Wells Score Calculator for Pulmonary Embolism (PE)

Background.

The Wells score for pulmonary embolism is a bedside pretest-probability aid for clinicians assessing a patient with suspected PE. It shares an author with the Wells DVT rule and nothing else — different items, different weights, different cut-points. Do not use one score's bands to read the other's total. This page is a clinical decision aid, not a diagnosis, and it does not replace clinical judgement or a local diagnostic pathway.

Seven findings carry a weight. Clinical signs and symptoms of DVT, and the judgement that no alternative diagnosis is more likely than PE, each score three points. A heart rate over 100, immobilisation or surgery in the previous four weeks, and a previous objectively diagnosed DVT or PE each score one and a half. Haemoptysis and malignancy each score one. Nothing subtracts, so the total can only rise as findings accumulate; the minimum is 0 and the maximum is 12.5.

Three published readings of that total exist, and this calculator reports all three rather than picking one silently. The two-level scheme splits at 4: a total above 4 is "PE likely", a total of 4 or below is "PE unlikely". That is the scheme most modern pathways are built around, because "PE unlikely" plus a negative D-dimer is what allows imaging to be withheld. The three-level scheme reports low, moderate or high probability. The simplified rule of Gibson and colleagues gives one point to each of the same seven variables and splits at 1, and was externally validated by Douma and colleagues, who found the proportion of patients classified "unlikely" was 70 percent with the simplified rule against 78 percent with the original, at PE prevalences of 12 and 13 percent respectively.

The three-level band edges deserve a warning, because reproductions of them disagree. The derivation paper defines them as below 2 for low, 2.0 to 6.0 for moderate, and over 6 for high. Some guideline tables render the same bands as 0 to 1, 2 to 6 and 7 or more; at least one published reproduction renders them as 0 to 2, 3 to 6 and over 6. Since every Wells PE total is a multiple of 0.5, those renderings are not equivalent — a total of 1.5, 2.5 or 6.5 lands differently under each. This calculator implements the derivation paper's own inequalities, which are the only rendering with no gaps, and says so beside the result.

The score does not rule PE in or out. A "PE unlikely" result excludes PE only in combination with a negative D-dimer; alone it excludes nothing. A "PE likely" result selects imaging, usually CT pulmonary angiography, and does not confirm a clot. In the ESC 2019 guideline's summary of the literature, roughly 10 percent of patients in the low three-level category, 30 percent in the intermediate and 65 percent in the high category had PE confirmed, with about 12 percent in the two-level "unlikely" group and 30 percent in the "likely" group. Those are cohort proportions, not a probability for one person, and they shift with the population's baseline prevalence.

Two scope limits matter more than the arithmetic. The rule was derived in patients presenting with suspected PE who were stable enough to be worked up: it is not a triage instrument for a haemodynamically unstable patient, who needs immediate assessment rather than a score. And it is not validated in pregnancy, where dedicated pathways exist. One item — "no alternative diagnosis is more likely than PE" — is an explicit clinician judgement worth three points, a quarter of the maximum. A calculator cannot make that judgement, and a score assembled without it is not the score the literature validated.

What is wells score calculator (pe)?

The Wells score for pulmonary embolism is a clinical prediction rule that turns seven history, examination and vital-sign findings into a pretest probability of PE. It was derived by Philip Wells and colleagues in 2000 from a logistic regression on forty candidate variables, and it exists to decide what test comes next rather than to make a diagnosis.

Its practical role is to gate D-dimer and imaging. A patient in the "PE unlikely" band with a negative D-dimer can have PE excluded without a CT scan, which spares contrast, radiation and time. A patient in the "PE likely" band goes to imaging regardless of the D-dimer, because a negative D-dimer in a high-probability patient is not reassuring enough.

The simplified version assigns one point to each of the same seven variables. It was introduced by Gibson and colleagues in 2008 and validated externally by Douma and colleagues in 2009 at a cut-off of 1. It exists because the weighted version's half-points are error-prone at the bedside, and its performance is close enough that either is acceptable in most pathways.

What the score is not: it is not a diagnostic test, it says nothing about clot burden or right-heart strain, it does not stratify severity once PE is confirmed — that is what the PESI score and echocardiography are for — and it is not a substitute for the clinician's overall assessment of a patient who looks unwell.

How to use this calculator.

1. 01Confirm the patient is stable enough for a diagnostic work-up. A haemodynamically unstable patient with suspected PE is outside this rule.
2. 02Answer each of the seven criteria from history, examination and observations. Heart rate is the measured rate at assessment.
3. 03Answer the "no alternative diagnosis is more likely than PE" item deliberately — it is worth three points, a quarter of the maximum, and it is a judgement the calculator cannot make for you.
4. 04Read the two-level band first if your local pathway is built around D-dimer gating; read the three-level band if your pathway uses low, moderate and high language.
5. 05Check the simplified score if your unit uses the one-point-per-variable version. It can disagree with the original — two items alone make PE "likely" on the simplified rule but "unlikely" on the original.
6. 06Remember that "PE unlikely" excludes PE only alongside a negative D-dimer, and that "PE likely" selects imaging rather than confirming a clot.

The formula.

The Wells PE score is a weighted sum of seven yes/no findings. Clinical signs and symptoms of DVT scores 3. No alternative diagnosis more likely than PE scores 3. Heart rate over 100 per minute scores 1.5. Immobilisation or surgery in the previous four weeks scores 1.5. Previous DVT or PE scores 1.5. Haemoptysis scores 1. Malignancy scores 1. There is no negative item, so the total runs from 0 to 12.5 in steps of 0.5.

Rounding stage: none. Every weight is an exact multiple of 0.5, the total is an exact sum on that lattice, and each band is decided by comparing the exact total against its threshold. The arithmetic runs in Decimal rather than raw floating point specifically so that 1.5 plus 1.5 plus 1.5 is exactly 4.5 at the "above 4" cut, rather than a value that prints as 4.5 but compares as slightly more.

The two-level scheme, as defined in the derivation paper, assigns "PE unlikely" to totals of 4.0 or below and "PE likely" to totals above 4.0. The three-level scheme, in the same paper's own words, assigns low probability below 2.0, moderate probability from 2.0 to 6.0, and high probability above 6.0. This calculator implements those inequalities literally. Other tables render the three-level bands as 0 to 1, 2 to 6 and 7 or more, or as 0 to 2, 3 to 6 and over 6; because totals of 1.5, 2.5 and 6.5 are all reachable, those renderings assign real patients differently, and the derivation paper's version is the only one without a gap.

The simplified rule replaces every weight with 1, giving a total from 0 to 7, and splits at 1: a total of 1 or below is "PE unlikely", 2 or more is "PE likely". The two rules can disagree. Haemoptysis and malignancy together give an original total of 2.0, which is "PE unlikely", but a simplified total of 2, which is "PE likely".

Worked through the example on this page: clinical signs of DVT (3) plus no more likely alternative diagnosis (3) plus heart rate over 100 (1.5) plus recent surgery (1.5) gives 3 + 3 + 1.5 + 1.5 = 9.0. Nine is above 4, so the two-level band is "PE likely"; nine is above 6, so the three-level band is high probability. The same four items give a simplified score of 1 + 1 + 1 + 1 = 4, which is 2 or more, so the simplified band is also "PE likely".

A worked example.

Frequently asked questions.

References& sources.

1. [1]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. Source of the seven item weights and of both banding schemes, quoted verbatim in the module header. (Abstract retrieved 2026-07-29; publisher full text paywalled.)
2. [2]Gibson NS, Sohne M, Kruip MJHA, et al. Further validation and simplification of the Wells clinical decision rule in pulmonary embolism. Thromb Haemost. 2008;99(1):229–234. Source of the simplified one-point-per-variable rule. (Abstract retrieved 2026-07-29; publisher full text paywalled.)
3. [3]Douma RA, Gibson NS, Gerdes VEA, et al. Validity and clinical utility of the simplified Wells rule for assessing clinical probability for the exclusion of pulmonary embolism. Thromb Haemost. 2009;101(1):197–200. Independent external validation; states the simplified cut-off of ≤1 and the 12% vs 13% "unlikely" prevalences. (Abstract retrieved verbatim 2026-07-29.)
4. [4]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. Source of the observed PE prevalence per category (≈10/30/65% three-level; ≈12/30% two-level). The Wells table itself sits in the Supplementary Data, which automated retrieval could not reach, so this reference is used only for the prevalences stated in the main text. (Retrieved 2026-07-29.)
5. [5]Cross-check used for the item weights and the conflicting three-level band rendering ("low (0-2 points), moderate (3-6 points), and high (> 6 points)"): PMC9490165, a Wells-rule comparison study. Weights agree with the derivation paper; band edges do not, and the disagreement is documented on this page. (Retrieved 2026-07-29.)