Development and Validation of a Nomogram for Lower Extremity Deep Venous Thrombosis in Patients after Acute Stroke

Leveraging machine learning for enhanced and interpretable risk prediction of venous thromboembolism in acute ischemic stroke care

BACKGROUND

Venous thromboembolism (VTE) is a life-threatening complication commonly occurring after acute ischemic stroke (AIS), with an increased risk of mortality. Traditional risk assessment tools lack precision in predicting VTE in AIS patients due to the omission of stroke-specific factors.

METHODS

We developed a machine learning model using clinical data from patients with acute ischemic stroke (AIS) admitted between December 2021 and December 2023. Predictive models were developed using machine learning algorithms, including Gradient Boosting Machine (GBM), Random Forest (RF), and Logistic Regression (LR). Feature selection involved stepwise logistic regression and LASSO, with SHapley Additive exPlanations (SHAP) used to enhance model interpretability. Model performance was evaluated using area under the receiver operating characteristic curve (AUC), sensitivity, specificity, positive predictive value (PPV), and negative predictive value (NPV).

RESULTS

Among the 1,632 AIS patients analyzed, 4.17% developed VTE. The GBM model achieved the highest predictive accuracy with an AUC of 0.923, outperforming other models such as Random Forest and Logistic Regression. The model demonstrated strong sensitivity (90.83%) and specificity (93.83%) in identifying high-risk patients. SHAP analysis revealed that key predictors of VTE risk included elevated D-dimer levels, premorbid mRS, and large vessel occlusion, offering clinicians valuable insights for personalized treatment decisions.

CONCLUSION

This study provides an accurate and interpretable method to predict VTE risk in patients with AIS using the GBM model, potentially improving early detection rates and reducing morbidity. Further validation is needed to assess its broader clinical applicability.

Machine learning-based prediction model of lower extremity deep vein thrombosis after stroke

This study aimed to apply machine learning (ML) techniques to develop and validate a risk prediction model for post-stroke lower extremity deep vein thrombosis (DVT) based on patients' limb function, activities of daily living (ADL), clinical laboratory indicators, and DVT preventive measures. We retrospectively analyzed 620 stroke patients. Eight ML models-logistic regression (LR), support vector machine (SVM), random forest (RF), decision tree (DT), neural network (NN), extreme gradient boosting (XGBoost), Bayesian (NB), and K-nearest neighbor (KNN)-were used to build the model. These models were extensively evaluated using ROC curves, AUC, PR curves, PRAUC, accuracy, sensitivity, specificity, and clinical decision curves (DCA). Shapley's additive explanation (SHAP) was used to determine feature importance. Finally, based on the optimal ML algorithm, different functional feature set models were compared with the Padua scale to select the best feature set model. Our results indicated that the RF algorithm demonstrated superior performance in various evaluation metrics, including AUC (0.74/0.73), PRAUC (0.58/0.58), accuracy (0.75/0.77), and sensitivity (0.78/0.80) in both the training set and test set. DCA analysis revealed that the RF model had the highest clinical net benefit. SHAP analysis showed that D-dimer had the most significant influence on DVT, followed by age, Brunnstrom stage (lower limb), prothrombin time (PT), and mobility ability. The RF algorithm can predict post-stroke DVT to guide clinical practice.

Venous Thromboembolism Prevention in Rehabilitation: A Review and Practice Suggestions

ABSTRACT

Venous thromboembolism is a frequent complication of acute hospital care, and this extends to inpatient rehabilitation. The timely use of appropriate thromboprophylaxis in patients who are at risk is a strong, evidence-based patient safety priority that has reduced clinically important venous thromboembolism, associated mortality and costs of care. While there has been extensive research on optimal approaches to venous thromboembolism prophylaxis in acute care, there is a paucity of high-quality evidence specific to patients in the rehabilitation setting, and there are no clinical practice guidelines that make recommendations for (or against) thromboprophylaxis across the broad spectrum of rehabilitation patients. Herein, we provide an evidence-informed review of the topic with practice suggestions. We conducted a series of literature searches to assess the risks of venous thromboembolism and its prevention related to inpatient rehabilitation as well as in major rehabilitation subgroups. Mobilization alone does not eliminate the risk of venous thromboembolism after another thrombotic insult. Low molecular weight heparins and direct oral anticoagulants are the principal current modalities of thromboprophylaxis. Based on the literature, we make suggestions for venous thromboembolism prevention and include an approach for consideration by rehabilitation units that can be aligned with local practice.

Implementation of early prophylaxis for deep-vein thrombosis in intracerebral hemorrhage patients: an observational study from the Chinese Stroke Center Alliance

BACKGROUND

There is substantial evidence to support the use of several methods for preventing deep-vein thrombosis (DVT) following intracerebral hemorrhage (ICH). However, the extent to which these measures are implemented in clinical practice and the factors influencing patients' receipt of preventive measures remain unclear. Therefore, we aimed to evaluate the rate of the early implementation of DVT prophylaxis and the factors associated with its success in patients with ICH.

METHODS

This study enrolled 49,950 patients with spontaneous ICH from the Chinese Stroke Center Alliance (CSCA) between August 2015 and July 2019. Early DVT prophylaxis implementation was defined as an intervention occurring within 48 h after admission. Univariate and multivariate logistic regression analyses were conducted to identify the rate and factors associated with the implementation of early prophylaxis for DVT in patients with ICH.

RESULTS

Among the 49,950 ICH patients, the rate of early DVT prophylaxis implementation was 49.9%, the rate of early mobilization implementation was 29.49%, and that of pharmacological prophylaxis was 2.02%. Factors associated with an increased likelihood of early DVT prophylaxis being administered in the multivariable model included receiving early rehabilitation therapy (odds ratio [OR], 2.531); admission to stroke unit (OR 2.231); admission to intensive care unit (OR 1.975); being located in central (OR 1.879) or eastern regions (OR 1.529); having a history of chronic obstructive pulmonary disease (OR 1.292), ischemic stroke (OR 1.245), coronary heart disease or myocardial infarction (OR 1.2); taking antihypertensive drugs (OR 1.136); and having a higher Glasgow Coma Scale (GCS) score (OR 1.045). Conversely, being male (OR 0.936), being hospitalized in tertiary hospitals (OR 0.778), and having a previous intracranial hemorrhage (OR 0.733) were associated with a lower likelihood of early DVT prophylaxis being administered in patients with ICH.

CONCLUSIONS

The implementation rate of early DVT prophylaxis among Chinese patients with ICH was subpar, with pharmacological prophylaxis showing the lowest prevalence. Various controllable factors exerted an impact on the implementation of early DVT prophylaxis in this population.

Systematic review and meta-analysis of studies comparing baseline D-dimer level in stroke patients with or without cancer: Strength of current evidence

Objectives

D-dimer levels are increased in stroke and cancer. Cancer patients are at a higher risk of stroke. However, the evidence is unclear if high D-dimer in stroke patients can suggest the diagnosis of concomitant cancer or the development of stroke in a cancer patient. The objective is to assess the evidence available on the baseline D-dimer level in stroke patients with and without cancer.

Materials and Methods

We conducted the systematic review and meta-analysis using the Preferred Reporting Items for Systematic Reviews and Meta-analysis guidelines. We searched PUBMED, Cochrane, ScienceDirect, and Scopus for potentially eligible articles published till June 2023. All the review steps were iterative and done independently by two reviewers. The Newcastle-Ottawa scale tool was used to assess the quality of included studies for case control and cohort studies and the Agency for Healthcare Research and Quality tool for cross-sectional studies. The qualitative synthesis is presented narratively, and quantitative synthesis is shown in the forest plot using the random effects model. I2 of more than 60% was considered as high heterogeneity.

Results

The searches from all the databases yielded 495 articles. After the study selection process, six papers were found eligible for inclusion in the qualitative and quantitative synthesis. In the present systematic review, 2651 patients with ischemic infarcts are included of which 404 (13.97%) patients had active cancer while 2247 (86.02%) did not. The studies included were of high quality and low risk of bias. There were significantly higher baseline D-dimer levels in stroke patients with cancer than in non-cancer patients with a mean difference of 4.84 (3.07-6.60) P < 0.00001.

Conclusion

D-dimer is a simple and relatively non-expensive biomarker that is increased to significant levels in stroke patients, who have cancer and therefore may be a tool to predict through screening for active or occult cancer in stroke patients.

Risk prediction models for deep venous thrombosis in patients with acute stroke: A systematic review and meta-analysis

BACKGROUND

The number of risk prediction models for deep venous thrombosis (DVT) in patients with acute stroke is increasing, while the quality and applicability of these models in clinical practice and future research remain unknown.

OBJECTIVE

To systematically review published studies on risk prediction models for DVT in patients with acute stroke.

DESIGN

Systematic review and meta-analysis of observational studies.

METHODS

China National Knowledge Infrastructure (CNKI), Wanfang Database, China Science and Technology Journal Database (VIP), SinoMed, PubMed, Web of Science, The Cochrane Library, Cumulative Index to Nursing and Allied Health Literature (CINAHL) and Embase were searched from inception to November 7, 2022. Data from selected studies were extracted, including study design, data source, outcome definition, sample size, predictors, model development and performance. The Prediction Model Risk of Bias Assessment Tool (PROBAST) checklist was used to assess the risk of bias and applicability.

RESULTS

A total of 940 studies were retrieved, and after the selection process, nine prediction models from nine studies were included in this review. All studies utilized logistic regression to establish DVT risk prediction models. The incidence of DVT in patients with acute stroke ranged from 0.4 % to 28 %. The most frequently used predictors were D-dimer and age. The reported area under the curve (AUC) ranged from 0.70 to 0.912. All studies were found to have a high risk of bias, primarily due to inappropriate data sources and poor reporting of the analysis domain. The pooled AUC value of the five validated models was 0.76 (95 % confidence interval: 0.70-0.81), indicating a fair level of discrimination.

CONCLUSION

Although the included studies reported a certain level of discrimination in the prediction models of DVT in patients with acute stroke, all of them were found to have a high risk of bias according to the PROBAST checklist. Future studies should focus on developing new models with larger samples, rigorous study designs, and multicenter external validation.

REGISTRATION

The protocol for this study is registered with PROSPERO (registration number: CRD42022370287).

Predictors and outcomes of deep venous thrombosis in patients with acute ischemic stroke: results from the Chinese Stroke Center Alliance

BACKGROUND

No large-scale, multicenter studies have explored the incidence rate and predictors of deep vein thrombosis (DVT) in patients with acute ischemic stroke (AIS). We aimed to determine the risk factors of DVT, and assess the association between DVT and clinical outcomes in AIS patients.

METHODS

In total, 106,612 patients with AIS enrolled in the Chinese Stroke Center Alliance between August 2015 and July 2019 were included. The predictors of DVT in AIS patients were screened based on the logistic regression analysis for the comparison of the characteristics and clinical outcomes of patients with and without DVT.

RESULTS

The overall incidence of DVT after AIS was 4.7%. Factors associated with increased incidence of DVT included advanced age, female sex, high admission National Institutes of Health Stroke Scale score, history of cerebral hemorrhage, transient ischemic attack (TIA), dyslipidemia, atrial fibrillation, and peripheral vascular disease, International Normalized Ratio (INR) <0.8 or >1.5, and blood uric acid >420 μmol/L. Ambulation and early antithrombotic therapy were associated with a lower incidence of DVT. Patients with DVT was associated with longer hospital stay (OR=1.44, 95% CI: 1.35-1.54), and higher in-hospital mortality (OR=1.68, 95% CI: 1.25-2.27).

CONCLUSIONS

This large-scale, multi-center study showed that the occurrence of DVT in AIS patients is associated with various modifiable and objective indicators, such as abnormal INR and uric acid >420 μmol/L. Ambulatory status and early antithrombotic therapy can reduce the occurrence of DVT in AIS patients. In AIS patients, DVT may prolong the hospital stay and increase the risk of in-hospital mortality. Future research should focus on the clinical implementation of existing evidence on DVT prevention in AIS patients.

Stroke Recovery Is a Journey: Prediction and Potentials of Motor Recovery after a Stroke from a Practical Perspective

Stroke recovery is a journey. Stroke survivors can face many consequences that may last the rest of their lives. Assessment of initial impairments allows reasonable prediction of biological spontaneous recovery at 3 to 6 months for a majority of survivors. In real-world clinical practice, stroke survivors continue to improve their motor function beyond the spontaneous recovery period, but management plans for maximal recovery are not well understood. A model within the international classification of functioning (ICF) theoretical framework is proposed to systematically identify opportunities and potential barriers to maximize and realize the potentials of functional recovery from the acute to chronic stages and to maintain their function in the chronic stages. Health conditions of individuals, medical and neurological complications can be optimized under the care of specialized physicians. This permits stroke survivors to participate in various therapeutic interventions. Sufficient doses of appropriate interventions at the right time is critical for stroke motor rehabilitation. It is important to highlight that combining interventions is likely to yield better clinical outcomes. Caregivers, including family members, can assist and facilitate targeted therapeutic exercises for these individuals and can help stroke survivors comply with medical plans (medications, visits), and provide emotional support. With health optimization, comprehensive rehabilitation, support from family and caregivers and a commitment to a healthy lifestyle, many stroke survivors can overcome barriers and achieve potentials of maximum recovery and maintain their motor function in chronic stages. This ICF recovery model is likely to provide a guidance through the journey to best achieve stroke recovery potentials.

A nomogram to predict postoperative deep vein thrombosis in patients with femoral fracture: a retrospective study

OBJECTIVE

The implementation of more active anticoagulant prevention and treatment measures has indeed led to a significant reduction in the incidence of perioperative deep vein thrombosis (DVT) among patients with bone trauma. However, it is important to note that despite these efforts, the incidence of DVT still remains relatively high. According to the Caprini score, all patients undergoing major orthopedic surgery were defined as the high-risk group for DVT. Stratifying the risk further within high-risk groups for DVT continues to present challenges. As a result, the commonly used Caprini score during the perioperative period is not applicable to orthopedic patients. We attempt to establish a specialized model to predict postoperative DVT risk in patients with femoral fracture.

METHODS

We collected the clinical data of 513 patients undergoing femoral fracture surgery in our hospital from May 2018 to December 2019. According to the independent risk factors of DVT obtained by univariate and multivariate logistic regression analysis, the corresponding nomogram model was established and verified internally. The discriminative capacity of nomogram was evaluated by receiver operating characteristic (ROC) curve and area under the curve (AUC). The calibration curve used to verify model consistency was the fitted line between predicted and actual incidences. The clinical validity of the nomogram model was assessed using decision curve analysis (DCA) which could quantify the net benefit of different risk threshold probabilities. Bootstrap method was applied to the internal validation of the nomogram model. Furthermore, a comparison was made between the Caprini score and the developed nomogram model.

RESULTS

The Caprini scores of subjects ranged from 5 to 17 points. The incidence of DVT was not positively correlated with the Caprini score. The predictors of the nomogram model included 10 risk factors such as age, hypoalbuminemia, multiple trauma, perioperative red blood cell infusion, etc. Compared with the Caprini scale (AUC = 0.571, 95% CI 0.479-0.623), the calibration accuracy and identification ability of nomogram were higher (AUC = 0.865,95% CI 0.780-0.935). The decision curve analysis (DCA) indicated the clinical effectiveness of nomogram was higher than the Caprini score.

CONCLUSIONS

The nomogram was established to effectively predict postoperative DVT in patients with femoral fracture. To further reduce the incidence, more specialized risk assessment models for DVT should take into account the unique risk factors and characteristics associated with specific patient populations.

The interaction effect of transfusion history and previous stroke history on the risk of venous thromboembolism in stroke patients: a prospective cohort study

BACKGROUND

Blood transfusion and previous stroke history are two independent risk factors of venous thromboembolism (VTE) in stroke patients. Whether the potential interaction of transfusion history and previous stroke history is associated with a greater risk of VTE remains unclear. This study aims to explore whether the combination of transfusion history and previous stroke history increases the risk of VTE among Chinese stroke patients.

METHODS

A total of 1525 participants from the prospective Stroke Cohort of Henan Province were enrolled in our study. Multivariate logistic regression models were used to explore the associations among transfusion history, previous stroke history and VTE. The interaction was evaluated on both multiplicative and additive scales. The odds ratio (95% CI), relative excess risk of interaction (RERI), attributable proportion (AP), and synergy index (S) of interaction terms were used to examine multiplicative and additive interactions. Finally, we divided our population into two subgroups by National Institutes of Health Stroke Scale (NIHSS) score and re-evaluated the interaction effect in both scales.

RESULTS

A total of 281 (18.4%) participants of 1525 complicated with VTE. Transfusion and previous stroke history were associated with an increased risk of VTE in our cohort. In the multiplicative scale, the combination of transfusion and previous stroke history was statistically significant on VTE in both unadjusted and adjusted models (P<0.05). For the additive scale, the RERI shrank to 7.016 (95% CI: 1.489 ~ 18.165), with the AP of 0.650 (95% CI: 0.204 ~ 0.797) and the S of 3.529 (95% CI: 1.415 ~ 8.579) after adjusting for covariates, indicating a supra-additive effect. In subgroups, the interaction effect between transfusion history and previous stroke history was pronouncedly associated with the increased risk of VTE in patients with NIHSS score > 5 points (P<0.05).

CONCLUSIONS

Our results suggest that there may be a potential synergistic interaction between transfusion history and previous stroke history on the risk of VTE. Besides, the percentage of VTE incidence explained by interaction increased with the severity of stroke. Our findings will provide valuable evidence for thromboprophylaxis in Chinese stroke patients.

A nomogram for individualized prediction of lower extremity deep venous thrombosis in stroke patients: A retrospective study

To develop and validate a nomogram for individualized prediction of lower extremity deep venous thrombosis (DVT) in stroke patients based on extremity function and daily living ability of stroke patients. In this study, 423 stroke patients admitted to the Rehabilitation Medical Center of the First Affiliated Hospital of Nanjing Medical University from December 2015 to February 2019 were taken as the subjects, who were divided into the DVT group (110) and No-DVT group (313) based on the existence of DVT. Inter-group comparison of baseline data was performed by 1-way Analysis of Variance, Kruskal-Wallis rank-sum test, or Pearson chi-square test. Data dimensions and predictive variables were selected by least absolute shrinkage and selection operator (LASSO); the prediction model was developed and the nomogram was prepared by binary logistics regression analysis; the performance of the nomogram was identified by the area under the receiver operating characteristic curve (AUC), Harrell's concordance index, and calibration curve; and the clinical effectiveness of the model was analyzed by clinical decision curve analysis. Age, Brunnstrom stage (lower extremity), and D-dimer were determined to be the independent predictors affecting DVT. The independent predictors mentioned above were developed and presented as a nomogram, with AUC and concordance index of 0.724 (95% confidence interval [CI]: 0.670-0.777), indicating the satisfactory discrimination ability of the nomogram. The P value of the results of the Hosmer-Lemeshow test was 0.732, indicating good fitting of the prediction model. Decision curve analysis showed that the clinical net benefit of this model was 6% to 50%. We developed a nomogram to predict lower extremity deep venous thrombosis in stroke patients, and the results showed that the nomogram had satisfactory prediction performance and clinical efficacy.

Association of D-dimer with short-term risk of venous thromboembolism in acutely ill medical patients: A systematic review and meta-analysis

BACKGROUND

D-dimer, a marker of ongoing procoagulant activity, has been widely used for the diagnosis of venous thromboembolism (VTE). The prognostic significance of D-dimer in stratifying VTE risk for acutely ill medical patients has not been well-established.

METHODS

A literature search was performed to collect studies that compared the incidence of short-term VTE between acutely ill medical patients with elevated or nonelevated D-dimer levels. The cutoff of D-dimer was 0.5 μg/mL or otherwise defined by included studies. The study endpoint was any occurrence of VTE (inclusive of deep vein thrombosis [DVT], pulmonary embolism, or VTE-related death) within 90 days of hospital presentation. A meta-analytic approach was employed to estimate the odds ratio (OR) with 95% CI by fitting random-effects models using the generic inverse variance weighted approach.

RESULTS

A total of 10 studies representing 31,119 acutely ill medical patients were included. Compared to those with nonelevated D-dimer levels, patients with elevated D-dimer had approximately threefold greater odds for short-term VTE within 90 days (OR, 3.28; 95% CI, 2.44 to 4.40; p < 0.0001). The association of elevated D-dimer with VTE composite (OR, 3.33; 95% CI, 2.20 to 5.02) and with DVT (OR, 3.26; 95% CI, 2.32 to 4.58) was comparable. The association was significant among patients who presented various acute medical illness (OR, 2.68; 95% CI, 2.01 to 3.58) and those who presented with acute stroke (OR, 3.25; 95% CI, 2.31 to 4.58).

CONCLUSION

Elevation of D-dimer was predictive of the occurrence of VTE within 90 days among acutely ill medical patients. PROSPERO Registration ID: CRD42021264555.

A Nomogram for Individualized Prediction of Calf Muscular Vein Thrombosis in Stroke Patients During Rehabilitation: A Retrospective Study

Objectives: To develop a nomogram for predicting calf muscle veins thrombosis (CMVT) in stroke patients during rehabilitation. Methods: We enrolled 360 stroke patients from the Rehabilitation Medicine Center from December 2015 to February 2019. Of the participants, 123 were included in the CMVT group and 237 in the no CMVT group. The least absolute shrinkage and selection operator (LASSO) regression model was applied to optimize feature selection for the model. Multivariable logistic regression analysis was applied to construct a predictive model. Performance and clinical utility of the nomogram were generated using the Harrell's concordance index, calibration curve, and decision curve analysis (DCA). Results: Age, Brunnstrom stage (lower extremity), D-dimer, and antiplatelet therapy were associated with the occurrence of CMVT. The prediction nomogram showed satisfactory performance with a concordance index of 0.718 (95% CI: 0.663-0.773) in internal verification. The Hosmer-Lemeshow test, P = .217, suggested that the model was of goodness-of-fit. In addition, the DCA demonstrated that the CMVT nomogram had a good clinical net benefit. Conclusions: We developed a nomogram that could help clinicians identify high-risk groups of CMVT in stroke patients during rehabilitation for early intervention.