Critical care outreach team's effect on patient outcome

Editor—The average hospital mortality in patients admitted to intensive care units in England is 30%, of which around 8% occurs after discharge from intensive care.¹ In the study by Ball et al the retrospective control group had a median predicted hospital mortality of 20.4%, and yet 19% died after discharge from intensive care.² No details are given of mortality in the intensive care unit, and only the results in the subgroup who survived to discharge from intensive care were reported. Clinicians can therefore only question why mortality after intensive care seems so much higher in a population of patients with an unremarkable predicted mortality.

Outcome prediction using the acute physiology and chronic health evaluation (APACHE II) score predicts hospital mortality rather than mortality in intensive care³ because discharge criteria from intensive care may vary between units and even within units as a function of demand (therefore influencing mortality in intensive care). By the same argument variations in mortality after intensive care might have been influenced by variation in discharge practices rather than variation in the quality of ward care. According to Ball et al, the difference between the two groups of survivors of critical care on admission to intensive care did not reach significance, but they cannot conclude that the groups were equivalent on discharge from intensive care.

If the outreach team has had a genuine effect on mortality, a reduction in the standardised mortality ratio to hospital discharge for all patients admitted to intensive care may be expected. If, however, the standardised mortality ratio has not changed, a reduction in ward mortality may simply reflect changes in discharge policy of the intensive care unit.