Wells Score for PE Calculator

The Wells score was derived by Philip S. Wells and the thrombosis group at the Ottawa Hospital in Canada. The original cohort study landed in Annals of Internal Medicine in 1998, and the simplified dichotomized version that powers most emergency department algorithms was published in Thrombosis and Haemostasis in 2000. Wells later co-developed the parallel Wells score for deep vein thrombosis. Across both rules the design goal was the same: turn a tangle of bedside findings into a small integer that safely separates patients who can be ruled out with a D-dimer from those who need imaging.

The three-tier scheme keeps the original 1998 cut-offs: under 2 points is low probability (roughly 2 percent prevalence of PE), 2 to 6 is moderate (around 20 percent), and over 6 is high (about 50 percent and up). The dichotomized scheme collapses the same score into a single cut at 4 points: at or below 4 is PE unlikely, above 4 is PE likely. The two-tier version is easier to remember and pairs cleanly with a single D-dimer step, which is why ACEP, ESC, and most North American emergency departments default to it. The three-tier version is still useful when you want a sharper read on a patient already known to be moderate or high risk.

PERC, the Pulmonary Embolism Rule-out Criteria of Jeff Kline, is the gate that sits in front of Wells. Apply PERC only when your clinical gestalt or Wells score already puts the patient in the low-probability bucket (under 2 points on the three-tier scheme, or PE unlikely on the dichotomized one). If all eight PERC criteria are negative, you can rule out PE without any blood test or imaging, because the post-test probability falls under 2 percent. If even one PERC criterion is positive, fall back to a high-sensitivity D-dimer. Wells and PERC are complements, not competitors: Wells stratifies, PERC short-circuits the workup in the lowest-risk slice.

D-dimer is a fibrin degradation product that rises in essentially every active thrombosis. The clinically useful property is its negative predictive value, not its positive predictive value: a normal high-sensitivity assay drives the post-test probability of PE under 1 to 2 percent in a Wells low or moderate patient, which is the same residual risk you accept when you discharge the patient. The PEGeD and YEARS trials further showed that an age-adjusted D-dimer (age x 10 ng/mL for patients over 50) or a Wells-tied two-cutoff strategy safely rules out PE without imaging in roughly a third more patients than a flat 500 ng/mL cutoff. A positive D-dimer is non-specific, which is why the workflow only uses it to rule out, never to rule in.

CTPA is indicated whenever the pre-test probability is high (Wells over 6 on the three-tier scheme or PE likely on the dichotomized scheme), or when a low-or-moderate patient has a positive high-sensitivity D-dimer. At high probability the false-negative rate of D-dimer climbs to around 5 percent, which is too high to act on, so most guidelines tell you to skip the blood test and go straight to imaging. Use a ventilation-perfusion (V/Q) scan instead of CTPA when the patient is pregnant, has severe iodinated contrast allergy, or has stage 4 to 5 chronic kidney disease without dialysis. In hemodynamically unstable patients with suspected massive PE, bedside echocardiography looking for right-ventricular strain often comes before any cross-sectional imaging.