Influence of critical illness on physicians' prognoses for underlying disease: a randomized study using simulated...

Area of Science:

• Medical Decision Making
• Critical Care Medicine
• Clinical Prognostication

Background:

• Physicians estimate disease severity to guide patient and family care directives.
• Estimates are crucial during critical illness but may be subject to bias.

Purpose of the Study:

• To investigate if acute illness context, prior patient prognoses, or question phrasing influence physician assessments of underlying disease severity.
• To identify potential biases in physician prognostication during critical care.

Main Methods:

• Internal medicine physicians evaluated case vignettes of cardiomyopathy and lung cancer.
• Vignettes were presented with either septic shock or an uneventful clinic visit context.
• Case order and survival question phrasing were randomized.

Main Results:

• Cardiomyopathy survival predictions were significantly affected by acute illness context (septic vs. clinic) and prior case presentation (lung cancer).
• Lung cancer survival predictions were sensitive to question phrasing but not context.
• Quality of life predictions were also influenced by context and case order.

Conclusions:

• Physician assessments of underlying disease severity are vulnerable to cognitive biases in the critical care setting.
• These biases can be influenced by acute illness, prior patient outcomes, and how questions are asked.
• The nature of the underlying disease may modulate the impact of these biases.