David Kriz1, Jürgen Schmidt, Rüdiger Nübling. 1. Privatinstitut für Evaluation und Qualitätssicherung im Gesundheits- und Sozialwesen, Hamburg. kriz@GfQG.de
Abstract
PURPOSE: To enable the measurement of satisfaction with the service delivery of residential homes, a practicable and statistical proven screening-test was supposed to educe. In order to identify critical quality-related situations a specific cut-off should be allocated. METHODS: Based on literature research a questionnaire was developed and responded by 3581 relatives of residents of 45 nursing homes for the elderly. The questionnaire was analysed and compressed based on scale, item and factor analysis. For validation, demographic variables and separate quality-ratings were available. Finally a ROC analysis was performed to recognize specificity and sensitivity of possible cut-offs. RESULTS: The Angehörigen-Screening-Fragebogen (relatives-screening-questionnaire) ZUF-A-7 was generated consisting of seven items. The scale is reliable (Cronbach's Alpha). Like other satisfaction measurements it displays a skew distribution. Middle size up to large correlations with the quality-ratings exists but there were no relevant coherences to the demographic variables. The ROC analysis suggested a critical cut-off of 21 (range = 7-28).
PURPOSE: To enable the measurement of satisfaction with the service delivery of residential homes, a practicable and statistical proven screening-test was supposed to educe. In order to identify critical quality-related situations a specific cut-off should be allocated. METHODS: Based on literature research a questionnaire was developed and responded by 3581 relatives of residents of 45 nursing homes for the elderly. The questionnaire was analysed and compressed based on scale, item and factor analysis. For validation, demographic variables and separate quality-ratings were available. Finally a ROC analysis was performed to recognize specificity and sensitivity of possible cut-offs. RESULTS: The Angehörigen-Screening-Fragebogen (relatives-screening-questionnaire) ZUF-A-7 was generated consisting of seven items. The scale is reliable (Cronbach's Alpha). Like other satisfaction measurements it displays a skew distribution. Middle size up to large correlations with the quality-ratings exists but there were no relevant coherences to the demographic variables. The ROC analysis suggested a critical cut-off of 21 (range = 7-28).