ICU Staffing in Michigan: The Role of Nonintensivist Providers and Hospitalists

1University of Michigan, Ann Arbor, MI
2University of California, San Francisco, San Francisco, CA

Meeting: Hospital Medicine 2007, May 23-25, Dallas, Texas

Abstract number: 20


The National Quality Forum has advocated the use of board‐certified intensivists to deliver critical care in intensive care units (ICUs). However, the extent of intensivist staffing, the use of nonintensivist providers, and the role different providers play in medical decision making and performance improvement in ICUs in Michigan are unknown.


We surveyed participants in the Michigan Hospital Association Keystone ICU Project, a statewide ICU quality and safety collaboration, regarding the nature of critical care in participating ICUs.


The ICUs participating in the Keystone project represented 85% of all ICU beds in Michigan. We sent surveys to all participating hospitals. Ninety‐six surveys from 115 ICUs in 72 hospitals were received, representing a response rate of 100%. Five institutions, representing 22 ICUs replied with only 1 survey. The average ICU size was 13.3 ± 7.0 beds, median 12, range 4‐42. The average hospital size was 280 ± 22 beds, median 249, range 40‐1031. Of all ICUs, 24 (25%) reported only intensivists acting as attendings of record, and 72 (75%) had nonintensivist attendings. In ICUs with nonintensivist providers, 61 (85%) had primary care physicians, 58 (81%) had cardiologists, and 36 (50%) had hospitalists serving as attendings of record. Intensivists were the primary decision makers in 30 (32%) ICUs and nonintensivists in 34 (36%), and decision making was shared in 31 (33%). Hospitals with intensivist‐only attendings and where intensivists made most clinical decisions were larger and had larger ICUs (Table 1). Board certification in critical care of attendings in each ICU was uncommon, even in ICUs with intensivist‐only attendings (100% of attendings board certified in 46% (11/24) of these ICUs) or in ICUs where intensivists made most decisions (100% certification in 39% [11/28]). Only 11 ICUs(15%) required additional training (such as the FCCS course) for nonintensivist providers. Interestingly, most institutions reported that nursing was primarily responsible for improving ICU performance rather than intensivists or hospitalists.



In Michigan, nonintensivist physician staffing is common and is especially predominant in smaller hospitals. Improvements in ICU quality and safety are unlikely to be achieved only by the use of board‐certified intensivists. Quality and safety efforts will need to partner with both nursing and physicians who are not board certified in intensive care medicine.

Author Disclosure:

S. A. Flanders, None; R. C. Hyzy, None; A. D. Auerbach, None.

To cite this abstract:

Flanders S, Hyzy R, Auerbach A. ICU Staffing in Michigan: The Role of Nonintensivist Providers and Hospitalists. Abstract published at Hospital Medicine 2007, May 23-25, Dallas, Texas Abstract 20. Journal of Hospital Medicine. 2007; 2 (suppl 2). Accessed April 25, 2019.

« Back to Hospital Medicine 2007, May 23-25, Dallas, Texas