Objective  

Mortality after many procedures is lower in centers where more procedures are done. It is controversial whether this is true for intensive care units, too. We examined the relationship between the volume of activity of intensive care units (ICUs) and mortality by a measure of risk-adjusted volume of activity specific for ICUs.

Design  

Prospective, multicenter, observational study.

Setting  

Eighty-nine ICUs in 12 European countries.

Patients  

During a 4-month study period, 12,615 patients were enrolled.

Interventions  

Demographic and clinical statistics, severity at admission and a score of nursing complexity and workload were collected.

Results  

Total volume of activity was defined as the number of patients admitted per bed per year, high-risk volume as the number of high-risk patients admitted per bed per year (selected combining of length of stay and severity of illness). A multi-step risk-adjustment process was planned. ICU volume corresponding both to overall [odds ratio (OR) 0.966] and 3,838 high-risk (OR 0.830) patients was negatively correlated with mortality. Relative mortality decreased by 3.4 and 17.0% for every five extra patients treated per bed per year in overall volume and high-risk volume, respectively. A direct relationship was found between mortality and the ICU occupancy rate (OR 1.324 and 1.351, respectively).

Conclusions  

Intensive care patients, whatever their level of risk, are best treated where more high-risk patients are treated. Moreover, the higher the ICU occupancy rate, the higher is the mortality.