Quality of life measured by OHIP-14 and GOHAI in elderly people from Bialystok, north-east Poland

Abstract

Background: The Oral Health Impact Profile-14 (OHIP-14) and the Geriatric/General Oral Health Assessment Index (GOHAI) have never been compared for a group of the same subjects in the Polish population. The aim of the study was to compare the OHIP-14 and GOHAI measures.Methods: 178 independently living people over the age of 55 were included in the study. The GOHAI and OHIP-14 measures were used. Other variables included age, gender, self-ratings of oral general health, education, number of missing teeth, chewing problems and dry mouth.Results: The mean age of respondents was 70.8 years. The internal reliability (Cronbach's alpha) showed a high internal consistency for both measures. Spearman's rank correlation coefficient between the GOHAI and OHIP-14 scores was 0.81. Using the additive method of creating scores, 1.1% of respondents had the GOHAI score of zero, indicating no impact from oral conditions, while 13.5% of them had an OHIP-14 score of zero. Dental status, partial dentures, chewing problems, dry mouth and self-rated oral health were significantly associated with the results of the GOHAI and the OHIP-14 (Kruskal-Wallis test, Mann-Whitney U test). The numbers of preserved and missing teeth significantly correlated with the GOHAI and the OHIP-14, while DMF was significantly associated with the GOHAI only. 6 individuals with discrepant results were revealed. After the exclusion of the abovementioned patients, the internal reliability (Cronbach's alpha) still showed a high internal consistency, and the correlation between the GOHAI and OHIP-14 scores using Spearman's rank-correlation coefficient increased to 0.87. This phenomenon was identified as a " fatigue effect".Conclusions: There was a strong correlation between the GOHAI and the OHIP-14. Both instruments demonstrated good discriminant properties and helped capture the respondents' oral health problems. The questionnaires should be randomly distributed to avoid the influence of " fatigue effect" on the results of a comparison of different measures.