Validity and reliability of the Nursing Minimum Data Set for the Netherlands (NMDSN)

The Nursing Minimum Data Set for the Netherlands (NMDSN) describes nursing care based on nursing phenomena, interventions and outcomes. The validity and reliability of its data collection has not been tested yet. PURPOSE: To report about the discriminative validity and the interrater reliability of the NMDSN. DESIGN: Data were collected in an intensive care ward, in a nursing home and in a residential home. The unit of measurement and analysis is the 'patient day'. The analysis for validity consisted of ridits calculations, and their graphical representations. Interrater reliability was measured by percentage agreement and Cohen's kappa. RESULTS: Graphs illustrate the differences on most nursing phenomena and interventions as expected beforehand. The percentage agreements for the residential home vary from 60.4 to 100%, and the kappa statistics from -0.09 to 0.85, indicating a poor to almost perfect interrater reliability. CONCLUSION: Intensive care patients and patients in the nursing home have more problems and need more nursing interventions compared with general hospital patients, while the patients in the residential home have lesser of both. This illustrates the discriminative validity of the NMDSN. The kappa values for various NMDSN variables are sufficient. A similar test in the general hospital is recommended.