The patient‐at‐risk team

We read the paper by Goldhill et al. (Anaesthesia 1999; 54: 853–60), describing their ‘Patient-at-risk team’ (PART) with interest. We would like to comment, based on data from the ‘Medical Emergency Team’ (MET) system which Liverpool Hospital has operated since 1990 [1]. The aim of the MET system is to identify patients early in the course of their physiological deterioration, allow prompt delivery of diagnostic and resuscitative procedures and improve outcome. The team consists of the intensive care and medical registrar, intensive care resident and a senior intensive care nurse. Frequently the intensive care senior registrar or specialist will also respond. The calling criteria have been much simplified during the evolution of the system (Table 1). A major difference compared with the PART criteria is the inclusion of a ‘worried’ criterion. Of the patients receiving a MET call on the basis of this criterion, 22% are eventually admitted to ICU/HDU. In the last 6 months of 1998, there were 429 MET calls compared with 63 PART calls over the same time period. The figure for MET calls represents 1.6% of all Liverpool Hospital discharges. The proportion of patients admitted to ICU/HDU following a MET call was 18%, lower than the PART admission rate of 44%. The median APACHE 2 for MET patients admitted to ICU/HDU was 17, similar to the PART figure of 14. Although one must be cautious in drawing conclusions from these crude data, a number of observations arise. Given the greater number of calls, the MET system may be more sensitive in detecting sick patients. This is supported by the fact that in the Royal London Hospital, patients admitted to ICU who did not receive a PART call had similar APACHE 2 scores to those who were admitted via the PART, whereas patients admitted to ICU in Liverpool hospital by routes other than via a MET call had a significantly lower mean APACHE 2 score than patients admitted by the MET team (13.1 vs. 17.4, p < 0.001). This is related to the calling criteria and the fact any hospital staff member can initiate the MET system without prior consultation with the admitting medical team. Nurses activate the MET system in the vast majority of cases. The lower proportional admission rate reflects a lower specificity for the MET calling criteria. Whilst a greater number of calls increases medical and nursing workload in terms of time spent attending MET calls and disruption of other tasks, admission to ICU/HDU should not be seen as the sole endpoint, however, given that the MET system frequently benefits sick patients who do not require admission. At the very least, a MET call focuses the attention of the admitting team on the patient. Another tangible benefit of both systems is the identification of patients who are dying and for whom palliative care is more appropriate. ‘Not for resuscitation’ orders were made in similar proportions of calls (3% of PART calls and 2.5% of MET calls). Sickness severity scores for patients admitted after a MET or PART call are similar, reflecting similar decision making in both hospitals. However, it is rare for admission to be delayed for deserving patients after a MET call, whereas only 64% of patients ultimately admitted to ICU within 48 h were admitted immediately following the PART call. This is despite similar APACHE 2 scores in the patients admitted immediately and in those whose admission was delayed. The delay in admission may have been for clinical reasons, but as Goldhill et al. state, shortage of beds was a major factor. It is clear that once an early warning system is in place, many critically ill patients are identified, many of whom would benefit from admission to a critical care environment. Liverpool Hospital has eight nominal ICU beds and 12 nominal HDU beds with roughly 600 inpatient beds overall. The MET system was a major factor in determining these numbers of critical care beds during the planning of the hospital. It is becoming clear that the traditional ‘four-walls’ model of intensive care is outdated. Admitting patients with established multi-organ failure is all too often futile. Physiological deterioration must be identified and treated promptly, by taking intensive care skills to where they are needed. Further research is required into the development of the ideal early warning system; such research may include the development of a common data set for evaluation between different sites. However, current systems and criteria are effective and can be applied in any acute hospital. Outcome data should be used in the argument for adequate critical care bed provision.