Deteriorating patients managed with end-of-life care following Medical Emergency Team calls

Aim: To describe the characteristics of patients whose end-of-life care was initiated in response to a Medical Emergency Team (MET) call and to develop a predictive score to aid prospective identification of these patients.

Methods: Retrospective cohort study of all MET calls in a tertiary teaching hospital between April 2010 and March 2011. All inpatients attended by the hospital MET. The main outcome measures were patient demographics, admission features and comorbidities in active and palliative patients, timing, frequency, physiology, and interventions in active and palliative MET calls.

Results: One thousand, five hundred and sixty-seven MET calls were called for 1073 patients. Sixty (5.6%) patients had at least one MET call resulting in initiation of end-of-life care. Palliative MET call patients compared with active patients were older (76.4 vs 65.9 years; P < 0.0001), had a shorter hospital stay (7.5 vs 12 days; P = 0.0002), had increased in-hospital mortality (73.3% vs 13.5%; P < 0.001), had higher Charlson comorbidity scores (3.1 vs 2.1; P = 0.0002) and were more likely to receive multiple MET calls (1.95/patient vs 1.43/patient; P = 0.011). Patient physiological parameters were worse at palliative MET calls. Prior history of malignancy, hemiplegia and peripheral vascular disease, and increasing age were independently associated with initiation of end-of-life care and were used to derive a 13-point predictive score. Patients with a score of 7 or more had a 20% chance of having a palliative MET call.

Conclusion: Prospective identification of patients requiring palliative care may be possible prior to MET involvement. This may allow more timely and appropriate end-of-life discussions.