BACKGROUND:Global epidemiological data regarding outcomes for patients in intensive care units (ICUs) are scarce, but are important in understanding the worldwide burden of critical illness. We, therefore, did an international audit of ICU patients worldwide and assessed variations between hospitals and countries in terms of ICU mortality.METHODS:730 participating centres in 84 countries prospectively collected data on all adult (>16 years) patients admitted to their ICU between May 8 and May 18, 2012, except those admitted for fewer than 24 h for routine postoperative monitoring. Participation was voluntary. Data were collected daily for a maximum of 28 days in the ICU and patients were followed up for outcome data until death or hospital discharge. In-hospital death was analysed using multilevel logistic regression with three levels: patient, hospital, and country.FINDINGS:10,069 patients were included from ICUs in Europe (5445 patients; 54·1%), Asia (1928; 19·2%), the Americas (1723; 17·1%), Oceania (439; 4·4%), the Middle East (393; 3·9%), and Africa (141; 1·4%). Overall, 2973 patients (29·5%) had sepsis on admission or during the ICU stay. ICU mortality rates were 16·2% (95% CI 15·5-16·9) across the whole population and 25·8% (24·2-27·4) in patients with sepsis. Hospital mortality rates were 22·4% (21·6-23·2) in the whole population and 35·3% (33·5-37·1) in patients with sepsis. Using a multilevel analysis, the unconditional model suggested significant between-country variations (var=0·19, p=0·002) and between-hospital variations (var=0·43, p<0·0001) in the individual risk of in-hospital death. There was a stepwise increase in the adjusted risk of in-hospital death according to decrease in global national income.INTERPRETATION:This large database highlights that sepsis remains a major health problem worldwide, associated with high mortality rates in all countries. Our findings also show a significant association between the risk of death and the global national income and suggest that ICU organisation has an important effect on risk of death.