PERC Rule Calculator for Pulmonary Embolism

The problem with testing everyone for PE

Pulmonary embolism is the diagnosis emergency clinicians fear missing. A clot that has travelled to the lungs can kill quickly, and the symptoms, chest pain, breathlessness, a fast heart rate, are shared with dozens of benign conditions. Faced with that asymmetry, the natural instinct is to test widely: order a D-dimer, and if that is positive, order a CT pulmonary angiogram. The trouble is that testing has its own harms. D-dimer is positive in a large fraction of unwell patients for reasons that have nothing to do with clots, CT pulmonary angiography exposes patients to radiation and iodinated contrast, and both tests generate incidental findings that lead to further invasive workups. Testing a patient whose real probability of PE is 1% does not make them safer; it mostly manufactures false alarms.

Decision science gives this problem a formal answer called the test threshold: the probability below which the harms of testing outweigh the benefits. For pulmonary embolism, that threshold sits at roughly 2%. Below it, the right move is to stop, reassure, and look for another explanation for the symptoms. The difficulty has always been identifying, rigorously and defensibly, which patients are truly below that line. Clinical gestalt alone is inconsistent between clinicians, and D-dimer-first strategies still subject low-risk patients to blood tests and the cascade that follows a positive result. The PERC rule was designed to solve exactly this: a short list of criteria that, when all absent in an already low-risk patient, puts the PE probability below the test threshold with evidence to back it up.

What the PERC rule is

PERC stands for pulmonary embolism rule-out criteria. The rule was derived by Jeffrey Kline and colleagues and published in 2004 in the Journal of Thrombosis and Haemostasis (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). The team analysed emergency department patients evaluated for suspected PE and identified eight clinical features that, when all absent, were associated with a very low risk of venous thromboembolism. The rule was then tested prospectively in a multicentre study published in 2008 (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), which confirmed that PERC-negative patients had a risk of PE under 2%, below the test threshold.

The logic of the rule is deliberately strict. Unlike scores that add points, PERC is binary and unforgiving: a single present criterion makes the patient PERC positive, and the rule offers nothing further. There is no partial credit and no intermediate band. That strictness is the point. The rule is engineered for sensitivity, to miss as few true PEs as possible, and the price of that sensitivity is that many low-risk patients will be PERC positive and go on to D-dimer testing anyway. The reward is a well-defined group of patients in whom the workup can stop entirely, with no blood test and no scan.

The gate: low pretest probability comes first

The single most important and most often violated aspect of PERC is the gate that precedes it. The rule applies ONLY to patients whose pretest probability of PE is already low. In practice that means a Wells PE score of 4 or below, or an experienced clinician’s gestalt estimate under 15%. The eight criteria were derived and validated in that low-risk population, and their performance outside it is unknown and almost certainly worse. Applying PERC to a patient with a high Wells score, or to a patient in whom the clinician is genuinely worried about PE, is a misuse of the rule that the original authors explicitly warned against.

This gate is also what makes PERC different from a D-dimer. D-dimer can be applied across a wider range of pretest probabilities (though not to high-risk patients as a rule-out either), whereas PERC is a rule for a specific, pre-selected population: the low-risk patient in whom the clinician is considering PE but does not strongly suspect it. If the patient was never low risk, skip this calculator entirely and follow the standard diagnostic pathway of D-dimer and imaging. This calculator enforces the gate as its first question for exactly this reason.

The eight criteria explained

1. Age 50 years or older. The risk of venous thromboembolism rises with age, so patients under 50 start from a lower baseline. This is the commonest reason otherwise low-risk patients fail PERC, and it is not a judgment on the patient: it simply reflects epidemiology.
2. Heart rate 100 beats/min or more. Tachycardia is a physiological response to the strain PE places on the right heart, and it is one of the most consistent vital-sign abnormalities in confirmed PE. The threshold is 100/min at rest.
3. Oxygen saturation below 95% on room air. Hypoxaemia reflects ventilation-perfusion mismatch from clot obstructing pulmonary blood flow. The measurement must be on room air: a saturation of 98% on supplemental oxygen does not count as passing this criterion.
4. Unilateral leg swelling. A swollen leg suggests deep vein thrombosis, the usual source of pulmonary emboli. The swelling must be unilateral; bilateral swelling points toward systemic causes such as heart failure.
5. Haemoptysis. Coughing up blood suggests pulmonary infarction, the death of lung tissue downstream of an embolus. It is uncommon but highly specific when present.
6. Surgery or trauma requiring hospitalisation in the past 4 weeks. Recent surgery and major trauma are among the strongest provoking factors for venous thromboembolism, combining vessel injury, immobility, and a prothrombotic inflammatory response.
7. Prior DVT or PE. A previous venous thromboembolism marks a patient with a demonstrated tendency to clot, whether from an inherited thrombophilia, persistent risk factors, or simply bad luck that may recur.
8. Oestrogen use. Combined oral contraceptives, hormone replacement therapy, and other systemic oestrogen preparations increase clotting factor levels and VTE risk. The criterion refers to current use.

If every one of these is absent, the patient is PERC negative. If any single one is present, the patient is PERC positive. There is no weighting, no summing, and no clinical override: one present criterion ends the rule.

How to use this calculator

Begin with the gate question: is the pretest probability of PE low? Answer honestly. If the Wells PE score is above 4, or if your gestalt estimate is 15% or more, or if you are frankly worried about PE, answer No: the calculator will tell you PERC does not apply and direct you to the D-dimer and imaging pathway. Only answer Yes if the patient is genuinely low risk.

Then enter the age, heart rate, and room-air oxygen saturation, and answer the five Yes or No questions. Measure the vitals rather than estimating them, confirm the saturation is on room air, and take a proper history for the surgery, prior VTE, and oestrogen items. Every field is required: an incomplete PERC assessment cannot rule anything out. Press Calculate to see whether PERC applies and, if so, whether the patient is PERC negative or PERC positive, with the criteria that drove the result listed explicitly.

Interpreting the result: PERC negative

A PERC negative result means the pretest probability was low AND all eight criteria are absent. In the prospective multicentre validation, patients meeting these conditions had a risk of venous thromboembolism under 2% (Kline JA et al., J Thromb Haemost. 2008;6(5):772-780). That is below the test threshold, so the recommendation is to stop the PE workup: no D-dimer, no CT pulmonary angiogram. This is the entire purpose of the rule, and it is a strong recommendation, not a suggestion to consider testing anyway. Ordering a D-dimer after a PERC negative result defeats the rule and reintroduces exactly the harms it was designed to avoid.

Stopping the PE workup is not the same as stopping the clinical encounter. The patient still has symptoms that brought them in, and those symptoms still need an explanation: musculoskeletal chest pain, anxiety, asthma, pneumothorax, and acute coronary syndrome all remain on the table. Document the PERC assessment clearly, including the low pretest probability judgment and each criterion, explain to the patient why no further PE testing is needed, and give safety-net advice about what should bring them back.

Interpreting the result: PERC positive

A PERC positive result means at least one of the eight criteria is present. The rule has nothing more to say: it cannot quantify the resulting probability, and it must not be used to argue the patient is high risk. The correct next step is the standard diagnostic pathway, beginning with a highly sensitive D-dimer in line with local protocols, followed by CT pulmonary angiography or ventilation-perfusion scanning if the D-dimer is positive or if the clinical situation demands imaging directly.

Remember that most PERC positive patients do not have PE. The rule is tuned for sensitivity, so failing it is common and uninformative: a 52-year-old with a heart rate of 102 and no other features is PERC positive on two criteria, yet their actual PE probability may still be only a few percent. That is precisely why the D-dimer exists as the next step. Do not let a PERC positive result, on its own, push you toward anticoagulation or urgent imaging without the rest of the pathway.

Interpreting the result: PERC does not apply

If the gate question is answered No, the calculator reports that PERC does not apply. This is not a negative result and not a positive result; it is the rule declining to answer. The patient was never in the population the rule was built for, so none of its safety guarantees hold. Proceed directly to the standard pathway: clinical assessment, D-dimer where appropriate, and definitive imaging. Do not attempt to salvage the situation by answering the gate question differently; the gate is a clinical judgment, and the rule is only as safe as the honesty of that judgment.

Accuracy and evidence

The key performance figure is the miss rate in the 2008 prospective multicentre validation: PERC-negative patients had a rate of venous thromboembolism under 2% at follow-up. That figure is what places the rule below the test threshold and justifies withholding both D-dimer and imaging. The rule was derived to be highly sensitive, and subsequent studies and meta-analyses have broadly confirmed that PERC-negative low-risk patients have very low event rates, though the exact numbers vary with the population and the rigour of follow-up.

It helps to understand what the rule is optimised for. PERC is not designed to find PE; it is designed to identify a group in whom PE is rare enough to stop looking. Its specificity is poor by design, which is why a PERC positive result carries no diagnostic weight. The efficiency of the rule, the proportion of low-risk patients it classifies as PERC negative and therefore spares testing, varies between settings but is typically in the range of one in five to one in three low-risk patients. In a busy emergency department, sparing even a fifth of suspected-PE patients a D-dimer and a possible CT scan is a meaningful reduction in harm, cost, and crowding.

Limitations

• The gate is everything. PERC has no validated performance in patients who were not already low risk. Applying it to moderate or high-risk patients, or to patients the clinician is genuinely worried about, is unsafe and unsupported.
• Not validated in all populations. The derivation and validation cohorts were emergency department outpatients. Hospitalised patients, pregnant patients, and patients already on anticoagulation were not the study population, and the rule should not be applied to them without specific evidence and senior input.
• Oxygen saturation must be on room air. Assessing this criterion while the patient receives supplemental oxygen invalidates it. If the patient needs oxygen, they are unlikely to be an appropriate PERC candidate in the first place.
• It does not diagnose PE. PERC positive means only that the rule cannot rule PE out. It says nothing about how likely PE is, and it must never be used to justify treatment or urgent imaging on its own.
• Gestalt is subjective. The low-risk gate often rests on clinician gestalt, which varies between clinicians. When in doubt about whether the patient is truly low risk, the safer course is to treat them as not low risk and follow the standard pathway.
• No score replaces clinical judgment. An unstable patient, a patient with signs of right heart strain, or a deteriorating patient needs urgent senior assessment and definitive imaging whatever any rule says.