Diagnostic concordance among nursing clinical decision support system users: a pilot study.

OBJECTIVE: To analyze the nursing diagnostic concordance among users of a clinical decision support system (CDSS), The Electronic Documentation System of the Nursing Process of the University of São Paulo (PROCEnf-USP®), structured according to the Nanda International, Nursing Intervention Classification and Nursing Outcome Classification (NNN) Taxonomy. MATERIALS AND METHODS: This pilot, exploratory-descriptive study was conducted from September 2017 to January 2018. Participants were nurses, nurse residents, and nursing undergraduates. Two previously validated written clinical case studies provided participants with comprehensive initial assessment clinical data to be registered in PROCEnf-USP®. After having registered the clinical data in PROCEnf-USP®, participants could either select diagnostic hypotheses offered by the system or add diagnoses not suggested by the system. A list of nursing diagnoses documented by the participants was extracted from the system. The concordance was analyzed by Light's Kappa (K). RESULTS: The research study included 37 participants, which were 14 nurses, 10 nurse residents, and 13 nursing undergraduates. Of the 43 documented nursing diagnoses, there was poor concordance (K = 0.224) for the diagnosis "Ineffective airway clearance" (00031), moderate (K = 0.591) for "Chronic pain" (00133), and elevated (K = 0.655) for "Risk for unstable blood glucose level" (00179). The other nursing diagnoses had poor or no concordance. DISCUSSION: Clinical reasoning skills are essential for the meaningful use of the CDSS. CONCLUSIONS: There was concordance for only 3 nursing diagnoses related to biological needs. The low level of concordance might be related to the clinical judgment skills of the participants, the written cases, and the sample size.

Usability Testing of PROCEnf-USP: A Clinical Decision Support System.

Decision support systems (DSSs) are recognized as important tools, capable of processing high volumes of data and increasing productivity. The usability of these tools affects their effectiveness. By evaluating the interactions between registered nurses (RNs) and the DSSs, this study explores how they impact RN decision-making. This study analyzed 24 months (2011-2012) of data collected in Brazil in two units of a large, public, urban hospital in São Paulo that uses a nurse documentation system with an embedded DSS based on NANDA-I. Using mixed effects logistic regression, this study analyzed the agreement between RNs and a DSS when selecting nursing diagnoses. Results suggest that the agreement is mediated by characteristics of the RNs (education and experience) as well as units and year of encounter. Surprisingly, disagreement between RN and DSS when selecting defining characteristics (DC) had positive effects on the odds of agreement on diagnoses. Our results suggest that DSSs support nurses' clinical decision making, but the nurse's clinical judgment is the mediating factor. More research is necessary.