Development of a German version of the Oral Health Assessment Tool.

Background: Limited assessment tools for estimating the oral health of nursing home residents are available in the German language.

Aims: To develop a German version of the Oral Health Assessment Tool (OHAT) and to evaluate test-retest and inter-examiner reliability in use for the assessment of nursing home residents' oral health by caregivers before and after dental training.

Methods: The original English version of the OHAT was translated into German by a forward-backward translation process. Reliability assessments were conducted in a nursing home (n = 18) by independent application of the OHAT by two trained dentists and four professional caregivers. After receiving dental training, the caregivers repeated the OHAT with the same participants. Reliability analyses of single items were performed using Cohen's kappa statistics. Intra-class correlations were compiled to assess reliability of the total scores. T tests were used to compare percentage agreement, and under- and overestimation of findings between a reference dentist and the caregivers before and after the training unit.

Results: Specificity and sensitivity of the German OHAT were 87% and 86%. Test-retest reliability of the total OHAT score as estimated by the dentists was excellent (ICC 0.910; 95% CI 0.776-0.965). Before training, mean κ values between the reference dentist and caregivers ranged between 0.155 and 0.912, whereas the inter-examiner reliability of most items was only fair. After training, overall agreement between the dentist and the caregivers improved significantly from 62.1 to 83.1% (p < 0.001), as well as satisfying inter-examiner agreement for the single items.

Conclusions: The German version of the OHAT is a reliable and valid tool for the assessment of oral health conditions of nursing home residents. When used by caregivers, instruction on the tool and practical training are mandatory to ensure reliable estimations. However, further studies with a larger sample size are encouraged to verify the outcomes of this study.