007)

Holzhauer et al. (2012)

Klaas et al. (2008)

Manske et al. (2002)

60

Relun et al. (2012/ 2013) Data from SRMA

Smith et al. (2014)

Prevention of the occurrence of active DD lesions (M1 or M2)

Prevention of active DD lesions (M1 and M2).

Prevention of new cases (scores 1 to 3) during the study period

Prevention of new cases (scores 1 to 3) during the study period

Prevention of new M2 lesions during the study period

Prevention of new cases during the study period

Probability matrix of M-stage transitions

Cows with O or H at start of trial and for at least 2 consecutive data points

Absence of deterioration of feet with score < 2 at the beginning of experiment

Success definition

1

1

1

1

1

1

1

1

2

1

Experiment number Hoofmat: water (within-cow control) Hoofmat 7% CuSO4, after every milking (2/d) No treatment (untreated control) Foam-system: tensides and paracetic acid, 2 milkings/d Footbath: 5% Formalin, once/day, 3 consecutive d/wk Footbath: experimental H, once per d, 5d/ wk Footbath: experimental P, once/d, 5d/wk Footbath: 5% CuSO4, once per day, 3d/wk Footbath: 1% Thymox, once per day, 3d/ wk Footbath: 5% CuSO4, once a day, 5d/wk (within-cow control) Footbath: 2% DoubleAction, once a day, 5d/wk Footbath: 4% Formalin, twice daily 1d every second week Footbath: Digiderm+, twice daily 5d/wk Footbath: empty (within-cow control; split footbath) Footbath: KlingonBlue (CuSO, ZnSO, and organic acid solution), once weekly for 17h Footbath: empty (within-cow control; split footbath) Footbath: 7.5% CuSO4, once weekly for 17 h Footbath: water (within-cow control) Footbath: 0.6% acidified ionized copper, 2 milkings/d for 10 d, with 5.5 d in between No footbath (untreated control) Footbath: 5% Hoof-Fit Bath, 4 consecutive milkings every 4 weeks Footbath: 5% Hoof-Fit Bath, 4 consecutive milkings every 2 weeks Footbath: 5% CuSO4 (within-cow positive control), once daily, 5 d/wk Footbath: 3% Provita Hoofsure Endurance, once daily, 5 d/wk

Footbath protocol

Table 2 Summary of evidence of footbath protocols in the prevention of digital dermatitis (DD); results are expressed as odds ratios (OR).

Reference 0.48 (0.27 – 0.84) 0.55 (0.32 – 0.92) Reference NS1

0.08 (NS1) 0.14 (NS1) 0.15 (NS1)

(5/38) (3/38) (4/34) (2/34)

0.39 (112/285) 0.41 (117/285)

0.18 (343/1,917) 0.22 (309/1,434) 0.12 (132/1,063)

0.31 (17/55) 0.30 (17/56)

0.13 0.08 0.12 0.06

0.20 (20/99) 0.07 (7/97)

0.50 (17/34) 0.51 (20/39)

Reference 1.26 (1.06 – 1.50) 0.65 (0.52 – 0.81) Reference 1.08 (0.76 – 1.53)

Reference 0.97 (0.40 – 2.36)

Reference 0.31 (0.10 – 0.81) Reference 0.57 (0.08 – 3.20) Reference 0.47 (0.04 – 3.59)

None

Only in ulcerative lesions

None

Only in ulcerative lesions

Only in ulcerative lesions

NS1

NS1

NS1

None

None

Additional individual treatment

(continued on next page)

Reference –

NS1 NS1

Reference 1.05 (0.38 – 2.92)

Reference –

NS1 NS1

NS1 NS1

OR (95% CI)

Outcome proportion (event/ no)

C. Jacobs et al.

Preventive Veterinary Medicine 164 (2019) 56–71

61

2

Not specified. Outcome reported was combined effectiveness of all footbath protocols.

1

New cases defined as active lesions (scores 1 to 4) in feet previously scored as non-active lesions (scores 0 or 5) (Manske et al., 2002)

Thomsen et al. (2008)

Prevention of new active lesions (scores 1 to 4) during the study.

Prevention of new active lesions during the study

Presence of active (painful upon palpitation) lesion

Teixeira et al. (2010)

Prevention of the occurrence of active DD lesions (M1 or M2)

Proportion of cows with M0 to M0 transitions between 2 consecutive weeks

Cows with M1 or M2 lesions on at least 1 foot were defined as having active lesions

Speijers et al. (2010)

Success definition

Speijers et al. (2012)

Case definition

Study

Table 2 (continued)

1

2

1

2

1

3

2

1

Experiment number

No footbath (within cow control; split footbath) Footbath: 1.5% glutaraldehyde (Virocid), 2 consecutive milkings, twice a week Footbath: 2% quaternary ammonium compounds (Hoofcare DA), 2 consecutive milkings, twice a week Footbath: 1% organic acids (Kickstart 2), 2 consecutive milkings, twice a week

Footbath: 10% CuSO4, twice weekly Footbath: 5% Dragonhyde, twice weekly

No footbath (untreated control) Footbath: 5% CuSO4, 4 consecutive milkings/wk Footbath: 2% ClO− hypochlorite sodium, 4 consecutive milkings/wk Footbath: 2% CuSO4, 4 consecutive milkings/wk Footbath: 5% CuSO4, 4 consecutive milkings/wk Footbath: 2% CuSO4, 4 consecutive milkings every 2wk Footbath: 5% CuSO4, 4 consecutive milkings every 2wk Footbath: 5% CuSO4, 4 consecutive milkings every 2wk Footbath: 5% CuSO4, 4 consecutive milkings alternating weekly with 10% NaCl Footbath: 5% CuSO4, 4 consecutive milkings alternating weekly with tap water Footbath: 5% CuSO4, 4 consecutive milkings/wk Footbath: 5% CuSO4, 4 consecutive milkings every 2wk Footbath: 5% CuSO4, 4 consecutive milkings every 2wk Footbath: 5% CuSO4, 4 consecutive milkings every 4wk Footbath: 5% Formalin, twice weekly Footbath: 5% Dragonhyde, twice weekly

Footbath protocol

(23/29) (26/36) (13/19) (11/19)

0.08 (43/535) 0.10 (55/535)2

(119/189) (111/167)

0.15 (20/130) 0.15 (23/151)

0.46 (12/26) 0.48 (13/27)

0.90 (26/29) 0.88 (28/32)

0.86 (25/29) 0.86 (24/28) 0.82 (23/28)

0.79 0.72 0.68 0.58

0.74 (29/39) 0.68 (25/37) 0.73 (27/37)

Outcome proportion (event/ no)

–

–

–

–

1.01 (0.50 – 2.04) Reference 0.86 (0.54 1.36) Reference Reference 1.31 (0.85 – 2.04)

0.92 (0.27 – 3.10) Reference

1.24 (0.19 – 9.24) Reference

Reference 1.04 (0.17 – 6.27) 0.74 (0.13 – 3.91)

Reference 0.72 (0.23 2.18) 0.93 (0.30 2.94) Reference 0.68 (0.17 2.47) Reference 0.63 (0.14 2.90)

OR (95% CI)

None

Not specified

Not specified

Only in ulcerative lesions

Only in ulcerative lesions

Only in ulcerative lesions

Only in ulcerative lesions

Only in ulcerative lesions

Additional individual treatment

C. Jacobs et al.

Preventive Veterinary Medicine 164 (2019) 56–71

62

Active lesions defined as early or acute DD (M1 or M2)

Presence of lesions defined as feet with active (M1 and M2) lesions

Smith et al. (2014)

Lesions scored according to M-stage system; M2 lesions were outcome of interest

Holzhauer et al. (2012)

Relun et al. (2012/ 2013)

Present/absent, scoring method not specified

Goossens and Hemling (2007)

Active lesions (scores 1 to 3)

Lesions scored according to M-stage system

Döpfer et al. (2013)

Manske et al. (2002)

Lesions scored using a graded scoring system (no lesion (O), hyperkeratosis (H), ulcerative (U), granulomatous (G) and papillomatous (P))

Cramer and Benavides (2012)

Lesions scored 0 (no lesion) to 2 (red, deep lesion)

Lesions were scored from 0 (no lesion) to 2 (severe lesion)

Bergsten et al. (2006)

Laven and Hunt (2002)

Case definition

Study

Transition of active lesions to nonactive lesions (M0, M3, M4)

Active lesions that transitioned to nonactive (M0 or M4) lesions

Transition of feet previously scored 1 to 3 (active) to scores 0, 4 or 5 (nonactive)

Lesions present at day 0 that were not present at day 21

Transition of M2 lesions to other Mstages

Symptoms of DD disappeared in feet previously scored as DD present

Probability matrix of M-stage transitions

Cows with U/G/P at start of trial and have 2 consecutive O or H evaluations

Improvement of feet scored > 0 at the beginning of the study

Success definition

1

1

1

1

1

1

1

1

2

1

Experiment number Hoofmat: water Hoofmat 7% CuSO4, after every milking (2/d) No treatment (untreated control) Foam-system: tensides and paracetic acid, 2 milkings/d Footbath: 5% Formalin, once/day, 3 consecutive d/wk Footbath: experimental H, once per d, 5d/wk Footbath: experimental P, once/d, 5d/ wk Footbath: 5% CuSO4, once per day, 3d/ wk Footbath: 1% Thymox, once per day, 3d/wk Footbath: 5% CuSO4, once a day, 5d/wk (within cow control) Footbath: 2% DoubleAction, once a day, 5d/wk Footbath: 4% Formalin, twice daily 1d every second week Footbath: Digiderm+ (concentration not specified), twice daily 5d/wk Footbath: 6% Formalin, once a day, 7d/ wk Footbath: 2% CuSO4, once a day, 7d/ week Footbath: 1% peracetic acid (Hyperox), once a day, 7d/wk Footbath: water (within-cow control) Footbath: 0.6% acidified ionized copper, 2 milkings/d for 10 d, with 5.5 d in between No footbath (untreated control) Footbath: 5% Hoof-Fit Bath, 4 consecutive milkings every 4 weeks Footbath: 5% Hoof-Fit Bath, 4 consecutive milkings every 2 weeks Footbath: 5% CuSO4 (within-cow positive control), once daily, 5 d/wk Footbath: 3% Provita Hoofsure Endurance, once daily, 5 d/wk

Footbath protocol

Table 3 Summary of evidence of footbath protocols in the treatment of digital dermatitis (DD); results are expressed as odds ratios (OR).

Reference – Reference NS1 NS1

Reference NS1

NS1 NS1 0.14 (NS1) 0.18 (NS1) 0.16 (NS1)

NS1 NS1

0.22 (63/285) 0.23 (65/285)

0.84 (249/296) 0.79 (200/252) 0.94 (186/197)

0.52 (12/23) 0.83 (20/24)

0.63 (40/64) 0.75 (33/44) 0.46 (32/70)

0.61 (19/31) 0.85 (17/20)

None

Only in ulcerative lesions

None

NS1

Only in ulcerative lesions

NS1

NS1

NS1

None

None

Additional individual treatment

(continued on next page)

Reference 0.73 (0.46 – 1.15) 3.19 (1.57 – 7.00) Reference 1.04 (0.69 – 1.57)

Reference 4.58 (1.02 – 23.58)

Reference 1.80 (0.72 – 4.68) 0.51 (0.24 – 1.07)

Reference 3.58 (0.76 – 22.57)

Reference 3.09 (0.87 – 11.73)

Reference –

NS1 NS1

0.22 (6/27) 0.47 (15/32)

OR (95% CI)

Outcome proportion (event/ no)

C. Jacobs et al.

Preventive Veterinary Medicine 164 (2019) 56–71

63

2

Transition of feet scored as active lesions (scores 1 to 4) to non-active lesions (scored 0 or 5)

Diagnosed as active (painful) lesion at the beginning of the study and free of DD at the end of the study

Success definition

Not specified. Outcome reported was combined effectiveness of all footbath protocols.

Active lesions (scores 1 to 4) (Manske et al., 2002)

Thomsen et al. (2008)

1

Presence of active (painful upon palpitation) lesion

Case definition

Teixeira et al. (2010)

Speijers et al. (2012)

Speijers et al. (2010)

Study

Table 3 (continued)

1

2

1

2

1

3

2

1

Experiment number

No footbath (within cow control; split footbath) Footbath: 1.5% glutaraldehyde (Virocid), 2 consecutive milkings, twice a week Footbath: 2% quaternary ammonium compounds (Hoofcare DA), 2 consecutive milkings, twice a week Footbath: 1% organic acids (Kickstart 2), 2 consecutive milkings, twice a week

Footbath: 10% CuSO4, twice weekly Footbath: 5% Dragonhyde, twice weekly

No footbath (untreated control) Footbath: 5% CuSO4, 4 consecutive milkings/wk Footbath: 2% ClO− hypochlorite sodium, 4 consecutive milkings/wk Footbath: 2% CuSO4, 4 consecutive milkings/wk Footbath: 5% CuSO4, 4 consecutive milkings/wk Footbath: 2% CuSO4, 4 consecutive milkings every 2wk Footbath: 5% CuSO4, 4 consecutive milkings every 2wk Footbath: 5% CuSO4, 4 consecutive milkings every 2wk Footbath: 5% CuSO4, 4 consecutive milkings alternating weekly with 10% NaCl Footbath: 5% CuSO4, 4 consecutive milkings alternating weekly with tap water Footbath: 5% CuSO4, 4 consecutive milkings/wk Footbath: 5% CuSO4, 4 consecutive milkings every 2wk Footbath: 5% CuSO4, 4 consecutive milkings every 2wk Footbath: 5% CuSO4, 4 consecutive milkings every 4wk Footbath: 5% Formalin, twice weekly Footbath: 5% Dragonhyde, twice weekly

Footbath protocol

(11/29) (19/36) (7/19) (8/19)

0.35 (28/79) 0.26 (21/79)2

0.24 (10/41) 0.18 (7/38)

0.17 (7/41) 0.31 (9/29)

0.58 (15/26) 0.52 (14/27)

0.76 (22/29) 0.69 (22/32)

0.69 (20/29) 0.64 (18/28) 0.71 (20/28)

0.38 0.53 0.37 0.42

0.10 (4/39) 0.35 (13/37) 0.14 (5/37)

Outcome proportion (event/ no)

–

–

–

–

0.46 (0.12 – 1.64) Reference 1.43 (0.42 – 5.02) Reference Reference 0.66 (0.32 – 1.37)

1.27 (0.38 – 4.30) Reference

1.43 (0.40 – 5.27) Reference

Reference 0.81 (0.23 – 2.81) 1.13 (0.31 – 4.11)

Reference 4.74 (1.24 21.95) 1.37 (0.27 7.50) Reference 1.83 (0.61 5.59) Reference 1.25 (0.28 5.60)

OR (95% CI)

None

NS1

NS1

Only in ulcerative lesions

Only in ulcerative lesions

Only in ulcerative lesions

Only in ulcerative lesions

Only in ulcerative lesions

Additional individual treatment

C. Jacobs et al.

Preventive Veterinary Medicine 164 (2019) 56–71

Preventive Veterinary Medicine 164 (2019) 56–71

C. Jacobs et al.

a)

b)

Fig. 2. Network diagram of trials examining footbath protocols for prevention (a) and treatment (b) of digital dermatitis (DD).

times/wk category (Speijers et al., 2010). The ORs associated with the prevention outcome ranged from 0.31 to 1.31 (Table 2) with only 1 study reporting a benefit of the footbath (Relun et al., 2012). The ORs associated with the treatment outcome ranged from 0.46 to 4.74 (Table 3), with 3 studies reporting a benefit (Manske et al., 2002; Speijers et al., 2010; Relun et al., 2012).

results of the first endpoint was used in the network meta-analysis for the study by Klaas et al. (2008); however, both endpoints were used for the study by Teixeira et al. (2010), because each endpoint studied effects of distinct footbath protocols.

3.3. Risk of Bias assessment

3.4.1. Prevention Eleven trials were included in the network meta-analysis for prevention of DD. Results of the network meta-analysis for the prevention of DD using Turner priors are shown (Table 5). None of the pair-wise active treatment comparisons reached statistical significance with either the Turner priors, fixed effect or random effect models, due to wide credible intervals.

3.4. Network meta-analyses

Results of the risk of bias assessment are presented in Table 4. There was only a low risk of bias in 6 of 14 studies for sequence generation and allocation concealment, with unclear risk of bias in these areas for the remaining 8 studies. The majority of studies (10) had a low risk of bias in the baseline characteristics domain, with unclear risk of bias for the remaining 4 studies. All studies were unclear in the blinding of caregivers and the majority of studies (12) were unclear in the blinding of the outcome assessor, with only 2 studies having a low risk of bias in this domain. All studies had a low risk of bias in the random outcome assessment, incomplete outcome data and selective outcome reporting domains. Three studies included crossover effects, with risk being low for 1 study (this effect was balanced across groups; Speijers et al. (2010)) and high in 2 of the studies (Klaas et al., 2008; Teixeira et al., 2010) which lacked a washout period. Due to this lack of washout, only

3.4.2. Treatment Ten trials were included in the network meta-analysis for treatment of DD. Results of the network meta-analysis for treatment of DD using Turner priors are shown (Table 6). Only CuSO4 ≥ 5% used ≥ 4 times/ wk was superior to no treatment (OR: 5.26; 95% Credibility Interval (CrI): 1.27–28.8), consistent with a direct comparison (Speijers et al., 2010) between CuSO4 ≥ 5% used ≥ 4 times/wk compared to no treatment (OR: 4.74; 95% CI: 1.24–21.95). Furthermore, CuSO4 ≥ 5%

64

65

no

no no

no no

all cows

all cows only cows with lesions all cows all lactating cows

3

2

1

randomly selected 100 cows per farm

yes

no

no

unclear

unclear

unclear

low unclear

low

low low

unclear

unclear

Not applicable. Selection of cows with DD lesions present for experiment 1. Selection of cows with DD lesions absent for experiment 2.

Thomsen et al. (2008)

Teixeira et al. (2010)

Speijers et al. (2012)

no no

all cows all cows

no no

low low

no

all cows

all cows all cows without heel horn erosion (HHE) DD lesions present2 or absent3 dry cows

unclear unclear

no

all cows

Sequence generation

unclear

unclear

low

low low

low low

unclear unclear

low

low low

low

low

Baseline characteris-tics

SYRCLE domains

Bergsten et al. (2006) Cramer and Benavides (2012) Döpfer et al. (2013) Goossens and Hemling (2007) Holzhauer et al. (2012) Klaas et al. (2008) Laven and Hunt (2002) Manske et al. (2002) Relun et al. (2012/ 2013) Smith et al.(2014) Speijers et al. (2010)

Sample size calculation

Inclusion/ exclusion criteria specified

Study

unclear

unclear

unclear

low unclear

low low

unclear unclear

low

low low

unclear

unclear

Allocation conceal-ment

NA

NA

NA

NA NA

NA NA

NA NA

NA

NA NA

NA

NA

Random housing1

unclear

unclear

unclear

unclear unclear

unclear unclear

unclear unclear

unclear

unclear unclear

unclear

unclear

Blinding of caregivers

Table 4 Risk of bias assessment of included trails using the SYRCLE’s risk of bias tool for animal studies (Hooijmans et al., 2014).

low

low

low

low low

low low

low low

low

low low

low

low

Random outcome assessment

low

unclear

unclear

low unclear

unclear unclear

unclear unclear

unclear

unclear unclear

unclear

unclear

Blinding of outcome assessor

low

low

low

low low

low low

low low

low

low low

low

low

Incomplete outcome data

low

low

low

low low

low low

low low

low

low low

low

low

Selective outcome reporting

carryover effects: high low

low carryover effects, balanced across treatments: low low

low low

carryover effect: high low

low

low low

low

low

Other biases

C. Jacobs et al.

Preventive Veterinary Medicine 164 (2019) 56–71

1.84 (0.06 – 93.1)

0.58 (0.09 – 3.54) 0.36 (0.06 – 2.29) 0.96 (0.30 – 3.08) 0.52 (0.07 – 4.20) 0.76 (0.09 – 6.85) 0.57 (0.05 – 5.45) 0.57 (0.04 – 8.32) 0.27 (0.02 – 3.19) 0.36 (0.03 – 4.83) 0.42 (0.06 – 3.45) 0.42 (0.04 – 4.39) 1.04 (0.09 – 11.55)

0.70 (0.22 – 2.29) 0.44 (0.10 – 1.70) 1.18 (0.21 – 6.44) 0.65 (0.10 – 3.93) 0.92 (0.25 – 3.39) 0.70 (0.11 – 4.07) 0.70 (0.07 – 6.43) 0.33 (0.03 – 2.87) 0.43 (0.04 – 4.30) 0.52 (0.09 – 2.67) 0.53 (0.07 – 3.83) 1.26 (0.36 – 4.35)

66

5% Hoof fit bath, 4 milkings every 4 wk

2% formalin, 2x/wk

3% Provita Hoofsure Endurance, 5x/ wk 5% CuSO4 alternating weekly with 10% NaCl, 4x/wk 5% CuSO4 alternating weekly with water, 4x/wk 5% Dragonhyde, 2x/ wk

2% sodium hypochlorite, 4x/ wk 2% Double Action, 5x/wk

≤ 5% CuSO4, ≤ 4x/ wk

≤ 5% CuSO4, ≥ 4x/ wk

≥ 5% CuSO4, ≤ 4x/ wk

Various (1.5% Virocid, 2% Hoofcare DA, 1% Kickstart 2), 4x/ wk ≥ 5% CuSO4, ≥ 4x/ wk

1.23 (0.15 – 9.77) 2.20 (0.15 – 66.7)

Water

Water

None

Intervention

Comparator

0.56 (0.01 – 12.72)

0.23 (0.01 – 6.79)

0.23 (0.01 – 5.73)

0.19 (0.00 – 6.84)

0.14 (0.00 – 4.86)

0.30 (0.01 – 11.09)

0.31 (0.01 – 7.80)

0.41 (0.01 – 8.83)

0.29 (0.01 – 7.42)

0.52 (0.01 – 13.87)

0.19 (0.01 – 4.07)

0.31 (0.01 – 6.10)

Various (1.5% Virocid, 2% Hoofcare DA, 1% Kickstart 2), 4x/wk

0.63 (0.22 1.83) 1.65 (0.43 7.07) 0.90 (0.22 4.14) 1.33 (0.37 4.75) 0.98 (0.26 3.86) 0.99 (0.12 7.23) 0.47 (0.06 3.37) 0.61 (0.08 4.88) 0.74 (0.18 2.90) 0.76 (0.13 4.25) 1.82 (0.34 9.87) –

–

–

–

–

–

–

–

–

–

–

≥ 5% CuSO4, ≥ 4x/ wk

2.67 (0.62 – 11.29) 1.43 (0.34 – 6.84) 2.12 (0.44 – 10.38) 1.55 (0.29 – 8.84) 1.60 (0.15 – 14.34) 0.75 (0.13 – 3.89) 0.98 (0.17 – 6.05) 1.17 (0.46 – 3.01) 1.20 (0.30 – 4.83) 2.84 (0.48 – 19.49)

≥ 5% CuSO4, ≤ 4x/ wk

0.53 (0.11 -2.92) 0.79 (0.13 – 4.94) 0.59 (0.08 – 3.94) 0.60 (0.05 – 6.42) 0.28 (0.03 – 2.58) 0.37 (0.03 – 3.84) 0.44 (0.08 – 2.42) 0.45 (0.06 – 3.26) 1.07 (0.13 – 8.75)

≤ 5% CuSO4, ≥ 4x/ wk

1.48 (0.20 – 9.76) 1.09 (0.14 – 7.42) 1.07 (0.08 – 11.90) 0.52 (0.04 – 4.72) 0.68 (0.06 – 6.75) 0.82 (0.14 – 4.41) 0.83 (0.11 – 6.00) 1.96 (0.22 – 18.19)

≤ 5% CuSO4, ≤ 4x/ wk

1.36 (0.22 – 8.40)

0.56 (0.07 – 4.67)

0.56 (0.09 – 3.51)

0.46 (0.04 – 5.13)

0.35 (0.03 – 3.58)

0.75 (0.07 – 7.41)

0.74 (0.12 – 4.81)

2% sodium hypochlorite, 4x/wk

1.01 (0.08 – 10.49) 0.48 (0.04 – 4.64) 0.63 (0.05 – 6.68) 0.75 (0.10 – 4.92) 0.77 (0.08 – 6.54) 1.80 (0.21 – 16.72)

2% Double Action, 5x/wk

1.80 (0.14 – 26.45)

0.73 (0.06 – 12.03)

0.74 (0.07 – 9.28)

0.60 (0.03 – 11.26)

0.47 (0.03 – 7.94)

3% Provita Hoofsure Endurance, 5x/wk

3.87 (0.30 – 56.44)

1.62 (0.18 – 16.04)

1.58 (0.23 – 12.16)

1.31 (0.24 – 7.67)

5% CuSO4 alternating weekly with 10% NaCl, 4x/wk

2.92 (0.22 – 42.20)

1.22 (0.13 – 12.57)

1.20 (0.16 – 9.44)

5% CuSO4 alternating weekly with water, 4x/wk

2.46 (0.32 – 20.56)

1.02 (0.36 – 2.91)

5% Dragonhyde, 2x/wk

Table 5 Pairwise comparisons (odds ratios with 95% credible interval in parentheses) and network meta-analysis of footbath protocols for the prevention of digital dermatitis.

2.39 (0.24 – 25.41)

2% formalin, 2x/wk

5% Hoof fit bath, 4 milkings every 2 wk

(continued on next page)

5% Hoof fit bath, 4 milkings every 4 wk

C. Jacobs et al.

Preventive Veterinary Medicine 164 (2019) 56–71

Preventive Veterinary Medicine 164 (2019) 56–71

0.80 (0.05 – 11.98)

used ≥ 4 times/wk was also superior to a water placebo (OR: 9.47; 95% CrI: 1.03–85.8). However, these pairwise comparison were no longer significant in the random-effects model, due to wide credibility intervals.

0.43 (0.03 – 6.15)

4. Discussion

0.84 (0.03 – 18.59)

1.71 (0.07 – 37.62)

1.37 (0.05 – 28.17)

1.10 (0.06 – 17.48)

1.06 (0.05 – 21.18)

This systematic review and network meta-analysis summarized the current body of literature for efficacy of footbath protocols on prevention or treatment of DD. Evidence consisted of studies with mostly a low or unclear risk of bias. CuSO4 ≥ 5% used ≥ 4 times/wk was superior to no treatment or water for treatment of DD lesions. However, no footbath protocol was superior to any other (including no intervention) for prevention of DD lesions. The search strategy and selection criteria necessary for the network meta-analysis resulted in targeting studies evaluating incidence and clinical cure of DD lesions; consequently, prevalence studies were not included, some of which are sources of evidence in favour of footbath use (Speijers et al., 2010; Relun et al., 2012; Fjeldaas et al., 2014). Unfortunately, duration of follow up differed among studies, which could result in multiple lesion transitions occurring between start and end of the study which could not be captured in the effect estimate for prevention or treatment, especially given that the reported interval before occurrence of a new DD lesion is 5 mo (Relun et al., 2013; Krull et al., 2016) and that a DD lesion can be completely healed within 1 mo (Holzhauer et al., 2008). Additionally, despite widespread use of footbath protocols and a wide variety of products available for footbaths, only 11 studies were identified that could be included in the network meta-analysis. This paucity of scientific evidence results in commercial footbath products with little to no scientific evidence of efficacy. An additional shortcoming to the studies included for the network meta-analysis was heterogeneity in dimensions of footbaths used. Cook et al. (2012) suggested dimensions should be 3.0–3.7 m long and 0.5 – 0.6 m wide, with a step-in height of 0.28 m; however, dimensions in the included studies were typically not specified or too short, with only one of the studies included in the network meta-analysis utilizing a footbath with a length of 3 m. Therefore, results of the footbath protocols studied may be limited due to inadequate footbath design. The impact of literature recommendations regarding footbath dimensions and protocols has been investigated by Solano et al. (2017) in a study involving an intervention trial consisting of a footbath protocol of 5% CuSO4 used for 4 consecutive milkings over 2 consecutive days, with cow passes not exceeding 200 or time not exceeding 24 h. Stainless steel split walkthrough footbaths were used, consisting of 2 baths separated by a grate with dimensions of 3 m long and 0.25 m wide, with a fluid depth of 0.25 m. Prevalence of active (M2 and M4.1) lesions decreased from intervention to end of trial by 50% for farms with a high (> 15%) baseline within-herd prevalence of active DD lesions. For farms with a low (< 15%) baseline within-herd prevalence of active DD lesions, the prevalence of active DD lesions did not change from intervention to the end of the trial period. It was speculated that footbath dimensions have a role in effectively preventing DD lesions, especially on farms with a high baseline prevalence of active DD lesions and that 5% CuSO4 used at least 4 times/wk was an effective protocol for controlling DD. Further heterogeneity existed between the studies in terms of the number of cow passes per footbath. It is recommended that footbath solutions be changed every 100–300 cow passes (Cook et al., 2012) and the majority of studies followed this recommendation with cow passes ranging from 45 to 300 cow passes per footbath with only 2 studies (Laven and Hunt, 2002; Döpfer et al., 2013) neglecting to report the number of cow passes per footbath. Because footbath efficacy is finite, variation in terms of number of cow passes may influence results if studies with relatively low numbers of cow passes would achieve contact with a solution that was more efficacious compared to studies with the maximum number of cow passes recommended. In this case, the

0.60 (0.04 – 8.12) 0.23 (0.00 – 8.45)

0.57 (0.04 – 6.93) 0.45 (0.02 – 9.88) 0.69 (0.15 – 2.73) 0.55 (0.05 – 5.28) 5% Hoof fit bath, 4 milkings every 2 wk KlingonBlue, once weekly for 17h

Table 5 (continued)

Comparator

0.31 (0.01 – 6.87)

0.99 (0.15 – 6.16) 0.81 (0.06 – 9.91)

1.57 (0.20 – 11.18) 1.29 (0.08 – 17.25)

0.58 (0.06 – 5.28) 0.47 (0.03 – 8.00)

1.08 (0.10 – 11.29) 0.87 (0.05 – 15.23)

0.75 (0.10 – 5.02)

0.98 (0.10 – 9.91) 0.82 (0.04 – 14.13)

0.98 (0.07 – 15.36)

2.13 (0.14 – 33.21)

1.59 (0.10 – 25.04)

1.36 (0.14 – 11.69)

1.34 (0.11 – 14.33)

0.55 (0.12 – 2.19)

C. Jacobs et al.

67

1.53 (0.01 – 194.90)

9.47 (1.03 – 85.84) 6.48 (0.68 – 61.88) 5.05 (0.54 – 50.81) 4.82 (0.98 – 27.02) 1.50 (0.12 – 20.80) 2.45 (0.16 – 33.98) 14.59 (0.79 – 228.95) 9.22 (0.34 – 259.32)

5.36 (0.36 – 70.88)

7.61 (0.52 – 105.63)

4.25 (0.29 – 68.92)

5.26 (1.27 – 28.80) 3.75 (0.55 – 24.62) 2.91 (0.40 – 25.33) 2.77 (0.36 – 26.44) 0.85 (0.08 – 10.52) 1.37 (0.26 – 8.00) 8.16 (0.86 – 88.28) 5.28 (0.27 – 108.57)

68

3.03 (0.29 – 34.75)

4.41 (0.41 – 48.15)

2.54 (0.22 – 28.93)

3% Provita Hoofsure Endurance, 5x/ wk 5% CuSO4 alternating weekly with 10% NaCl, 4x/ wk 5% CuSO4 alternating weekly with water, 4x/wk 5% Dragonhyde, 2x/wk

1% Hyperox (peracetic acid), 7 x/wk 2% sodium hypochlorite, 4x/wk 2% Double Action, 5x/wk

≤ 5% CuSO4, ≤ 4x/wk

≤ 5% CuSO4, ≥ 4x/wk

≥ 5% CuSO4, ≤ 4x/wk

Various (1.5% Virocid, 2% Hoofcare DA, 1% Kickstart 2), 4x/wk ≥ 5% CuSO4, ≥ 4x/wk

0.56 (0.04 – 10.05) 0.87 (0.02 – 50.65)

Water

Water

None

Intervention

Comparator

2.81 (0.03 – 232.67)

4.66 (0.05 – 445.31)

3.42 (0.03 – 282.94)

6.18 (0.04 – 811.37)

9.04 (0.09 – 829.24)

1.61 (0.02 – 105.48)

0.96 (0.01 – 99.72)

3.31 (0.03 – 273.10)

3.13 (0.04 – 278.69)

4.07 (0.05 – 279.39)

6.07 (0.08 – 422.53)

Various (1.5% Virocid, 2% Hoofcare DA, 1% Kickstart 2), 4x/wk

0.46 (0.08 – 2.86)

0.80 (0.13 – 4.49)

0.55 (0.10 – 3.10)

0.69 (0.24 – 1.87) 0.54 (0.14 – 2.03) 0.52 (0.13 – 2.27) 0.16 (0.03 – 1.01) 0.26 (0.06 – 1.12) 1.50 (0.28 – 8.51) 0.97 (0.08 – 12.49)

≥ 5% CuSO4, ≥ 4x/ wk

0.68 (0.16 – 2.91)

1.18 (0.25 – 4.89)

0.82 (0.19 – 3.35)

0.81 (0.20 – 3.19) 0.78 (0.17 – 3.40) 0.24 (0.03 – 1.56) 0.38 (0.06 – 2.32) 2.22 (0.31 – 16.87) 1.45 (0.09 – 22.33)

≥ 5% CuSO4, ≤ 4x/ wk

0.84 (0.12 – 6.59)

1.50 (0.18 – 10.33)

1.04 (0.13 – 7.53)

0.97 (0.21 – 4.24) 0.30 (0.08 – 1.06) 0.48 (0.06 – 3.59) 2.76 (0.35 – 26.20) 1.78 (0.10 – 32.88)

≤ 5% CuSO4, ≥ 4x/ wk

0.87 (0.11 – 7.50)

1.53 (0.19 – 11.68)

1.07 (0.13 – 8.01)

0.30 (0.04 – 2.31) 0.51 (0.06 – 3.88) 2.91 (0.31 – 26.57) 1.86 (0.11 – 34.11)

≤ 5% CuSO4, ≤ 4x/ wk

2.82 (0.27 – 32.64)

5.03 (0.40 – 52.89)

3.47 (0.31 – 37.74)

5.90 (0.25 – 146.73)

9.31 (0.82 – 122.53)

1.62 (0.14 – 17.54)

1% Hyperox (peracetic acid), 7 x/ wk

1.76 (0.18 – 19.79)

3.10 (0.30 – 31.56)

2.18 (0.22 – 21.67)

3.75 (0.19 – 76.26)

5.99 (0.61 – 57.05)

2% sodium hypochlorite, 4x/wk

0.30 (0.03 – 3.67)

0.53 (0.04 – 5.90)

0.36 (0.03 – 4.27)

0.65 (0.03 – 12.17)

2% Double Action, 5x/wk

0.46 (0.02 – 10.88)

0.78 (0.04 – 20.46)

0.56 (0.03 – 12.86)

3% Provita Hoofsure Endurance, 5x/wk

0.81 (0.11 – 6.59)

1.42 (0.34 – 6.09)

5% CuSO4 alternating weekly with 10% NaCl, 4x/wk

0.57 (0.08 – 4.70)

5% CuSO4 alternating weekly with water, 4x/ wk

Table 6 Pairwise comparisons (odds ratios with 95% credible interval in parentheses) and network meta-analysis of footbath protocols for the treatment of digital dermatitis.

5% Dragonhyde, 2x/wk

2% formalin, 2x/wk

5% Hoof fit bath, 4 milkings every 2 wk

(continued on next page)

5% Hoof fit bath, 4 milkings every 4 wk

C. Jacobs et al.

Preventive Veterinary Medicine 164 (2019) 56–71

Preventive Veterinary Medicine 164 (2019) 56–71

C. Jacobs et al.

2.33 (0.06 – 88.06) 1.64 (0.09 – 25.10) 0.72 (0.06 – 7.49) 0.40 (0.04 – 5.03) 0.58 (0.05 – 6.90) 0.34 (0.01 – 8.21) 0.22 (0.02 – 2.55) 1.25 (0.12 – 14.84) 0.32 (0.05 – 2.11)

This systematic review and network meta-analysis revealed limited publications (despite the widespread use of footbaths and availability of various footbath products) on footbath protocols suitable for inclusion in a network meta-analysis; therefore, there remains limited strength of evidence for use. Studies assessing herd level outcomes (prevalence) should report prevention and treatment outcomes so as to be included in future meta-analysis methods. Standardized protocols with large sample sizes are needed to further investigate effectiveness of footbath protocols for control of DD.

Not estimable; too sparse data.

1.90 (0.02 – 209.54)

Acknowledgements The authors thank all those who translated documents: Drs. Ivana Simkova (University of Veterinary and Pharmaceutical Sciences Brno, Czech Republic), Masahito Oba (University of Alberta, AB, Canada), Caroline Ritter (University of Calgary, AB, Canada), and Laura Solano (Farm Animal Care Associates, AB, Canada). We also appreciate Lorraine Toews (University of Calgary, AB, Canada) for her assistance

1

3.03 (0.24 – 42.88)

5% Hoof fit bath, 4 milkings every 4 wk 5% Hoof fit bath, 4 milkings every 2 wk 6% formalin, 7x/ wk

1.73 (0.16 – 22.12)

NE1

0.81 (0.01 – 124.65)

5. Conclusions

0.48 (0.07 – 3.27)

0.61 (0.09 – 4.57)

NE1 NE1

0.60 (0.16 – 2.21)

0.54 (0.02 – 11.12) 0.85 (0.02 – 34.30)

2.01 (0.64 – 7.01)

NE1 NE1 NE1 NE1 NE1

0.29 (0.01 – 11.01) 0.17 (0.01 – 6.26) 0.24 (0.01 – 9.00) 0.14 (0.00 – 7.79)

0.79 (0.05 – 13.13)

0.38 (0.03 – 5.19) 0.26 (0.01 – 7.60) NE1 1.87 (0.10 – 41.14) 1.31 (0.03 – 55.98) NE1 1.11 (0.07 – 18.89) 0.78 (0.05 – 10.09) NE1 2% formalin, 2x/ wk

Table 6 (continued)

Comparator

1.25 (0.01 – 137.93)

0.20 (0.02 – 2.08) 0.14 (0.01 – 2.94) NE1

0.30 (0.04 – 2.37) 0.20 (0.01 – 5.30) NE1

0.37 (0.03 – 4.52) 0.25 (0.01 – 7.42) NE1

1.24 (0.09 – 20.83)

0.13 (0.01 – 2.35) 0.09 (0.00 – 2.92) NE1

0.21 (0.01 – 6.52)

0.36 (0.03 – 4.82)

0.25 (0.02 – 3.55)

0.44 (0.10 – 1.87)

0.66 (0.01 – 32.05)

NE1

NE

treatment and prevention outcomes may favour protocols with low numbers of cow passes. All studies included contained an unclear risk of bias in at least 1 of the SYRCLE domains suggesting that certain biases are not being accounted for or discussed in current research. For example, all studies had an unclear risk of bias in the “blinding of caregivers” domain, with 12 of the studies also having an unclear risk of bias in the “blinding of outcome assessor” domain. Clear definitions of DD lesion stages, as well as having scoring done by a trained observer, may minimize impacts of these domains of bias. Various other domains contained unclear risk of bias for various studies, with allocation concealment and sequence generation as the most prevalent domains. Unfortunately, without more detail, it is difficult to assess impacts of unclear risk of bias for each study and doing so without the relevant information would be speculative. The importance of including and discussing risk of bias in studies should not be ignored as it provides insight on the conclusions drawn from each study. Future research regarding footbath protocols for the control of DD should include protocols standardized for frequency, concentration, footbath dimensions, number of cow passes and clear descriptions of DD lesions with case and success definitions. Additionally, based on the timeframe of lesion development and healing, multiple observations of DD lesions at intervals no longer than 1 month, as well as longer follow up periods (of at least 5 months) should be included to correctly assess the success of the footbath protocol and the product being tested. Unfortunately, these requirements implicate a significant financial contribution and may not always be feasible. Ariza et al. (2017) suggested 7 key elements for assessing future protocols in the prevention and treatment of DD lesions, which included: 1) proper randomization or comparable methods; 2) inclusion of negative untreated controls; 3) objective measure of DD lesion evolution; 4) lesion assessments at no longer than 1 month intervals; 5) follow up period at least 5 months; 6) sample size determined for statistical power; and 7) adjustment for co-interventions of other confounding variables in the analysis. These are all appropriate recommendations; however, including untreated controls may be considered unethical and with the ability to statistically compare various protocols to each other, may not be necessary. A strength of this network meta-analysis was the ability to compare all protocols to each other, enabling comparisons among products, concentrations and frequencies; however, many studies had small sample sizes, which when adjusted for clustering decreased the sample size considerably resulting in limited power to detect differences between protocols.

69

Preventive Veterinary Medicine 164 (2019) 56–71

C. Jacobs et al.

with search criteria for the systematic review. Thank you also to Dr. John Kastelic (University of Calgary, AB, Canada) for editing the manuscript. Dr. Glen Hazelwood is supported by a CIHR New

Investigator Salary Award and The Arthritis Society Young Investigator Salary Award. No additional funding sources were utilized for this project and none of the authors declare any conflicts of interest.

Appendix A

Table A1 PICO search strategy for CAB Abstracts database. Search Focus

Search Headings and Keywords

Population (P) Dairy cattle Intervention (I) Footbaths Comparator (C) Outcome (O) Digital Dermatitis

TX "dairy cattle" OR "dairy cow*" OR "dairy herd*" TX footbath* OR "foot bath*" OR "hoof bath*" OR hoof bath* Not included in search strategy TX "digital dermatitis" OR "hairy heel" OR "Mortellaro*"

cows. Smith, R.A. (Ed.), Proceedings of the Fortieth Annual Conference, American Association of Bovine Practitioners, Vancouver, British Columbia, Canada, 20-22 September, 2007. Higgins, J.P.T., S, G. (Eds.), 2011. Cochrane Handbook for Systematic Reviews of Interventions. The Cochrane Collaboration. Hoff, B., Boermans, H.J., Baird, J.D., 1998. Retrospective study of toxic metal analyses requested at a veterinary diagnostic toxicology laboratory in Ontario (1990-1995). Can. Vet. J. 39, 39–43. Holzhauer, M., Bartels, C.J.M., Borne, B.H.Pvd., Schaik, Gv., 2006a. Intra-class correlation attributable to claw trimmers scoring common hind-claw disorders in Dutch dairy herds. Prev. Vet. Med. 75, 47–55. Holzhauer, M., Hardenberg, C., Bartels, C.J.M., Frankena, K., 2006b. Herd- and cow-level prevalence of digital dermatitis in the Netherlands and associated risk factors. J. Dairy Sci. 89, 580–588. Holzhauer, M., Bartels, C.J.M., Döpfer, D., Schaik, Gv., 2008. Clinical course of digital dermatitis lesions in an endemically infected herd without preventive herd strategies. Vet. J. 177, 222–230. Holzhauer, M., Bartels, C.J., Bergsten, C., Riet, M.M.Jv., Frankena, K., Lam, T.J.G.M., 2012. The effect of an acidified, ionized copper sulphate solution on digital dermatitis in dairy cows. Vet. J. 193, 659–663. Hooijmans, C.R., Rovers, M.M., de Vries, R.B.M., Leenaars, M., Ritskes-Hoitinga, M., Langendam, M.W., 2014. SYRCLE’s risk of bias tool for animal studies. BMC Med. Res. Methodol. 14 43-43. Hutton, B., Salanti, G., Caldwell, D.M., Chaimani, A., Schmid, C.H., Cameron, C., Ioannidis, J.P.A., Straus, S., Thorlund, K., Jansen, J.P., Mulrow, C., Catala-Lopez, F.G., Gøtzsche, P.C., Dickersin, K., Boutron, I.A., Altman, D.G., Moher, D., 2015. The PRISMA extension statement for reporting of systematic reviews incorporating network meta-analyses of health care interventions: checklist and explanations. Ann. Intern. Med. 162 (777), 784. Klaas, I.C., Krogh, K., Sorensen, J.T., 2008. The effect of two different footbath solutions on digital dermatitis. 15th International Symposium & Confrence on Lameness in Ruminants 186, 189. Krull, A.C., Shearer, J.K., Gorden, P.J., Cooper, V.L., Phillips, G.J., Plummer, P.J., 2014. Deep sequencing analysis reveals temporal microbiota changes associated with development of bovine digital dermatitis. Infect. Immun. 82, 3359–3373. Krull, A.C., Shearer, J.K., Gorden, P.J., Scott, H.M., Plummer, P.J., 2016. Digital dermatitis: natural lesion progression and regression in Holstein dairy cattle over 3 years. J. Dairy Sci. 99, 3718–3731. Laven, R.A., 1999. The environment and digital dermatitis. Cattle Pract. 7, 349–354. Laven, R.A., Hunt, H., 2002. Evaluation of copper sulphate, formalin and peracetic acid in footbaths for the treatment of digital dermatitis in cattle. Vet. Rec. 151, 144–146. Laven, R., Logue, D., 2006. Treatment strategies for digital dermatitis for the UK. Vet. J. 171, 79–88. Manske, T., Hultgren, J., Bergsten, C., 2002. Topical treatment of digital dermatitis associated with severe heel-horn erosion in a Swedish dairy herd. Prev. Vet. Med. 53, 215–231. Orsel, K., Plummer, P., Shearer, J., De Buck, J., Carter, S.D., Guatteo, R., Barkema, H.W., 2018. Missing pieces of the puzzle to effectively control Digital Dermatitis. Transbound. Emerg. Dis. 65 (Suppl. 1), 186–198. Read, D.H., Walker, R.L., 1998. Papillomatous digital dermatitis (footwarts) in California dairy cattle: clinical and gross pathologic findings. J. Vet. Diagn. Invest. 10, 67–76. Relun, A., Lehebel, A., Bareille, N., Guatteo, R., 2012. 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. J. Dairy Sci. 95, 3722–3735. Relun, A., Lehebel, A., Bruggink, M., Bareille, N., Guatteo, R., 2013. Estimation of the relative impact of treatment and herd management practices on prevention of digital dermatitis in French dairy herds. Prev. Vet. Med. 110, 558–562.

References Ariza, J.M., Relun, A., Bareille, N., Oberle, K., Guatteo, R., 2017. Effectiveness of collective treatments in the prevention and treatment of bovine digital dermatitis lesions: a systematic review. J. Dairy Sci. 100, 7401–7418. Bergsten, C., Hultgren, J., Hillström, A., 2007. Using copper sulphate, peracetic acid or a combination of both in foot bath for the control of digital dermatitis and heel horn erosion in dairy cows. Aland, A. (Ed.), Animal Health, Animal Welfare and Biosecurity. Proceedings of 13th International Congress in Animal Hygiene, Tartu, Estonia, 17-21 June, 2007. Volume 1. Blowey, R., Carter, S., White, A., Barnes, A., 1994. Borrelia burgdorferi infections in UK cattle: A possible association with digital dermatitis. Vet. Rec. 135, 577–578. Blowey, R., Sullivan, L., Evans, N., Carter, S., 2013. Detection of bovine digital dermatitis on hoof knives and hands and their removal by disinfection. Whay, B., Hockenhull, J. (Eds.), Proceedings of the 17th International Symposium and 9th International Conference on Lameness in Ruminants. Brooks, S.P., Gelman, A., 1998. General methods for monitoring convergence of iterative simulations. J. Comput. Graph. Stat. 7, 434–455. Cook, N.B., Rieman, J., Gomez, A., Burgi, K., 2012. Observations on the design and use of footbaths for the control of infectious hoof disease in dairy cattle. Vet. J. 193, 669–673. Cramer, G., Benavides, M.L., 2012. An evaluation of the preventative and curative properties of 3 foot bath products used to control digital dermatitis. Smith, R.A. (Ed.), Proceedings of the 45th Annual Conference of the American Association of Bovine Practitioners, Montreal, Canada, 20-22 September, 2012. American Association of Bovine Practitioners 258-258. Cramer, G., Lissemore, K.D., Guard, C.L., Leslie, K.E., Kelton, D.F., 2008. Herd- and cowlevel prevalence of foot lesions in Ontario dairy cattle. J. Dairy Sci. 91, 3888–3895. Doane, M., Sarenbo, S., 2014. Exposure of farm laborers and dairy cattle to formaldehyde from footbath use at a dairy farm in New York State. Sci. Total Environ. 487, 65–71. Döpfer, D., Dandurand, U.D., Desrochers, A., Letellier, A., Quessy, S., Bourgeois, A., Dussault, C.F., Allard, M., 2013. Comparison of hoof bath solutions for the prevention and control of digital dermatitis in dairy cows [Conference poster]. Whay, B., Hockenhull, J. (Eds.), Proceedings of the 17th International Symposium and 9th International Conference on Lameness in Ruminants, Bristol, UK, 11-14 August 2013 228–229. Edwards, A., Dymock, D., Jenkinson, H., 2003. From tooth to hoof: treponemes in tissuedestructive diseases. J. Appl. Microbiol. 94, 767–780. Epperson, B., Midla, L., 2007. Copper sulfate for footbaths - issues and alternatives. Eastridge, M.L. (Ed.), Proceedings of the 2007 Tri-State Dairy Nutrition Conference, Fort Wayne, Indiana, USA, 24-25 April, 2007 51–54. Evans, N.J., Timofte, D., Isherwood, D.R., Brown, J.M., Williams, J.M., Sherlock, K., Lehane, M.J., Murray, R.D., Birtles, R.J., Hart, C.A., Carter, S.D., 2012. Host and environmental reservoirs of infection for bovine digital dermatitis treponemes. Vet. Microbiol. 156, 102–109. Fjeldaas, T., Knappe-Poindecker, M., Boe, K.E., Larssen, R.B., 2014. Water footbath, automatic flushing, and disinfection to improve the health of bovine feet. J. Dairy Sci. 97, 2835–2846. Flemming, C., Trevors, J., 1989. Copper toxicity and chemistry in the environment: a review. Water Air Soil Pollut. 44, 143–158. Gelman, A., Rubin, D.B., 1992. Inference from iterative simulation using multiple sequences. Stat. Sci. 457–472. Gomez, A., Cook, N.B., Rieman, J., Dunbar, K.A., Cooley, K.E., Socha, M.T., Döpfer, D., 2015. The effect of digital dermatitis on hoof conformation. J. Dairy Sci. 98, 927–936. Goossens, X., Hemling, T., 2007. Comparison of the efficacy of double action and copper sulphate footbath solutions for the control of digital dermatitis in lactating dairy

70

Preventive Veterinary Medicine 164 (2019) 56–71

C. Jacobs et al. Richardson, W.S., Wilson, M.C., Nishikawa, J., Hayward, R.S., 1995. The well-built clinical question: a key to evidence-based decisions. ACP J. Club 123, A12–A13. Rodríguez Lainz, A., Hird, D.W., Walker, R.L., Read, D.H., 1996. Papillomatous digital dermatitis in 458 dairies. J. Am. Vet. Med. Assoc. 209, 1464–1467. Smith, A.C., Wood, C.L., McQuerry, K.J., Bewley, J.M., 2014. Effect of a tea tree oil and organic acid footbath solution on digital dermatitis in dairy cows. J. Dairy Sci. 97, 2498–2501. Solano, L., Barkema, H.W., Pajor, E.A., Mason, S., LeBlanc, S.J., Zaffino Heyerhoff, J.C., Nash, C.G., Haley, D.B., Vasseur, E., Pellerin, D., Rushen, J., de Passille, A.M., Orsel, K., 2015. Prevalence of lameness and associated risk factors in Canadian HolsteinFriesian cows housed in freestall barns. J. Dairy Sci. 98, 6978–6991. Solano, L., Barkema, H.W., Mason, S., Pajor, E.A., LeBlanc, S.J., Orsel, K., 2016. Prevalence and distribution of foot lesions in dairy cattle in Alberta. Canada. J. Dairy Sci. 99, 6828–6841. Solano, L., Barkema, H.W., Pickel, C., Orsel, K., 2017. Effectiveness of a standardized footbath protocol for prevention of digital dermatitis. J. Dairy Sci. 100, 1295–1307. Somers, J.G.C.J., Frankena, K., Noordhuizen-Stassen, E.N., Metz, J.H.M., 2005. Risk factors for digital dermatitis in dairy cows kept in cubicle houses in the Netherlands.

Prev. Vet. Med. 71, 11–21. Speijers, M.H.M., Baird, L.G., Finney, G.A., McBride, J., Kilpatrick, D.J., Logue, D.N., O’Connell, N.E., 2010. Effectiveness of different footbath solutions in the treatment of digital dermatitis in dairy cows. J. Dairy Sci. 93, 5782–5791. Speijers, M.H.M., Finney, G.A., McBride, J., Watson, S., Logue, D.N., O’Connell, N.E., 2012. Effectiveness of different footbathing frequencies using copper sulfate in the control of digital dermatitis in dairy cows. J. Dairy Sci. 95, 2955–2964. Teixeira, A.G.V., Machado, V.S., Caixeta, L.S., Pereira, R.V., Bicalho, R.C., 2010. Efficacy of formalin, copper sulfate, and a commercial footbath product in the control of digital dermatitis. J. Dairy Sci. 93, 3628–3634. Thomsen, P.T., 2015. Efficacy of copper sulfate hoof baths against digital dermatitis where is the evidence? J. Dairy Sci. 98, 2539–2544. Thomsen, P.T., Sørensen, J.T., Ersbøll, A.K., 2008. Evaluation of three commercial hoofcare products used in footbaths in Danish dairy herds. J. Dairy Sci. 91, 1361–1365. Turner, R.M., Davey, J., Clarke, M.J., Thompson, S.G., Higgins, J.P.T., 2012. Predicting the extent of heterogeneity in meta-analysis, using empirical data from the Cochrane Database of Systematic Reviews. Int. J. Epidemiol. 41, 818–827.

71