Risk factors for prolonged stay in cardiac surgery intensive care units

Authors:
Publisher website
Google Scholar
Abstract
Aims and objectives: To identify the factors that might affect the length of stay in the intensive care unit (ICU‐LOS) among cardiac surgery patients. Background: ICU‐LOS forms an important factor for assessing the effectiveness of the provided nursing care. A number of factors can be accused for increasing patient hospitalization. The nursing workload (NWL), among others, was found to play a significant role as it is closely associated with the quality of care. Design: An observational cohort study among 313 consecutive patients who were admitted to the cardiac surgery intensive care unit of a general, tertiary hospital of Athens, Greece from November 2008 to November 2009. Methods: Data collection was performed by using a short questionnaire (for basic demographic information) and two instruments, the Nursing Activities Score (NAS) and the logistic EuroSCORE, for assessing the NWL and the perioperative risk for each patient respectively. Results: ICU‐LOS of more than 2 days increased with age and was more common among females (p < 0·001 and p = 0·02, respectively). Multivariate logistic regression analysis revealed a positive association between increased perioperative risk and the increased ICU‐LOS [odd ratio (OR) 1·9, 95% confidence interval (CI) 1·0–3·5, p = 0·04], while patients with a first day NAS of more than 61·6% had an almost 5·2 times greater probability to stay in the cardiac surgery unit for more than 2 days (OR 5·2, 95% CI 3·0–8·8, p < 0·001). Conclusions: Increased level of NWL and patient perioperative risk are closely associated with increased ICU‐LOS. Relevance to clinical practice: The correlation between patient perioperative risk and ICU‐LOS encourages the early identification of high‐risk patients for prolonged hospitalization. Furthermore, the relationship between NWL and ICU‐LOS allows the early identification of these patients with the use of an independent nursing tool.

This publication has 44 references indexed in Scilit: