Participants were consecutively recruited from social media and through posters at the buildings of Universidade Nove de Julho (São Paulo, Brazil). The examinations were conducted at the Musculoskeletal Research Center, Universidade Nove de Julho.

Brazilian men and women between 18 and 70 years old15; with 15 points or more on the short form Fonseca Anamnestic Index16; with Brazilian cultural identity (born in Brazil; with at least one parent born in Brazil; Brazilian Portuguese as the first language; and self-identified as a member of Brazilian culture)8; and with the cognitive ability to answer questionnaires as assessed with the Mini Mental State Examination (cut-off score > 23)17 and Cloze Test (≥60% correct answers).18

Participants who used: analgesics during the last 24 h prior to the examination; muscle relaxants or steroids in the last seven days prior to the examination; non-steroidal anti-inflammatory drugs (NSAIDs) in the last three days prior to the examination; anti-depressants or anxiolytics unless the dose was stable for at least 60 days. Additional exclusion criteria were illiteracy; pregnancy; fibromyalgia; neurological disease; TMD treatment history in the last six months; facial trauma in the last two months; ongoing dental treatment; wearing dentures; and TMJ surgery in the last six months.15

This study was conducted according to the Guidelines for Establishing Cultural Equivalency of Instruments, the Consensus-based Standards for the Selection of Health Measurement Instruments (COSMIN), and the Guidelines for Reporting Reliability and Agreement Studies (GRRAS).8,13,14 The validation process included the phases of translation, cross-cultural adaptation, and testing of the measurement properties of the Brazilian Portuguese version of the DC/TMD Axis I.

The present study is part of a larger study that received approval from the institutional review board of Universidade Nove de Julho (certificate number: 29,673,920.9.0000.5511) and is in compliance with Helsinki Declaration. All study participants signed an informed consent form.

The translation and cross-cultural adaptation of Axis I of the DC/TMD into Brazilian Portuguese followed the Guidelines for Establishing Cultural Equivalency of Instruments of the INfORM, which included: forward-translation conducted independently by two translators (who were native speakers of Brazilian Portuguese and specialists in orofacial pain, aware of the content of the instrument); synthesis and resolution of discrepancies; back-translation conducted by one English native speaker (unaware of the content and lay to the field); independent review; revision related to disparities; consolidation of all translation; review by an expert committee (including TMD and orofacial pain specialists, doctors of dental science, physical therapists, psychologist, and language translator) and cultural validity revision; and preparation of the pre-final instrument.8 Logs were used for documentation process, according to INfORM recommendations, and the team leader completed the records at the end of each phase.8,21

After an independent review of the translation process by the INfORM, the pre-final version was posted on the INfORM website, followed by a formal assessment in a Brazilian sample.

Two physical therapists with more than eight years of experience each were trained and calibrated in DC/TMD by self-instruction.22-24 They downloaded the instructional video and read the DC/TMD documentation to learn and memorize all mandatory commands.25 In addition, a calibration examination with feedback by a previously calibrated researcher (DA) was performed.

All potentially eligible participants answered a screening questionnaire to assess the inclusion and exclusion criteria, including the cognition tests and the short form Fonseca Anamnestic Index for TMD assessment.16-18 After this first stage, the included individuals reported their pain intensity during last week for the assessment of orofacial pain, answered the Axis I's instruments of the DC/TMD on paper – Demographics, TMD Pain Screener and Symptom Questionnaire – and subsequently underwent clinical examination by examiner 01. After 10 mins of rest, an independent and identical clinical examination was performed by examiner 02, who was blinded to the results of the first examination (Fig. 1).23 This 10-minute interval was defined based on the Guidelines for DC/TMD Training and Calibration.23

Fig. 1.

Formal assessment. NPRS, numeric pain rating scale.

(0.18MB).

In the clinical examination, the mandibular range of motion was measured using a digital caliper (Starrett®, São Paulo, BR), and palpation calibration – as recommended for the clinical Axis I examination protocol – was made with the assistance of a digital algometer (DD-200 model, Instrutherm®, São Paulo, Brazil).24 The calibration was performed immediately before each palpation site, which ensured a more precise finger pressure (1 kg or 0.5 kg) in each specific tissue, according to the clinical form.24

For the assessment of intra-observer agreement, an interval from six to 14 days between the assessment days was considered long enough to prevent recall and short enough to guarantee the participant's condition stability.13 The second assessment day was conducted in the same location and conditions as the first day, and only participants with a maximum of 2 points of difference (plus or minus) in average pain intensity during the last week from the first evaluation were included.20 The participants answered the same questionnaires – pain intensity, TMD Pain Screener, and Symptom Questionnaire – and underwent an independent clinical examination performed by examiner 01.

The descriptive data are presented as median and interquartile interval for continuous data, and frequencies and percentages for categorical variables. The Kolmogorov-Smirnov analysis was used to test data normality.

According to COSMIN, criterion validity is defined as the degree to which the tested instrument reflects a ‘gold standard’. When analyzing a short form, the long form of the questionnaire can be considered as the gold standard. Thus, the criterion validity of the short form TMD Pain Screener was verified with accuracy analysis using the long version as the reference, through the analysis of the ROC curve, sensitivity, and specificity. For this purpose, the criterion validity of both versions of the TMD Pain Screener was also previously analyzed using the DC/TMD as the gold standard, through accuracy analysis. The area under the curve (AUC) was considered suitable and sufficient if ≥0.70.13 Acceptable values were defined as ≥0.70 for sensitivity and ≥0.95 for specificity.5

The inter and intrarater reliability of the DC/TMD diagnoses were analyzed using kappa statistics (κ) and interpreted as: <0.00 poor, 0.00 – 0.20 slight, 0.21 – 0.40 fair, 0.41 – 0.60 moderate, 0.61 – 0.80 substantial, and 0.81 – 1.00 almost perfect reliability.26 The 95% confidence interval (CI) was also calculated. Only diagnoses prevalent in at least 5% of the total sample (≥0.05) were considered adequate for the kappa analysis.10,27 Overall agreement (OA) was also calculated and expressed as a percentage, and the cut-off of 75% was considered a good agreement.9,28

The inter and intrarater reliability of the measurements of mandibular range of motion and the TMD Pain Screener scores were analyzed using intraclass correlation coefficient (ICC2,1) and their respective 95% CI. ICC values were classified as excellent (>0.75), moderate (0.40–0.75), or poor (<0.4) reliability.29

SAS for Windows v.9.4 and SPSS for Windows v.21 software were used for data analysis. Missing data were excluded from the analysis.

The Translation Logs of each stage were created in separate files as follows: TMD Pain Screener, Symptom Questionnaire, Demographics, Examination-Related Pain Interview, and Examiner Commands. All stages were followed as recommended, and the necessary number of iterations was made to achieve consensus regarding language equivalency.8 The expert panel evaluated the translation for semantic, idiomatic, experiential, and conceptual equivalency; and suggested some modifications when necessary.

The Brazilian Portuguese version is available on the website https://ubwp.buffalo.edu/rdc-tmdinternational/tmd-assessmentdiagnosis/dc-tmd/. The dubbed version of the instructional video following the translated standardized protocol was created and will be published online. The documentation of Brazilian Portuguese translation and adaptation is detailed in the Translation Logs available with the team leader and the INfORM.

Three hundred and thirty-eight participants completed the screening questionnaire. One hundred and twenty-two were excluded for not meeting inclusion criteria or for having one or more exclusion criteria, and 99 declined to participate (Fig. 2). The specific reasons for excluding participants are presented in the Appendices – Table 2.

Fig. 2.

Flow diagram. NPRS, numeric pain rating scale. *Specific reasons for excluding participants are presented in the Appendices – Table 2.

(0.2MB).

The total sample comprised 117 participants with a median age of 33 years (IQR: 25.5 – 41.5), median orofacial pain intensity of 6.0 (IQR: 3.5 – 7.5), 76.1% women, 0.9% with basic school, 27.4% with a high school degree, and 65% graduated college. A subgroup of participants (n = 51) was recruited to the interrater reliability study and another subgroup (n = 72) to the intrarater reliability study. Of this sample, 19 participants were excluded due to lack of condition stability - which is considered inappropriate for reliability analysis overtime - resulting in 53 participants for the intrarater study (Fig. 2).13 The average duration between assessments was 9 days (SD 3).

TMD Pain Screener

One hundred and seventeen individuals with TMD answered the TMD Pain Screener, pain intensity, and Symptom Questionnaire, and underwent clinical examination. According to DC/TMD diagnostic algorithms, 104 participants were diagnosed with pain-related TMD and 13 with nonpainful TMD. The specific diagnoses are listed in the Appendices – Table 3.

Both long and short versions were compared to the DC/TMD for criterion validity assessment through accuracy analysis. The long version's total score demonstrated a sensitivity of 0.92, specificity of 0.46, and an AUC of 0.87. The short version showed a sensitivity of 0.90, specificity of 0.53, and an AUC of 0.73 with the pain-related DC/TMD diagnoses as the reference.

When considering the COSMIN methodology for criterion validity of the short version (with the long version as the ‘gold standard’), we found a sensitivity of 0.94, specificity of 0.64, and an AUC of 0.95 (Fig. 3).

Fig. 3.

ROC curve TMD-Pain Screener short form vs long form.

(0.1MB).

Fifty-three participants who had stable orofacial pain intensity answered the TMD Pain Screener on two different days, and it demonstrated excellent reliability (intrarater) for both the long (ICC = 0.91; 95% CI 0.85, 0.95) and short (ICC = 0.88; 95% CI 0.79, 0.93) versions.

Despite the large sample size and standardized methods in our study, some limitations should be pointed out. We did not perform the pre-test phase of the Brazilian Portuguese version of the DC/TMD Axis I. Therefore, future studies are required to conduct this phase and verify the instrument's interpretability, adaptation, and understanding by the professionals trained to apply and use the DC/TMD Axis I. Furthermore, the reliability assessment was not performed by an INfORM Reference Examiner or with operators calibrated by a Reference Examiner, as recommended by the INfORM. However, this may not be a factor of importance in this particular study because Vilanova et al.22 showed that self-teaching DC/TMD did not differ regarding TMD pain diagnoses. Also, there are no official Brazilian Reference Examiners recognized by the INfORM at the present time. In addition, subjects with other types of orofacial pain were not included and a dental examination was not conducted to verify the presence of odontogenic pain. Nevertheless, we consider that this did not influence the validity of our results.