University of Twente Student Theses


Detection of deterioration in continuously monitored surgical ward patients : analysis of a new risk prediction model combining nurse intuition and physiological changes.

Kalsbeek, M.R. (2020) Detection of deterioration in continuously monitored surgical ward patients : analysis of a new risk prediction model combining nurse intuition and physiological changes.

[img] PDF
Abstract:BACKGROUND Due to the ageing population, the demand for intensive care is increasing (1). As a result, critically ill patients will be transferred to the general surgical ward earlier. These critically ill patients are often complex cases and early detection of deterioration is critical in preventing major complications that can result in impairment, function loss, longer days spend in the hospital, higher hospital costs and even mortality. Nurses monitor patients by manually measuring vital signs. Literature shows that changes in vital signs can detect deterioration of a patient. Nurses use an instrument called MEWS to score deviating vital signs and detect deterioration. However, it turns out that nurse intuition plays an essential role for nurses in the identification of deterioration and that vital signs are mostly used to validate their intuitive feeling. Literature even shows that nurse intuition alone is better at predicting deterioration than deviating vital signs. Unfortunately the current method of measuring and recording vital signs is suboptimal and importance of nurse intuition is often not acknowledged sufficiently, resulting in an underestimation of physiological deterioration in patients. Continuous monitoring of vital functions can potentially resolve the shortcomings of the current practice and has already shown promising results in the early detection of deterioration. Though, it has not yet been established whether combining nurse intuition and continuous monitoring will yield even more benefits. OBJECTIVE The research objective is firstly, to investigate the predictive value of MEWS and nurse worry indictors in detecting patient deterioration. Secondly, it will be investigated whether continuous vital sign monitoring combined with nurse worry results in earlier and better detection of deterioration in surgical patients compared to current practice. METHOD An exploratory study was conducted. In the first part, the current performance of the MEWS and nurse worry indicators as composite scores were investigated and compared by calculating sensitivity, specificity, positive- and negative predictive values and Area under the Curve values on a discretely measured dataset. In the second part, logistic regression was used to determine the predictive performance of the separate variables in predicting adverse events. The variables with highest individual predictive values were combined in a multi logistic regression model to construct a new risk scoring instrument that was then tested on continuously recorded vital signs data of the patients. RESULTS Results show that MEWS and nurse worry as composite scores have poor predictive power with an AUC of 0.6 and 0.5, respectively. The most important variables that make up MEWS and Nurse worry, were blood pressure (P=0.006) , heart rate (P≤0.000), respiratory rate (P=0.095), nurse worry about temperature (P=0.168), nurse feeling (P=0.270) and patient feeling (P=0.169). These variables were used in the new model (CWS). The CWS with a time interval of 60 minutes had a sensitivity of 77% , specificity of 72%, PV+ 44 % , a PV- of 92%, and an AUC of 0.77. These results are significantly better than the results observed for traditional MEWS on both manually obtained data and on the continuous data. The CWS detected adverse events on average a day earlier than the traditional MEWS on manually obtained data. CONCLUSION Introducing continuous monitoring on the surgical wards and using vital signs and nurse worry indicators to detect deterioration in patients’ condition both show great potential towards earlier and better detection of adverse events. Better and earlier detection of an adverse event can results in more patient safety, lower mortality rates, reduce workload of nurses and improve communication between health professionals. Future research should focus on establishing more certainty about scoring ranges for the vital signs in continuous monitoring solutions, automated pattern recognition in vital signs in combination with nurse worry indicators and should look into the added value of relative scores besides absolute scores to indicate deterioration.
Item Type:Essay (Master)
Faculty:BMS: Behavioural, Management and Social Sciences
Subject:44 medicine
Programme:Health Sciences MSc (66851)
Link to this item:
Export this item as:BibTeX
HTML Citation
Reference Manager


Repository Staff Only: item control page