Translation and cross-cultural adaptation of the manchester orofacial pain disability scale.

OBJECTIVES: The purpose of the present study was to translate and perform a cross-cultural adaptation of Manchester Orofacial Pain Disability Scale to the Portuguese language.
MATERIAL AND METHODS: A synthesis of two independent translations done by bilingual translators whose mother tongue was the Portuguese language began the process of translation. From the synthesis of the translated version and totally blind to the original version, two different non-native English language teachers without dental knowledge translated the questionnaire back to English. The pre-final version was done by an Expert committee: the researchers, two other non-native English language teachers and one native English language speaker. The new questionnaire was then piloted among 8 patients from the target setting that were interviewed to probe it on their perceived meaning of each question. The Manchester Orofacial Pain Disability Scale (MOPDS) thus translated was called Brasil-MOPDS and was validated in 50 patients with Orofacial pain from TMJ and Occlusion clinic ambulatory of São Paulo University School of Dentistry. The Brasil-MOPDS was administered twice by an interviewer (15 - 20 day interval) and once by a second independent interviewer. The Brazilian version of the short form oral health impact profile (OHIP-14) questionnaire and the visual analogue pain scale (VAS) were applied on the same day.
RESULTS: Internal consistency (Cronbach's α = 0.9), inter-observer (ICC = 0.92) and intra-observer (ICC = 0.98) correlations presented high scores. Validity of Brasil-MOPDS compared to OHIP-14 (r = 0.85) and VAS (r = 0.75) shown high correlations.
CONCLUSIONS: Brasil-MOPDS was successfully translated and adapted to be applied to Brazilian patients, with satisfactory internal and external reliability.

INTRODUCTION

Pain affects all aspects of a person's life including sleep, work, leisure and relationships. It impacts on the lives of sufferers by causing limitations in daily life and this in turn imposes a huge burden on society, due to the high direct costs of treatment and indirect costs from lost productivity [1-4].

Pain is a personal, subjective experience. Approaches to the measurement of pain include verbal and numeric self-rating scales, behavioural observation scales, and physiologic responses. Because of complex nature of the experience of pain and its subjective characteristic, patients' self-reports provide the most valid measure [5].

Orofacial pain has a variety of consequences on physical and psychosocial functions [6]. Self-reported measures of oral health quality of life relate altered functions and symptoms to social and psychological well-being. However, different oral conditions may affect functioning and physical and emotional problems in different ways and therefore disease specific instruments may be invaluable in investigating the impact of these conditions on individuals.

Brazilian studies about disabilities caused by orofacial pain use RDC/TMD (Research Diagnostic Criteria for Temporomandibular Disorders) Axis II, OIDP (Oral Impact Daily Performance), OHIP (Oral Health Impact Profile), GHQ-12 (General Health Questionnaire) and Brazilian version of McGill Pain Questionnaire. Although these tools were not specifically developed to assess orofacial pain related disability, they all concluded that oro-facial pain imposes a huge burden on the daily life of sufferers [7-9].

The Manchester Orofacial Pain Disability Scale (MOPDS) construction and validation has been previously demonstrated. This tool has been shown to be robust in measuring orofacial pain related disability and had good construct validity [10].

With the increasing number of international research projects, the need to adapt measures of health status to use the language of origin has grown rapidly. Most questionnaires were developed in English-speaking countries, but even within those countries, researchers should consider translating these for immigrants, especially when their exclusion may lead to systematic bias in studies of health care related quality of life. Thus Beaton et al. [11] recommend cross-cultural adaptation in addition to simple translation for questionnaires of health status that are going to be used in a different language and culture from which they were originally developed.

This paper therefore aims to adapt the Manchester Orofacial Pain Disability Scale to a Portuguese language version through a formal translation/back-translation process and summarize available data about its psychometric properties.

Specific objectives were to determine the internal consistency and reliability of the translated scale and to validate its use for measuring orofacial pain specific disability in Brazilian patients.

MATERIAL AND METHODS

The guidelines for translation and cross-cultural adaptation process (Beaton et al. [11] and Wild et al. [12] were used following the stages above:

Stage I: Initial Translation

Two independent translations from English to Portuguese were made by two bilingual translators whose mother tongue is the Portuguese language in order to accurately reflect the nuances of the language. Both translators had different profiles or backgrounds - one with academic vinculum (with theme knowledge) and the second one a clinical dentist.

Stage II: Synthesis of the Translations

A synthesis of these translations was first conducted (producing one common translation) by the authors of the research and two non-native English language teachers.

Stage III: Back Translation

Working from the synthesis of the translated versions and totally blind to the original version, two different non-native English language teachers without dental knowledge translated the questionnaire back to the English language.

Stage IV: Expert Committee

An Expert committee reviewed the pre-final version: the researchers, two other non-native English language teachers and one native English speaking professor.

Stage V: Test of the Prefinal Version

This field test of the new questionnaire was used in 8 patients from the target setting. Each subject completed the questionnaire and was interviewed to probe what he or she believed to be the meaning of each questionnaire item and its response. In this stage these patients suggested to exclude the question "I am irritable, angry and easily frustrated", justifying that its items are repeated in other questions. The second question suggested to exclude was "I have lost earnings" because its semantic equivalence with the item "I have had to take time off work".

Stage VI: Submission of Documentation to the Coordinating Committee for Appraisal of the Adaptation Process

The final stage in the adaptation process was the submission of all the reports to a committee formed by the authors of the research and another Portuguese speaking professor. As related in the item above this committee decided to remove the questions "I am irritable, angry and easily frustrated "and "I have lost earnings".

The Brazil-MOPDS was applied in fifty patients with symptoms of orofacial pain who consulted at the TMJ and Occlusion clinic of the Prosthetic Department of São Paulo University School of Dentistry, Brazil. All patients were recruited before any type of treatment, clinical consultation or other information about orofacial pain was collected and all of approached patients agreed to participate. Initially they were 75 but only 50 concluded the steps of the designed study (Figure 1). VAS and OHIP-14 were applied to evaluate validity of Brazil-MOPDS. The Brasil-MOPDS was administered twice by an interviewer (15 - 20 day interval) and once by a second independent interviewer. The Brazilian version of the short form oral health impact profile (OHIP-14) questionnaire and the visual analogue scale (VAS) were applied on the same day.

Figure 1

Study design.

OHIP-14 score were based on its seven dimensions (physiological discomfort, pain, physiologic inability, physical inability, function limitation, disability and social inability) and scored from 0 - 4 based on a Likert scale: 0 - never, 1 - rarely, 2 - sometimes, 3 - frequently and 4 - always. This value was then multiplied by the weight of each question to give a score ranging from 0 (less oral health impact on daily profile) to 28 points (more impact) [13].

All participants have read and signed informed consent form. The use of human subjects in this study has been reviewed and approved by University of São Paulo Dentistry School Ethics Committee. This study was conduced from August 1, 2008 to March 1, 2009).

Statistical analysis

Data were tabulated and analyzed in STATA 10 [14] and for all tests it were used Confidence Interval of 95%. Internal consistency was examined by Cronbach α. In order to find the correlation between each question and the overall outcome of the test, the Spearman Correlation Coefficient (SCC) was performed, using data from the first interview (main test). Psychometric Properties: reliability was estimated by assessing the internal consistency (indicated by Cronbach α) and reproducibility (test-retest). The statistical value of Cronbach α was also calculated for each excluded question from the Brazil-MOPDS as the objective of statistical investigation was to numerically represent the interests of uniformity or trend in the responses within each item of the questionnaire.

Reliability was checked by the application of Brazil-MOPDS on two occasions by the same rater, called internal reliability. External reliability was the comparison of the answers obtained by the first and the second evaluator. For these analysis interclass correlation coefficient (ICC) and Bland-Altman test [15] were used.

RESULTS

The samples were distributed by gender and age. Most of the participants were women (86%) with the average age of 40.7 (SD 14.03) years old (Table 1).

Table 1

Sample distribution by gender and age

	N	%
Gender
Female	43	86
Male	7	14
Total	50	100
Age (years)
18 – 29	12	24
30 – 49	23	46
40 – 69	15	30
Mean (SD)	40.7 (14.03)

N = number of participants.

SD = standard deviation.

The answers of the questionnaire obtained by the first rate (first and second interview) and the second rate are described in Table 2. The correlation between each question, the whole result and Cronbach α is shown in Table 3.

Table 2

Answers distribution by the three interviews

Question/Answer	First interviewer/ 1st interview		First interviewer/ 2nd interview		Second interviewer
Because of pain in my face, jaws or mouths:	N	%	N	%	N	%
I cannot open my mouth as wide as I could
None of the time	17	34	17	34	17	34
On some days	18	36	13	26	16	32
On most/everyday(s)	15	30	20	40	17	34
I find it difficult to talk for long periods of time
None of the time	26	52	23	46	24	48
On some days	11	22	11	22	12	24
On most/everyday(s)	13	26	16	32	14	28
I find it difficulty to smile or laugh
None of the time	26	52	22	44	27	54
On some days	16	32	16	32	12	24
On most/everyday(s)	8	16	12	24	11	22
I cannot touch my face
None of the time	30	60	28	56	30	60
On some days	15	30	14	28	11	22
On most/everyday(s)	5	10	8	16	9	18
I cannot find a comfortable position in which to sleep
None of the time	25	50	25	50	23	46
On some days	13	26	14	28	14	28
On most/everyday(s)	12	24	11	22	13	26
I wake up at night in pain
None of the time	25	50	25	50	26	52
On some days	18	36	18	36	16	32
On most/everyday(s)	7	14	7	14	8	16
I have difficulty falling asleep
None of the time	23	46	26	52	27	54
On some days	18	36	16	32	11	22
On most/everyday(s)	9	18	8	16	12	24
I cannot eat hard foods like apples or toast
None of the time	12	34	11	22	14	28
On some days	17	24	18	36	10	32
On most/everyday(s)	21	42	21	42	20	40
I take longer to finish my meals
None of the time	32	64	32	64	33	66
On some days	7	14	7	14	3	6
On most/everyday(s)	11	22	11	22	14	28
I am unable to eat out in restaurants
None of the time	39	78	41	82	43	86
On some days	6	12	4	8	4	8
On most/everyday(s)	5	10	5	10	3	6
I no longer enjoy my food
None of the time	40	80	41	82	40	80
On some days	5	10	3	6	5	10
On most/everyday(s)	5	10	6	12	5	10
I find it sore to kiss
None of the time	30	60	33	66	34	68
On some days	17	34	13	26	11	22
On most/everyday(s)	3	6	4	8	5	10
I have had to take time off work
None of the time	36	72	36	72	40	80
On some days	10	20	10	20	6	12
On most/everyday(s)	4	8	4	8	4	8
People find me difficult to live with
None of the time	33	66	31	62	33	66
On some days	14	28	15	30	12	24
On most/everyday(s)	3	6	4	8	5	10
I have had to take time off work
None of the time	23	46	23	46	22	44
On some days	21	14	20	40	18	36
On most/everyday(s)	6	12	7	14	10	20
I have problems performing normal household tasks
None of the time	33	66	32	64	35	70
On some days	15	30	15	20	12	24
On most/everyday(s)	2	4	3	6	3	6
I would rather be by myself
None of the time	28	56	27	54	25	50
On some days	11	22	11	22	12	24
On most/everyday(s)	11	22	12	24	13	26
I have cancelled social activities and holidays
None of the time	35	70	34	68	37	74
On some days	12	24	12	24	9	18
On most/everyday(s)	3	6	4	8	4	8
I feel weary/tired
None of the time	12	24	13	26	10	20
On some days	22	44	24	48	25	50
On most/everyday(s)	16	32	13	26	15	30
I cannot stop crying
None of the time	36	72	34	68	33	66
On some days	8	16	11	22	10	20
On most/everyday(s)	6	12	5	10	7	14
I am worried that I may have a serious illness
None of the time	31	62	32	64	30	60
On some days	12	24	8	16	6	12
On most/everyday(s)	7	14	10	20	14	28
I feel embarrassed and self conscious
None of the time	38	76	38	76	40	80
On some days	10	20	11	22	8	16
On most/everyday(s)	2	4	1	2	2	4
I feel depressed
None of the time	30	60	32	64	32	64
On some days	14	28	12	24	12	24
On most/everyday(s)	6	12	6	12	6	12
I feel I no longer take any pleasure in life
None of the time	37	74	37	74	37	74
On some days	13	26	11	22	9	18
On most/everyday(s)	0	0	2	4	4	8

N = number of participants.

Table 3

Correlation between questions and whole Brazil - MOPDS and Cronbach α if the question were excluded

Because of pain in my face, jaws or mouths	Correlation between each questionand whole resulta	Cronbach α with questionexclusion
I cannot open my mouth as wide as I could	0.482	0.917
I cannot touch my face	0.293	0.92
I have difficulty falling asleep	0.659	0.913
I wake up at night in pain	0.746	0.911
I cannot find a comfortable position in which to sleep	0.568	0.915
I cannot eat hard foods like apples or toast	0.526	0.916
I take longer to finish my meals	0.578	0.915
I no longer enjoy my food	0.68	0.913
I find it sore to kiss	0.639	0.913
I find it difficulty to smile or laugh	0.551	0.915
People find me difficult to live with	0.535	0.915
I have had to take time off work	0.446	0.917
I have found it difficult to concentrate	0.559	0.915
I have problems performing normal household tasks	0.577	0.914
I would rather be by myself	0.739	0.911
I find it difficult to talk for long periods of time	0.635	0.914
I have cancelled social activities and holidays	0.615	0.914
I am unable to eat out in restaurants	0.631	0.913
I feel weary/tired	0.654	0.913
I cannot stop crying	0.563	0.915
I am worried that I may have a serious illness	0.374	0.919
I feel embarrassed and self conscious	0.727	0.912
I feel depressed	0.787	0.91
I feel I no longer take any pleasure in life	0.744	0.914

aPearson correlation, question-whole result.

Reproducibility (test-retest): interobserver correlation data obtained in questionnaire administration indicated an excellent agreement with ICC = 0.924 (CI 0.46 - 0.98) (Figure 2). In order to evaluate graphically the agreement or discrepancy between the sums of numerical responses given by patients questioned on the same day by two different interviewers, the results of applications of Brazil - MOPDS were plotted in a Bland-Altman Plot (Figure 3).

Figure 2

Bland-Altman Plot with inter observer reproducibility.

Figure 3

Bland-Altman Plot with intra observer reproducibility.

This enabled us to recognize the magnitude of variation between the responses as well as the existence of systematic bias between the two interviewers. The value of the arithmetic mean was equal to - 0.48 (CI 1.08 - 0.12), showing a strong correlation [15].

Data obtained in the intra-observer correlation in the administration of the questionnaire, also showed an excellent agreement with ICC = 0.982 (CI 0.967 - 0.997) (P < 0.001) (Figure 3.

Criterion validity was established by comparing the data of the scale with the results of the OHIP-14, VAS and the three Brazil-MOPDS interviews (Table 4).

Table 4

OHIP-14, VAS and Brazil-MOPDS answers

ParticipantsN=50	OHIP-14	VAS	Brazil-MOPDSFirst interviewer(1st interview)	Brazil-MOPDSFirst interviewer(2nd interview)	Brazil-MOPDSSecond interviewer
Mean	10.28	6.67	13.66	14.14	14.12
SD	6.16	2.03	9.93	10.26	10.17

SD = standard deviation.

The correlation between the scores of Brazil-MOPDS and OHIP-14 was high, r = 0.857 (CI 0.765 - 0.915) (P < 0.001) and the correlation of scores of Brazil - MOPDS with VAS was also strong - r = 0.758 (CI 0.615 - 0.852) (P < 0.001).

DISCUSSION

We have successfully adapted the MOPDS into the Portuguese language. Not only were we able to translate it, but also validate and culturally adapt it. Previous instruments have not been specifically designed to measure the impact associated with orofacial pain.

Orofacial pain has not only biological but also psychological and social effects on patients [16]. Psychosocial effects are difficult to objectively measure and certainly deserve attention in the evaluation of therapeutic measures for chronic orofacial pain. In fact they may be the only measures to assess patient improvement particularly where there is no underlying pathology to explain reported symptoms [17-21].

Further, indicators of oral health related quality of life are often used to complement clinical data in cross-sectional and longitudinal studies [22]. Orofacial pain is a commonly observed symptom of dental disease in Brazil. Borges et al. [23] estimated the prevalence of dental and gingival pain and associated factors among 16,126 young Brazilians (15 - 19 years) who participated in the epidemiological survey of Oral Health (2002 - 2003) in Brazil for six months. Through simple and multiple Poisson regression there was a high prevalence of dental and gingival pain: 35.6% (CI: 34.8 - 36.4). The increased prevalence of pain was associated with female public school students, low income and education gap. Adolescents with high levels of caries and dental calculus also reported a higher prevalence of dental pain.

The successful translation and cross-cultural adaptation of the MOPDS [10], which had its development based on people who had just this specific condition will therefore be extremely valuable to measure the specific effects of orofacial pain in Brazilian populations.

Other questionnaires like GOHAI (Geriatric Oral Health Assessment Index, RDC/TMD, OHIP, McGill pain questionnaire and OIDP have been translated and validated for use in Brazilian researches [24-29]. However, these instruments have not been specifically designed to measure the impact associated with orofacial pain.

Although some English terms do not have equivalent descriptors in Portuguese, trans-cultural adjustments proved to be a valid and efficient alternative to overcome this difficulty, which can be confirmed in the application of the final version of the questionnaire. Adaptation to local culture is essential for a correct evaluation of the process of pain [30].

Therefore the Brazil-MOPDS was obtained from translation and cultural adaptation as performed according to current and internationally accepted guidelines [11,12]. It showed excellent reproducibility, validity and practicality application with high Cronbach's α scores, and good correlation co-efficient. The Bland-Altmann test, associated with the ICC was used for providing more information than the use of a test alone. One of the main advantages of this method, in which differences between the scores of the first and second evaluation are plotted on the mean values ​​is to allow the recognition of both the amplitude of variation as to the existence of systematic bias between the two interviews It is necessary at least 50 individuals for the use of this method.

However, our study was conducted on patients referred into a tertiary setting who may represent the most severe and intractable cases of orofacial pain. Further research is needed in establishing the validity of the Brazil-MOPDS in population settings and also its responsiveness as a measurement of treatment outcomes in clinical trials of chronic orofacial pain. We envisage that it will be an important tool for such patients where improvement cannot be assessed in terms of structural improvement (elimination of infection or removal of diseased tissue) as there is no underlying cause for reported symptoms. Rather, measurements of disability before and after treatment may provide important information on treatment outcomes and therefore influence further management of the patient [8].

CONCLUSIONS

The data showed that the process of translation and cross-cultural adaptation of Manchester Orofacial Pain disability Scale was successful and that Brazil Orofacial Pain disability Scale seems to be a valid and reliable instrument for describing pain-related impact among patients with symptoms of orofacial pain.