Embargoed Until Sunday, May 21, 9:15 a.m. ET


Session:  A15 Critical Care: The Future of Critical Care? Regionalization and Closing the Gender Gap
Abstract Presentation Time: Sunday, May 21, 10 a.m. ET
Location: Room 146 C (Middle Bldg, Street Level), Walter E. Washington Convention Center


Newswise — ATS 2017, WASHINGTON, DC ─ In a new study that analyzed data from the large French CUB-REA database, high volume intensive care units (ICUs) were found to have lower death rates from acute respiratory distress syndrome (ARDS) than low and moderate volume ICUs.  The researchers also found that, overall, among the 35 ICUs in the CUB-REA registry, ICU mortality decreased despite an increase in ARDS severity. The study was presented at the 2017 American Thoracic Society International Conference.

ARDS is a common disease affecting 10 percent of patients admitted to ICUs.  ARDS occurs when fluid builds up in the tiny, elastic air sacs in the lungs.  This buildup reduces oxygen levels in the bloodstream, depriving organs of the oxygen they need to function.

“Our study is the first to look at a large and regional database that suggests patients with ARDS admitted to centers caring for large numbers of people with ARDS may have better prognoses than in centers with a lower case volume,” said lead author Martin Dres, MD, of the AP-HP, Groupe Hospitalier Pitié-Salpêtrière Charles Foix, Service de Pneumonologie et Réanimation Médicale, Paris, France. “We hope our research fuels further discussion regarding global quality of care and global management in the setting of ARDS.”

Dr. Dres and colleagues looked at CUB-REA data for 316,000 ICU stays taking place between 2000 and 2014.  They assessed the difference in ARDS severity using the SAPS2 scoring method, which measures the severity of disease for patients admitted to the ICU.  They also looked at ARDS-related mortality for low, moderate and high patient volume ICUs.  They statistically adjusted their analyses to account for severity of illness and other confounding factors.

The researchers found that 18,022 ICU stays between 2000 and 2014 were related to ARDS.  Overall ICU prevalence of ARDS was 8.3 percent, while its prevalence progressively increased from 6.5 percent in 2000 to 10.5 percent in 2008.  Prevalence then decreased to 7.2 percent in 2014.

During this time, SAPS2 severity scores increased from an average of 52 to 58 (with 52 representing a predicted 50 percent mortality rate).   ICU mortality gradually decreased over these years, from 57.9 percent to 45.5 percent, and hospital mortality decreased from 59.1 percent to 48.4 percent.  SAPS2 scores were highest in high-volume ICUs, while high volume ICUs had 51.5 percent mortality; moderate volume ICUs had 54 percent mortality; and low volume ICUs had 55.2 percent mortality rates.  Length of stay was also lowest in high volume ICUs.

“Further studies are warranted to confirm and delineate our results and determine whether regionalization of patients with ARDS should be recommended,” said Dr. Dres.

He added:  “Since submitting our research abstract, we have confirmed our findings with several methods of case volume calculation for a shorter duration of time (2000-2012).  This allowed us to confirm our preliminary results with a more homogeneous population.”

Contact for Media: [email protected]



Abstract 6524
Effect of High Volume ICUs on Mortality in ARDS Over 15 Years

 Authors: M. Dres1, T. Pham2, P. Aegerter3, B. Guidet4, L.J. Brochard2, A. Demoule5, G. Geri3; 1AP-HP, Groupe Hospitalier Pitié-Salpêtrière Charles Foix, Service de Pneumologie et Réanimation Médicale - Paris/FR, 2St Michael's Hospital - Toronto, ON/CA, 3Assistance Publique Hôpitaux de Paris - Boulogne Billancourt/FR, 4Assistance Publique Hôpitaux de Paris - Paris/FR, 5AP-HP, Groupe Hospitalier Pitié-Salpêtrière Charles Foix - Paris/FR; CUB-REA investigators


Introduction: Acute Respiratory Distress Syndrome (ARDS) is a common disease affecting 10% of patients admitted in the Intensive Care Unit (ICU) and associated with a crude ICU mortality ranging from 30 to 40% (1). As the management of ARDS patients requires a complex, well organized, and technically sophisticated level of care, we aimed at investigating whether there is or there is not a relationship between patient’s outcomes and case-volume ICUs. A secondary objective was to assess trends in patients’ characteristics, outcomes, and management between 2000 and 2014.


Material and Methods: The French CUB-REA database includes 35 ICUs, which collected 316,000 individual ICUs stays between 2000 and 2014. ARDS were identified from the hospital coding system. We assessed the difference in severity (SAPS2) and ICU mortality between tertiles of case volume. To reach this goal, multivariate analysis using mixed models was performed to adjust for severity of illness and other confounding factors.


Results: Between 2000 and 2014, 18,022 stays were related to ARDS. Overall, ICU prevalence of ARDS was 8.3%. ICU prevalence of ARDS progressively increased from 6.5% in 2000 to 10.5% in 2008 and then decreased to 7.2% in 2014 (p<0.001).


During this period, SAPS II increased from 52 [37-71] to 58 [41-75] (p<0.001) whereas both, ICU and hospital mortality, gradually decreased, from 57.9% to 45.5% (p<0.0001) and from 59.1% to 48.4% (p<0.0001) respectively.


The three ARDS case volume tertiles were: low volume (≤37 admissions per year), moderate volume (38-67), and high volume (>68). SAPS2 was significantly higher in high volume ICUs as compared to low and moderate volume ICUs: 61[43-80] vs. 53 [39-72] and 56 [41-74] respectively (p<0.001). As compared to low and moderate volume ICUs, high volume ICUs were characterised by lower ICU mortality: 51.5% vs. 54.3% and 55.2% respectively (p<0.0001) and lower hospital length of stay: 15 [5-31] days vs. 18 [7-36] and 17 [6-35] (p<0.0001). Multivariate analysis confirmed that ICU mortality in high volume ICUs was lower than in low and moderate volume ICUs (Odd ratio 0.72, 95%CI, 0.65-0.79).


Conclusion: Between 2000 and 2014, among ICUs of the CUB-REA registry, despite an increase in ARDS severity, ICU mortality decreased. There was a significantly lower ICU mortality in high volume ICUs.



  1. Bellani G, Laffey JG, Pham T, et al.: Epidemiology, Patterns of Care, and Mortality for Patients With Acute Respiratory Distress Syndrome in Intensive Care Units in 50 Countries. JAMA 2016; 315:788–800