Prognostic Value of 2026 AHA/ACC Clinical Categorization in Acute Pulmonary Embolism Patients Presenting to the Emergency Department (APEX-2026)

STUDY DESIGN: Single-center prospective observational cohort. No intervention applied. Patients managed per standard of care. SETTING: Marmara University Pendik Training and Research Hospital Emergency Department, Istanbul, Turkey. Annual ED census approximately 200,000 visits. STUDY POPULATION: Adult patients (18 years or older) presenting to the ED with confirmed acute PE on computed tomography (CT) pulmonary angiography, enrolled consecutively from April 2026 to January 2027 (anticipated enrollment period 10 months). EXPOSURE: 2026 AHA/ACC PE clinical category assigned independently by two trained emergency physicians at presentation. Categories: A (subclinical/incidental), B (symptomatic, low severity; Pulmonary Embolism Severity Index (PESI) I-II / simplified PESI (sPESI)=0), C1/C2/C3 (elevated severity; stratified by right ventricular (RV) dysfunction and cardiac biomarkers), D1/D2 (incipient cardiopulmonary failure), E1/E2 (established cardiopulmonary failure). R modifier applied when hypoxemia/tachypnea/escalating oxygen requirement present without meeting higher category criteria. PRIMARY OUTCOME - MAPEE (Major Adverse Pulmonary Embolism Event): composite endpoint defined as occurrence of any of the following within 30 days: (1) All-cause mortality, (2) Hemodynamic deterioration (systolic blood pressure (SBP) less than 90 mmHg for at least 15 minutes or vasopressor requirement), (3) Treatment escalation (systemic thrombolysis, catheter-directed therapy, surgical embolectomy, or extracorporeal membrane oxygenation (ECMO)), (4) Cardiopulmonary resuscitation. Expected event rate: 12% (n approximately 59 events out of 492 analyzable patients after 18% attrition from 600 gross enrollment). SAMPLE SIZE RATIONALE: Based on the Hanley-McNeil variance formula (Hanley and McNeil, Radiology 1982;143:29-36). H0: AUC=0.65 (minimum clinically meaningful discrimination); H1: AUC=0.80 (supported by published composite-outcome AUC values for comparable systems: PESI 0.84, Bova score 0.82, Hestia criteria SROC 0.81); two-sided alpha=0.05. At the planned n_gross=600 enrollment (n_net=492 analyzable, 59 expected events, 433 non-events), the design provides approximately 97.6% power for H1=0.80 using the standard single-proportion Hanley-McNeil test (a conservative pooled-variance sensitivity approach yields approximately 79.2%; the sample size is considered adequate under either method). Minimum detectable AUC at 80% power is approximately 0.761. Events per variable (EPV) = 59/5 = 11.8, supporting multivariable models with up to 5 predictor variables (EPV of at least 10 recommended per Peduzzi et al. 1996). DeLong AUC comparison (AHA/ACC 2026 vs ESC 2019, secondary outcome S1) will require substantially larger samples for adequate power given the two correlated ROC curves; this comparison is designated exploratory/hypothesis-generating and will not be used for confirmatory inference. STROBE COMPLIANCE: This study follows the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) 2007 checklist for reporting of observational studies.