- First, estimate the patient's pre-test probability of pulmonary embolism (PE) using clinical gestalt. PERC is validated only when the pre-test probability is low (< 15%).
- Toggle the "Clinician's pre-test probability for PE" input to Low (< 15%) or Not low.
- Check each of the eight criteria. Leave the switch OFF (0) if the criterion is absent; turn it ON (+1) if present.
- If pre-test probability is low AND all eight criteria are absent (score = 0), PE is ruled out and no further workup (D-dimer, computed tomography pulmonary angiogram or CTPA, ventilation–perfusion or V/Q scan) is required.
- If pre-test probability is not low, OR if any criterion is present, PERC does not apply. Move on to age-adjusted D-dimer or imaging guided by Wells' or Geneva scores.
All computation runs in your browser; no values are stored or transmitted.
When to Use
Apply PERC in the emergency department or outpatient setting to a patient with chest pain, dyspnea, syncope, or tachycardia whose clinician-estimated pre-test probability of PE is already low (< 15% by unstructured gestalt). PERC is designed to spare low-risk patients the harms of unnecessary D-dimer testing and imaging — the small but real risks of contrast-induced acute kidney injury (AKI), radiation, incidental findings, and anticoagulation for a false-positive test.
Appropriate population
Adult emergency department (ED) or ambulatory patients with symptoms suggestive of PE in whom the treating clinician has first judged the probability of PE to be low by clinical gestalt. The PROPER trial confirmed that in this population, a PERC-negative result allows safe discharge without further testing (non-inferior thromboembolic event rate at 3 months vs. standard workup).
When NOT to apply it
Do not use PERC if pre-test probability is intermediate or high, in pregnant patients, in patients already on therapeutic anticoagulation (which lowers D-dimer and biases the rule), in patients with known thrombophilia or active cancer, or in hospitalized patients (the derivation and validation were in ED populations). In a high-prevalence population (> ~15% pre-test probability), PERC's sensitivity is not sufficient to safely rule out PE.
Pearls & Pitfalls
Two gates: gestalt first, then PERC
PERC is not a stand-alone rule. The first gate is the clinician's low-probability gestalt (< 15%); only then is PERC applied to formally rule out PE without any further testing. Applying PERC to intermediate-risk or high-risk patients will miss real PEs — its 97% sensitivity translates to an acceptably low miss rate only when the pre-test probability is already very low.
A mnemonic for the eight items
"HAD CLOTS": Hormones (exogenous estrogen, including oral contraceptive pill or OCP and hormone replacement therapy or HRT); Age ≥ 50; DVT or PE history; Coughing blood (hemoptysis); Leg swelling (unilateral); Oxygen (SpO₂ < 95% on room air); Tachycardia (heart rate ≥ 100 beats per minute or bpm); Surgery or trauma within 4 weeks requiring general anesthesia. Every item is scored 0 or 1; the only "rule-out" output is a total of 0 in a low pre-test probability patient.
Pitfalls
(1) Pulse oximetry on supplemental oxygen is not valid for PERC — the SpO₂ < 95% criterion requires a room-air reading. (2) The unilateral leg swelling criterion means visibly asymmetric calf diameter on exam, not patient-reported swelling. (3) PERC does not account for BMI, immobility beyond "recent surgery requiring general anesthesia", or active cancer — all patient-specific risk factors the clinician must weigh when setting pre-test probability. (4) Chest pain that is reproducible on palpation, pleuritic, or clearly musculoskeletal does not on its own make a patient PERC-eligible; the gestalt assessment comes first.
Why Use It
Pulmonary embolism is overdiagnosed and overworked-up in the emergency department, where chest pain and dyspnea are extremely common and PE prevalence in the overall symptomatic population is only ~5–15%. D-dimer testing has a high false-positive rate in ED patients, and the reflex to image everyone with a positive D-dimer drives contrast load, radiation exposure, and incidental findings without changing outcomes in truly low-risk patients. PERC gives a validated structure to the clinical decision not to test: in a population whose pre-test probability is already low, an all-negative eight-item PERC supports discharge without further workup, with a residual PE prevalence below the test threshold at which anticoagulation's harms outweigh its benefits (roughly 1.8%). The PROPER randomized trial confirmed that this approach is non-inferior to systematic D-dimer testing for 3-month thromboembolic outcomes while reducing hospital stay, chest imaging, and ED length of stay.
PERC Rule for Pulmonary Embolism
Set the clinician's pre-test probability, then toggle each of the eight PERC criteria. The rule-out verdict appears below.
⚕ Kline JA et al. J Thromb Haemost. 2004;2(8):1247–1255. PERC is validated only in patients whose pre-test probability of PE is already low (< 15%) by clinical gestalt. A PERC score of 0 in a low-risk patient supports ruling out PE without D-dimer or imaging (sensitivity ~97%); a positive score (≥ 1) means PERC does not rule out PE — proceed to D-dimer or imaging guided by Wells' or Geneva scores. For licensed clinicians; not a substitute for individualized clinical judgment.
Next Steps
Interpret the PERC result together with the clinician's pre-test probability.
- Low pre-test probability AND PERC = 0: PE is ruled out. No D-dimer, no imaging, no anticoagulation. Discharge home with reassurance and clear return precautions.
- Low pre-test probability AND PERC ≥ 1: PERC does not rule out PE. Proceed to age-adjusted D-dimer; if positive, go to computed tomography pulmonary angiography (CTPA) or a ventilation–perfusion (V/Q) scan when iodinated contrast is contraindicated.
- Intermediate or high pre-test probability: PERC is not applicable. Use the Wells' or Geneva scores plus D-dimer or imaging as indicated.
- When PE is confirmed, use the PESI or simplified PESI to stratify 30-day mortality risk and identify low-risk patients who may be candidates for outpatient management.
Evidence & References
The Eight PERC Criteria
| Criterion | Points |
|---|---|
| Age ≥ 50 years | +1 |
| Heart rate ≥ 100 bpm | +1 |
| Pulse oximetry < 95% on room air | +1 |
| Prior DVT or PE | +1 |
| Recent trauma or surgery within 4 weeks requiring general anesthesia | +1 |
| Hemoptysis | +1 |
| Exogenous estrogen use (OCP, HRT) | +1 |
| Unilateral leg swelling (visibly asymmetric) | +1 |
Interpretation
| Pre-test probability | PERC score | Action |
|---|---|---|
| Low (< 15% gestalt) | 0 | PE ruled out; no further testing (residual PE < 1.8%) |
| Low (< 15% gestalt) | ≥ 1 | PERC does not apply; proceed to D-dimer or imaging |
| Intermediate or high | Any | PERC does not apply; follow Wells'/Geneva + D-dimer or imaging |
Kline and colleagues derived PERC from a multicenter cohort of 3,148 ED patients, then prospectively validated an eight-rule set in 8,138 patients with pooled sensitivity 97.4% (95% CI 95.8–98.5%) and specificity 21.9%; the PROPER randomized trial confirmed non-inferiority of a PERC-based rule-out strategy vs. systematic D-dimer in low-risk ED patients.
