Wells Score for DVT and PE Assessment
The Wells score is a validated clinical prediction rule that should be used to stratify pretest probability of deep vein thrombosis (DVT) and pulmonary embolism (PE) in emergency department and outpatient settings, guiding decisions about D-dimer testing and imaging to safely exclude venous thromboembolism while reducing unnecessary testing. 1, 2
Purpose and Clinical Application
Risk Stratification Framework
The Wells score categorizes patients into distinct risk groups that directly correlate with disease probability:
For Pulmonary Embolism:
- Low risk (Wells score <2): 3-3.6% PE probability 3, 1
- Intermediate risk (Wells score 2-6): 13-20.5% PE probability 3, 1
- High risk (Wells score >6): 36-66.7% PE probability 3, 1
For Deep Vein Thrombosis:
Diagnostic Algorithm Integration
For suspected PE ("PE unlikely" category):
- Proceed with D-dimer testing first 3, 2
- Negative D-dimer safely excludes PE with 99.5% negative predictive value 1
- Positive D-dimer or "PE likely" category requires CT pulmonary angiography 3, 2
For suspected DVT (low-to-moderate risk):
- Perform high-sensitivity D-dimer testing first 3, 2
- Negative D-dimer excludes DVT with 99% negative predictive value 3, 2
- High-risk patients should proceed directly to compression duplex ultrasound without D-dimer testing 2
Clinical Efficiency and Safety
The combined Wells score plus D-dimer approach can safely reduce CT pulmonary angiography in approximately 30% of patients with suspected PE. 1 When using age-adjusted D-dimer cutoffs (age × 10 μg/L for patients >50 years), diagnostic accuracy improves further in older patients. 1
The Wells score performs comparably to gestalt clinical assessment, with both methods showing equivalent PE rates in low-risk groups (3% for both Wells and gestalt). 3
Important Limitations and Pitfalls
Subjective Components and Interrater Reliability
The Wells score contains subjective elements that affect reproducibility. 3, 1 Interrater agreement is:
- Moderate for DVT symptoms (κ=0.54), immobilization (κ=0.41), and "PE more likely than alternative diagnosis" (κ=0.5) 3
- Good to very good for hemoptysis (κ=0.76), previous DVT (κ=0.90), malignancy (κ=0.87), and tachycardia (κ=0.94) 3
Setting-Specific Performance
Critical caveat: The Wells score performs poorly in hospitalized inpatients. 2, 4 In the inpatient setting, the Wells score shows:
- Area under ROC curve of only 0.60 (barely better than chance) 4
- Failure rate of 5.9% in low-risk category (compared to <3% in outpatients) 4
- Efficiency of only 11.9% for excluding DVT 4
The Wells score was validated primarily in outpatient and emergency department populations and should not be relied upon for risk stratification in hospitalized patients. 5, 4
When D-dimer Testing Has Limited Utility
Bypass D-dimer testing and proceed directly to imaging in patients with: 3, 2
- Recent surgery or major trauma
- Active malignancy
- Hospitalization
- Other comorbid conditions likely to elevate D-dimer
Simplified vs. Original Wells Score
Both the original and simplified Wells rules demonstrate comparable performance when combined with age-adjusted D-dimer testing, with similar efficiency (30% vs 33%) and failure rates (0.8% vs 0.9%). 6 Given its ease of use, the simplified Wells rule is preferred over the original Wells rule in clinical practice. 6
Comparison with Alternative Approaches
The Wells score performs equivalently to gestalt clinical assessment in most comparative studies, though gestalt may have slightly better specificity in some populations. 3 The Pulmonary Embolism Rule-out Criteria (PERC) can be used in conjunction with Wells criteria for patients with very low clinical suspicion to potentially avoid unnecessary D-dimer testing. 2