Statistics from Altmetric.com
The paper by Young et al is a useful contribution to the question of work disability related to rheumatoid arthritis.1
However, one of the areas in which intervention is theoretically possible to reduce disability—namely, that of work place intervention, was not discussed. In the study by Young et al the types of work performed by the study cohort are grouped into four categories—namely, manual, semi-manual, semi-sedentary, and mainly sedentary. Presumably, the patients were allocated to a group based on job title using the British classification of occupations and coding index.2
It is well known to occupational physicians and others taking occupational histories that a job title does not adequately reflect the true nature of work.3 In addition, problems likely specifically to affect patients with rheumatoid arthritis are not usefully classified in a subjective ordinal scale such as that used in the study. Examples of such problems might be fine repetitive movements of the hand or work starting times.
To gain an accurate insight into work factors affecting work disability, a study is required in which information on the nature of the work tasks is obtained. Ideally this should be gathered prospectively by direct analysis of the work place by suitably trained observers. Tools have been developed that can assist with this type of data gathering.4,5
This would reduce recall and misclassification bias of previous studies. Once this information is obtained, work place intervention as an approach to minimising disability can be implemented and assessed for efficacy and effectiveness.
Dr Smith and colleagues make a valid point about studies on work disability in general, although I am not sure that this is highly relevant to our report. We do not dispute any of the other points made. Job title in our patients was based on the Office of National Statistics classification.
We agree that a very detailed account of work tasks taken at the onset by specially trained observers, and repeated regularly until work loss, might reduce possible recall and misclassification bias. However, ERAS, which was started in 1986, aimed at recording outcomes in several quite different dimensions in ordinary busy clinical settings and not in the degree of detail outlined. In the same way, very detailed accounts of home and social circumstances, factors also known to affect work disability, were not included. Despite this, as we make clear in our report, we feel we have adequately highlighted the importance of work disability in RA. Although the above authors do not say as such, the sort of study they describe needs to be set up with the primary aim of investigating possible specific interventions.