Spatio-temporal Structure of US Critical Care Transfer Network

Area of Science:

• Healthcare quality research
• Health services research
• Medical informatics

Background:

• Intensive Care Units (ICUs) are critical care settings with variable quality.
• Many deaths in ICUs could be preventable with appropriate hospital care.
• Existing patient transfer policies aim to improve outcomes but face inefficiencies.

Purpose of the Study:

• To identify and analyze inefficient patient transfer patterns in Intensive Care Units (ICUs).
• To discover data mining schemes for detecting cascading patient transfers.
• To provide evidence for improving patient transfer systems to enhance survival outcomes.

Main Methods:

• Analysis of Medicare critical care transfer data spanning nearly 5,000 U.S. hospitals over 10 years.
• Application of data mining schemes and significance tests to identify transfer inefficiencies.
• Examination of multi-step transfer cascades and hospital involvement.

Main Results:

• Evidence of recurrent patient transfers to secondary hospitals, bypassing more optimal choices.
• Identification of cascading transfers across multiple hospital transitions.
• Some hospitals are disproportionately involved in these inefficient transfer cascades.

Conclusions:

• The current patient transfer system for Intensive Care Units (ICUs) exhibits significant inefficiencies.
• Inefficient transfer cascades negatively impact patient outcomes and survival.
• Targeted interventions are needed to optimize hospital transfers and improve critical care quality.