British Journal of Anaesthesia, 2000, Vol. 84, No. 5 683
© 2000 The Board of Management and Trustees of the British Journal of Anaesthesia
Abstract |
Sequential organ scoring as a measure of effectiveness of critical care
1 Critical Care Complex, Norfolk and Norwich Acute NHS Trust, Brunswick Road, Norwich NR1 3SR, UK
Abstract
Introduction: Our objective was to establish whether sequential organ scoring could be used on the intensive care unit (ICU) as a measure of effectiveness of care. There is no current data to support the assumption that intensive care is of benefit to patients and no established method of measurement. Randomized controlled trials are ethically and practically difficult to perform. In-hospital mortality is influenced by case-mix, and reflects institutional rather than specifically ICU performance.
Design: Observational non-randomized study of 100 consecutively admitted patients over a 5 month period between September 12, 1998 and February 12, 1999 whose duration of stay on the ICU exceeded 48 h.
Methods: Simple demographic data were collected (age, source and type of referral, diagnosis, premorbid condition). Organ dysfunction was assessed daily using the Logistic Organ Dysfunction System (LODS) as described by Le Gall1 — measuring twelve physiological variables reflecting cardiovascular, respiratory, neurological, hepatic, renal and haematological status — using the worst physiological values for the preceding 24 h. Scores ranged from 0 (normal) to 22 (maximal derangement).
Main outcome measures: Change between score on admission and 72 h later, and comparison with eventual hospital mortality. The percentage of patients in whom there was no change or a decrease in organ score over this period.
Results: Seventy-five patients’ stay exceeded 72 h. A decrease or no change in LODS score over this period was associated with eventual hospital survival. The specificity (true prediction of survival) was excellent (0.95). We were able to achieve no change or an improvement in organ dysfunction in 80% of our patients.
However, the ability to correctly predict death from a change in score was poor (sensitivity 0.36). In general, the change in score overestimated the chance of survival.
In terms of individual organ function, intensive care was able to consistently improve scores relating to the cardiovascular, respiratory and renal systems over the first 72 h of care, but not the neurological, hepatic or haematological systems.
Conclusions: The hallmark of modern intensive care is general organ support. Physiological improvement over the first 72 h of care is largely due to reversal of cardiovascular, respiratory and renal dysfunction. Daily organ scoring usefully reflects the ability of an intensive care unit to stabilise or reverse physiological dysfunction, and this has been shown to predict eventual survival. As a means of comparison between units, we suggest that a figure of 80% improvement or no change in score over the first 72 h of admission could be a standard of care, whilst accepting that this may need to be reviewed based on other institutions’ findings.