Unexpected deaths and referrals to intensive care of patients on general wards. Are some cases potentially avoidable?

(i) To determine the incidence of unexpected deaths occurring on general wards, and whether any were potentially avoidable; (ii) to assess whether the quality of care on general wards prior to admission to intensive care affected subsequent outcome.

Six-month audit in teaching hospital. Review of medical, nursing and physiotherapy notes, bedside charts and laboratory data in ward patients either dying unexpectedly (i.e. not having a prior 'do not resuscitate' order) or requiring intensive care unit (ICU) admission. Panel assessment of quality of ward care prior to unexpected ward death or ICU admission.

Adult general ward patients admitted to ICU or dying unexpectedly.

ICU and hospital mortality.

(i) In the six-month study period, 317 of the 477 hospital deaths occurred on the general wards, of which 20 (6%) followed failed attempts at resuscitation. Thirteen of these unexpected deaths were considered potentially avoidable: gradual deterioration was observed in physiological and/or biochemical variables, but appropriate action was not taken; (ii) in the same period, 86 hospital inpatients were admitted on 98 occasions to the ICU, 31 of whom received suboptimal care pre-ICU admission due either to non-recognition of (the severity of) the problem or to inappropriate treatment. Both ICU (52% vs 35%) and hospital (65% vs 42%) mortality was significantly higher in these patients compared to well managed patients (p < 0.0001).

Patients with obvious clinical indicators of acute deterioration can be overlooked or poorly managed on the ward. This may lead to potentially avoidable unexpected deaths or to a poorer eventual outcome following ICU admission. Early recognition and correction of abnormalities may result in outcome benefit, but this requires further investigation.