Safety of Age-Adjusted and Clinical Probability-Adjusted D-dimer Cut-offs in Suspected Pulmonary Embolism: A Systematic Review and Meta-Analysis.

BACKGROUND: Adaptive D-dimer strategies for rule out of pulmonary embolism (PE) have been proposed to improve the efficiency of conventional fixed '500' D-dimer thresholds. Validations of adaptive D-dimer strategies post-hoc in cohorts managed according to the fixed D-dimer threshold seems to questions their safety. Research Question What is the proportion of diagnostic failures in patients with suspected PE managed according to an adaptive strategy who have a D-dimer level >500 ng/mL but below the adaptive threshold, defined as the "adaptive window"? STUDY DESIGN AND METHODS: We conducted a systematic review and meta-analysis of prospective diagnostic management studies that recruited at least 50 patients with suspected PE and in which an adaptive D-dimer threshold was used to guide chest imaging. Diagnostic failure was defined as venous thromboembolism or PE-related or unexplained death occurring within 3 months. Exclusion criteria were: i) data required to calculate the primary outcome were unavailable; or ii) the study included only children or pregnant women. Study-specific proportions were pooled using a random-effects meta-analysis. RESULTS: Six studies (12,194 patients) were included, with a PE prevalence ranging from 7.0% to 19.2%. Among all patients in whom PE was ruled out without CTPA using adaptive D-dimer strategies, failure rates ranged from 0.00% to 0.27%, with a pooled overall failure rate of 0.12% (95%CI 0.06-0.26%; I2 0%). The pooled failure rate in the adaptive D-dimer window was 0.50% (95%CI 0.20-1.11%; I2 0%). Pooled algorithm-based efficiency was 60.8% (95%CI 49.2-71.2; I2 99.3%). INTERPRETATION: The diagnostic failure rate of adaptive strategies for the diagnosis of acute PE is acceptably low, also in patients with a D-dimer in the adaptive window.