Full Citation: Kearon C et al. Diagnosis of pulmonary embolism with D-dimer adjusted to clinical probability. N Engl J Med 2019;381(22):2125-34.
Abstract Link: https://www.nejm.org/doi/full/10.1056/NEJMoa1909159?url_ver=Z39.88-2003&rfr_id=ori:rid:crossref.org&rfr_dat=cr_pub%20%200pubmed
Article Type: Diagnosis
Ratings: Methods – 4/5 Usefulness – 4/5
INTRODUCTION
Background
Clinical diagnosis of pulmonary embolism (PE) can be challenging and relies on objective testing, but gold-standard testing with computed tomography (CT) pulmonary angiography or ventilation-perfusion scan (“chest imaging”) exposes patients to potential harms. Thus, it is important to identify low-risk patients who can be discharged safely without the need for extensive testing.
Objectives
To assess whether a variable D-dimer threshold can be used based on clinical probability to exclude the diagnosis of PE safely without further imaging.
METHODS
Design
Prospective multi-centre cohort study.
Setting
Nine university-affiliated hospitals in Canada between 2015 and 2018.
Eligibility criteria
Outpatients (emergency department [ED] or outpatient clinics) with signs or symptoms of PE.
Intervention
Implementation of the Pulmonary Embolism Graduated D-Dimer (PEGeD) algorithm using three-tier clinical pre-test probability (C-PTP) risk stratification: low (Wells 0–4), moderate (Wells 4.5–6), or high (Wells >6), followed by D-dimer testing. Patients with low C-PTP and a D-dimer <1000 or moderate C-PTP and a D-dimer <500 did not undergo further diagnostic testing. All other patients underwent chest imaging and treatment for PE, if necessary.
Outcomes
Primary outcome was incidence of venous thromboembolism (VTE) at a 90-day clinical follow-up. Outcome assessors were unaware of the results of testing at initial presentation.
MAIN RESULTS
Of 3,133 patients with signs or symptoms of PE, 2,017 were included. The mean age was 52, and two-thirds were women. On clinical assessment, 87% of patients had low, 11% had moderate, and 2% had high C-PTP. The overall incidence of PE was 7.4%. Further, 67% of patients with low (1,285 patients) or moderate (40 patients) C-PTP had a negative D-dimer and were discharged without further imaging. None of these patients had VTE on clinical follow-up at 90 days (95% confidence interval [CI] 0.00–0.29%). This included 315 patients who had a D-dimer level between 500 and 999 ng/mL. Only two patients were diagnosed with PE at follow-up, both of whom had positive initial D-dimers and subsequent negative imaging on initial presentation. Chest imaging was used in 34.3% of patients, a 17.6% reduction (95% CI −19.2 to −15.9), as compared with a strategy that rules out PE in patients with low C-PTP and a D-dimer level <500 ng/mL.
APPRAISAL
Strengths
• Clear, focused, and clinically important question
• Patient-centred outcome with an adequate follow-up period
• Applicable to an ED setting
• Uses Wells score for risk assessment, internationally used and recognized
• Multiple D-dimer assays improve external validity
• No industry funding
• Few patients lost to follow-up
• Blinded outcome assessment
Limitations
• Non-consecutive patient enrolment
• Inadequate sample size to reach conclusions on a moderate risk population
• Only 315 of 2,017 patients in the low C-PTP group had a D-dimer between 500 and999 ng/mL, limiting the strength of conclusions in this important subgroup. It did not assess how many of these patients might have been excluded using pulmonary embolism rule-out criteria (PERC; Kline 2008)
• No direct comparison to usual practice or alternate diagnostic strategy
• Cannot apply to inpatients, pregnant patients, or those who had recent major surgery
• The 7.4% incidence of PE is low compared with prior studies
BOTTOM LINE
The results of this study demonstrate a safe approach to exclude the diagnosis of PE using variable D-dimer thresholds based on C-PTP. Specifically, the evidence is strong for use of this method in patients with low C-PTP and a D-dimer <1000 ng/mL. Further studies with larger sample sizes are needed to validate the safety of this approach for patients with moderate C-PTP. This approach can be adopted by EDs to reduce chest imaging in the workup for PE.
Full Citation: Kearon C et al. Diagnosis of pulmonary embolism with D-dimer adjusted to clinical probability. N Engl J Med 2019;381(22):2125-34.
Abstract Link: https://www.nejm.org/doi/full/10.1056/NEJMoa1909159?url_ver=Z39.88-2003&rfr_id=ori:rid:crossref.org&rfr_dat=cr_pub%20%200pubmed
Article Type: Diagnosis
Ratings: Methods – 4/5 Usefulness – 4/5
INTRODUCTION
Background
Clinical diagnosis of pulmonary embolism (PE) can be challenging and relies on objective testing, but gold-standard testing with computed tomography (CT) pulmonary angiography or ventilation-perfusion scan (“chest imaging”) exposes patients to potential harms. Thus, it is important to identify low-risk patients who can be discharged safely without the need for extensive testing.
Objectives
To assess whether a variable D-dimer threshold can be used based on clinical probability to exclude the diagnosis of PE safely without further imaging.
METHODS
Design
Prospective multi-centre cohort study.
Setting
Nine university-affiliated hospitals in Canada between 2015 and 2018.
Eligibility criteria
Outpatients (emergency department [ED] or outpatient clinics) with signs or symptoms of PE.
Intervention
Implementation of the Pulmonary Embolism Graduated D-Dimer (PEGeD) algorithm using three-tier clinical pre-test probability (C-PTP) risk stratification: low (Wells 0–4), moderate (Wells 4.5–6), or high (Wells >6), followed by D-dimer testing. Patients with low C-PTP and a D-dimer <1000 or moderate C-PTP and a D-dimer <500 did not undergo further diagnostic testing. All other patients underwent chest imaging and treatment for PE, if necessary.
Outcomes
Primary outcome was incidence of venous thromboembolism (VTE) at a 90-day clinical follow-up. Outcome assessors were unaware of the results of testing at initial presentation.
MAIN RESULTS
Of 3,133 patients with signs or symptoms of PE, 2,017 were included. The mean age was 52, and two-thirds were women. On clinical assessment, 87% of patients had low, 11% had moderate, and 2% had high C-PTP. The overall incidence of PE was 7.4%. Further, 67% of patients with low (1,285 patients) or moderate (40 patients) C-PTP had a negative D-dimer and were discharged without further imaging. None of these patients had VTE on clinical follow-up at 90 days (95% confidence interval [CI] 0.00–0.29%). This included 315 patients who had a D-dimer level between 500 and 999 ng/mL. Only two patients were diagnosed with PE at follow-up, both of whom had positive initial D-dimers and subsequent negative imaging on initial presentation. Chest imaging was used in 34.3% of patients, a 17.6% reduction (95% CI −19.2 to −15.9), as compared with a strategy that rules out PE in patients with low C-PTP and a D-dimer level <500 ng/mL.
APPRAISAL
Strengths
• Clear, focused, and clinically important question
• Patient-centred outcome with an adequate follow-up period
• Applicable to an ED setting
• Uses Wells score for risk assessment, internationally used and recognized
• Multiple D-dimer assays improve external validity
• No industry funding
• Few patients lost to follow-up
• Blinded outcome assessment
Limitations
• Non-consecutive patient enrolment
• Inadequate sample size to reach conclusions on a moderate risk population
• Only 315 of 2,017 patients in the low C-PTP group had a D-dimer between 500 and999 ng/mL, limiting the strength of conclusions in this important subgroup. It did not assess how many of these patients might have been excluded using pulmonary embolism rule-out criteria (PERC; Kline 2008)
• No direct comparison to usual practice or alternate diagnostic strategy
• Cannot apply to inpatients, pregnant patients, or those who had recent major surgery
• The 7.4% incidence of PE is low compared with prior studies
CONTEXT
This study builds upon a strong body of evidence using clinical criteria and D-dimer testing to reduce advanced chest imaging in patients with a possible diagnosis of PE, including the Wells score with D-dimer, PERC rule,Reference Kline, Courtney and Kabrhel1 YEARS criteria,Reference van der Hulle, Cheung and Kooij2 and age-adjusted D-dimer.Reference Righini, Van Es and Den Exter3 In particular, this study demonstrates the ability to exclude PE safely in patients with low C-PTP and a D-dimer level between 500 and 999 ng/mL.
BOTTOM LINE
The results of this study demonstrate a safe approach to exclude the diagnosis of PE using variable D-dimer thresholds based on C-PTP. Specifically, the evidence is strong for use of this method in patients with low C-PTP and a D-dimer <1000 ng/mL. Further studies with larger sample sizes are needed to validate the safety of this approach for patients with moderate C-PTP. This approach can be adopted by EDs to reduce chest imaging in the workup for PE.