• S Ramani Moonesinghe, Michael G Mythen, and Michael P W Grocott.
• Centre for Anaesthesia, 3rd floor Maples Link Corridor, University College London Hospital, 235 Euston Road, London NW1 2BU, UK. rmoonesinghe@gmail.com
• Curr Opin Crit Care. 2009 Aug 1; 15 (4): 320-7.

Purpose Of ReviewThe risk of adverse outcome in patients undergoing major surgery is affected both by cardiorespiratory fitness, and the presence and severity of comorbidities. Accurate risk stratification is essential for the identification of patients who may benefit from specific perioperative management strategies or from an augmented level of perioperative care. Risk stratification techniques include risk prediction models, assessment of functional capacity and novel biochemical markers. This review examines the evidence for the use of these different techniques in perioperative patients.Recent FindingsThere remains considerable variation in the predictive ability of risk stratification models, in part due to the subjective nature of some of the component variables. Whereas a basic assessment of functional capacity using structured questionnaires may be helpful, in patients thought to be at high risk, the most accurate technique may be cardiopulmonary exercise testing, although the strength of the hypothesized relationship between functional capacity and perioperative outcome has not been fully defined. There have been advances in the identification and refinement of biochemical markers for risk prediction, in particular, brain natriuretic peptide and C-reactive protein. Currently, few centres routinely systematically utilize these strategies to risk stratify perioperative patients.SummaryThe development of improved risk stratification techniques would be assisted by large-scale epidemiological studies. Improvements to currently used risk prediction models are likely to result from the use of variables which more objectively measure patient health and fitness than current tools, and may use a combination of all the above techniques to improve predictive accuracy.

7 keywords...

hide…