Search

Introduction In a time of increasingly complex health care, the provision of care for the - surgical -critically ill patient is of vital importance. The Intermediate Care Unit (IMCU) fulfils a growing need in the provision of this critical care and is therefore increasingly being implemented in hospitals world-wide. Through admitting the right patient at the right place, the IMCU has the potential to reduce the burden on ICU beds, provide specialty-specific critical care, improve patient outcome and decrease health care expenditures. However, whether IMCUs achieve these goals is highly debated, and therefore a scientific approach to support or contradict these viewpoints is justified and much desired. Aims This thesis first defines the concept of the IMCU and analyses how to assess its quality. Second, it analyses whether, and to what extent, IMCUs are beneficial for patients, hospitals and society. Third, it seeks to improve the safety of IMCUs through the development of hands-on tools and integrated computer algorithms. Outline Section 1 identifies differently used IMCU formats and definitions, and provides a guide for quality assessment of these units. Section 2 assesses the stand-alone surgical IMCU by describing and comparing various aspects of the IMCU, such as its management format, triage criteria, role in acute trauma critical care, possibilities and limitations and cost-efficiency. Section III provides supportive triage tools to optimize patient outcome before and during admission. Conclusions When properly formatted, staffed and managed, the surgical IMCU we investigated efficiently and safely reduced ICU admissions and healthcare expenditures. Hence it tailors the provision of critical care to the individual patient’s needs. The described surgical IMCU can herein serve as a benchmark for further research, comparison and constant quality improvement. The safety of IMCUs could be further improved through the use of hands-on tools at admission and automated computer algorithms during admission, using repeatedly measured covariates in a neural network.