Scoring Systems for Pulmonary Embolism Assessment
The Wells score and the revised Geneva score are the two primary validated clinical decision rules for assessing pretest probability of pulmonary embolism, with both demonstrating comparable performance when combined with D-dimer testing to safely exclude PE. 1
Primary Validated Scoring Systems
Wells Score
The Wells score is a 7-variable risk stratification model that can be interpreted using either a three-level or two-level classification system 1:
Three-level interpretation:
- Low risk (0-1 points): 3.6% PE prevalence 1
- Moderate risk (2-6 points): 20.5% PE prevalence 1
- High risk (>6 points): 66.7% PE prevalence 1
Two-level interpretation (dichotomized):
- PE unlikely (0-4 points): 7.8% PE prevalence - safe for D-dimer exclusion 1
- PE likely (>4 points): 40.7% PE prevalence 1
The Wells score achieved a negative predictive value of 99.5% when combined with D-dimer testing in low-probability patients 1. A major limitation is the subjective variable "alternative diagnosis less likely than PE," which is worth 3 points and represents physician judgment override 1.
Revised Geneva Score
The revised Geneva score is a fully objective clinical decision rule using 8 parameters based solely on risk factors, symptoms, and clinical signs without requiring diagnostic testing 1:
Three-level interpretation:
- Low risk (0-3 points): 7.9% PE prevalence 1
- Intermediate risk (4-10 points): 28.5% PE prevalence 1
- High risk (≥11 points): 73.7% PE prevalence 1
Simplified Revised Geneva Score assigns 1 point to each parameter (except heart rate ≥95 bpm which receives 2 points total) 1:
The 2019 ESC guidelines confirm that both Wells and Geneva scores demonstrate similar diagnostic performance, with approximately 12% PE prevalence in the "unlikely" category and 30% in the "likely" category when using two-level classification 1.
Comparative Performance
The Wells score demonstrated superior discriminatory performance compared to the simplified revised Geneva score in direct comparison studies, with area under the curve of 0.85 versus 0.76 (p=0.005) 2. However, the simplified Wells rule and original Wells rule show comparable performance when combined with age-adjusted D-dimer testing, with similar efficiency (30% vs 33%) and failure rates (0.8% vs 0.9%) 3.
Additional Clinical Tools
PERC Rule (Pulmonary Embolism Rule-out Criteria)
The PERC rule comprises 8 clinical variables designed to identify patients whose PE likelihood is so low that diagnostic workup should not be initiated 1:
- Age <50 years
- Pulse <100 bpm
- SaO₂ >94%
- No unilateral leg swelling
- No hemoptysis
- No recent trauma or surgery
- No history of VTE
- No oral hormone use
All criteria must be met in patients with low clinical probability to safely exclude PE without further testing 1. However, the low overall PE prevalence in validation studies limits generalizability 1.
Kline (Charlotte) Criteria
The Kline rule creates a binary partition identifying patients safe for D-dimer testing 1. Unsafe patients (requiring imaging regardless of D-dimer) have either shock index >1.0 OR age >50 years PLUS any of: unexplained hypoxemia (SaO₂ <95%), unilateral leg swelling, recent major surgery, or hemoptysis 1. This identified 21% as "unsafe" with 42.1% PE prevalence, while 79% were "safe" with 13.3% PE prevalence 1.
Pisa Model
The Pisa model uses regression coefficients for multiple clinical and ECG variables, with the highest weights for sudden-onset dyspnea (coefficient 2.00) and acute cor pulmonale on ECG (coefficient 1.96) 1. In validation, PE prevalence was 2% when predicted probability was low (0-10%), 28% when moderate (11-50%), 67% when substantial (51-80%), and 94% when high (81-100%) 1. The Pisa model showed superior area under ROC curve (0.94) compared to Wells (0.75) and Geneva (0.54) in one small study, though this had significant patient selection bias with 43% PE prevalence 1.
Clinical Application Algorithm
For hemodynamically stable patients with suspected PE 1:
Calculate Wells or Geneva score - both are acceptable, though Wells may perform slightly better in high-risk patients 1, 2
Low/intermediate probability with negative D-dimer: PE excluded, no imaging needed 1
Low/intermediate probability with positive D-dimer: Proceed to CT pulmonary angiography 1
High probability: Proceed directly to CT pulmonary angiography regardless of D-dimer 1
Critical caveat: The Wells score showed 100% sensitivity and negative predictive value when combined with age-adjusted D-dimer in hospitalized patients at risk for DVT 4, but inpatients have higher PE prevalence (36%) than emergency department populations, and the score alone is insufficient to rule out PE in this population 5.