Reliability of a German Questionnaire about General Practitioners? Handling of Female Urinary Incontinence

Background: Recently our group performed a cross-sectional study in which 930 general practitioners (GP) in Germany and Denmark received a newly developed questionnaire concerning lower urinary tract symptoms. We developed the questionnaire on the basis of cognitive interviews with GPs and tested the reliability of the German version of the questionnaire in a test-retest process. Methods: 16 GPs took part in the test-retest process and completed the questionnaire twice with a time period of about four weeks between each attempt. The questionnaire consists of 28 questions. The given-answer categories and description fields sum up to a total of 60 items (answers). We assessed the reliability of answers by calculating and interpreting the absolute agreement and Cohen´s Kappa with a 95%-confidence interval for data that were nominal scaled and Pearson´s correlation coefficient for data that were interval scaled, respectively. Results: A total of 27 questions with 59 answer items were included in the analysis. Of them, 13 questions dealt with “management of UI”, six questions addressed the “communication about UI”, four questions asked for the “structure of the practice”, and five questions assessed personal data of the GP. Each more than 50% of the items in the subject areas “management of UI” (53.1%), “communication about UI” (66.6%), “structure of the practice” (57%), and “personal data” (100%) were rated as having high reliability. In summary, 35 of the analyzed items were rated a having a high reliability and 22 items were rated as having a moderate reliability. Conclusion: Given the low number of study participants our results have to be interpreted with caution, but is seems that the developed questionnaire is – for the vast majority of items – a reliable tool for assessing the communication about and the management of female urinary incontinence during a general practitioners' consultation hour. Before application in future studies we recommend revising one item of the questionnaire in order to gain a higher reliability of this item.