Night-time care routine interaction and sleep disruption in adult cardiac surgery.

AIMS AND OBJECTIVES: To explore the context and the influence of night-time care routine interactions (NCRIs) on night-time sleep effectiveness (NSE) and daytime sleepiness (DSS) of patients in the cardiac surgery critical-care and progressive-care units of a hospital. BACKGROUND: There exists a paucity of empirical data regarding the influence of NCRIs on sleep and associated outcomes in hospitalised adult cardiac surgery patients. METHODS: An exploratory repeated-measures research design was employed on the data provided by 38 elective cardiac surgery patients (mean age 60.0 ± 15.9 years). NCRI forms were completed by the bedside nurses and patients completed a 9-item Visual Analogue Sleep Scale (100-mm horizontal lines measuring NSE and DSS variables). All data were collected during postoperative nights/days (PON/POD) 1 through 5 and analysed with IBM SPSS software. RESULTS: Patient assessment, medication administration and laboratory/diagnostic procedures were the top three NCRIs reported between midnight and 6:00 a.m. During PON/POD 1 through 5, the respective mean NSE and DSS scores ranged from 52.9 ± 17.2 to 57.8 ± 13.5 and from 27.0 ± 22.6 to 45.6 ± 16.5. Repeated-measures ANOVA showed significant changes in DSS scores (p < .05). NSE and DSS were negatively correlated (r = -.44, p < .05), but changes in NSE scores were not significant (p > .05). Finally, of 8 NCRIs, only 1 (postoperative exercises) was significantly related to sleep variables (r > .40, p < .05). CONCLUSION AND RELEVANCE TO CLINICAL PRACTICE: Frequent NCRIs are a common occurrence in cardiac surgery units of a hospital. Further research is needed to make a definitive conclusion about the impact of NCRIs on sleep/sleep disruptions and daytime sleepiness in adult cardiac surgery. Worldwide, acute and critical-care nurses are well positioned to lead initiatives aimed at improving sleep and clinical outcomes in cardiac surgery.

Estimating Patient's Health State Using Latent Structure Inferred from Clinical Time Series and Text.

Modern intensive care units (ICUs) collect large volumes of data in monitoring critically ill patients. Clinicians in the ICUs face the challenge of interpreting large volumes of high-dimensional data to diagnose and treat patients. In this work, we explore the use of Hierarchical Dirichlet Processes (HDP) as a Bayesian nonparametric framework to infer patients' states of health by combining multiple sources of data. In particular, we employ HDP to combine clinical time series and text from the nursing progress notes in a probabilistic topic modeling framework for patient risk stratification. Given a patient cohort, we use HDP to infer latent "topics" shared across multimodal patient data from the entire cohort. Each topic is modeled as a multinomial distribution over a vocabulary of codewords, defined over heterogeneous data sources. We evaluate the clinical utility of the learned topic structure using the first 24-hour ICU data from over 17,000 adult patients in the MIMIC-II database to estimate patients' risks of in-hospital mortality. Our results demonstrate that our approach provides a viable framework for combining different data modalities to model patient's states of health, and can potentially be used to generate alerts to identify patients at high risk of hospital mortality.