Assessing the psychometric characteristics of the Macedonian version of the Oral Health Impact Profile questionnaire (OHIP-MAC49).

AIM

The aim of this study was to adapt the Oral Health Impact Profile-49 (OHIP-49) for use by the Macedonian-speaking population and to assess its psychometric properties.

METHODS

After piloting a back-translated version of the OHIP-49 questionnaire in Macedonian, the resulting Macedonian version of the Oral Health Impact Profile questionnaire (OHIP-MAC49) was administered to 247 patients who had been recruited in four groups: Group 1 was composed of 163 randomly chosen blood donors representing the general population; Group 2 was a convenience sample of 20 patients who attended a clinic for relief of dental pain; Group 3 was a sample of 29 prosthodontic patients; and Group 4 was composed of 35 students. In order to ensure that all questions were answered, data were gathered from participants in the four groups by trained and experienced interviewers. The internal reliability of the OHIP-MAC49 scale and its constituent seven subscales was calculated for Groups 1, 3, and 4 by using Cronbach's alpha coefficient and average inter-item correlations. The test-retest stability of the instrument was estimated by calculating the intra-class correlation coefficients and the limits of agreement of the scores obtained from the participants in Groups 3 and 4 in a repeat interview three to four weeks after the first interview. Convergent validity was evaluated by comparing OHIP-MAC49 scores-both total (0-4) and subscores (2-4) with self-reported health for Group 3 patients-using the Spearman coefficient of correlation. For the purpose of evaluating group validity, the OHIPMAC49 scores of Group 3 patients were compared by using Spearman's correlation coefficient. In addition, comparisons were made between patients with and without burning-mouth symptoms, temporomandibular pain, joint clicking, and oral habits (point-biserial correlation). The OHIP-MAC49 scores of subjects with and without dentures in Group 1 were compared by using point-biserial correlation. The responsiveness of the instrument was assessed by calculating the difference in OHIP-MAC49 means (total subscores) in Group 2 patients prior to and after treatment for relief of pain.

RESULTS

The internal consistency of each subscale and the whole scale estimated was excellent. Cronbach's alpha coefficients for whole scale ranged from 0.92 to 0.95. The intra-class correlation coefficients ranged from 0.83 to 0.99, suggesting that the instrument had satisfactory reliability in terms of time stability. The questionnaire had acceptable responsiveness, confirmed by a significant differences (P<0.01) between the mean OHIP-MAC49 score at baseline and follow-up, both for the total sum (0-4) and the sum scores (2-4). Convergent validity, evaluated by comparing OHIP-MAC49 scores with self-reported oral health in Group 1 and Group 3 patients, was confirmed, because all correlation coefficients were significant (P<0.01). The results from testing the anticipated differences on the basis of seven self-reported indicators of oral health in Group 3 patients, as well as the differences in OHIP-MAC49 scores between subjects in general population who wore and did not wear dentures, strongly suggest that the instrument has satisfactory group validity.

CONCLUSION

The Macedonian version of the OHIP-49 demonstrated satisfactory validity, excellent reliability, and sufficient responsiveness and therefore can be used for assessing the impact of oral health on different aspects of quality of life in Macedonia.