Developing an outcome measure in temporomandibular disorders
Assessment of treatment outcome in clinical trials of temporomandibular disorders (TMD) lacks validated patient-based measures of the condition and its improvement. Such measures are questionnaire-based. The questions need to be chosen and framed appropriately. In line with the guidelines of the National Institute for Health and Clinical Excellence in the U. K., this project aimed to start the development of a patient-derived quality of life instrument to measure TMD treatment outcome. Three studies were conducted for this purpose. Study One aimed to establisht he optimal time to assesstr eatmento utcomef or TMD (the Review Period), and the best Reference Period to determine the time-frame of questions. Data consisted of daily diaries of pain intensity from 72 patients who had participated in a clinical trial of conservative TMD treatment. The study involved two stages. The first aimed to identify patientsr espondingto treatment,s o-called" Improvers" andt o excludet he "Non- Improvers" whosed atam ight dilute the results.T his was accomplishedin two ways: firstly by visually assessingin dividual plots of pain versust ime; secondlyb y mathematicallyc alculating the reduction of pain scores during treatment. Criteria for improvement were set and tested for both methods.D efinite improvers were selectedi f they met criteria of both assessmentsIn. the second stage, only improvers' data were examined to analyse the general trend of improvement and establish the review and reference periods. Study One determined 15 weeks as the best review period and four weeks as the best referencep eriod when assessinga commonly used TMD treatment. Study Two was intended to triangulate with Study One and had the same aim. Study Two explored, using a qualitative approach, patients' perception of improvement of TMD symptoms and rates of their recovery. Ten TMD patients showing improvement to conservativet reatmentw ere interviewed. Data were collected using a pre-designedt opic guide and analysed using the Framework approach. The interviewees consistently reported that pain was the symptom with the most impact on their quality of life, and the one to which improvement was linked the most. Other TMD symptoms were also important, but were less linked to patients' suffering and to perceived improvement. Different symptoms followed various rates of recovery. It was not possible to determine a definite time when all symptoms are considered improved. This study confirmed that the approach used in Study One, i. e. assessing TMD recovery based on measurement of pain intensity, was reasonable. It also confirmed that sufficient time (in a frame of some months) is needed before outcome for TMD conservative treatment can be assessed meaningfully. In addition, this study identified important new themes related to patient's journey throughout illness and their perception of received care. Study Three was the core study in this project. It aimed to establish a quality of life measure of TMD treatment outcome using the Oral Health Impact Profile (OHIP) as a starting point, using the reference period identified in studies one and two. A short form of OHIP was derived by identifying OHIP items with the largest impact on TMD patients' quality of life. A case-control design was used and 110 patients (PG) undergoing a variety of conservative treatments and matching controls (CG) were included. Candidate questions (items) for the intendeds hort form (OHIP/TMD-I) were selectedb asedo n three criteria: 1- showings tatistical differencesb etweenP G and CG; 2- showing largests cored ifference (PG:C G) as measuredb y four analysesM: ean, Median, Prevalencea ndI tem-Impact;3 - representinga ll domainso f OHIP's theoretical framework. This was followed by testing aspects of validity and reliability of the short form and comparing them to those of the full OHIP. Twenty items were identified. Their psychometric properties were comparable or better than those of the mother instrument. Future work is neededt o analyset he responsivenestso changew ith treatmento f the short form, to investigatew hetheri temso utwith OHIP aren eededto furnisht he intendedT M) outcome instrument, and to test the psychometric properties of the final instrument in a new sample.