Survey of Contemporary Cardiac Surgery Intensive Care Unit Models in the United States

Ann Thorac Surg. 2020 Mar;109(3):702-710. doi: 10.1016/j.athoracsur.2019.06.077. Epub 2019 Aug 15.


Background: Intensive care unit (ICU) structure and intensive care physician staffing (IPS) models are thought to influence outcomes after cardiac surgery. Given limited information on staffing in the cardiothoracic ICU, The Society of Thoracic Surgeons Workforce on Critical Care undertook a survey to describe current IPS models. We hypothesized that variability would exist throughout the United States.

Methods: A survey was sent to The Society of Thoracic Surgeons centers in the United States. Center case volume, ICU census, procedure profiles, and the primary specialties of consultants were queried. Definitions of IPS models were open (managed by cardiac surgeons), closed (all decisions made by dedicated intensivists 7 days a week), or semiopen (intensivist attends 5-7 days a week with surgeons cosharing management). Experience level of bedside providers and after-hours provider coverage were also assessed.

Results: Of the 965 centers contacted, 148 (15.3%) completed surveys. Approximately 41% of reporting centers used a dedicated cardiothoracic ICU for immediate postoperative management. The most common IPS model was open (47%), followed by semiopen (41%) and closed (12%). The primary specialties of intensivists varied, with pulmonary medicine/critical care being predominant (67%). Physician assistants were the most common after-hours provider (44%). More than one-third of responding centers described having no house staff, other than bedside nurses, for nighttime coverage.

Conclusions: Cardiothoracic ICU models vary widely in the United States, with almost half being open, often with no in-house coverage. In-house nighttime coverage was (1) not driven by case complexity and (2) most commonly provided by a physician assistant. Clinical outcomes associated with different ISPS models require further evaluation.

Publication types

  • Research Support, Non-U.S. Gov't

MeSH terms

  • Adult
  • Databases, Factual
  • Health Care Surveys
  • Health Workforce / statistics & numerical data
  • Humans
  • Intensive Care Units / organization & administration*
  • Intensive Care Units / statistics & numerical data
  • Medical Staff, Hospital / statistics & numerical data*
  • Nursing Staff, Hospital / statistics & numerical data
  • Thoracic Surgery / education
  • United States