Prognostic clinical prediction rules to identify a low‐risk pulmonary embolism: a systematic review and meta‐analysis

Summary. Background: Prognostic assessment is important for the management of patients with a pulmonary embolism (PE). A number of clinical prediction rules (CPRs) have been proposed for stratifying PE mortality risk. The aim of this systematic review was to assess the performance of prognostic CPRs in identifying a low‐risk PE. Methods: MEDLINE and EMBASE databases were systematically searched until August 2011. Derivation and validation studies that assessed the performance of prognostic CPRs in predicting adverse events‐risk in PE patients were included. Weighted mean proportion and 95% confidence intervals (CIs) of adverse events were then calculated and pooled using a fixed and a random‐effects model. Statistical heterogeneity was evaluated through the use of I² statistics. Results: Of 1125 references in the original search, 33 relevant articles were included. Nine CPRs were assessed in 37 cohorts, for a total of 35 518 patients. Pulmonary Embolism Severity Index and prognostic Geneva CPR were investigated in 22 and 6 cohorts, respectively. Eleven (29.7%) cohorts were of high quality. The median follow‐up was 30 days. In low‐risk PE patients, pooled short‐term mortality (within 14 days or less) was 0.7% (95% CI 0.3–1.1%, random‐effects model; I² = 49.6%), 30‐day mortality was 1.7% (95% CI 1.1–2.3%, random‐effects model; I² = 82.4%) and 90‐day mortality was 2.2% (95% CI 1.2–3.4%, random‐effects model; I² = 59.8%). Conclusions: Prognostic CPRs efficiently identify PE patients at a low risk of mortality. Before implementing prognostic CPRs in the routine care of PE patients, well‐designed management studies are warranted.