Article Text

Download PDFPDF

Effects of parenteral amoxicillin on recovery rates and new infection rates for contagious ovine digital dermatitis in sheep
  1. J. S. Duncan, BSc, BVM&S, PhD, MRCVS,
  2. D. Grove-White, BVSc, DBR, MSc, DLSHTM, PhD, DECBHM, FRCVS,
  3. J. W. H. Oultram, BVScCert, CHP, MRCVS,
  4. C. J. Phythian, BSc, BVSc, MRCVS,
  5. J. van Dijk, DVM, PhD, MRCVS,
  6. S. D. Carter, BSc, PhD, FIMLS, FRCPath,
  7. P. J. Cripps, BSc, BVSc, MSc, PhD, MRCVS and
  8. H. J. Williams, BVScCert, CHP, MRCVS
  1. School of Veterinary Science, University of Liverpool, Leahurst Campus, Neston, Wirral, Cheshire CH64 7TE, UK
  1. E-mail for correspondence jsduncan{at}

The present study is a randomised split-flock treatment trial, which compared the effect of foot bathing in a 1 per cent solution of chlortetracycline alone with a treatment protocol that added a single injection of a long-acting amoxicillin. Overall, the prevalence of contagious ovine digital dermatitis in the examined flock was 22 per cent, while 45.7 per cent of affected sheep had infections in two or more feet. Parenteral antibiotic treatment increased the odds of a recovery by 3.8 times (95 per cent confidence interval 1.05 to 14.0) (P=0.008). Moreover, the amoxicillin injection may also have had a preventative effect, reducing the rate of establishment of new infections from 2.5 per cent for foot bathing alone compared with 1.0 per cent with the addition of parenteral amoxicillin.

View Full Text

Statistics from

CONTAGIOUS ovine digital dermatitis (CODD) is a serious, painful and common foot disease of sheep in the UK, presenting an important challenge to both sheep welfare and the economics and sustainability of sheep farming. This cause of lameness was first reported in the UK in 1997 (Harwood and others 1997). To date, there are no reports of the condition occurring in sheep outside the UK. Although not formally quantified, for example, by lameness scoring, it is widely accepted that lameness caused by CODD is particularly severe, with affected animals often non-weightbearing on affected limbs. This is likely to be a consequence of the pathology associated with the disease that is characterised by under-running of the hoof wall starting at the coronary band, exposure of the sensitive laminae and ultimately avulsion of the entire hoof capsule (Fig 1A) (Winter 2008).

Fig 1 A

Photograph of contagious ovine digital dermatitis lesion, lesion score 1

The reported response of CODD to antimicrobials has been taken to indicate a bacterial aetiology for the disease (Davies and others 1999). Indeed, Treponema species related to isolates from bovine digital dermatitis and human periodontitis have been isolated from clinical cases of CODD (Collighan and others 2000, Demirkan and others 2001, Moore and others 2005, Naylor and others 1998, Sayers and others 2009). Moore and others (2005) reported that Dichelobacter nodosus (the cause of foot rot) and Treponema species were isolated from ≥70 per cent of CODD lesions found on six UK sheep farms.

Evidence suggests that CODD has already become established as a common cause of sheep lameness in the UK. A postal survey in 2003, of 209 UK sheep farmers (Wassink and others 2003), suggested that CODD was likely to be present on between 11 and 17 per cent of sheep farms in the UK, with an average flock prevalence of 25.5 per cent (range 0.5 to 100 per cent) in adult sheep and 14.6 per cent (range 0 to 80 per cent) in lambs. While a more recent survey in England identified a farmer-reported prevalence of 2.4 per cent, assuming that these farmer-made diagnoses were correct, these results of the present study suggested that CODD was almost as common a cause of lameness as was foot rot (3.7 per cent) (Kaler and Green 2008). Currently, there is no scientific evidence base to inform treatment and control strategies for the condition, nor are there any products authorised for the treatment of CODD in the UK. Tilmicosin (Micotil; Elanco Animal Health) is used by some veterinary surgeons to treat the condition. In addition, anecdotal evidence suggests that antibiotic foot baths with products containing lincomycin, spectinomycin or tylosin can be an effective treatment for CODD (Winter 2004). However, there are no controlled treatment studies to support their use. Therefore, the aim of the present study was to compare the curative effects of two treatments for CODD in a split-flock treatment trial.

Materials and methods

The study was conducted in accordance with the regulations of the Animals (Scientific Procedures) Act and was approved by the University of Liverpool Ethics Committee, Reference Number RETH 000291.

Farm background

The study was undertaken on a 1300 ewe, mixed-breed, lowland sheep farm in Cheshire, UK. CODD had first been diagnosed on the farm, by a veterinary surgeon, in 2007. The farmer reported that CODD prevalence tended to be highest in the lambs at housing.

The outbreak reported here occurred in December 2009 in a group of 320, approximately seven- to eight-month-old crossbred fattening lambs. Lameness was first observed by the farmer in a small number of lambs at pasture, one week before the lambs were housed. When housed for finishing, the lambs were managed on straw-bedded pens, at a stocking rate of approximately 30 lambs per 45m2 pen, and fed on a ration of barley straw and concentrate feed. Within approximately two weeks of housing, the prevalence of lameness in the group increased dramatically and veterinary advice was sought. CODD was confirmed as the main cause of lameness by the veterinarian, based on clinical signs.

Examination and treatment of sheep

All four feet of all lambs were individually examined by one of five veterinarians and each foot scored according to the following system: score 1 Presence of CODD lesion (Figs 1 A and B), score 0 Absence of CODD lesion or healed CODD lesion (Fig 2). Foot lesions other than CODD were also recorded. The ear tag number and pen number of each sheep were recorded after examination.

Fig 1 B

Photograph of contagious ovine digital dermatitis lesion, lesion score 1

Fig 2

Photograph of healing contagious ovine digital dermatitis lesion, lesion score 0

Sheep were allocated to one of two treatment protocols according to ear tag; sheep with an odd numbered ear tag were allocated to treatment protocol A and sheep with an even numbered ear tag were allocated to treatment protocol B. Treatment protocol A consisted of careful hoof trimming where required, followed by spraying any CODD-affected feet with 1 per cent chlortetracycline hydrochloride (aureomycin-soluble powder; Fort Dodge Animal Health). All sheep in the group were then foot bathed daily for three days with 1 per cent chlortetracycline hydrochloride. A foot bath of 80 litres volume was used and antibiotic foot bathing solution was topped up so that it remained at a depth that covered the accessory digits. Sheep stood in the foot bathing solution for two minutes and the footbath was completely changed after every 100 sheep. After foot bathing, the sheep stood on dry concrete for 30 minutes before returning to the housing. Treatment protocol B was identical, except for the addition of an intramuscular injection of 15 mg/kg long-acting amoxicillin trihydrate (Bimoxyl LA; Bimeda).

Following treatment, the sheep were returned to their original pens and maintained there until the follow-up visit three weeks later. At the follow-up visit, all feet were examined and classified (score 1 Presence of CODD, score 0 Absence of or healed CODD) by the same examiners. To ensure lesion scoring was not influenced by knowledge of treatment protocol, ear tag numbers were recorded after the sheep had been examined.

Following the second visit, further treatment of remaining lame lambs using protocol B was undertaken by the farmer until clinical recovery was established for the entire group.

Statistical methods

Data were entered in an Excel spreadsheet and transferred to Stata 10 (StataCorp) for analysis. Due to the hierarchical nature of the data with feet clustered within sheep and sheep clustered within pens, random effects logistic regression models were fitted with random terms for both sheep identity and pen. Likelihood ratio testing was used to determine which variables to keep in the final models with a P value <0.2 being taken to imply that retention of a variable in the model significantly improved model fit. Inclusion of the variable ‘pen’ did not improve model fit and thus was removed from all models. Two models were fitted: in the first, the binary outcome variable was whether clinical resolution of a diseased foot occurred while in the second model, the binary outcome variable was whether an animal acquired a new CODD infection during the study period. In both models, the explanatory variable under consideration was systemic antibiotic therapy.


One hundred and fifty-four lambs, of which 37 were affected by CODD (59 feet), were randomly allocated to treatment protocol A (foot trim and chlortetracycline foot bathing). One hundred and sixty-six lambs, of which 33 were affected by CODD (53 feet), were randomly allocated to treatment protocol B (foot trim, chlortetracycline foot bathing and systemic long-acting amoxicillin).

Before treatment, 70 sheep (21.9 per cent) were affected by CODD. A total of 112 feet (8.8 per cent) were affected. Of the 70 affected sheep, 38 (54.3 per cent) sheep were affected in one foot only, 24 sheep (34.2 per cent) in two feet, six sheep (8.6 per cent) in three feet and two sheep (2.9 per cent) in all four feet. Of the 112 affected feet, 70 (62.5 per cent) feet were hind feet while 42 (37.5 per cent) were front feet (P<0.001).

Nineteen sheep did not complete the study, of which six sheep had lesions of CODD at the first visit. Therefore, of the total 112 affected feet at the first visit, 104 were examined at the second visit, that is, eight affected feet were lost to follow-up. Fifty-five of the original 59 CODD-affected feet were in treatment protocol A group, while 49 of the original 53 CODD-affected feet were in treatment protocol B group. Twenty-nine (52.7 per cent; 95 per cent confidence interval [CI] 39.1 to 66.3) of the 55 feet treated by protocol A were considered to be recovered at the second visit while 38 (77.5 per cent; 95 per cent CI 65.4 to 89.7) of the 49 feet treated by protocol B were considered recovered at the second visit (Table 1). There was a significant effect (P=0.041) of systemic amoxicillin antibiotic treatment on recovery rate (OR 3.84; 95 per cent CI 1.05 to 14.0).

Table 1

Summary outcomes of treatment protocols A and B

Nineteen new CODD lesions from 1100 previously unaffected feet were observed at the second visit representing an overall new infection rate of 1.7 per cent (95 per cent CI 1.0 to 2.1). Thirteen of these feet were from sheep treated with treatment protocol A, while six were from animals also treated with treatment protocol B (addition of systemic amoxicillin) (Table 1). The baseline odds, that is, the odds of a foot in the non-intervention group (footbath only) acquiring a new CODD lesion were 0.015 (95 per cent CI 0.004 to 0.053). Additional treatment with systemic antibiotic reduced the odds of a foot acquiring a new CODD lesion by almost half (odds ratio 0.42; 95 per cent CI 0.15 to 1.14; P=0.09).


The within-group prevalence of sheep affected by CODD (21.9 per cent) was comparable with the levels reported by Wassink and others (2003). Most affected sheep had lesions in just one or two feet, but a small number had three or four affected feet, indicating very severe welfare problems for such affected animals.

Significantly more CODD lesions were found in hind feet (62.5 per cent) than front feet (37.5 per cent). Interestingly, digital dermatitis in cattle predominantly affects the hind feet. In one study of digital dermatitis, 82 per cent of cases involved only the hind limbs, compared with 13 per cent being exclusive to the forelimb (Read and Walker 1998). In the current study, the authors could not identify any factors that could explain this similar observation in sheep. Anecdotal evidence suggests that parenteral antibiotic therapies such as tilmicosin or topical treatment with tylosin or lincospectin can be effective against CODD (Winter 2004). The antibiotic selection of parenteral amoxicillin and topical chlortetracycline in this study was guided principally by published antibiotic sensitivity data of Treponema species isolated from a CODD lesion (Evans and others 2009). Additional practical considerations were the length of the withdrawal periods for fattening lambs and finally that the farmer should be legally permitted to administer any follow-up treatments. Clinical recovery of CODD was identified in both treatment protocols; however, the addition of systemically administered long-acting amoxicillin significantly improved the clinical recovery rate and tended to reduce the odds of a sheep acquiring a new infection. In the present study, both treatment groups were foot bathed daily for three days as compared with normal practice on sheep farms whereby sheep will be foot bathed on one occasion only (Grove-White; unpublished observations).

The decision to undertake the therapeutic approach of whole group treatment with amoxicillin (protocol B) was based on experience of treatment of CODD in previous outbreaks and the apparently contagious nature of the disease. The recovery rates observed go some way to justifying this decision in terms of animal welfare benefits. Although not quantified here, additional economic and therefore sustainability benefits such as time to finishing and feed costs for these fattening lambs were also reported by the farmer. However, a comparative study of the efficacy of targeted versus whole group antibiotic treatments for CODD should be undertaken to further inform clinical decision making.

In conclusion, CODD may severely affect large proportions of sheep in a flock and constitutes a very significant welfare problem. It can be treated with foot bathing (three times in three days) in chlortetracycline alone but adding a single injection of long-acting amoxicillin to the treatment protocol significantly improved recovery rates. Moreover, the injection may also reduce the risk of development of new infections'. The results of this work are promising but more work quantifying and optimising CODD treatment regimens is urgently needed.


The authors thank the farmer and the University of Liverpool Veterinary School final-year veterinary students for their assistance in this study.


View Abstract


  • Provenance not commissioned; externally peer reviewed

Request Permissions

If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.