Safety profile and resource utilisation associated with systemic alteplase versus unfractionated heparin in intermediate-high risk pulmonary embolism: a prospective ICU cohort study

Full-dose systemic thrombolysis reduces haemodynamic decompensation in intermediate-high risk pulmonary embolism (PE) but increases bleeding risk, and its real-world safety profile in ICU settings remains incompletely characterised. We describe treatment selection patterns, safety outcomes, and resource utilisation associated with systemic alteplase versus unfractionated heparin (UFH) in ICU-managed intermediate-high risk PE. Prospective single-centre cohort; 80 adults with CTPA-confirmed intermediate-high risk PE (2020-2024), 40 per arm. Outcomes included mortality, bleeding, clinical deterioration, and ICU and hospital length of stay (LOS). Firth penalised logistic regression was used for binary outcomes with complete separation; negative binomial regression with covariate adjustment (age, APACHE II, Charlson Comorbidity Index, PESI class) for LOS. Clinicians preferentially selected alteplase for younger patients with greater PE-specific severity markers, and UFH for older, more comorbid patients with higher physiological derangement. Major bleeding was rare in both groups (UFH 2/40; alteplase 1/40; P = 0.781), with no intracranial or fatal events. Clinical deterioration occurred exclusively in the UFH group (17.5% vs 0%; P = 0.012), though most events represented rescue thrombolysis, structurally unobservable in the alteplase arm; no residual signal persisted after its exclusion (adjusted OR 0.77; P = 0.89). ICU LOS (rate ratio 0.70; P = 0.007) and hospital LOS (rate ratio 0.73; P = 0.019) were shorter with alteplase after covariate adjustment. In this prospective ICU cohort of carefully selected intermediate-high risk PE patients, systemic alteplase was associated with a favourable safety profile and shorter hospitalisation. These hypothesis-generating findings support prospective studies targeting ICU-managed intermediate-high risk PE.