About the Wells Score for PE
The Wells Score for Pulmonary Embolism is the most widely used clinical prediction rule for estimating pre-test probability of PE. Derived by Wells et al. in 1998 (Thromb Haemost) and validated across multiple large prospective cohorts including the Christopher study (van Belle et al., JAMA 2006, n=3,306), the score addresses a core clinical problem: PE presents with nonspecific symptoms — dyspnea, pleuritic chest pain, hemoptysis, hypoxia — that overlap with pneumonia, pneumothorax, myocardial infarction, and musculoskeletal causes. Ordering CTPA empirically for every suspected case is both high-radiation and diagnostically inefficient.
The Wells score stratifies patients into low (<2), moderate (2–6), or high (>6) probability tiers in the 3-level schema, and PE Unlikely (≤4) versus PE Likely (>4) in the 2-level dichotomized schema. The result directly determines the diagnostic pathway: low-probability patients are routed to high-sensitivity D-dimer first, and high-probability patients proceed directly to CTPA. Implementation studies show CTPA utilization drops approximately 20–30% without missing clinically significant PE.
CT pulmonary angiography is the definitive diagnostic test for PE, with sensitivity exceeding 95% for central and segmental emboli and acceptable specificity for subsegmental PE. When the Wells score routes a patient to CTPA, contrast-nephrotoxicity risk must also be assessed — clinicians should verify eGFR (via the eGFR Calculator) and maximum safe contrast volume via the 5×CrCl rule (via the CrCl Calculator) before ordering. PE accounts for an estimated 60,000–100,000 deaths annually in the United States, making prompt and accurate risk stratification one of the most common decisions in emergency medicine.
Interpretation
The Wells score maps to two useful probability schemas. The 3-level schema is descriptive; the 2-level schema is the one most algorithms reference for the D-dimer-to-CTPA decision.
3-Level Stratification
| Score | Category | PE Prevalence | Recommended Workup |
|---|---|---|---|
| <2 | Low Probability | ~2% | High-sensitivity D-dimer; if negative, PE excluded (NPV >99%) |
| 2–6 | Moderate Probability | ~17% | D-dimer preferred; if positive or unavailable, proceed to CTPA |
| >6 | High Probability | ~40–66% | CTPA directly (do not rely on D-dimer) |
2-Level (Dichotomized) Stratification
| Score | Category | PE Prevalence | Recommended Workup |
|---|---|---|---|
| ≤4 | PE Unlikely | ~10% | High-sensitivity D-dimer first; if negative, PE excluded |
| >4 | PE Likely | ~38% | CTPA directly (do not rely on D-dimer) |
Limitations
The Wells PE score has several important limitations. The criterion "PE is the most likely diagnosis" — worth 3 points — is inherently subjective and accounts for significant interobserver variability. Studies measuring interrater agreement for this item show kappa values of 0.5–0.7, reflecting moderate to substantial agreement. In centers with liberal application, this criterion inflates CTPA rates.
The score is not validated for pregnant patients, in whom PE risk is substantially elevated and D-dimer is physiologically elevated throughout pregnancy. The British Thoracic Society recommends ventilation-perfusion (V/Q) scintigraphy as the preferred initial imaging test in pregnancy to minimize fetal radiation exposure from CTPA.
In hemodynamically unstable patients with suspected massive PE (sustained hypotension, cardiac arrest, obstructive shock), bedside echocardiography and clinical gestalt should guide immediate management rather than formal scoring. The Wells score is designed for stable patients in whom a structured diagnostic algorithm can be safely completed.
The score does not account for the severity of PE once diagnosed — for risk stratification after diagnosis, tools such as the PESI or sPESI are used to guide decisions about outpatient treatment, ICU admission, or thrombolysis candidacy.