Beyond the d-dimer – Machine-learning assisted pre-test probability evaluation in patients with suspected pulmonary embolism and elevated d-dimers

Gawlitza, Joshua; Ziegelmayer, Sebastian; Wilkens, Heinrike; Jagoda, Philippe; Raczeck, Paul; Buecker, Arno; Stroeder, Jonas

doi:10.1016/j.thromres.2021.07.001

Cited by 6 publications

(3 citation statements)

References 28 publications

Supporting

Mentioning

Contrasting

Order By: Relevance

“…Kwon et al found that D-dimer levels alone were not sufficient to negate the need for CTPA ( 27 ). Gawlitza et al discovered that the predictive value of D-dimer for PE could be improved by machine-learning ( 28 ). In our study, all participants were elderly cancer patients with VTE and D-dimer levels were generally high.…”

Section: Discussionmentioning

confidence: 99%

Risk factors for pulmonary embolism in lung cancer patients with lower limb deep venous thrombosis: a case-control study

Dong¹,

Tong²,

Wang³

et al. 2023

Transl Lung Cancer Res

View full text Add to dashboard Cite

Background There is growing evidence that misdiagnosis contributes to the high mortality rate in lung cancer patients complicated with pulmonary embolism (PE). This current study analyzed predictors of PE in lung cancer patients with lower extremity deep venous thrombosis (DVT) with the aim of personalizing the treatment and management of patients with PE. Methods This retrospective case-control study included lung cancer patients with DVT at the emergency department of Shanghai Chest Hospital from January 2018 to December 2019. Patients were classified as having DVT with or without PE. The following characteristics were examined, including age, gender, smoking, hypertension, surgical trauma, hyperlipidemia, long-term bedridden status, calf swelling, coronary heart disease, chronic pulmonary disease, DVT location, DVT type, prothrombin time (PT), international normalized ratio (INR), activated partial thromboplastin time (APTT), thrombin time (TT), fibrinogen, and D-dimer, and univariate and multivariate analyses were performed. Results A total of 90 patients with lung cancer and DVT were analyzed, of whom 60% (54/90) had PE. Those variables independently associated to PE were hypertension [odds ratio (OR): 7.883, 95% confidence interval (CI): 2.038–30.495, P=0.003], long-term bedridden status (OR: 4.166, 95% CI: 1.236–14.044, P=0.021), and D-dimer levels (OR: 2.123, 95% CI: 1.476–3.053, P=0.000) were identified as independent risk factors for PE. The cut-off value of the receiver operating characteristic (ROC) curve for predicting PE by presented scoring system according to the risk factors was 1.5 and the area under the curve (AUC) was 0.84 (P<0.001). Conclusions Hypertension, being bedridden for an extended period, and elevated serum D-dimer levels were independent risk factors of PE in lung cancer patients with lower extremity DVT. Novel strategies for patient management should be developed to decrease the risk of PE.

show abstract

Section: Discussionmentioning

confidence: 99%

Risk factors for pulmonary embolism in lung cancer patients with lower limb deep venous thrombosis: a case-control study

Dong¹,

Tong²,

Wang³

et al. 2023

Transl Lung Cancer Res

View full text Add to dashboard Cite

show abstract

“…Overall, 15 studies explored ML-based models to assist VTE diagnosis, either in the form of a pretest probability or to assist diagnosis after clinical presentation (Supplementary Table 4) [79][80][81][82][83][84][85][86][87][88][89][90][91][92][93] . Most of them do not describe clearly preprocessing steps, splitting/cross-validation, hyperparameters, and other performance metrics.…”

Section: Diagnosis Of Vtementioning

confidence: 99%

Machine Learning-Based Predictive Models for Patients with Venous Thromboembolism: A Systematic Review

Danilatou,

Dimopoulos,

Kostoulas

et al. 2024

Thromb Haemost

View full text Add to dashboard Cite

Background: Venous thromboembolism (VTE) is a chronic disorder with a significant health and economic burden. Several VTE-specific Clinical Prediction Models (CPMs) have been used to assist physicians in decision-making but have several limitations. This systematic review explores if machine learning (ML) can enhance CPMs by analyzing extensive patient data derived from electronic health records (EHRs). We aimed to explore ML-CPMs applications in VTE for risk stratification, outcome prediction, diagnosis, and treatment. Methods: Three databases were searched, PubMed, Google Scholar, and IEEE electronic library. Inclusion criteria focused on studies using structured data, excluding non-English publications, studies on non-humans, and certain data types such as natural language processing and image processing. Studies involving pregnant women, cancer patients, and children were also excluded. After excluding irrelevant studies, a total of 77 studies were included. Results: Most studies report that ML-CPMs outperformed traditional CPMs in terms of receiver operating area under the curve in the four clinical domains that were explored. However, the majority of the studies were retrospective, monocentric, and lacked detailed model architecture description and external validation, which are essential for quality audit. This review identified research gaps and highlighted challenges related to standardized reporting, reproducibility, and model comparison. Conclusion: ML-CPMs show promise in improving risk assessment and individualized treatment recommendations in VTE. Apparently, there is an urgent need for standardized reporting and methodology for ML models, external validation, prospective and real-world data studies, as well as interventional studies to evaluate the impact of AI in VTE.

show abstract

“…Other studies indicated that initial blood parameters seem to enable further differentiation of patients with suspected PE and elevated d-dimers to raise pre-test probability of PE. Machine-learning-based prediction models might help to further narrow down CT indications in future ( 13 ). Other findings showed that syncope, systolic blood pressure, oxygen saturation, white blood cell, neutrophil percentage, and others, are crucial for the feature selection to assess the severity of PE ( 14 ).…”

Section: Introductionmentioning

confidence: 99%

Developing a nomogram-based scoring model to estimate the risk of pulmonary embolism in respiratory department patients suspected of pulmonary embolisms

Lanfang,

Xu,

Jun

et al. 2023

Front. Med.

View full text Add to dashboard Cite

ObjectivePulmonary embolisms (PE) are clinically challenging because of their high morbidity and mortality. This study aimed to create a nomogram to accurately predict the risk of PE in respiratory department patients in order to enhance their medical treatment and management.MethodsThis study utilized a retrospective method to collect information on medical history, complications, specific clinical characteristics, and laboratory biomarker results of suspected PE patients who were admitted to the respiratory department at Affiliated Dongyang Hospital of Wenzhou Medical University between January 2012 and December 2021. This study involved a total of 3,511 patients who were randomly divided into a training group (six parts) and a validation group (four parts) based on a 6:4 ratio. The LASSO regression and multivariate logistic regression were used to develop a scoring model using a nomogram. The performance of the model was evaluated using receiver operating characteristic curve (AUC), calibration curve, and clinical decision curve.ResultsOur research included more than 50 features from 3,511 patients. The nomogram-based scoring model was established using six predictive features including age, smoke, temperature, systolic pressure, D-dimer, and fibrinogen, which achieved AUC values of 0.746 in the training cohort (95% CI 0.720–0.765) and 0.724 in the validation cohort (95% CI 0.695–0.753). The results of the calibration curve revealed a strong consistency between probability predicted by the nomogram and actual probability. The decision curve analysis (DCA) also demonstrated that the nomogram-based scoring model produced a favorable net clinical benefit.ConclusionIn this study, we successfully developed a novel numerical model that can predict the risk of PE in respiratory department patients suspected of PE, which can not only appropriately select PE prevention strategies but also decrease unnecessary computed tomographic pulmonary angiography (CTPA) scans and their adverse effects.

show abstract

Beyond the d-dimer – Machine-learning assisted pre-test probability evaluation in patients with suspected pulmonary embolism and elevated d-dimers

Cited by 6 publications

References 28 publications

Risk factors for pulmonary embolism in lung cancer patients with lower limb deep venous thrombosis: a case-control study

Risk factors for pulmonary embolism in lung cancer patients with lower limb deep venous thrombosis: a case-control study

Machine Learning-Based Predictive Models for Patients with Venous Thromboembolism: A Systematic Review

Developing a nomogram-based scoring model to estimate the risk of pulmonary embolism in respiratory department patients suspected of pulmonary embolisms

Contact Info

Product

Resources

About