External validation of the YEARS diagnostic algorithm for suspected pulmonary embolism

Diagnostic management of acute pulmonary embolism

Straightforward, accurate diagnostic management in patients presenting with clinically suspected pulmonary embolism (PE) is essential, since starting anticoagulant treatment may give important adverse effects of bleeding, while false exclusion of the disease may lead to recurrent VTE, with associated morbidity and mortality. In the past three decades, considerable improvement in the diagnostic management of PE has been made. Computed tomography pulmonary angiography (CTPA) has largely replaced conventional pulmonary angiography and ventilation-perfusion lung scanning as the imaging methods of choice. Several diagnostic algorithms, all able to minimize the need for radiological imaging have been developed and validated. Lastly, within the diagnostic algorithms, varying D-dimer cut-off levels have successfully been introduced to further downsize the need for radiological imaging.

Comparison of the accuracy of four diagnostic prediction rules for pulmonary embolism in patients admitted to the emergency department

INTRODUCTION AND OBJECTIVES

Ruling out pulmonary embolism (PE) through a combination of clinical assessment and D-dimer level can potentially avoid excessive use of computed tomography pulmonary angiography (CTPA). We aimed to compare the diagnostic accuracy of the standard approach based on the Wells and Geneva scores combined with a standard D-dimer cut-off (500 ng/ml), with three alternative strategies (age-adjusted and the YEARS and PEGeD algorithms) in patients admitted to the emergency department (ED) with suspected PE.

METHODS

Consecutive outpatients admitted to the ED who underwent CTPA due to suspected PE were retrospectively assessed. Sensitivity, specificity, positive and negative predictive values, likelihood ratios and diagnostic odds ratios were calculated and compared between the different diagnostic prediction rules.

RESULTS

We included 1402 patients (mean age 69±18 years, 54% female), and PE was confirmed in 25%. Compared to the standard approach (p<0.001), an age-adjusted strategy increased specificity with a non-significant decrease in sensitivity only in patients older than 70 years. Compared to the standard and age-adjusted approaches, the YEARS and PEGeD algorithms had the highest specificity across all ages, but were associated with a significant decrease in sensitivity (p<0.001), particularly in patients aged under 60 years (sensitivity of 81% in patients aged between 51 and 60 years).

CONCLUSION

Compared to the standard approach, all algorithms were associated with increased specificity. The age-adjusted strategy was the only one not associated with a significant decrease in sensitivity compared to the standard approach, enabling CTPA requests to be reduced safely.

A machine learning model for diagnosing acute pulmonary embolism and comparison with Wells score, revised Geneva score, and Years algorithm

BACKGROUND

Acute pulmonary embolism (APE) is a fatal cardiovascular disease, yet missed diagnosis and misdiagnosis often occur due to non-specific symptoms and signs. A simple, objective technique will help clinicians make a quick and precise diagnosis. In population studies, machine learning (ML) plays a critical role in characterizing cardiovascular risks, predicting outcomes, and identifying biomarkers. This work sought to develop an ML model for helping APE diagnosis and compare it against current clinical probability assessment models.

METHODS

This is a single-center retrospective study. Patients with suspected APE were continuously enrolled and randomly divided into two groups including training and testing sets. A total of 8 ML models, including random forest (RF), Naïve Bayes, decision tree, K-nearest neighbors, logistic regression, multi-layer perceptron, support vector machine, and gradient boosting decision tree were developed based on the training set to diagnose APE. Thereafter, the model with the best diagnostic performance was selected and evaluated against the current clinical assessment strategies, including the Wells score, revised Geneva score, and Years algorithm. Eventually, the ML model was internally validated to assess the diagnostic performance using receiver operating characteristic (ROC) analysis.

RESULTS

The ML models were constructed using eight clinical features, including D-dimer, cardiac troponin T (cTNT), arterial oxygen saturation, heart rate, chest pain, lower limb pain, hemoptysis, and chronic heart failure. Among eight ML models, the RF model achieved the best performance with the highest area under the curve (AUC) (AUC = 0.774). Compared to the current clinical assessment strategies, the RF model outperformed the Wells score ( P = 0.030) and was not inferior to any other clinical probability assessment strategy. The AUC of the RF model for diagnosing APE onset in internal validation set was 0.726.

CONCLUSIONS

Based on RF algorithm, a novel prediction model was finally constructed for APE diagnosis. When compared to the current clinical assessment strategies, the RF model achieved better diagnostic efficacy and accuracy. Therefore, the ML algorithm can be a useful tool in assisting with the diagnosis of APE.

D-dimer Levels for the exclusion of pulmonary embolism: making sense of international guideline recommendations

Several international guidelines provide recommendations around the use of D-dimer testing for exclusion of pulmonary embolism, including the appropriate D-dimer threshold (or cutoff), but there is no consensus among them. We briefly discuss guideline variation, performance characteristics, and limitations of commercially available D-dimer assays in this setting, referencing the Clinical and Laboratory Standards Institute guidelines that recommend immunoassays with high sensitivity (≥97%) and negative predictive value (≥98%). While age-adjusted D-dimer and pretest-adjusted D-dimer are considered a safe strategy across predefined patient subgroups, clinicians need to recognize the different performance characteristics of D-dimer assays to enable safe clinical decisions for their patients. Importantly, D-dimer values must be correlated not only to clinical findings but also interpreted within the context of the accuracy and precision of the specific testing modality, adhering to manufacturer specifications that are approved by regulatory authorities.

External validation of the PEGeD diagnostic algorithm for suspected pulmonary embolism in an independent cohort

Sequential diagnostic algorithms are used in the case of suspected pulmonary embolism (PE). The PEGeD study proposed a new diagnostic strategy to reduce the use of computed tomography pulmonary angiography (CTPA). We aimed to externally validate this diagnostic strategy in an independent cohort. We analyzed data from 3 prospective studies of outpatients with suspected PE. As per the PEGeD algorithm, patients were classified as having a low, moderate, or high clinical pretest probability (C-PTP). PE was excluded with a D-dimer <1000 ng/mL in case of low C-PTP and <500 ng/mL in case of moderate C-PTP. We assessed the yield and safety of this approach and compared them with those of previously validated algorithms. Among the 3308 evaluated patients, 1615 (49%) patients could have had PE excluded according to the PEGeD algorithm, without the need for imaging. Of these patients, 38 (2.3%; 95% confidence interval [CI], 1.7-3.2) were diagnosed with a symptomatic PE at initial testing or during the 3-month follow-up. On further analysis, 36 patients out of these 38 patients had a positive age-adjusted D-dimer. The risk of venous thromboembolic events among the 414 patients with a D-dimer <1000 ng/mL but above the age-adjusted D-dimer cut-off was 36 of 414 (8.7%; 95% CI, 6.4-11.8). We provide external validation of the PEGeD algorithm in an independent cohort. Compared with standard algorithms, the PEGeD decreased the number of CTPA examinations. However, caution is required in patients with a low C-PTP and a D-dimer <1000 ng/mL but above their age-adjusted D-dimer cut-off.

Diagnostic Approach for Venous Thromboembolism in Cancer Patients

Venous thromboembolic disease (VTE) is a common complication in cancer patients. The currently recommended VTE diagnostic approach involves a step-by-step algorithm, which is based on the assessment of clinical probability, D-dimer measurement, and/or diagnostic imaging. While this diagnostic strategy is well validated and efficient in the noncancer population, its use in cancer patients is less satisfactory. Cancer patients often present nonspecific VTE symptoms resulting in less discriminatory power of the proposed clinical prediction rules. Furthermore, D-dimer levels are often increased because of a hypercoagulable state associated with the tumor process. Consequently, the vast majority of patients require imaging tests. In order to improve VTE exclusion in cancer patients, several approaches have been developed. The first approach consists of ordering imaging tests to all patients, despite overexposing a population known to have mostly multiple comorbidities to radiations and contrast products. The second approach consists of new diagnostic algorithms based on clinical probability assessment with different D-dimer thresholds, e.g., the YEARS algorithm, which shows promise in improving the diagnosis of PE in cancer patients. The third approach uses an adjusted D-dimer threshold, to age, pretest probability, clinical criteria, or other criteria. These different diagnostic strategies have not been compared head-to-head. In conclusion, despite having several proposed diagnostic approaches to diagnose VTE in cancer patients, we still lack a dedicated diagnostic algorithm specific for this population.

Implementation of the YEARS algorithm to optimise pulmonary embolism diagnostic workup in the emergency department

BACKGROUND

Excessive use of CT pulmonary angiography (CTPA) to investigate pulmonary embolism (PE) in the emergency department (ED) contributes to adverse patient outcomes. Non-invasive D-dimer testing, in the context of a clinical algorithm, may help decrease unnecessary imaging but this has not been widely implemented in Canadian EDs.

AIM

To improve the diagnostic yield of CTPA for PE by 5% (absolute) within 12 months of implementing the YEARS algorithm.

MEASURES AND DESIGN

Single centre study of all ED patients >18 years investigated for PE with D-dimer and/or CTPA between February 2021 and January 2022. Primary and secondary outcomes were the diagnostic yield of CTPA and frequency of CTPA ordered compared with baseline. Process measures included the percentage of D-dimer tests ordered with CTPA and CTPAs ordered with D-dimers <500 µg/L Fibrinogen Equivalent Units (FEU). The balancing measure was the number of PEs identified on CTPA within 30 days of index visit. Multidisciplinary stakeholders developed plan- do-study-act cycles based on the YEARS algorithm.

RESULTS

Over 12 months, 2695 patients were investigated for PE, of which 942 had a CTPA. Compared with baseline, the CTPA yield increased by 2.9% (12.6% vs 15.5%, 95% CI -0.06% to 5.9%) and the proportion of patients that underwent CTPA decreased by 11.4% (46.4% vs 35%, 95% CI -14.1% to -8.8%). The percentage of CTPAs ordered with a D-dimer increased by 26.3% (30.7% vs 57%, 95% CI 22.2% 30.3%) and there were two missed PE (2/2695, 0.07%).

IMPACT

Implementing the YEARS criteria may safely improve the diagnostic yield of CTPAs and reduce the number of CTPAs completed without an associated increase in missed clinically significant PEs. This project provides a model for optimising the use of CTPA in the ED.

Comparison of YEARS and Adjust-Unlikely D-dimer Testing for Pulmonary Embolism in the Emergency Department

STUDY OBJECTIVE

We prospectively assessed the diagnostic accuracy of YEARS and a modified age-adjusted clinical decision rule ("Adjust-Unlikely") for pulmonary embolism (PE) testing in the emergency department.

METHODS

This study was conducted in tertiary care Canadian emergency departments. When the D-dimer was <500 ng/ml, PE was excluded. Pulmonary imaging for PE was performed when the D-dimer was ≥500 ng/ml. Patients were followed for 30 days, and PE outcomes were independently adjudicated. Physicians systematically recorded the presence or absence of YEARS items (PE most likely, hemoptysis, signs of deep venous thrombosis) prior to D-dimer testing and imaging. We analyzed the diagnostic accuracy of YEARS and the "Adjust-Unlikely" rule. Age adjustment (age x 10 in those >50 years old) was applied in patients where PE was not the most likely diagnosis and 500 ng/ml threshold when PE was most likely.

RESULTS

One thousand seven hundred three patients were included, median age 62 (50, 74), 58% female, PE prevalence 8.0%. YEARS sensitivity for PE diagnosis was 92.6% (87.0, 96.0%) and specificity 45.0% (42.5, 47.5%). Adjust-Unlikely sensitivity was 100.0% (97.2, 100.0%) and specificity 32.4% (30.1, 34.8%). Posttest probability of PE in the group of patients with PE excluded by D-dimer between 500 ng/ml and the adjusted limit was 2.8% (1.6, 5.1%) for YEARS and 0.0% (0.0, 2.6%) for the "Adjust-Unlikely" rule.

CONCLUSION

The "Adjust-Unlikely" rule would modestly reduce imaging and identify all cases of PE. YEARS would substantially reduce imaging but miss 1 in 14 cases of PE.

Systematic screening versus clinical gestalt in the diagnosis of pulmonary embolism in COVID-19 patients in the emergency department

BACKGROUND

Diagnosing concomitant pulmonary embolism (PE) in COVID-19 patients remains challenging. As such, PE may be overlooked. We compared the diagnostic yield of systematic PE-screening based on the YEARS-algorithm to PE-screening based on clinical gestalt in emergency department (ED) patients with COVID-19.

METHODS

We included all ED patients who were admitted because of COVID-19 between March 2020 and February 2021. Patients already receiving anticoagulant treatment were excluded. Up to April 7, 2020, the decision to perform CT-pulmonary angiography (CTPA) was based on physician's clinical gestalt (clinical gestalt cohort). From April 7 onwards, systematic PE-screening was performed by CTPA if D-dimer level was ≥1000 ug/L, or ≥500 ug/L in case of ≥1 YEARS-item (systematic screening cohort).

RESULTS

1095 ED patients with COVID-19 were admitted. After applying exclusion criteria, 289 were included in the clinical gestalt and 574 in the systematic screening cohort. The number of PE diagnoses was significantly higher in the systematic screening cohort compared to the clinical gestalt cohort: 8.2% vs. 1.0% (3/289 vs. 47/574; p<0.001), even after adjustment for differences in patient characteristics (adjusted OR 8.45 (95%CI 2.61-27.42, p<0.001) for PE diagnosis). In multivariate analysis, D-dimer (OR 1.09 per 1000 μg/L increase, 95%CI 1.06-1.13, p<0.001) and CRP >100 mg/L (OR 2.78, 95%CI 1.37-5.66, p = 0.005) were independently associated with PE.

CONCLUSION

In ED patients with COVID-19, the number of PE diagnosis was significantly higher in the cohort that underwent systematic PE screening based on the YEARS-algorithm in comparison with the clinical gestalt cohort, with a number needed to test of 7.1 CTPAs to detect one PE.

Role of Physical Therapists in the Management of Individuals at Risk for or Diagnosed With Venous Thromboembolism: Evidence-Based Clinical Practice Guideline 2022

No matter the practice setting, physical therapists work with patients who are at risk for or who have a history of venous thromboembolism (VTE). In 2016, the first clinical practice guideline (CPG) addressing the physical therapist management of VTE was published with support by the American Physical Therapy Association's Academy of Cardiovascular and Pulmonary Physical Therapy and Academy of Acute Care, with a primary focus on lower extremity deep vein thrombosis (DVT). This CPG is an update of the 2016 CPG and contains the most current evidence available for the management of patients with lower extremity DVT and new key action statements (KAS), including guidance on upper extremity DVT, pulmonary embolism, and special populations. This document will guide physical therapist practice in the prevention of and screening for VTE and in the management of patients who are at risk for or who have been diagnosed with VTE. Through a systematic review of published studies and a structured appraisal process, KAS were written to guide the physical therapist. The evidence supporting each action was rated, and the strength of statement was determined. Clinical practice algorithms based on the KAS were developed that can assist with clinical decision-making. Physical therapists, along with other members of the health care team, should implement these KAS to decrease the incidence of VTE, improve the diagnosis and acute management of VTE, and reduce the long-term complications of VTE.

Pulmonary embolism diagnosis: clinical assessment at the front door

This first of two practice reviews addresses pulmonary embolism (PE) diagnosis considering important aspects of PE clinical presentation and comparing evidence-based PE testing strategies. A companion paper addresses the management of PE. Symptoms and signs of PE are varied, and emergency physicians frequently use testing to 'rule out' the diagnosis in people with respiratory or cardiovascular symptoms. The emergency clinician must balance the benefit of reassuring negative PE testing with the risks of iatrogenic harms from over investigation and overdiagnosis.

Derivation and Validation of a 4-Level Clinical Pretest Probability Score for Suspected Pulmonary Embolism to Safely Decrease Imaging Testing

Importance

In patients with suspected pulmonary embolism (PE), overuse of diagnostic imaging is an important point of concern.

Objective

To derive and validate a 4-level pretest probability rule (4-Level Pulmonary Embolism Clinical Probability Score [4PEPS]) that makes it possible to rule out PE solely on clinical criteria and optimized D-dimer measurement to safely decrease imaging testing for suspected PE.

Design, Setting, and Participants

This study included consecutive outpatients suspected of having PE from US and European emergency departments. Individual data from 3 merged management studies (n = 11 114; overall prevalence of PE, 11%) were used for the derivation cohort and internal validation cohort. The external validation cohorts were taken from 2 independent studies, the first with a high PE prevalence (n = 1548; prevalence, 21.5%) and the second with a moderate PE prevalence (n = 1669; prevalence, 11.7%). A prior definition of pretest probability target values to achieve a posttest probability less than 2% was used on the basis of the negative likelihood ratios of D-dimer. Data were collected from January 2003 to April 2016, and data were analyzed from June 2018 to August 2019.

Main Outcomes and Measures

The rate of PE diagnosed during the initial workup or during follow-up and the rate of imaging testing.

Results

Of the 5588 patients in the derivation cohort, 3441 (61.8%) were female, and the mean (SD) age was 52 (18.5) years. The 4PEPS comprises 13 clinical variables scored from -2 to 5. It results in the following strategy: (1) very low probability of PE if 4PEPS is less than 0: PE ruled out without testing; (2) low probability of PE if 4PEPS is 0 to 5: PE ruled out if D-dimer level is less than 1.0 μg/mL; (3) moderate probability of PE if 4PEPS is 6 to 12: PE ruled out if D-dimer level is less than the age-adjusted cutoff value; (4) high probability of PE if 4PEPS is greater than 12: PE ruled out by imaging without preceding D-dimer test. In the first and the second external validation cohorts, the area under the receiver operator characteristic curves were 0.79 (95% CI, 0.76 to 0.82) and 0.78 (95% CI, 0.74 to 0.81), respectively. The false-negative testing rates if the 4PEPS strategy had been applied were 0.71% (95% CI, 0.37 to 1.23) and 0.89% (95% CI, 0.53 to 1.49), respectively. The absolute reductions in imaging testing were -22% (95% CI, -26 to -19) and -19% (95% CI, -22 to -16) in the first and second external validation cohorts, respectively. The 4PEPS strategy compared favorably with all recent strategies in terms of imaging testing.

Conclusions and Relevance

The 4PEPS strategy may lead to a substantial and safe reduction in imaging testing for patients with suspected PE. It should now be tested in a formal outcome study.

Suspected Acute Pulmonary Embolism: Gestalt, Scoring Systems, and Artificial Intelligence

Pulmonary embolism (PE) remains a diagnostic challenge in 2021. As the pathology is potentially fatal and signs and symptoms are nonspecific, further investigations are classically required. Based on the Bayesian approach, clinical probability became the keystone of the diagnostic strategy to rule out PE in the case of a negative testing. Several clinical probability assessment methods are validated: gestalt, the Wells score, or the revised Geneva score. While the debate persists as to the best way to assess clinical probability, its assessment allows for the good interpretation of the investigation results and therefore directs the correct diagnostic strategy. The wide availability of computed tomography pulmonary angiography (CTPA) resulted in a major increase in investigations with a moderate increase in diagnosis, without any notable improvement in patient outcomes. This leads to a new challenge for PE diagnosis which is the limitation of the number of testing for suspected PE. We review different strategies recently developed to achieve this goal. The last challenge concerns the implementation in clinical practice. Two approaches are developed: simplification of the strategies versus the use of digital support tools allowing more sophisticated strategies. Artificial intelligence with machine-learning algorithms will probably be a future tool to guide the physician in this complex approach concerning acute PE suspicion.