Validation of the Bonn test for knowledge in palliative care (BPW)

Background. Palliative care in nursing homes has become an increasing focal point of healthcare in Germany and the evaluation of the specific quality of palliative care in this setting is under discussion. The assessment of quantitative data has the advantage of allowing pre-post comparisons of different interventions and implementation procedures and can therefore be used for evaluation of effectiveness. However, no assessment tool in German is available yet. Recent research on the assessment of healthcare professionals' knowledge indicated that knowing about technical and psychosocial aspects of palliative care is an easily measurable index for the quality of care. In consequence this index will be used as the core of the test instrument. The specific self-efficacy related to palliative care should be additionally included. Method. An analysis of the two most common tests of palliative care knowledge by three experts revealed that these instruments were only partly useful in Germany because of differences in the healthcare systems. The experts also criticised that the tests included only a few items concerning psychosocial care and that these items often showed ceiling effects. In addition, the instruments were not suitable for use in nursing homes. For specific self-efficacy only one test is available in Dutch. The items of these three instruments were face validated by 5 experts and new items were generated where needed. The content validity was verified again by five experts with the think-aloud method. Items were validated in 23 calliative care trainers whose expert knowledge served as an external criterion. Items that scored low in this group were eliminated. A second sample of 36 assistant nurses working in nursing homes who took part in a workshop for palliative care was used to assess change sensitivity before and after training. Results. A total of 23 items for the knowledge test and 15 items for specific self-efficacy were selected. The level of difficulty was suitable because an expert group could answer the questions while for assistant nurses the items were sufficiently difficult to show a difference between groups. Items were very sensitive to change after a training course. Conclusion. The BPW is the first standardized test instrument for palliative care in the German language. In comparison to the other tests it includes items from the psychosocial domain without demonstrating a ceiling effect for these items. Content and criterion validity were good. Additional tests with larger samples and with other groups of nurses working in other settings would be needed to improve the data on reliability and to extend the validation to other settings.