Study Shows VA Hospitals Caring for Lower Volumes of Mechanically Ventilated Patients Do Not Have Worse Mortality
BACKGROUND:
The association between the volume of patients treated in a hospital and patient outcomes is one of the most consistent relationships observed in modern medicine. Across a broad array of high-risk surgical and medical diagnoses, care in centers with higher patient volume is associated with improved survival. Moreover, at least two large, observational studies examining patients cared for in non-federal hospitals demonstrated that mechanically ventilated (MV) non-surgical patients who receive care in higher-volume hospitals have 5-10% lower absolute risk of hospital death compared to lower volume centers. This retrospective cohort study sought to assess the relationship between volume of non-operative MV patients receiving VA ICU care and 30-day mortality. Using VA data, investigators identified all Veterans (n=5,131) who received mechanical ventilation in an ICU at one of 119 VA hospitals during 2009, excluding those who had undergone a surgical procedure during hospitalization. Risk-adjusted variables included: demographics, diagnosis on admission, comorbid conditions, and laboratory values (e.g., glucose, white blood count). For each VAMC, volume was defined as the total number of non-surgical admissions receiving mechanical ventilation in an ICU during 2009. VA hospitals in this study cared for a median of 40 non-post-operative MV patients in 2009.
FINDINGS:
- VA hospitals caring for lower volumes of mechanically ventilated patients do not have worse mortality compared to hospitals with higher volumes. The relationship between hospital volume of MV and 30-day mortality was not statistically significant: each 50-patient increase in volume was associated with a non-signicant 2% decrease in the odds of death within 30 days, By comparison, the published civilian hospital literature suggests a 10% decrease in odds per 50 patient increase over the same range of volume observed in the present study.
- There were no substantive differences in the primary results after excluding either repeat admissions or patients who were transferred into the VA from a non-VA facility.
- Compared to hospitals in the lowest quartile of volume, those in the highest quartile had more beds, were more often located in the South, were more often teaching hospitals, and offered greater complexity of ICU services.
- Veterans treated in the top quartile of hospitals by volume were slightly younger, more often African-American, and were more likely to be admitted via the ER. There were no consistent differences in sex or number of comorbid conditions across quartiles of hospital MV volume.
LIMITATIONS:
- Investigators were unable to test whether ICU organizational characteristics modified or mediated study results.
- Investigators were unable to account for patient referral outside of VA. Low-volume centers may selectively refer their sickest patients to non-VA hospitals, effectively reducing their mortality rates.
AUTHOR/FUNDING INFORMATION:
This study was funded by HSR&D (IIR 11-109). Drs. Wiitala, Sales, and Iwashyna and Mr. Kennedy are part of HSR&D's Center for Clinical Management Research, Ann Arbor, MI.
Cooke C, Kennedy E, Wiitala W, Almenoff P, Sales A, and Iwashyna T. Despite Variation in Volume, VA Hospitals Show Consistent Outcomes among Patients with Non-Post-Operative Mechanical Ventilation. Critical Care Medicine 2012 Sep;40(9):2569-2575.
