Identifying level one patients. A cross-sectional survey on an in-patient hospital population.

The purpose of this study was to identify the characteristics of level one patients and to explore how these differed from the other levels of care (zero and two). The study was conducted in two parts. Firstly, general adult in-patients (n=351) on the day of study were classified using the Intensive Care Society (ICS) Levels of Care. Secondly, a sample (n=67) of level zero, one and two patients were compared using physiological and demographic variables. Additionally, each patient was studied using three validated tools: EWS, TISS-28 and APACHE II. TISS-28 showed statistically significant results (p=0001) when correlated to level of care. When all three levels were analysed, EWS (p=0.001), APACHE II (p=0.0001) and variance in respiratory rate (p=0.001) showed significant differences in score according to level of care. However, no statistically significant differences were found between levels zero and one using the same data, allowing the deduction that ICS level two criteria are well defined and patients easily identifiable. The findings suggest that existing measurements of patient acuity, including the ICS criteria, are not sensitive enough to differentiate patients 'at risk' of deterioration (level one) from normal ward patients (level zero). This also suggests that level zero and one patients, based on the ICS classification, may not be from distinct populations but, in reality, one homogenous group.