Systematic review of prognostic scores and individual predictor variables for short-term mortality after acute pulmonary embolism.

Clinical Snapshot

Key Findings

• P Value: Not reported in abstract

• Effect Size: Pooled odds ratios for individual predictor variables ranged from OR 1.43 (D-dimer) to OR 2.65 (right heart thrombi); 24 prognostic models identified, all with c-index ≥ 0.7 in development studies

• Primary Outcome: Short-term mortality after acute symptomatic pulmonary embolism

• Nnt Or Sensitivity: C-index ≥ 0.7 for all 24 model-development studies (discrimination threshold); sPESI evaluated in 127 studies (most validated model); original PESI in 79 studies; ESC risk schema in 34 studies. Prognostic OR range: D-dimer OR 1.43 to right heart thrombi OR 2.65

• Confidence Interval: Not reported in abstract; full paper required for assessment

Clinical Application

The sPESI and PESI are freely available, require no additional investigations beyond standard clinical assessment, and are feasible in all acute care settings including resource-limited environments. RV dysfunction markers (echocardiography, CT-PA RV:LV ratio) and cardiac biomarkers (troponin, BNP/NT-proBNP) are routinely available in Australian hospitals. Right heart thrombus detection requires echocardiography, which may not be immediately available in all centres. The sPESI and PESI are endorsed in Australian clinical practice and referenced in Therapeutic Guidelines (Cardiovascular) and RACGP resources for acute PE management. Anticoagulation options for PE in Australia include PBS-listed direct oral anticoagulants (rivaroxaban, apixaban — both PBS-listed for PE treatment) and low-molecular-weight heparin. TGA-approved thrombolytics (alteplase) are available for high-risk PE. Risk stratification tools identified in this review directly inform decisions regarding outpatient versus inpatient management — a clinically and economically significant decision in the Australian healthcare context. The ESC risk schema, which integrates haemodynamic status, imaging, and biomarkers, is increasingly used in Australian tertiary centres with PE response teams (PERTs). This review's findings support the continued use of sPESI as the most extensively validated tool and highlight the additive prognostic value of right heart thrombi and RV dysfunction markers, relevant to Australian echocardiography-capable centres. Adults presenting with acute symptomatic pulmonary embolism across emergency, respiratory, cardiology, and intensive care settings. Applicable to both haemodynamically stable and unstable patients requiring risk stratification for triage, level-of-care decisions, and treatment intensity.

Abstract

BACKGROUND: For patients with acute pulmonary embolism (PE), assessment of prognosis helps with risk stratification, triage for level of care, management strategy, and communication among healthcare workers and patients. We sought to identify prognostic models and individual factors associated with short-term outcomes after acute symptomatic PE. METHODS: We performed a systematic review of prognostic factors for PE, searching MEDLINE, Embase, and Web of Science for records up to 1 June 2024. Studies of any design evaluating potential prognostic models or individual variables (not contained in the models) associated with short-term mortality after acute PE were included. RESULTS: We identified 314 studies that included 2,495,115 patients. Of these, 225 studies included 2,267,952 patients and evaluated 24 prognostic models for patients with acute PE. The most frequently used validated models were the simplified Pulmonary Embolism Severity Index (sPESI) (127 studies), the original PESI (79 studies), and the European Society of Cardiology risk schema (34 studies). Each model-development study had a c-index ≥ 0.7. Individual factors associated with prognosis included older age, presence of coexisting conditions, abnormal clinical signs and symptoms, clot burden, markers of right‑ventricle dilatation/dysfunction and myocardial injury, altered laboratory results indicating impaired haemodynamics, and tests that assess for systemic inflammation. Pooled odds ratios for variables not contained in any eligible prognostic model ranged from 1.43 (for D-dimer) to 2.65 (for right heart thrombi). CONCLUSIONS: This systematic review identified 24 prognostic models and 18 individual variables distinct from the prognostic models associated with short-term mortality after acute PE.

References

1. 1.Jimenez, D., Bikdeli, B., Tapson, V., Becattini, C., Moores, L., Barnes, G. D., Monreal, M., Konstantinides, S., Piazza, G., Morillo, R., Alvarez-Diaz, N., & Yusen, R. D. (2026). Systematic review of prognostic scores and individual predictor variables for short-term mortality after acute pulmonary embolism. Pulmonology. https://doi.org/10.1080/25310429.2026.2703889