Wells Score for Pulmonary Embolism
Estimates the clinical pre-test probability of pulmonary embolism using 7 bedside criteria. Includes the original 3-tier model and the simplified 2-tier “PE unlikely / likely” interpretation to guide D-dimer, CTPA, and V/Q scan decisions.
Calculate Wells Score for PE
Select the clinical features present in your patient. The score sums weighted points from 7 criteria (range 0–12.5). Results display both the traditional 3-tier and the simplified 2-tier interpretation. For patients with a very low pre-test probability, PERC rule applicability is also assessed.
The Wells score was developed for ambulatory patients presenting with suspected PE. It has limited validation in hospitalised inpatients, pregnant patients, and patients already receiving anticoagulation. The “PE is #1 diagnosis” criterion is inherently subjective — it reflects the clinician’s overall gestalt and is the most powerful single discriminator in the model.
Understanding the Wells Score for PE
Pulmonary embolism is a potentially fatal condition that is notoriously difficult to diagnose clinically. Classic presentations (sudden dyspnoea, pleuritic chest pain, haemoptysis) are present in a minority of cases, and symptoms overlap significantly with other cardiopulmonary conditions. The Wells score, published by Philip Wells and colleagues in 2000, was developed to standardise pre-test probability assessment and reduce both missed PE diagnoses and unnecessary imaging.
The model uses 7 weighted clinical criteria: two high-weight items (clinical DVT signs and “PE most likely diagnosis,” each +3 points), three moderate-weight items (tachycardia, immobilisation/surgery, and prior VTE, each +1.5 points), and two lower-weight items (haemoptysis and malignancy, each +1 point). The total ranges from 0 to 12.5.
Weighted Scoring System
+3 points:
• Clinical signs/symptoms of DVT
• PE is #1 diagnosis or equally likely
+1.5 points:
• Heart rate >100 bpm
• Immobilisation ≥3 d / surgery ≤4 wk
• Previous DVT or PE
+1 point:
• Haemoptysis
• Active malignancy
Worked Example
A 58-year-old woman presents with acute-onset dyspnoea 10 days after knee arthroplasty. HR 108 bpm. No leg swelling. No haemoptysis. No cancer history. No prior VTE. PE considered most likely diagnosis.
DVT signs = 0
PE most likely = +3
HR >100 = +1.5
Surgery ≤4 wk = +1.5
Previous VTE = 0
Haemoptysis = 0
Malignancy = 0
Total = 6 → Moderate (3-tier) / PE Likely (2-tier)
Key distinction — Wells vs. Geneva: The Wells score includes a subjective criterion (“PE most likely diagnosis”), which is its most powerful predictor but also its main limitation. The revised Geneva score uses entirely objective criteria and may be preferred in settings where inter-rater reproducibility is critical (e.g., research protocols). In clinical practice, both scores perform similarly, and most guidelines accept either.
Score Interpretation & Risk Stratification
Traditional 3-Tier Model (Wells 2000)
| Score | Pre-test Probability | PE Prevalence | Recommended Action |
|---|---|---|---|
| 0–1 | Low | ~1.3% | Consider PERC rule; if PERC-negative, no testing needed. If PERC-positive, D-dimer |
| 2–6 | Moderate | ~16.2% | D-dimer → if negative, PE excluded; if positive, CTPA or V/Q scan |
| ≥7 | High | ~40.6% | CTPA (or V/Q scan) directly — D-dimer insufficient to exclude at this probability |
Simplified 2-Tier Model (Wells 2001)
| Score | Category | PE Prevalence | Recommended Action |
|---|---|---|---|
| ≤4 | PE Unlikely | ~8% | D-dimer → if negative, PE excluded; if positive, CTPA or V/Q scan |
| >4 | PE Likely | ~34% | CTPA (or V/Q scan) directly as first-line investigation |
For patients with a low pre-test probability (Wells ≤1 on the 3-tier model or gestalt <15%), the Pulmonary Embolism Rule-Out Criteria (PERC) can eliminate the need for any testing — including D-dimer. If all 8 PERC criteria are met, PE can be excluded without further workup (miss rate <2%).
PERC criteria — ALL must be true to rule out: Age <50, HR <100, SpO₂ >94% on room air, no unilateral leg swelling, no haemoptysis, no recent surgery/trauma (≤4 weeks), no prior DVT/PE, no oestrogen use (OCP or HRT).
Important: PERC should only be applied when the clinician’s gestalt already places PE as a low-probability diagnosis. It is a “rule-out” for patients you already think are unlikely to have PE — not a screening tool for the general population.
Differential Diagnosis of Acute Dyspnoea & Chest Pain
The “PE is #1 diagnosis” criterion requires the clinician to consider alternative diagnoses systematically. PE symptoms — dyspnoea, chest pain, tachycardia, hypoxia — are shared by many other conditions. A structured differential approach reduces both over-testing and missed diagnoses.
ACS presents with chest pain (typically pressure-like, substernal, radiating to jaw or arm), dyspnoea, diaphoresis, and nausea. Unlike the pleuritic, sharp, lateralised pain of PE, ACS pain is often described as a tightness or heaviness that may be provoked by exertion. ECG changes (ST-segment deviation, T-wave inversion, new Q waves) and elevated troponin support ACS, though troponin may also be mildly elevated in massive PE due to right ventricular strain.
Key distinguishing features: ACS typically does not cause hypoxia unless complicated by pulmonary oedema or cardiogenic shock. Pleuritic pain that worsens with inspiration is more suggestive of PE, pneumothorax, or pleuritis than ACS. Both conditions can cause tachycardia and ECG changes — the classic S1Q3T3 pattern in PE is neither sensitive nor specific.
Community-acquired pneumonia presents with cough (productive), fever, pleuritic chest pain, dyspnoea, and crackles on auscultation. A productive cough with purulent sputum and fever favour pneumonia, while PE more often presents with acute-onset dyspnoea, clear lungs, and absent or low-grade fever. However, peripheral PE can cause pulmonary infarction with consolidation mimicking pneumonia radiographically.
Chest X-ray may show lobar consolidation (pneumonia) or wedge-shaped peripheral opacity (Hampton’s hump in PE), but normal CXR does not exclude either. Inflammatory markers (CRP, procalcitonin) are typically more elevated in pneumonia. If clinical doubt persists, CTPA can diagnose PE while also identifying alternative pathology such as pneumonia, empyema, or effusion.
Spontaneous pneumothorax presents with sudden-onset pleuritic chest pain and dyspnoea, closely mimicking PE. The classic clinical signs — reduced breath sounds and hyper-resonance on the affected side — may be subtle in small pneumothoraces. Tall, thin young men (primary spontaneous) and patients with COPD or connective tissue disorders (secondary) are at higher risk.
Chest X-ray (erect, in expiration) is usually diagnostic. In the emergency setting, point-of-care lung ultrasound showing absent lung sliding can rapidly identify pneumothorax before CXR. Tension pneumothorax (haemodynamic compromise, tracheal deviation) requires immediate needle decompression — do not delay for imaging.
Decompensated heart failure causes dyspnoea (particularly orthopnoea and paroxysmal nocturnal dyspnoea), bilateral lower limb oedema, elevated JVP, and bilateral crackles. Unlike PE, heart failure dyspnoea is typically bilateral, progressive over days, and worsened by lying flat. However, acute PE can precipitate acute right heart failure with similar haemodynamic features.
BNP/NT-proBNP is elevated in heart failure but may also rise in massive PE due to right ventricular strain. CXR showing cardiomegaly, bilateral pleural effusions, and upper lobe venous diversion supports heart failure. Echocardiography is key — right ventricular dilation and dysfunction suggest PE, while biventricular failure with reduced LVEF suggests decompensated cardiomyopathy.
Panic attacks are the most common diagnosis in young patients investigated for PE who are ultimately found to be negative. Symptoms include sudden-onset dyspnoea, chest tightness, palpitations, tingling, and a sense of impending doom — closely overlapping with PE. A history of recurrent similar episodes, absence of risk factors, and normal examination findings suggest panic disorder.
Caution: Anxiety and hyperventilation should be diagnoses of exclusion, particularly in patients with VTE risk factors. Studies show that PE is found in approximately 3–5% of patients initially diagnosed with panic attacks in the emergency department. If the clinician has any uncertainty about PE, objective testing should be performed before attributing symptoms to anxiety.
Aortic dissection can mimic PE with acute-onset chest or back pain, dyspnoea, and haemodynamic instability. Classic “tearing” pain radiating to the back and unequal blood pressures in the arms suggest dissection. Type A dissection (ascending aorta) can cause aortic regurgitation, pericardial tamponade, and coronary occlusion. Type B (descending) may present with abdominal pain and lower-limb ischaemia.
CTPA performed for suspected PE may incidentally identify aortic dissection — the CT should always be reviewed for aortic pathology even when PE is the primary indication. If dissection is strongly suspected, a dedicated CT aortogram with arterial-phase timing is preferred over a standard CTPA protocol.
When scoring the “PE is #1 diagnosis” criterion, systematically consider: (1) Could this be cardiac? (ACS, heart failure, pericarditis), (2) Could this be pulmonary? (pneumonia, asthma exacerbation, pneumothorax), (3) Could this be musculoskeletal? (costochondritis, rib fracture), (4) Could this be functional? (anxiety, hyperventilation). If PE remains equally or more likely than all alternatives, score +3.
Special Populations & Considerations
The Wells score was validated primarily in ambulatory outpatients. Its performance and the subsequent diagnostic pathway may differ in several important clinical scenarios.
Contrast allergy and renal impairment: When CTPA is contraindicated (iodinated contrast allergy, severe renal impairment with eGFR <30), V/Q scintigraphy is the preferred alternative imaging modality. A normal perfusion scan effectively excludes PE. An intermediate or indeterminate V/Q result may require further evaluation with lower limb compression ultrasound, repeat imaging, or clinical observation with serial assessment.
Systematic Diagnostic Approach
The following step-by-step pathway integrates the Wells score with PERC, D-dimer, and imaging in the evaluation of suspected PE. This approach follows ESC, ACEP, BTS, and ACCP guideline recommendations.
Before applying any clinical decision rule, determine if the patient is haemodynamically stable. If the patient has sustained systolic BP <90 mmHg, obstructive shock, or cardiac arrest with suspected PE, this is a high-risk (massive) PE pathway — skip the Wells score and proceed to emergency CTPA or bedside echocardiography. Empiric systemic thrombolysis should be considered if there is haemodynamic collapse and PE is suspected.
If the patient is haemodynamically stable, proceed to Step 2.
For haemodynamically stable patients with suspected PE, calculate the Wells score using all 7 criteria. Pay particular attention to the “PE is #1 diagnosis” criterion — this requires a conscious, documented consideration of alternative diagnoses. The score should be calculated before ordering any laboratory or imaging investigations.
Classify the result using the 2-tier model: PE Unlikely (≤4) or PE Likely (>4). Optionally, use the 3-tier model for finer stratification: Low (0–1), Moderate (2–6), High (≥7).
If the Wells score is ≤1 (low probability on the 3-tier model) and your clinical gestalt places the pre-test probability below ~15%, apply the PERC rule. If all 8 PERC criteria are met (age <50, HR <100, SpO₂ >94%, no leg swelling, no haemoptysis, no recent surgery/trauma, no prior VTE, no oestrogen use), PE can be excluded without D-dimer or imaging. The miss rate is <2%, which is below the accepted test threshold.
If any PERC criterion fails, proceed to D-dimer testing. PERC should not be applied to moderate- or high-probability patients (Wells ≥2).
For patients with Wells ≤4 (“PE unlikely”) who fail PERC or are in the moderate range (Wells 2–6 on the 3-tier model), a high-sensitivity D-dimer test is the next step. Use age-adjusted thresholds for patients over 50 (age × 10 µg/L).
If D-dimer is negative: PE is safely excluded — the 3-month VTE rate is <1%. No imaging is needed. If D-dimer is positive: proceed to CTPA (or V/Q scan if CTPA is contraindicated). A positive D-dimer does not diagnose PE — it indicates that imaging is required. D-dimer has many causes of false-positive elevation, including infection, surgery, malignancy, pregnancy, age, and hospitalisation.
For patients with Wells >4 (“PE likely”) or Wells ≥7 (high probability), CTPA should be performed as the first investigation. D-dimer alone cannot safely exclude PE at this pre-test probability. CTPA has a sensitivity of ~95–98% and specificity of ~97% for PE and also identifies alternative diagnoses (pneumonia, aortic dissection, pneumothorax).
If CTPA is negative in a “PE likely” patient, consider: (a) lower limb compression ultrasound to detect DVT that would indicate VTE regardless of PE, (b) V/Q scan if CTPA quality was suboptimal, or (c) clinical reassessment — a negative CTPA in the context of high clinical suspicion may warrant specialist review, repeat imaging, or short-term anticoagulation with follow-up.
Once PE is confirmed on CTPA, risk-stratify using haemodynamic status, RV function (echocardiography or CT RV/LV ratio), and cardiac biomarkers (troponin, BNP/NT-proBNP). The ESC classification divides PE into high-risk (massive), intermediate-high-risk (submassive with RV dysfunction + elevated biomarkers), intermediate-low-risk, and low-risk categories.
Treatment: anticoagulation should begin immediately. For most haemodynamically stable patients, a DOAC (apixaban or rivaroxaban) is the first-line choice. For high-risk PE with haemodynamic instability, systemic thrombolysis (alteplase) is the standard treatment, with catheter-directed therapy or surgical embolectomy as alternatives. Duration of anticoagulation depends on whether the PE was provoked (3 months) or unprovoked (extended/indefinite).
Common Pitfalls & Limitations
Despite being one of the most validated clinical decision rules in emergency medicine, the Wells PE score is frequently misapplied. The following pitfalls represent the most common sources of diagnostic error.
The +3 point criterion for “PE is #1 diagnosis or equally likely” is by far the most powerful and most controversial element of the Wells score. It effectively gives the clinician’s gestalt a quantitative weight — which is both its strength (incorporating pattern recognition) and its weakness (introducing subjectivity and inter-rater variability).
Studies show that the same patient may be scored 0 or 3 on this criterion by different clinicians, depending on their experience, cognitive biases, and risk tolerance. Junior clinicians may be more likely to assign +3 out of defensive practice, while experienced clinicians may anchor prematurely on an alternative diagnosis. Best practice: explicitly document your reasoning for or against assigning this criterion, and list the specific alternative diagnoses you have considered.
This is a critical patient safety error. When the Wells score is >4 (“PE likely”), the pre-test probability is approximately 34%. At this prevalence, even a high-sensitivity D-dimer has a clinically unacceptable false-negative rate — meaning a negative D-dimer could falsely reassure the clinician while the patient has PE. These patients require imaging (CTPA or V/Q scan) regardless of D-dimer.
This error commonly occurs when D-dimer is ordered reflexively before the Wells score is calculated — and the negative result is then used to discharge the patient without imaging. Always calculate Wells before ordering tests so results are interpreted in the correct clinical framework.
The PERC rule was validated exclusively for patients with a low pre-test probability (<15% clinical gestalt, roughly corresponding to Wells 0–1 on the 3-tier model). Applying PERC to moderate-risk patients (Wells 2–6) is not validated and can miss PE. The 2% miss rate that makes PERC acceptable in low-risk patients becomes 5–8% in moderate-risk patients — above the accepted safety threshold.
A common error: a patient with Wells 4 (moderate/PE unlikely on the 2-tier model) meets all PERC criteria. The clinician incorrectly concludes PE is excluded. But PERC should not have been applied at this pre-test probability — D-dimer was required, and its result determines whether imaging is needed.
The standard D-dimer cut-off of 500 µg/L produces a false-positive rate exceeding 60% in patients over 80 years, leading to unnecessary CTPA with its associated radiation exposure, contrast-related risks, and incidental findings. Age-adjusted D-dimer thresholds (age × 10 µg/L for patients >50) were validated in the prospective ADJUST-PE study involving over 3,000 patients and reduced unnecessary CTPA by approximately 12–30% in elderly patients without increasing the PE miss rate.
Example: in a 70-year-old with Wells ≤4, a D-dimer of 650 µg/L is negative by the age-adjusted threshold (70 × 10 = 700 µg/L) but positive by the fixed 500 µg/L threshold. This single change in practice can prevent a significant number of unnecessary CT scans, incidental findings, and downstream investigations in elderly patients.
Modern multi-detector CTPA has increased the detection of small, subsegmental PE (isolated clots in fifth-order pulmonary arteries or beyond). The clinical significance of many subsegmental PEs is debated — some may represent false-positive findings, in-situ thrombosis, or clinically insignificant emboli. Studies suggest that subsegmental PE without concomitant DVT has a very low risk of recurrent VTE if left untreated.
Current ESC guidelines suggest that in patients with isolated subsegmental PE, negative bilateral leg ultrasound, and no additional risk factors for recurrence, clinical surveillance without anticoagulation may be reasonable. However, this decision should involve specialist input and shared decision-making with the patient. The Wells score cannot differentiate between clinically significant and subsegmental PE — it assesses the probability of any PE.
Quick Reference Summary
| 2-Tier Score | Category | Next Step | If Negative |
|---|---|---|---|
| 0–1 + PERC- | Very Low | No testing needed | PE excluded (miss rate <2%) |
| ≤4 | PE Unlikely | D-dimer (age-adjusted if >50) | PE excluded (3-month VTE <1%) |
| >4 | PE Likely | CTPA (or V/Q scan) | Reassess clinically; consider leg US, serial imaging |
The Golden Rule: Calculate Wells before any testing. Score ≤4? D-dimer first (use age-adjusted thresholds in patients >50). Score >4? CTPA first — never rely on D-dimer alone. Score 0–1 with low gestalt? Apply PERC — if all 8 criteria met, no testing needed. Haemodynamically unstable? Skip scores — go directly to CTPA or bedside echo.
Disclaimer & References
For Educational Purposes Only. This calculator and the accompanying clinical information are intended as educational tools for healthcare professionals. They do not replace clinical judgement. Results should be interpreted in the full clinical context. Lab reference ranges vary by institution — verify with your own laboratory. Drug dosages should be confirmed against current prescribing information.
References
- Wells PS, Anderson DR, Rodger M, et al. Derivation of a simple clinical model to categorize patients probability of pulmonary embolism: increasing the models utility with the SimpliRED D-dimer. Thromb Haemost. 2000;83(3):416–420. DOI: 10.1055/s-0037-1613830
- Wells PS, Anderson DR, Rodger M, et al. Excluding pulmonary embolism at the bedside without diagnostic imaging: management of patients with suspected pulmonary embolism presenting to the emergency department by using a simple clinical model and D-dimer. Ann Intern Med. 2001;135(2):98–107. DOI: 10.7326/0003-4819-135-2-200107170-00010
- Kline JA, Mitchell AM, Kabrhel C, Richman PB, Courtney DM. Clinical criteria to prevent unnecessary diagnostic testing in emergency department patients with suspected pulmonary embolism. J Thromb Haemost. 2004;2(8):1247–1255. DOI: 10.1111/j.1538-7836.2004.00790.x
- Kline JA, Courtney DM, Kabrhel C, et al. Prospective multicenter evaluation of the pulmonary embolism rule-out criteria. J Thromb Haemost. 2008;6(5):772–780. DOI: 10.1111/j.1538-7836.2008.02944.x
- Righini M, Van Es J, Den Exter PL, et al. Age-adjusted D-dimer cutoff levels to rule out pulmonary embolism: the ADJUST-PE study. JAMA. 2014;311(11):1117–1124. DOI: 10.1001/jama.2014.2135
- Konstantinides SV, Meyer G, Becattini C, et al. 2019 ESC Guidelines for the diagnosis and management of acute pulmonary embolism developed in collaboration with the European Respiratory Society (ERS). Eur Heart J. 2020;41(4):543–603. DOI: 10.1093/eurheartj/ehz405
- Howard LSGE, Barden S, Condliffe R, et al. British Thoracic Society Guideline for the initial outpatient management of pulmonary embolism (PE). Thorax. 2018;73(Suppl 2):ii1–ii29. DOI: 10.1136/thoraxjnl-2018-211539
- van der Hulle T, Cheung WY, Kooij S, et al. Simplified diagnostic management of suspected pulmonary embolism (the YEARS study): a prospective, multicentre, cohort study. Lancet. 2017;390(10091):289–297. DOI: 10.1016/S0140-6736(17)30885-1
- Kearon C, de Wit K, Parpia S, et al. Diagnosis of pulmonary embolism with D-dimer adjusted to clinical probability. N Engl J Med. 2019;381(22):2125–2134. DOI: 10.1056/NEJMoa1909159
- Raja AS, Greenberg JO, Qaseem A, et al. Evaluation of patients with suspected acute pulmonary embolism: best practice advice from the Clinical Guidelines Committee of the American College of Physicians. Ann Intern Med. 2015;163(9):701–711. DOI: 10.7326/M14-1772