Measuring intensive care performance to attain collaborative quality improvement: Work in progress but much to be...

Intensive Care Medicine must be viewed as an organized system of care that ensures delivery of timely and expert treatment to critically ill patients. Like any system, it needs a matrix of continuous quality improvement. The development of robust indicators covering all domains of quality is crucial for that matrix. The combined use of standardized mortality ratio and of standardized resource use ratio may allow the assessment of ICU performance and efficiency. ICU length of stay is a good proxy of resource use. Data driven models, possibly fueled by machine learning, may predict length of stay in all and in subpopulations of ICU patients.