Detecting unfinished nursing care among at risk and patients with delirium: a development and validation study

Background: Unfinished nursing care (UNC) refers to those interventions that are required by patients but are delayed or omitted by nurses. No instruments to date have been developed to measure UNC among patients at risk or with delirium that may help in understanding the wide variations in the prevalence of delirium documented thus far. Methods: The Unfinished Nursing Care Survey for Patients at Risk of and with Delirium (UNCSD) is an adaptation of the Unfinished Nursing Care Survey. It consists of Part A (35 unfinished interventions) and Part B (23 reasons for UNC). Validation took place in 2023-2024 in accordance with the COnsensus-based Standards for the selection of health Measurement INstruments guidelines; data were collected online from 296 nurses. Acceptability and psychometric properties, including construct validity, hypothesis testing and criterion validity, were assessed using Mokken analysis, exploratory factor analysis and analysis of variance. Results: The UNCSD showed a high level of acceptance (Part A, 100% and Part B, 93.7 %). Part A showed strong scalability (H = 0.58), which indicates the one-dimensional structure of the scale. Part B yielded four factors with 65.2 % explained variance and a Cronbach's alpha (internal consistency) of 0.775. Registered nurses (RNs) with graduate degrees, with more experience, and those providing individualized care reported significantly lower scores on the UNCSD, suggesting that the instruments are capable of detecting differences between groups. The statistically significant differences between the UNCSD and the UNCS Part A indicate that the new instrument is better able to detect unfinished care in patients with delirium and at-risk patients. Conclusions: The UNCSD is valid in terms of acceptability, construct validity, hypothesis testing and criterion validity. (c) 2025 Elsevier Inc. All rights are reserved, including those for text and data mining, AI training, and similar technologies.