Derived for haemodynamically stable outpatients in emergency or acute care with suspected acute pulmonary embolism, where a D-dimer–based diagnostic strategy is appropriate.

4PEPS integrates variables from existing tools (such as Wells, revised Geneva, PERC, YEARS and age-adjusted D-dimer approaches) into a single 4-level clinical pretest probability (CPP) score.

CPP band	4PEPS score	Approximate PE prevalence (derivation cohort)	Intended rule-out strategy
Very low	<0	~1%	PE may be excluded on clinical criteria alone in suitable patients.
Low	0–5	~6%	Use a high-sensitivity D-dimer; PE can be ruled out if D-dimer is <1.0 µg/mL (1000 µg/L).
Moderate	6–12	~31%	Use a high-sensitivity D-dimer; PE can be ruled out if D-dimer is <0.5 µg/mL (500 µg/L) or below the age-adjusted cut-off (age ×0.01 µg/mL, i.e. age ×10 µg/L for patients >50 years).
High	≥13	~74%	D-dimer testing alone cannot safely exclude PE; imaging (usually CTPA or V/Q scan) is required.

The original derivation and validation study used individual patient data from three prospective emergency-department management studies (combined n≈11 100; overall PE prevalence ~11%) from France, Belgium and the US, then tested the score in two further multinational ED cohorts (n≈1 550 and n≈1 670, with PE prevalence ~22% and ~12%). Discrimination was good (area under the ROC curve around 0.8 across cohorts) and calibration acceptable.

When the full 4PEPS strategy (score plus CPP-specific D-dimer thresholds) was retrospectively applied to the two external validation cohorts, the estimated failure rate (missed venous thromboembolism at baseline or follow-up) was <1% in each, with an absolute reduction in CTPA use of about 19–22% compared with a standard revised Geneva plus fixed 0.5 µg/mL D-dimer strategy.

A later external validation in the YEARS diagnostic management cohort (n=3 465; VTE prevalence ~14%) suggested that a 4PEPS-based pathway could have ruled out PE without imaging in ~58% of patients, with an overall failure rate around 1.3%, still within commonly accepted safety margins but higher than in the original study.

All derivation and validation work was conducted in emergency-department outpatients; performance in inpatients, peri-operative patients, pregnancy and other high-risk groups is uncertain. Prospective implementation studies, including the SPEED&PEPS trial protocol, aim to confirm safety and real-world impact on imaging use.

4-Level Pulmonary Embolism Clinical Probability Score (4PEPS)