Item selection for a Temporomandibular Disorders (TMD) specific outcome measure

Objectives: To combine qualitative and quantitative measures to identify candidate items from the Oral Health Impact Profile (OHIP) most suited to forming a TMD specific outcome measure.

Methods: Data were combined from two studies. From the first, data from 110 pre-treatment TMD patients and 110 age/sex matched controls were compared to identify differences in mean, median, prevalence and item impact scores for each OHIP-49 item. Independently, in-depth interviews were conducted with a separate purposive sample of pre-treatment TMD patients (n=29). Data collection and analysis was iterative and continued until saturation. The 49 items of OHIP were pre-set as a priori themes for analysis so any new (non-OHIP) themes could be identified whilst the relevance of existing items to TMD sufferers could be cross-referenced with the quantitative data.

Results: Eighteen items met the quantitative selection criteria. These were the highest ranking items from each of the seven OHIP domains plus a further 11 items from across the whole measure which met a pre-defined “clinically meaningful difference” between cases and controls (one full unit of difference on OHIP's 5 point scale for both mean and median differences). The content analysis of the qualitative data triangulated well to these 18 quantitatively determined candidate items. It also identified two existing items in the social disability and handicap domains, excluded by quantitative analysis due to “floor effects” but with an important impact on sufferers. Two additional areas were missed by OHIP and for these additional items may need to be generated and tested, giving twenty-two candidate items for a condition specific OHIP for TMD.

Conclusions: A combination of quantitative and qualitative methods worked well to allow the provisional selection of 22 items for what should be a robust, condition specific measure of the impact of TMD on the sufferer.