Introduction

Many objective measures of disease in dogs with OA are largely "point-in-time" assessments. Body mounted activity monitors provide a potential means to gather objective data about a dog's activity over a period of days to weeks. It is intuitive that states of chronic orthopaedic pain could affect levels and/or patterns of activity, and accelerometer-based devices have been used in studies of OA in people, dogs and cats. Their use in home-residing domestic animals, however, is in its infancy and many questions remain regarding the best way to process and interpret the data gathered. Our aim was to evaluate the usefulness of these devices as an outcomes measure, and to explore methods of analysing the gathered data.

Methods

Dogs with OA were recruited to take part in a clinical trial of two licensed NSAIDs via a local community-based awareness campaign and a staged screening process. Recruited dogs wore an activity monitor (Actical, Philips Respironics) mounted at the ventral aspect of their collar for a total of 14 weeks: two weeks initially whilst not receiving treatment (baseline), and for 12 weeks whilst receiving a licensed NSAID (on label). Epoch length was set at one minute, i.e., an activity "count" was returned for every minute the device was worn. An explorative approach to data analysis was taken, considering the findings of published reports. On treatment activity data were compared to baseline data using Wilcoxon Ranked Sum Tests.

Results

A significant change (p < 0.05) in the average activity count for the period midnight–6 am was found when comparing the baseline with the sixth week of treatment, but no other significant changes were found over this period.

Discussion

Previous reports on these devices suggest that "Total Weekly Counts" and activity levels for the period 6am to noon may be useful, but our results thus far do not support these recommendations. Owner compliance and device failure were limiting factors in our study: these problems may have been exacerbated by the prolonged period (six weeks) between visits. Data collection and analysis continues, and data will also be compared against other outcomes measures.

Conclusions

If activity monitors are to be a useful measure of canine OA issues surrounding owner compliance and device reliability should be considered, and alternative methods of data analysis should continue to be explored.