Direct and delayed admission to an intensive care or high dependency unit following discharge from the emergency department: associated patient characteristics and hospital outcomes.

OBJECTIVE To compare patients admitted from the emergency department (ED) directly to a ward (EDWard), the intensive care unit (EDICU) or stepdown (high dependency) unit (EDSDU) with patients admitted via the ED, but whose admission to an ICU (EDWardICU) or SDU (EDWardSDU) was preceded by a ward stay. DESIGN, SETTING AND PARTICIPANTS Administrative and clinical data linkage; 650-bed, tertiary referral hospital, whose ED has about 60 000 patient presentations per annum; adult patients admitted via the ED to a ward, ICU or SDU and whose ED length of stay (LOS) was < 24 h. MAIN OUTCOME MEASURE Hospital outcome and stay. RESULTS From January 2004 to December 2007, there were 43 484 patients, of whom 40 609 (93.4%) were EDWard, 1020 (2.3%) were EDICU, 873 (2.0%) were EDSDU, 503 (1.2%) were EDWardSDU, and 479 (1.1%) were EDWardICU. Hospital mortality for EDWardICU patients exceeded that of EDICU patients (34.9% v 23.3%; P < 0.01), as did EDWardSDU exceed EDSDU (12.3% v 7.8%; P < 0.01). Median ward stay for EDWardICU patients was 47 h 37 min (IQR, 14 h 48 min - 131 h 53 min) and for EDWardSDU patients, 46 h 18min (IQR, 18h 28 min - 140h 12 min) (P=0.75). Compared with patients admitted to the ICU from the operating theatre, EDWardICU patients had a longer median ward stay (58 h 35 min v 34 h 36 min; P = 0.03) and hospital mortality (42.8% v 20.2%; P < 0.01). CONCLUSION Patients discharged from the ED to a general ward and subsequently to an ICU or SDU had a mortality that exceeded that of ED patients admitted directly to the ICU or SDU. Further investigations are warranted to explain this excess mortality and ascertain the extent of potential preventability.