Rationale: The US has seven times as many intensive care unit (ICU) beds per capita as the UK; the effect on care of critically ill patients is unknown.
Objective: Compare medical ICU admission in the US and UK.
Methods: Retrospective (2002-2004) cohort study of 172,785 admissions (137 US ICUs, Project IMPACT database; 160 UK ICUs, UK Case Mix Programme), with patients followed until initial hospital discharge.
Results: UK (vs US) admissions were less likely to be admitted directly from the emergency room (ER), (33.4 vs 58.0%), had longer hospital stays before ICU admission (mean days 2.6±8.2 vs 1.0±3.6), and fewer were ≥85 years (3.2% vs 7.8%). UK patients were more frequently mechanically ventilated within 24h after ICU admission (68.0% vs 27.4%), were sicker (mean Acute Physiology Score 16.7±7.6 vs 10.6±6.8), and had higher primary hospital mortality (38.0% vs 15.9%; adjusted Odds Ratio (OR) 1.73, 95%CI 1.50-1.99). There was no mortality difference for mechanically ventilated patients admitted from the ER (adjusted OR 1.09, 0.89-1.33). Comparisons of hospital mortality were confounded by differences in casemix, hospital length of stay (UK median 10 days (IQR 3-24) vs US 6 (3-11)), and discharge practices: more US patients were discharged to skilled care facilities (29.0% of survivors vs 6.0% in the UK).
Conclusions: Lower UK ICU bed availability is associated with fewer direct admissions from the ER, longer hospital stays before ICU admission, and higher severity of illness. Interpretation of between-country hospital outcomes is confounded by differences in casemix, processes of care and discharge practices.