Objective: To investigate the impact of various facets of ICU organization on outcome in a large cohort of ICU patients from different geographic regions.
Design: International, multicenter, observational study.
Setting: All 1,265 ICUs in 75 countries that contributed to the 1-day point prevalence Extended Prevalence of Infection in Intensive Care study.
Patients: All adult patients present on a participating ICU on the study day.
Interventions: None.
Measurements and main results: The Extended Prevalence of Infection in Intensive Care study included data on 13,796 adult patients. Organizational characteristics of the participating hospitals and units varied across geographic areas. Participating North American hospitals had greater availability of microbiologic examination and more 24-hour emergency departments than did the participating European and Latin American units. Of the participating ICUs, 82.9% were closed format, with the lowest prevalence among North American units (62.7%) and the highest in ICUs in Oceania (92.6%). The proportion of participating ICUs with 24-hour intensivist coverage was lower in North America than in Latin America (86.8% vs 98.1%, p = 0.002). ICU volume was significantly lower in participating ICUs from Western Europe, Latin America, and Asia compared with North America. In multivariable logistic regression analysis, medical and mixed ICUs were independently associated with a greater risk of in-hospital death. A nurse:patient ratio of more than 1:1.5 on the study day was independently associated with a lower risk of in-hospital death.
Conclusions: In this international large cohort of ICU patients, hospital and ICU characteristics varied worldwide. A high nurse:patient ratio was independently associated with a lower risk of in-hospital death. These exploratory data need to be confirmed in large prospective studies that consider additional country-specific ICU practice variations.