Knowledge-based clinical decision support engine for psychiatric nursing process: design and evaluation (Preprint)

BACKGROUND Nursing assessments used in the psychiatry department considerably differ from those used in other departments. OBJECTIVE We developed a psychiatric knowledge-based clinical decision support system (Psy-KBCDSS) to help nurses in assessing patients’ problems. In addition, we evaluated the sensitivity and specificity of the Psy-KBCDSS and determined whether the Psy-KBCDSS can accurately formulate nursing diagnoses to assist nurses in providing care for psychiatric patients. METHODS Visual Studio 2005 was adopted as the primary software development tool. C# was used as the main development language, and a graphical concept was applied to develop the interface. We established a clinical diagnostic validity inference engine (CDVIE) to calculate the actual nursing assessment scores of nurses engaging in clinical tasks and to compute the nursing diagnosis data registered in the psychiatric nursing process system (Psy-NPS). The sensitivity and specificity of the nursing diagnoses formulated by the senior nurses and junior nurses regarding the same patient were extracted from the Psy-NPS and Psy-KBCDSS databases to conduct effectiveness assessment. RESULTS This study involved 22 nursing diagnoses commonly encountered in psychiatric wards. Of these diagnoses, the top 8 most common diagnoses formulated by the participants were altered thought processes, ineffective coping, sensory and perceptual alterations, insomnia, risk for other-directed violence, anxiety, impaired social interaction, and risk for suicide in Psy-NPS and Psy-KBCDSS. However, the diagnoses that showed significant increase in sensitivity between the Psy-NPS and Psy-KBCDSS were sensory and perceptual alterations, ineffective coping, and insomnia. The specificity of ineffective coping also increased considerably. CONCLUSIONS The Psy-KBCDSS is an empirical patient-oriented nursing clinical decision-making support system that may be used in patients’ individual assessment and helps nurses in formulating appropriate nursing diagnoses according to the nursing process.