Psychological cue use and implications for a clinical decision support system

Area of Science:

• Psychology
• Clinical Decision Making
• Artificial Intelligence

Background:

• Effective clinical decision-making relies on accurate probability judgments of patient outcomes.
• Current models often struggle to explain how clinicians process cues and assess likelihoods, particularly regarding base rates.

Purpose of the Study:

• To present a novel psychological classification model for understanding clinical judgment.
• To explain how clinicians utilize conditional probabilities and cue competition in decision-making.
• To provide a framework for improving mental-health risk assessment tools.

Main Methods:

• Development of a new psychological classification model based on conditional probabilities.
• Analysis of cue competition in determining outcome likelihoods.
• Application of the model to a clinical example of suicide risk prediction.

Main Results:

• The model explains how clinicians' responses to conditional probabilities influence classification.
• It accounts for apparent inappropriate responses to base rates, such as overestimating rare categories.
• The model effectively represents expert clinical judgments and is psychologically valid.

Conclusions:

• The proposed model offers a comprehensible representation of clinical judgment for mental-health risk assessment.
• It is suitable for integration into decision support systems, linking with statistical and pattern recognition tools.
• This approach can foster a web-based resource combining empirical data and clinical expertise for risk assessment and training.