Prescriptive models to support decision making in genetics

Area of Science:

• Clinical Genetics
• Decision Analysis
• Medical Informatics

Background:

• Formal prescriptive models aid in understanding risks and uncertainties.
• Bayes rule is crucial for interpreting clinical data and providing individualized probabilities.
• Screening programs necessitate consideration of prior disease probabilities.

Purpose of the Study:

• To demonstrate how formal prescriptive models improve decision-making in clinical genetics.
• To outline the application of Bayes rule, decision trees, and utility theory in genetic counseling.
• To enhance communication and integration of patient preferences into clinical decisions.

Main Methods:

• Utilizing Bayes rule for interpreting patient data (pedigrees, history, labs).
• Employing decision trees to structure diagnostic, therapeutic, and reproductive choices.
• Applying utility theory to quantify patient preferences for reproductive outcomes.

Main Results:

• Individualized probabilities of diagnoses and outcomes can be calculated.
• Decision trees facilitate structured evaluation of medical and reproductive options.
• Sensitivity analyses reveal the impact of patient attitudes on decision-making.

Conclusions:

• Formal clinical decision analytic models deepen understanding in clinical genetics.
• These models support well-reasoned decisions by integrating probabilities and patient utilities.
• Improved patient-clinician communication and decision-making are achieved through these formalisms.