Diamond-Forrester Score for Chest Pain
The Diamond-Forrester score is a clinical risk assessment tool that estimates the pretest probability of obstructive coronary artery disease (CAD) in patients with chest pain based on three variables: age, sex, and type of chest pain (typical angina, atypical angina, or nonanginal chest pain). 1
Purpose and Clinical Application
The Diamond-Forrester method serves to stratify patients into low, intermediate, or high probability categories for having significant coronary stenosis (≥50% luminal narrowing) before proceeding with diagnostic testing. 1 This pretest probability assessment helps clinicians determine which patients require further cardiac imaging or invasive evaluation versus those who can be safely managed without additional testing. 1
How the Score Works
The model generates probability estimates by combining:
- Patient age (stratified by decade: 30-39,40-49,50-59,60-69 years)
- Sex (male or female)
- Chest pain classification: typical angina (substernal chest discomfort provoked by exertion/emotion and relieved by rest/nitroglycerin), atypical angina (meeting 2 of 3 typical criteria), or nonanginal chest pain (meeting ≤1 typical criteria) 1
For example, a 64-year-old man with typical angina has a 94% probability of significant CAD, while a 32-year-old woman with nonanginal chest pain has only a 1% probability. 1
Critical Limitations in Contemporary Practice
The Diamond-Forrester model substantially overestimates the actual prevalence of obstructive CAD in modern populations, particularly in women and lower-risk patients. 1, 2, 3 Multiple recent studies demonstrate this overestimation:
- The PROMISE trial showed the actual prevalence of CAD ≥50% was only 13.9%, compared to an average Diamond-Forrester prediction of 40.6% (only one-third of predicted). 3
- The model performs particularly poorly in women, where validation studies show significant overestimation of disease probability. 1, 2
- In contemporary low-risk acute chest pain populations, typical angina as defined by Diamond-Forrester was not predictive of CAD on coronary CT angiography or exercise testing. 4
Superior Alternative Risk Stratification Methods
The coronary calcium score (CCS) more accurately stratifies stable CAD risk than the Diamond-Forrester method and should be preferentially used when available. 1 The ACR Appropriateness Criteria explicitly state that CCS provides better risk stratification than Diamond-Forrester for stable CAD patients. 1
Additional considerations:
- The Duke Clinical Score performs better than Diamond-Forrester in low-risk patients. 1
- The Framingham Risk Score demonstrates superior ability to predict CAD compared to Diamond-Forrester, with lower rates of significant CAD in the low-risk category. 5
- For acute chest pain evaluation, the HEART score was specifically designed for emergency department use and provides more accurate contemporary risk stratification than historical models. 1
When Diamond-Forrester May Still Be Applied
Despite its limitations, the Diamond-Forrester method remains referenced in guidelines as one available tool among multiple risk assessment options (including Framingham, PROCAM, CCS, and Duke Clinical Score) that use different criteria such as age, sex, family history, chest pain type, lipid levels, and previous cardiovascular events. 1 However, clinicians must recognize that risk assessment using various methods can lead to variable pretest probabilities and may stratify the same patient into different risk categories. 1
Common Pitfalls to Avoid
- Do not rely solely on Diamond-Forrester predictions without considering contemporary validation data showing systematic overestimation. 2, 3
- Do not assume the model performs equally across all demographic groups—it is particularly inaccurate in women and patients over 70 years of age. 1, 2
- Do not use Diamond-Forrester alone in patients with intermediate pretest probability—combine with coronary calcium scoring for improved accuracy. 1, 6
- In patients classified as high probability by Diamond-Forrester with a calcium score of zero, the prevalence of obstructive CAD and adverse events remains elevated (7-10%), so calcium scoring cannot exclude disease in this group. 6