Assessing patients with possible heart disease using scores.

Area of Science:

• Cardiology
• Medical Informatics
• Biostatistics

Background:

• General practitioners act as gatekeepers for specialist cardiac care.
• Optimal use of basic clinical tools (history, physical exam, exercise test) is crucial.
• Current diagnostic methods for coronary artery disease (CAD) can be improved.

Purpose of the Study:

• To evaluate the utility of multivariable analysis of clinical and exercise test data for CAD diagnosis and prognosis.
• To explore the effectiveness of derived scores compared to traditional methods.
• To identify areas for improvement in prediction algorithms for cardiac patients.

Main Methods:

• Multivariable statistical techniques applied to clinical and exercise test data.
• Development and representation of scores as nomograms or additive tables for physician use.
• Comparison of score-derived estimates with physician judgment.

Main Results:

• Multivariable scores show superior discriminating power over simple ST response classification.
• Scores effectively stratify patients by disease probability and prognosis.
• Physician-reluctant computer-derived scores, when presented accessibly, perform as well as or better than expert cardiologists.

Conclusions:

• Multivariable analysis of clinical and exercise data offers a valuable tool for CAD diagnosis, prognosis assessment, and cost reduction.
• Further research is needed to clarify the discriminating power of specific variables and improve prediction algorithms, especially for women.
• Validated, portable prediction equations can empower clinicians to ensure appropriate and cost-effective cardiac care.