Reproducibility and construct validity of the utrecht mixing ability Test to obtain masticatory performance outcome in patients with condylar mandibular fractures

Abstract Objective This study assessed the test‐retest reproducibility of the Utrecht mixing ability test (MAT) and the construct validity of the MAT in relation to the Mandibular Function Impairment Questionnaire (MFIQ) in patients with mandibular condylar fractures. Material and methods Twenty‐six patients treated for a mandibular condylar fracture participated in this clinimetric study; all patients performed the MAT twice. Simultaneously the MFIQ was conducted. Test‐retest reliability and construct validity were assessed using the intra‐class correlation coefficient (ICC) and Spearman correlation, respectively. Results The ICC of the MAT was 0.906 (95% CI: 0.801‐0.957), which indicates an excellent reliability. A weak correlation of 0.386 (P = .052) between the first MAT and the overall outcome of the MFIQ was found. A significant moderate correlation of 0.401 (P = .042) was found between the retest of the MAT and the overall outcome of the MFIQ. One question on the MFIQ (about yawning) showed a moderate positive correlation of 0.569 (P = .002) and 0.416 (P = .034) for the MAT test and retest, respectively. Conclusion The MAT is an easy test to use in follow‐up of patients. The test‐retest reliability of this test is excellent in condylar trauma patients. As the validity of the MAT and the MFIQ could not be confirmed, the MFIQ may be an addition to patient's feedback about the rehabilitation process of their mandibular functioning.

a role in mastication. 2 The mandible is fractured in 36% to 54% of all patients with maxillofacial trauma. 3,4 In 36% to 44% of mandibular fractures, the mandibular condyle is involved. 3,4 Such a fracture can influence masticatory functioning due to anatomical change to the mandible or injury to the nerve or musculature. 5,6 However, the therapy received may influence mastication by complications due to open reduction and internal fixation, such as fistulas of the parotid gland and/or facial nerve damage or hardware problems. Similarly, ankyloses of the temporomandibular joint or limited mouth opening can occur due to long-term immobilisation in the case of conservative treatment. 7 Thereby, malocclusion, limited range of motion of the mouth or chronic pain can disturb the mastication process. 8 Masticatory performance is the objective efficiency of this mastication process, which can be measured by different methods (fi comminution or mixing ability methods). 9 The Utrecht mixing ability test (MAT) with two-coloured wax was described as a reliable test for patients with cerebral palsy syndrome. 10 For patients with mandibular trauma, such as condylar fracture, the reproducibility and validity of this test have not yet been investigated. 2,9 Masticatory ability is the subjective testing of the mastication process, which reflects the expectations of the patients and their quality of life by taking the psychological and emotional adjustment of the patient in their daily life into account. This can be an advantage over measurement of objective outcomes alone. Subjective efficiency of the mastication process is tested in various departments with several questionnaires, such as the Oral Health Impact Profile-14 questionnaire (OHIP-14) 11 and the Mandibular Function Impairment Questionnaire (MFIQ). 12,13 In recent studies, the MFIQ has been used to measure subjective masticatory ability in condylar trauma patients. 5,14 It is important to the rehabilitation of the patient to get insight into mastication after mandibular injury by performing reliable and valid tests. 15 Therefore, the aim of this study is to determine, on the one hand, the test-retest reproducibility (reliability, measurement error and agreement) of the MAT, and, on the other hand, the construct validity of the MAT in relation to the MFIQ in patients with mandibular condylar fractures. We hypothesise that the reproducibility of the MAT will be sufficient (ICC ≥ 0.7) and that the construct validity would be at least moderately correlated (≥ 0.60).

| Subjects
Patients treated for a mandibular condylar fracture at the Department All patients had to be stable on the interim period of measurements, and the test conditions and test instructions were kept similar for all subjects.
Study power was calculated based on sample size calculation for reliability studies by Walter et al 16 An ICC of at least 0.7 (ρ 0 = 0.7 and ρ 1 = 0.9) showed a sample size of 18.4 patients. 16 A second power analysis based on the MAT reproducibility and validity in a comparable study in children with cerebral palsy showed a sample size of 25-30 patients. 10 Therefore, we choose to include 26 patients.
The study protocol was approved by the Ethics Committee of UMC Utrecht (NL59658.041.16). All subjects received a written explanation of the study, and informed consent was obtained from each subject before the start of the tests. A repetition of 15 times was chosen for this trauma group because the authors assume that this group has no problems with tongue mobility or dentition, in contrast to oncological patients, for whom this test was originally designed. 2 A ceiling in outcome will be received when chewing more strokes. 2 This procedure was repeated with a second wax tablet, with an appropriate time interval of 15 minutes minimum. Thereafter, the chewed tablets were removed, flattened between foil to a thickness of 2.0 mm and photographed on both sides using a high-quality scanner (Epson V750, Long Beach). The retrieved images were analysed and processed using Adobe Photoshop, CS3 extended (Adobe, San Jose), a commercially available program for image analysis. The MAI was obtained by measuring the intensity distributions of the red and blue colouring on the combined image on both sides of the flattened wax.

| Mandibular function impairment questionnaire
The Mandibular Function Impairment Questionnaire (MFIQ) is designed to assess the masticatory ability, or, in other words, the patient's perception of mandibular function impairment. The MFIQ has been proven reliable in patients with painfully restricted temporomandibular joints by a moderate to good test-retest reliability (Spearman correlation of 0.69 to 0.96). 13 The minimal amount of change to be detected is 14 units on a scale of 0 to 68, 13

| Reproducibility of the test-retest
The test-retest reproducibility is divided into reliability and agreement parameters. Reliability (the proportion of the total variance in the measurements that is due to "true" differences among patients) of the MAT was calculated with an intra-class correlation coefficient (ICC) with corresponding 95% confident intervals, based on a mean rating (k = 2), absolute agreement, two-way random-effects model and single measures (ICC 2.1). This is calculated as: , with MS R = mean square for rows; MS E = mean square for error; MS C = mean square for columns; and k = number of raters/measurements. Cut-off points for the ICC were chosen as <0.5 = poor, 0.5 to 0.75 = moderate, 0.75 to 0.90 = good and >0.90 = excellent reliability. 19 A threshold of 0.75 for the ICC was taken as an acceptable level of test-retest reliability. 19 The measurement error consists of the systematic and random error of a patient's score, which is not attributed to true changes in the construct of disability. Agreement was assessed by calculating the standard error of measurements (SEM) of the MAT. The SEM is a measure of how much measured test scores are spread around a "true" score. This is calculated from the ICC as SEM agreement = SD * √(1 -ICC), with SD meaning "standard deviation of the differences of the MAT." The SEM agreement was additionally used to calculate the smallest de- These variables were integrated into a scatter plot where the difference between test and retest values was put on the y-axis, and the average of the test and retest values was put on the x-axis. 19,20

| RE SULTS
Twenty-six patients were included in this study and are depicted in Table 1. Eighteen patients (69%) were male, and the mean age was 41 years with a range of 18 to 69. Twenty-two subjects underwent the mixing ability test and retest six weeks after treatment for

| Reproducibility of the test-retest
The Bland-Altman plot with corresponding LoA can be found in Table 2 and Figure 1. Systemic bias was visually assessed by Bland-Altman.
It showed a consistent variability across the graph.

| Construct validity
At the first measurement moment, there was a weak-positive corre- for the MAT test and retest, respectively (see also Table 3).  The authors expected the retest to generate a better outcome than the first test for two reasons. The first reason was that when a patient had just finished treatment of the condylar fracture, and the first thing the patient was allowed to chew on was the MAT tablet, the patient's chewing performance might be limited by fear (of f.i.

| D ISCUSS I ON
pain), also known as kinesiophobia. Our second idea involved the presence of a learning curve in the method of chewing the wax tablet. As seen in Table 1, this difference in outcome is limited to a minimum. This minimal difference could be explained by possible fatigue of the masticatory muscles when taking the retest, despite the set time between taking the retest.

| Strengths and limitations
All results were written down according to the COnsensus-based Standards for the selection of health Measurement INstruments (COSMIN) to ensure methodological quality. 25 The data of this study were collected with a prospective design. All data were collected by the same author (FMW). The MATs were evaluated by the same observer (CMS).
In general, this study was conducted with a fair sample size, with two different follow-up periods. The participants came from three different hospitals, resulting in a heterogeneous sample.
One limitation of this study is that measurements on inter-rater reliability are missing. An additional measurement was judged to be too time-consuming for participants. In a usual care or research setting, most evaluative measurements would be performed by the same person.
As the subjects in this study were patients with condylar trauma, we have to be careful to generalise these results to general oral and maxillofacial trauma patients, and, in particular, all mandibular traumas.

| Future research
As the treatment modality of patients with condylar trauma is still subject to debate, investigations like the MAT and MFIQ could help determine whether open surgery is preferable to conservative treatment, or vice versa. Based on the results of this study, we expect the outcome of the MAT to be of excellent reliability, and therefore, reliable conclusions can be made.

| Conclusion
The test-retest reliability of the MAT is excellent in condylar trauma patients and may be used in follow-up in prospective studies. As the validity of the MAT and the MFIQ is not convincing, the MFIQ could be an addition to patient feedback about the rehabilitation process of their mandibular functioning.

CO N FLI C T S O F I NTE R E S T
The authors report no financial conflicts of interest.

E TH I C A L A PPROVA L
No ethical approval was needed.

CO N S E NT CO -AUTH O R S
All authors have viewed and agreed to this submission.