Derivation and external validation of the SHIeLD score for predicting outcome in normotensive pulmonary embolism

Abstract

Identifying patients with normotensive pulmonary embolism (PE) who may benefit from thrombolysis remains challenging. We sought to develop and validate a score to predict 30-days PE-related mortality and/or rescue thrombolysis.

METHODS: We retrospectively assessed 554 patients with normotensive PE. Independent predictors of the studied endpoint were identified from variables available at admission in the emergency department and were used to create a score. The model was validated in 308 patients from a separate hospital.

RESULTS: A total of 64 patients died or needed rescue thrombolysis (44 in the derivation cohort). Four independent prognostic factors were identified: Shock index ≥ 1.0 (OR 3.33; 95% CI 1.40-7.93; P = 0.006), HypoxaemIa by the PaO2/FiO2 ratio (OR 0.92 per 10 units; 95% CI 0.88-0.97; P < 0.001), Lactate (OR 1.38 per mmol/L; 95% CI 1.09-1.75; P = 0.008) and cardiovascular Dysfunction (OR 5.67; 95% CI 2.60-12.33; P < 0.001) - SHIeLD score. In the development cohort, event rates for each risk tercile were 0.0%, 2.2%, and 21.6%. In the validation cohort, corresponding rates were 0.0%, 1.9%, and 14.3%. The C-statistic was 0.90 (95% CI 0.86-0.94, P < 0.001) in the derivation cohort and 0.82 (95% CI 0.75-0.89, P < 0.001) in the validation cohort. Decision curve analysis showed that the SHIeLD score is able to accurately identify more true positive cases than the European Society of Cardiology decision criteria.

CONCLUSIONS: A risk score to predict 30-days PE-related mortality and/or rescue thrombolysis in patients with normotensive PE was developed and validated. This score may assist physicians in selecting patients for closer monitoring or aggressive treatment strategy.