Pulmonary Embolism Workup: From Risk Scoring to CTPA

A pulmonary embolism workup is a stepwise process that moves from clinical suspicion through risk scoring, blood testing, and imaging to either confirm or safely rule out a blood clot in the lungs. No single test does the job alone. Instead, clinicians layer tools on top of each other, starting with the least invasive and escalating only when the prior step cannot give a clear answer. The reason for all this staging is that PE symptoms overlap heavily with other common conditions, and both missing a clot and over-testing healthy patients carry real consequences.

Why Pulmonary Embolism Is So Often Missed

The classic triad of sudden shortness of breath, chest pain, and a fast heart rate shows up in dozens of conditions besides PE. A systematic review pooling data across multiple studies found that roughly 28% of PE patients in emergency departments are initially given a wrong diagnosis, and in inpatient hospital settings that figure climbs to about 54%.¹ The most common misdiagnoses are chest infections, heart failure, and heart attacks. Among patients who die in intensive care and undergo autopsy, about 38% were found to have a PE that had been entirely overlooked.²

The nonspecific nature of PE symptoms and the limitations of routine tests like chest X-rays and basic EKGs drive these numbers.³ A young patient with chest pain and mild EKG changes can easily be labeled with pericarditis or myocarditis before anyone considers PE.⁴ The structured workup that follows exists precisely because gut feeling alone fails too often.

Step One: Clinical Probability Scoring

Before any blood draw or scan, clinicians estimate how likely PE is using a standardized scoring system. The two most widely used are the Wells score and the revised Geneva score. Both assign points based on risk factors and symptoms: things like recent surgery, a history of blood clots, leg swelling, an elevated heart rate, and whether PE is the single most likely explanation for what you are experiencing. The total determines whether you fall into a low, intermediate, or high probability category.

A meta-analysis comparing the two found that the Wells score was somewhat more accurate, with sensitivity between roughly 64% and 79% depending on the threshold used, compared with 55% to 74% for the revised Geneva score.⁵ That said, a separate head-to-head study found no statistically significant difference in overall performance between the two when measured by area under the curve, and both safely excluded PE when combined with a normal D-dimer result and three months of follow-up showed no clots.⁶ In practice, whichever one your hospital has adopted works well enough. The point of these scores is not to make a diagnosis but to sort patients into lanes that determine what happens next.

When You Can Skip Everything Else

For the lowest-risk patients, there is an even faster off-ramp: the Pulmonary Embolism Rule-out Criteria, or PERC rule. It is an eight-item checklist covering age, heart rate, oxygen level, leg swelling, recent surgery or trauma, hemoptysis (coughing blood), hormone use, and prior clot history. If a patient already has a low clinical suspicion and meets all eight criteria, PE can be considered excluded without further testing.⁷

A randomized trial tested whether applying the PERC rule early led to missed clots. After three months, only one thromboembolic event occurred in the group managed by the PERC rule, a rate of 0.1%, with none in the control group that received the usual workup.⁸ The appeal of PERC is speed and cost: it lets the very lowest-risk patients go home without a blood draw. Its limitation is that the criteria are strict, and it only applies when your clinical probability is already low.

D-Dimer Testing and the Age-Adjusted Cutoff

For patients who are not PERC-negative but have a low or intermediate probability of PE, the next step is usually a D-dimer blood test. D-dimer is a fragment released when blood clots break down. A normal result is very good at ruling PE out, but an elevated result does not mean you have a clot. D-dimer rises with age, inflammation, infection, pregnancy, cancer, recent surgery, and many other conditions, which means false positives pile up the older or sicker you are.

The conventional cutoff has long been 500 micrograms per liter. The problem is that in patients over 70, D-dimer is elevated so often that the test becomes nearly useless at that threshold. The age-adjusted cutoff solves this by raising the normal limit to your age multiplied by ten (so a 75-year-old’s cutoff becomes 750 instead of 500). The ADJUST-PE study, which included nearly 3,500 patients, showed that using the age-adjusted threshold allowed about 12% more patients to be classified as “D-dimer negative” and avoid CT scanning, with a failure rate at three months of just 0.3% among those extra patients.⁹ Earlier work in multiple large cohorts showed the same pattern: the biggest jump in the proportion of patients safely excluded came in those over 70, where the age-adjusted cutoff ruled out PE in 13% to 16% more people than the traditional number.¹⁰

A meta-analysis confirmed that the age-adjusted cutoff is both effective and safe for excluding PE and reduces false positives compared with the fixed threshold in patients with non-high clinical probability.¹¹ If you are over 50, ask whether your hospital is using the age-adjusted number. Many now do, but the older cutoff persists in some settings.

CT Pulmonary Angiography

When the D-dimer is elevated or the clinical probability is high enough that a blood test alone is not sufficient, the gold-standard next step is CT pulmonary angiography, or CTPA. You receive an injection of contrast dye, and the scanner takes rapid cross-sectional images of your chest timed to capture the dye flowing through the lung arteries. Clots show up as dark filling defects against the bright contrast. Modern multi-detector scanners are highly accurate. In a head-to-head trial comparing standard and reduced-dose protocols, sensitivity was around 97% to 100% and specificity around 97% to 98%, with no PE or PE-related death in the 90-day follow-up of either group.¹²

Radiation exposure from CTPA is a legitimate concern, particularly for younger patients and anyone who might need repeated scans. Research on newer scanners with advanced reconstruction algorithms has shown that accurate PE diagnosis remains possible at as low as a quarter of the standard radiation dose.¹³ Low-dose CTPA protocols are becoming standard at centers with up-to-date equipment, though availability varies. If radiation is a concern, it is reasonable to ask whether a reduced-dose protocol is an option.

When CT Is Not an Option

Not everyone can have a CTPA. Severe kidney disease, a true contrast allergy, or pregnancy may push clinicians toward alternatives. The ventilation-perfusion (V/Q) scan is the oldest imaging alternative: you inhale a small amount of radioactive tracer and receive another tracer intravenously, then a gamma camera compares airflow with blood flow. Mismatched areas, where air gets in but blood does not, suggest a clot.¹⁴ V/Q scans work best in patients with otherwise normal lungs. In someone with underlying lung disease, the results are often indeterminate, which forces further testing anyway.

Compression ultrasound of the legs is another indirect approach. If a patient suspected of PE also has a deep vein thrombosis (DVT) visible on ultrasound, that finding essentially confirms the diagnosis and treatment can start without chest imaging. A meta-analysis found that proximal leg ultrasound has a sensitivity of only about 41% for PE, so a negative result does not rule anything out, but its specificity is around 96%, meaning a positive finding is highly reliable.¹⁵ Ultrasound is most useful when CT is off the table: finding a leg clot confirms the clinical picture, and treatment for DVT and PE overlaps heavily.

Bedside Echocardiography in Unstable Patients

When someone with suspected PE is crashing, there is no time to wait for a CT scanner. Bedside echocardiography fills that gap. It cannot directly visualize most clots in the lungs, but it can show the downstream effect: the right side of the heart straining against the sudden blockage. Key signs include a dilated right ventricle, abnormal motion of the wall between the two ventricles, a specific pattern called McConnell’s sign where the right ventricle tip still contracts while the rest of the free wall does not, and elevated pressure in the pulmonary artery.¹⁶ In a hemodynamically unstable patient with these findings, clinicians may start life-saving clot-busting therapy before confirmatory imaging.

Biomarkers That Gauge Severity

Once PE is confirmed, the workup is not over. Clinicians need to know how dangerous the particular clot is, and two blood markers help with that assessment: troponin and brain natriuretic peptide (BNP or its precursor, NT-proBNP).

Troponin leaks from heart muscle cells when they are injured. A meta-analysis found that elevated troponin in PE is associated with roughly a fivefold increase in in-hospital mortality and nearly a sixfold increase in ICU admission.¹⁷ In one observational study, mortality at 30 days and three years was significantly higher in PE patients with positive troponin, and the risk of dying increased about 2.5 times after adjusting for age, sex, and right ventricular dysfunction.¹⁸ BNP, released when the heart’s chambers are stretched, also helps identify patients at higher risk, with one study showing pro-BNP had a sensitivity of about 85% and specificity of about 80% for detecting at-risk patients at an optimal cutoff.¹⁹

These markers do not change the diagnosis, but they shape what happens next. A low-risk PE with normal biomarkers may be treated at home. A PE with sky-high troponin and right ventricular strain on echocardiography may warrant intensive care or advanced interventions like catheter-directed therapy.

Deciding Between Hospital and Home

Not every confirmed PE requires a hospital stay. The Hestia criteria are a checklist of practical questions: Is the patient hemodynamically stable? Is there active bleeding? Is there a reason they cannot manage anticoagulation safely at home? A meta-analysis found that among PE patients classified as low-risk by the Hestia rule, rates of death and serious complications were low, around 2.3% overall mortality and 4.4% for combined poor outcomes.²⁰

The original Hestia study followed 297 patients managed entirely as outpatients. During three months, recurrent blood clots occurred in 2%, and no one died from PE.²¹ A later randomized trial tested whether adding an NT-proBNP blood test to the Hestia criteria improved safety; the Hestia criteria alone already performed well, with only about 1% of directly discharged patients meeting the primary adverse outcome endpoint.²² Outpatient PE management is growing, but it works only with reliable follow-up and clear patient instructions about when to return to the emergency room.

The PE Workup in Pregnancy

Pregnancy complicates every step. D-dimer rises naturally as pregnancy progresses, making the standard cutoff nearly useless by the third trimester. Clinicians also want to minimize radiation to the fetus. The pregnancy-adapted YEARS algorithm addresses both issues by combining three clinical items (signs of DVT, hemoptysis, and whether PE is the most likely diagnosis) with a D-dimer threshold that changes based on pregnancy trimester. In a prospective study of nearly 500 pregnant women, this algorithm safely ruled out PE across all trimesters and avoided CT scanning in 32% to 65% of patients.²³ Only one case of popliteal deep vein thrombosis was diagnosed in the three-month follow-up, and no PE was missed.

A systematic review of individual patient data confirmed that both the Wells rule and the YEARS algorithm can safely exclude PE in pregnancy when paired with adapted D-dimer thresholds, with failure rates between about 0.4% and 1.4%.²⁴ If you are pregnant and undergoing a PE workup, the goal is the same as for anyone else, but the tools are calibrated differently to account for the changes your body is already going through.

PE Workup in Children

Pediatric PE is rare, which paradoxically makes it more dangerous: because clinicians are not looking for it, the diagnosis often comes late. In one study from China, the median time from symptom onset to diagnosis was 12 days, with a range stretching from 2 to 45 days.²⁵ Adult scoring tools like the Wells score and PERC have not been validated in children and perform poorly in younger patients. CTPA remains the primary diagnostic imaging when PE is suspected, but the threshold for ordering it is higher because radiation exposure matters more in growing bodies.²⁶ There is active research into better pediatric decision tools, but for now the workup in children relies heavily on clinical vigilance rather than standardized algorithms.

The Subsegmental PE Debate

Modern CT scanners are so sensitive that they sometimes find very small clots in the tiny arteries at the edges of the lungs, called subsegmental pulmonary emboli (SSPE). Whether these matter is genuinely controversial. Some researchers argue that the smallest clots may be clinically irrelevant and that treating them with anticoagulation, which carries a bleeding risk, may cause more harm than good. Others counter that even small clots could be the tip of a larger problem.²⁷

The current pragmatic approach is often to check the legs for DVT. If a patient has an isolated subsegmental PE and no leg clot, some guidelines now suggest that clinical observation without anticoagulation can be reasonable in low-risk individuals. This is one of those situations where the workup’s answer, “yes, you have a clot,” does not automatically dictate the treatment.

Artificial Intelligence as a Safety Net

AI tools are increasingly being deployed as a second reader for CTPA scans. One study evaluating an AI algorithm found per-case sensitivity of about 94% and specificity of about 95%. More strikingly, among the positive PE cases in that study, about 16% had not been mentioned in the original clinical radiology report. The algorithm caught roughly three-quarters of those missed cases, cutting the overall miss rate from 16% down to about 4%.²⁸ A separate study of an AI algorithm applied to over 1,400 scans showed similar numbers: sensitivity around 93% and specificity around 96%.²⁹

These tools do not replace radiologists. They flag cases that might otherwise be overlooked, particularly small or peripheral clots and scans read during overnight shifts when fatigue is a factor. The technology is still being refined, but it represents a genuine addition to the workup chain rather than a novelty.

Speed and Cost Gains From Streamlined Algorithms

The layered approach to PE diagnosis is not just about accuracy; it also saves time and money. A study measuring the real-world impact of implementing the YEARS algorithm, which combines clinical criteria with D-dimer thresholds to reduce unnecessary CT scans, found that patients were discharged from the emergency department about 54 minutes faster when CTPA was not needed and about 60 minutes faster overall. Patients who did have PE and needed a scan received their first dose of anticoagulation about 53 minutes sooner. Total costs dropped by an average of €123 per visit.³⁰ These efficiency gains matter in busy emergency departments where both scanner time and clinician bandwidth are finite.

What Happens After the Acute Workup

Surviving a PE is not the end of the story. A small but important fraction of patients develop chronic thromboembolic pulmonary hypertension (CTEPH), a condition where residual clot material scars the lung arteries and drives up pressure in the right side of the heart over months to years. Registry data from over 2,700 PE patients found a cumulative CTEPH diagnosis rate of about 2.3% at three years, with risk factors including being female, having low oxygen levels at initial diagnosis, showing right heart strain, and having an unprovoked PE with no obvious trigger.³¹

CTEPH is treatable, sometimes even curable with surgery, but it is frequently missed because patients attribute their lingering breathlessness to deconditioning or anxiety. A structured symptom screening program dramatically improved detection. Before the screening was introduced, CTEPH was found in only about 0.5% of PE survivors through routine care. After symptom-based screening was added, the detection rate roughly tripled. Among patients who reported persistent breathlessness of moderate severity or worse, the CTEPH rate was about 8.5%.³² If you have had a PE and still feel short of breath months later, that is worth bringing up with your doctor rather than assuming it is normal recovery.