**3. The study findings**

396 A Bird's-Eye View of Veterinary Medicine

the rest (frequency and amount of concentrate feeding , mineral supplementation, type of fodder, and frequency of manure removal from the walk-alleys were collected by the author interviewing the farmers or the persons managing the zero-grazing units. All the data both for animal-and unit-level risk factors were entered into the questionnaires using codes adopted at the pre-coding stage during an initial pilot study in 5 sample smallholder zero-

Each cow was examined for general lameness using universally described locomotion scoring system (Sprecher et al., 1997). Data on claw disorders were collected by examining each claw on each cow. Each cow was restrained in a standing posture in a restraining area (crush). The limbs were lifted one at a time by tying with a rope and fastening the rope to an overhead pole or cross-bar. In that raised position, each claw was thoroughly washed with soap and water and examined for any lesions particularly on the weight-bearing surface. About 1-2 mm thickness of the horn of the sole was trimmed-off using a sharp quitter knife to expose any underlying lesions particularly sole haemorrhages. Trimming did not reach the level of corium and therefore no pain was caused on the cows. In cases with painful claw conditions, foot anaesthesia using 2% lignocaine hydrochloride was applied to allow

The data were entered and stored in Microsoft Office Excel 2003. After verification and validation of data, they were imported into SAS© 2002-2003 (SAS Institute Inc., Cary, NC, USA). Descriptive statistics were computed for cow-level and unit-level risk factors. The prevalence rate of each claw disorder was calculated independent of other claw disorders. The prevalence of each claw disorder was calculated as the number of cows affected by the specific claw disorder or lesion (nCL) divided by the total number of cows (300) examined,

Chi-square (χ2) statistics were used to determine unconditional associations between all risk factors and the claw lesions. An association was considered significant at the level of p<0.05. Chi square values were determined using 2 x 2 contingency tables constituting 2 rows and 2 columns. In these associations the Chi square calculations were determined by evaluating each risk factor (variable) against each claw disorder (outcome) on the cow. The degrees of

Therefore, df was 1 for each association test. Multiple logistic regressions were done through a step-down regression in which the risk factors that made the least variation to the occurrence of the claw lesions were eliminated one at a time through consideration of their odds ratios. Only the factors that were found to influence the occurrence of claw lesions significantly were retained in the model. The effects of confounding risk factors were dealt

[(row-1) x (columns-1)], hence [(2-1) x (2-1)] (2)

*nCL <sup>x</sup>* (1)

painless manipulation. The lesions found on each cow were recorded.

Prevalence (%) 100 <sup>100</sup>

freedom (df) in each case was standard, being calculated by:

grazing dairy units.

**2.4 Examination of the cows for claw lesions** 

**2.5 Data management and analysis** 

and then multiplied by 100.
