Construction and validation of a risk prediction model for venous thromboembolism post-VATS in simultaneous multicentric primary lung cancers

Introduction

Lung cancer is the most commonly diagnosed cancer globally, accounting for 12.4% of the total new cancer cases. It is also the leading cause of cancer-related deaths, responsible for 18.7% of the total cancer mortality (1). With the widespread adoption of lung cancer screening and advancements in imaging technology, particularly the clinical application of high-resolution chest computed tomography (HRCT) and positron emission tomography (PET), the detection rate of synchronous multiple primary lung cancer (sMPLC) in patients has been increasing. sMPLC refers to the presence of more than one malignant lesion in the same patient’s lung, with each lesion originating from a different source (2,3). The concept was first proposed by Beyreuther in 1924, and clinical reports indicate that the proportion of sMPLC among new lung cancer cases ranges from 0.8% to 20% (4-8).

Venous thromboembolism (VTE) encompasses deep venous thrombosis (DVT) and pulmonary embolism (PE), which are two clinical manifestations of the same disease at different locations and stages. Despite clinical efforts to reduce the incidence of VTE in postoperative lung cancer patients through the promotion of minimally invasive thoracoscopic surgery, standardized perioperative anticoagulant therapy, shortened preoperative and postoperative fasting times, and early postoperative ambulation, current research still indicates that the incidence of VTE in postoperative lung cancer patients ranges from 0.2% to 26% (9-11). VTE is one of the most common causes of death within 30 days post-lung cancer surgery, and it can increase the mortality rate after lung resection from 1.2% to 8% (12). Hospitalized lung cancer patients who develop VTE are associated with a longer length of stay, higher in-hospital mortality rates, increased costs, and greater discharge disability (13).

Current studies mainly examine the risks associated with VTE following surgery for single primary lung cancer, whereas research on VTE risk after VATS in patients with sMPLC is less prevalent. This study aims to formulate a risk prediction model for VTE following VATS in sMPLC patients, offering a basis for early diagnosis and prevention of postoperative VTE in these patients. We present this article in accordance with the TRIPOD reporting checklist (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-558/rc).

Methods

Study population

A retrospective cohort analysis was conducted on 1,984 adult patients with sMPLC.

All patients underwent their first video-assisted thoracoscopic surgery (VATS) at The Second Affiliated Hospital of Army Medical University in Chongqing, China, between November 2017 and December 2024. Data were retrospectively collected from the Hospital Information System (HIS). Inclusion criteria were: (I) age 18 years or older; (II) clinically diagnosed with sMPLC; (III) resection of ≥2 primary lung cancer lesions; (IV) expected survival of more than 12 months; (V) capable of using smart devices and completing electronic questionnaires; (VI) clear consciousness, without reading, communication, or cognitive impairments, and agree to participate in the study. Exclusion criteria were: (I) preoperative Eastern Cooperative Oncology Group (ECOG) performance status score >1; (II) pregnant or breastfeeding patients; (III) continuous use of anticoagulant drugs; (IV) a history of mental illness, hearing impairment, communication disorders, or other conditions that would prevent completion of the questionnaire; (V) patients from whom complete information could not be collected; (VI) a history of surgery, chemotherapy, or radiotherapy within the 3 months prior to study participation. This study was reviewed and approved by the Ethics Committee of The Second Affiliated Hospital of Army Medical University of the Chinese People’s Liberation Army, with the ethics review number 2024-S-119-01. Individual consent for this retrospective analysis was waived. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Outcomes

The outcome indicators were the clinical diagnosis of postoperative PE confirmed by computed tomography (CT) pulmonary angiography or DVT detected by Doppler ultrasound of the extremities. All included patients were informed about the symptoms of VTE and were asked to report the timing of these symptoms. At the first postoperative month follow-up visit, DVT was diagnosed using extremity vascular ultrasound. PE was diagnosed through CT angiography, pulmonary angiography, or ventilation-perfusion scanning (used for patients with renal insufficiency or those allergic to contrast agents). All VTE events were submitted to the department’s VTE management team, which comprised experts in vascular surgery and radiology, and the final outcomes were determined by a review committee.

Statistical analysis

In this study, data were systematically obtained via incorporating the HIS, structured questionnaires, WeChat messaging, and telephonic follow-ups. Statistical analysis of the data was performed using SPSS 23.0 and R 4.2.0. Categorical variables were analyzed using χ² tests, and continuous variables were assessed with t-tests or Mann-Whitney U test for univariate analysis. Independent variables with statistical significance (P<0.05) identified through univariate analysis were further analyzed by least absolute shrinkage and selection operator (LASSO) regression. The selected variables were included in the Logistic regression analysis to establish a risk prediction model.

The Bootstrap resampling method was used for internal validation of the model on the construction dataset. The area under the receiver operating characteristic (ROC) curve was used to assess the discriminative ability of the model, calibration curves were used to evaluate the calibration of the model, and clinical decision curves were used to assess the clinical effectiveness of the model, with α=0.05 as the level of significance for testing.

Results

Patient characteristics

A total of 2,189 newly diagnosed non-small cell lung cancer (NSCLC) patients were included in this study. A total of 190 patients were excluded due to incomplete information and missing data, including 102 patients who lacked materials from Doppler ultrasound examination of the extremities, 88 patients who missed complete follow-up information, and 15 patients who had a history of DVT or PE within the 3 months prior to recruitment. Ultimately, 1,984 eligible patients were included in our study (Figure 1).

Figure 1 Flow chart of the study design and analysis. DCA, decision curve analysis; LASSO, least absolute shrinkage and selection operator; ROC, receiver operating characteristic; sMPLC, synchronous multiple primary lung cancer; VTE, venous thromboembolism.

Risk of VTE

Univariate analysis of VTE in sMPLC patients

Patients were divided into two groups based on whether they developed VTE within 1 month after discharge: the occurrence group and the non-occurrence group. Previous studies have demonstrated that thoracoscopic radical lung cancer surgeries were predominantly completed within 3 hours (180 minutes) (14). Prolonged surgical duration (≥180 minutes) has been associated with a significant elevation in postoperative inflammatory biomarkers, including C-reactive protein (CRP) and procalcitonin (PCT) levels (15). Based on the surgical time distribution characteristics in our cohort, we stratified the operative duration into two clinically relevant categories: ≥180 and <180 minutes, for comparative statistical analysis.

Univariate analysis was performed on these groups. The results showed that there were statistically significant differences (P<0.05) between the two groups in terms of age, preoperative D dimer levels, smoking history, hypertension, coronary artery disease, cerebrovascular disease, diabetes mellitus, chronic obstructive pulmonary disease (COPD), atherosclerotic plaques in the extremities, hyperlipidemia, BMI, surgical method, pathological type, intraoperative transfusion, preoperative Caprini score, postoperative Caprini score, number of primary lesions, and receipt of anticoagulant therapy (Table 1).

LASSO-logistic regression based screening variables and model construction

Using VTE as the dependent variable and 18 statistically significant indicators identified in the univariate analysis as independent variables, the optimal penalty coefficient λ was determined through a 10-fold cross-validated LASSO regression model. Cross-validation was used to find the best λ value that reduces the model’s prediction error, and the λ value within the standard error range (λ.1se) was also considered to balance model complexity and prediction performance.

At λ and λ.1se, the paths of variable coefficients with λ were plotted respectively to show the selection process of variables intuitively. Eleven potential influencing factors were selected at λ.1se, with these variables retaining non-zero coefficients under this penalty intensity. These factors were age, smoking history, coronary artery disease, cerebrovascular disease, COPD, atherosclerotic plaques in the extremities, D dimer, surgical method, intraoperative transfusion, postoperative Caprini score, and number of primary lesions. The selection process was shown in Figure 2.

Figure 2 The process of feature variable selection using LASSO regression. LASSO, least absolute shrinkage and selection operator.

Variables with statistically significant coefficients identified through LASSO regression analysis were subsequently incorporated into the multivariable logistic regression model. The assignment of values for the independent variables was as follows: smoking history: never smoked =1, quit smoking (≥1 year) =2, current smoking =3; coronary artery disease: none =0, present =1; cerebrovascular disease: none =0, present =1; COPD: none =0, present =1; atherosclerotic plaques in the extremities: none =0, present =1; surgical method: single wedge resection =1, multiple wedge resection =2, segmentectomy + wedge resection =3, single segmentectomy =4, multiple segmentectomy =5, lobectomy =6; intraoperative transfusion: none =0, present =1; postoperative Caprini score: postoperative Caprini score (low risk) =1, postoperative Caprini score (medium risk) =2, postoperative Caprini score (high risk) =3; number of primary lesions, number of primary lesions [2] =1, number of primary lesions [3–5] =2, number of primary lesions [≥6] =3; age and preoperative D dimer were entered as original values. The results showed that age, smoking history, coronary artery disease, cerebrovascular disease, COPD, Atherosclerotic plaques in the extremities, lobectomy, intraoperative transfusion, postoperative Caprini score (medium risk), postoperative Caprini score (high risk), and number of primary lesions were factors influencing the occurrence of VTE in patients (P<0.05), as shown in Table 2.

In the final model construction process, although D-dimer did not reach statistical significance in the multivariate logistic regression [P=0.063, Exp(B) =4.472], which may reflect insufficient sample size or small effect size, it does not equate to ‘no association’. According to ICH E9 guidelines, such situations require a comprehensive judgment combining clinical prior evidence (16). It was still included in the predictive model based on the following considerations: (I) existing evidence-based medical evidence indicates a dose-response relationship between D-dimer levels and venous thrombosis (17-22); (II) this indicator has been widely incorporated into thrombotic risk assessment systems in clinical practice (23-25).

Using logistic regression analysis, we derived the following equation: Logit P = age × 0.053+ smoking history (former, quit smoking ≥1 year) × 0.816 + smoking history (current) × 1.879 + coronary artery disease × 1.192 + cerebrovascular disease × 2.084 + COPD × 1.593 + atherosclerotic plaques in the extremities × 0.988 + surgical method (multiple wedge resection) × 0.078 + surgical method (segmentectomy + wedge resection) × 1.254 + surgical method (a single segmentectomy) × 1.568 + surgical method (multiple segmentectomy) × 1.417 + surgical method (lobectomy) × 1.902 + D-dimer × 1.599 + transfusion × 2.023 + postoperative Caprini score (high risk) × 0.496 + nodules × 0.899. Subsequently, we developed a nomogram (Figure 3) using the R language, which visually illustrates the influence of various risk factors on the risk of postoperative thrombosis in patients. The area under the curve (AUC) for this model was 0.917 (Figure 4), demonstrating excellent discriminative ability. The 95% confidence interval (CI) ranged from 0.894 to 0.941, with a maximum Youden index of 0.703. The optimal cutoff value was determined to be 0.061, corresponding to a prediction score of 234.262, which further validated the stability of the model’s predictive performance. With a sensitivity of 0.885 and a specificity of 0.818, the model exhibited robust performance in accurately identifying both true positives and true negatives.

Figure 3 Thrombosis risk prediction scoring system. Binary variables: 0= absent, 1= present; pre_Caprini: 1= low risk, 2= moderate risk, 3= high risk; Smoking history: A = never smoked, B = quit smoking for ≥1 year, C = current smoker; Surgical approach: A = single wedge resection, B = multiple wedge resection, C = segmentectomy plus wedge resection, D = single segmentectomy, E = multiple segmentectomy, F = lobectomy. CAD, coronary artery disease; CD, cerebrovascular disease; COPD, chronic obstructive pulmonary disease; DVT, deep venous thrombosis.

Figure 4 ROC curve of the thrombosis risk prediction model. AUC, area under the curve; ROC, receiver operating characteristic.

Validation of VTE prediction models in sMPLC patients

Using the bootstrap method with 1,000 repetitions, we assessed the discriminative ability and calibration of the model. The calibration curve (Figure 5) demonstrates a close alignment between the observed and predicted probabilities, thereby further validating the accuracy and reliability of the model.

Figure 5 Calibration curve of model prediction versus actual thrombosis occurrence.

Decision curve analysis (Figure 6) was employed to evaluate the clinical utility of the predictive model. The graph illustrates the standardized net benefit for our model, the preoperative Caprini score (pre Caprini), the postoperative Caprini score (pos Caprini), and the scenarios of either taking no action or adopting all preventive interventions. The results demonstrate that our model provides superior clinical benefits compared to the Caprini scores and can assist healthcare providers in making more informed and rational treatment decisions.

Figure 6 Comparison of standardized net benefits among different models in predicting thrombosis risk.

Discussion

The present study successfully developed and validated a robust risk prediction model for VTE following video-assisted thoracoscopic surgery (VATS) in patients with sMPLC. This model integrates 11 variables—age, smoking history, coronary artery disease, cerebrovascular disease, COPD, atherosclerotic plaques in the extremities, surgical method, intraoperative transfusion, postoperative Caprini score, and the number of primary lesions—into a clinically practical nomogram. The model demonstrated exceptional discriminative power, with an AUC of 0.917 (95% CI: 0.894–0.941), and exhibited strong calibration and superior clinical utility compared to the conventional Caprini score.

Current research has consistently identified age as a significant risk factor for VTE, with older age associated with a higher risk of VTE (26,27). Our binary logistic regression analysis revealed a 4.4% increase in VTE risk per additional year of age [odds ratio (OR) =1.044, 95% CI: 1.012–1.077, P=0.007]. The escalation in the risk of VTE associated with age may be attributed to two principal factors. First, the progressive deterioration of vascular endothelial function with increasing age can lead to the formation of atherosclerotic plaques, which in turn results in a diminished vascular wall compliance and aberrant blood flow shear stress. Second, the prolonged immobilization following surgery, which induces venous stasis, constitutes one of the pathophysiological elements within the triad of Virchow that was implicated in the genesis of thromb (28,29). The importance of thromboprophylaxis in senior patients was emphasized by this finding.

Smoking, both current and former, serves as a significant independent predictor of VTE risk in patients with lung cancer (30). Current smokers exhibited a significantly higher risk of VTE compared to never smokers (P<0.001, OR =6.509, 95% CI: 4.012–10.562) and those who had quit smoking for ≥1 year (P=0.011, OR =2.227, 95% CI: 1.198–4.139). This finding affirms the positive significance of smoking cessation in reducing VTE risk. The impact of smoking on vascular health encompasses endothelial damage and increased blood viscosity. Cessation of smoking can be considered a crucial preventive measure against VTE. Advanced age and smoking were associated with a higher risk of VTE, which was aligning with the conclusions of our study (31). Clinical healthcare providers should give more attention and guidance to VTE interventions for elderly and smoking patients within the sMPLC patient population.

In this study, we found that sMPLC patients with coronary artery disease, cerebrovascular disease, COPD, and atherosclerotic plaques in the extremities were more prone to developing VTE. Among these, a history of COPD was a risk factor for VTE in lung cancer patients, which was consistent with the conclusion drawn by Chen (32) in their study. In this study, we observed that patients who underwent lobectomies had a significantly increased risk of postoperative VTE (P<0.05). This finding was consistent with the conclusions reported in existing studies, particularly where previous research has established an association between lobectomy and an increased risk of postoperative VTE in lung cancer patients (33). Our study results not only provide additional empirical support for this association but also underscore the importance of considering the surgical approach and its potential impact on VTE risk in the clinical decision-making process. For patients undergoing lobectomy or segmentectomy with a moderate or high risk of thrombosis, it was recommended to use non-oral anticoagulants for VTE prophylaxis, with the duration of prevention extended to 28–35 days after discharge, rather than limiting the prophylaxis to the in-hospital period only (34).

Intraoperative blood transfusion was one of the risk factors for the occurrence of VTE after cardiac surgery, bariatric surgery, rectal cancer surgery, and gastric cancer surgery (35-38). The research indicated that receiving a blood transfusion during surgery was a risk factor for developing VTE after surgery in patients with sMPLC. We also found that the risk of VTE in patients with sMPLC was associated with the number of nodules removed surgically, and that the greater number of nodules removed intraoperatively could lead to longer surgery time and increased scope of surgical trauma, both of which were known risk factors for VTE formation in the past.

Additionally, it was found that patients with a high Postoperative Caprini score (high risk) should receive more attention. Because the early postoperative period was a high-risk time for VTE (39,40), For sMPLC patients, intervention for postoperative VTE can be initiated early based on the patient’s physical condition. These results were crucial for enhancing postoperative VTE prevention strategies and better managing sMPLC patients clinically.

The VTE risk prediction model constructed for sMPLC patients after VATS surgery in this study demonstrates good discrimination, calibration, and clinical validity. An AUC higher than 0.9 was usually interpreted as excellent model discrimination, whereas an AUC from 0.7 to 0.9 was considered to indicate good discrimination, and an AUC less than 0.7 indicates that the model has poor discriminative ability (41,42). AUC of 0.917 indicates the model has a high level of discriminative ability. With a sensitivity of 0.885 and a specificity of 0.818, the model exhibits strong predictive capabilities. There was a good match between the observed actual outcomes and the model’s predicted results on the calibration curve, which further validated the model’s accuracy and reliability. The decision curve analysis indicates that when the predicted risk falls between 10% to 20%, the use of the model to intervene in patients results in a substantial standardized net benefit, showing it has a higher clinical benefit than the Caprini score. The research presents a nomogram tool for managing VTE risk, offering a visual representation of how various factors contribute to the risk. This risk assessment model predicts the risk of VTE by considering a variety of clinical factors of the patient, calculating a total score. Medical staff can formulate personalized prevention measures for postoperative patients based on the level of the total score. For patients at high risk, low-dose anticoagulant drugs and other pharmaceutical preventive measures may be recommended; whereas for patients with contraindications to medication, mechanical prevention methods such as elastic stockings or limb air pumps may be employed. It offers clinical healthcare professionals an easy-to-use, intuitive, and effective approach for risk evaluation.

Close monitoring of VTE events in sMPLC patients was crucial for promoting early diagnosis and treatment of lung cancer (43). From a clinical practice perspective, this model provides strong support for personalized intervention in sMPLC patients. By identifying high-risk patients early on, healthcare professionals can more effectively develop management strategies for preventing VTE, such as enhancing VTE monitoring and optimizing anticoagulation therapy regimens. Additionally, this predictive model also helps to improve the allocation of healthcare resources by focusing on patients who were most in need of intervention, thereby enhancing the efficiency and quality of healthcare delivery.

Several limitations must be acknowledged. First, the retrospective single-center design is inherently susceptible to selection bias. However, this potential bias was mitigated through the application of stringent inclusion criteria and robust statistical adjustments. Second, the lack of external validation in diverse populations represents a significant gap. Given the known racial and regional variations in the incidence of VTE, external validation is a critical step to ensure the generalizability of our findings. Third, although the model incorporates key clinical variables, the inclusion of intraoperative factors (such as the cardiac arrest or other complications) and postoperative factors (such as the length of hospital stay, time to first mobilization after surgery) could further enhance the predictive performance of the model. Finally, the proficiency of the surgeon in performing the procedure is another factor that may influence postoperative thrombosis formation. Future research should consider incorporating these factors to develop a more comprehensive and effective VTE risk prediction model.

Finally, the operationalization of anticoagulant therapy as a binary variable (received/not received) precludes analysis of regimen-specific effects, meriting investigation in future studies.

Conclusions

The model developed in this study has shown outstanding performance in forecasting the risk of VTE following VATS surgery in sMPLC patients, highlighting its high accuracy, reliability, and clinical relevance. It can be used as a novel method for forecasting VTE risk in sMPLC patients and supplies data and theoretical backing for improving healthcare interventions. In the future, prospective multicenter validation and integration into electronic health record systems will facilitate clinical utility in this high-risk cohort.

Acknowledgments

None.

Footnote

Reporting Checklist: The authors have completed the TRIPOD reporting checklist. Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-558/rc

Data Sharing Statement: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-558/dss

Peer Review File: Available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-558/prf

Funding: This work was supported by the Chongqing Science and Health Joint Medical Science and Technology Innovation Key Technology R&D Program (2025GGXM001); Chongqing Technology Innovation and Application Development Key Projects (CSTB2022TIAD-CUX0019); the Key Projects of Talent Incubation Plan of Xinqiao Hospital (2023YQB010); and the Nursing Development Project of The Second Affiliated Hospital of Army Medical University (2023HLPY14).

Conflicts of Interest: All authors have completed the ICMJE uniform disclosure form (available at https://jtd.amegroups.com/article/view/10.21037/jtd-2025-558/coif). The authors have no conflicts of interest to declare.

Ethical Statement: The authors are accountable for all aspects of the work in ensuring that questions related to the accuracy or integrity of any part of the work are appropriately investigated and resolved. This study was reviewed and approved by the Ethics Committee of The Second Affiliated Hospital of Army Medical University of the Chinese People’s Liberation Army, with the ethics review number 2024-S-119-01. Individual consent for this retrospective analysis was waived. This study was conducted in accordance with the Declaration of Helsinki and its subsequent amendments.

Open Access Statement: This is an Open Access article distributed in accordance with the Creative Commons Attribution-NonCommercial-NoDerivs 4.0 International License (CC BY-NC-ND 4.0), which permits the non-commercial replication and distribution of the article with the strict proviso that no changes or edits are made and the original work is properly cited (including links to both the formal publication through the relevant DOI and the license). See: https://creativecommons.org/licenses/by-nc-nd/4.0/.

References

1. Bray F, Laversanne M, Sung H, et al. Global cancer statistics 2022: GLOBOCAN estimates of incidence and mortality worldwide for 36 cancers in 185 countries. CA Cancer J Clin 2024;74:229-63. [Crossref] [PubMed]
2. He X, Yang Z, Wu F, et al. Confronting synchronous multiple primary lung cancers: Navigating the intersection of challenges and opportunities. Lung Cancer 2024;197:107994. [Crossref] [PubMed]
3. Dong H, Tian Y, Xin S, et al. Diagnosis and management of multiple primary lung cancer. Front Oncol 2024;14:1392969. [Crossref] [PubMed]
4. Xiao F, Liu D, Guo Y, et al. Survival rate and prognostic factors of surgically resected clinically synchronous multiple primary non-small cell lung cancer and further differentiation from intrapulmonary metastasis. J Thorac Dis 2017;9:990-1001. [Crossref] [PubMed]
5. Lv J, Zhu D, Wang X, et al. The Value of Prognostic Factors for Survival in Synchronous Multifocal Lung Cancer: A Retrospective Analysis of 164 Patients. Ann Thorac Surg 2018;105:930-6. [Crossref] [PubMed]
6. Tie H, Luo J, Shi R, et al. Characteristics and prognosis of synchronous multiple primary lung cancer after surgical treatment: A systematic review and meta-analysis of current evidence. Cancer Med 2021;10:507-20. [Crossref] [PubMed]
7. Samadzadeh Tabrizi N, Gallant B, Harris E, et al. Contemporary Incidence of Synchronous Multiple Primary Lung Cancers and Survival in the Era of Lung Cancer Screening. Innovations (Phila) 2024;19:23-9. [Crossref] [PubMed]
8. Tabrizi NS, Harris ES, Gallant BP, et al. Clinical and pathologic staging accuracy in patients with synchronous multiple primary lung cancers. J Thorac Dis 2024;16:491-7. [Crossref] [PubMed]
9. Agzarian J, Hanna WC, Schneider L, et al. Postdischarge venous thromboembolic complications following pulmonary oncologic resection: An underdetected problem. J Thorac Cardiovasc Surg 2016;151:992-9. [Crossref] [PubMed]
10. Akhtar-Danesh GG, Akhtar-Danesh N, Shargall Y. Venous Thromboembolism in Surgical Lung Cancer Patients: A Provincial Population-Based Study. Ann Thorac Surg 2022;114:890-7. [Crossref] [PubMed]
11. Mørkved AL, Søgaard M, Skjøth F, et al. Risk and Timing of Venous Thromboembolism After Surgery for Lung Cancer: A Nationwide Cohort Study. Ann Thorac Surg 2024;117:289-96. [Crossref] [PubMed]
12. Axtell AL, Gaissert HA, Bao X, et al. Predictors of Venous Thromboembolism After Lung Cancer Resection. Ann Thorac Surg 2024;117:998-1005. [Crossref] [PubMed]
13. Steuer CE, Behera M, Kim S, et al. Predictors and outcomes of venous thromboembolism in hospitalized lung cancer patients: A Nationwide Inpatient Sample database analysis. Lung Cancer 2015;88:80-4.
14. Zhang J, Gao Y, Zou W, et al. Simultaneous resection of coexisting pulmonary and mediastinal lesions by video-assisted thoracic surgery: a case-series study. BMC Surg 2022;22:237. [Crossref] [PubMed]
15. RETRACTION. Comprehensive Analysis of Risk Factors for Surgical Site Infections Following Thoracoscopic Radical Resection in Patients With Lung Cancer. Int Wound J 2025;22:e70593. [Crossref] [PubMed]
16. Kahan BC, Hindley J, Edwards M, et al. The estimands framework: a primer on the ICH E9(R1) BMJ 2024;384:e076316. addendum. [Crossref] [PubMed]
17. Posch F, Riedl J, Reitter EM, et al. Dynamic assessment of venous thromboembolism risk in patients with cancer by longitudinal D-Dimer analysis: A prospective study. J Thromb Haemost 2020;18:1348-56. [Crossref] [PubMed]
18. Wang P, Zhao H, Zhao Q, et al. Risk Factors and Clinical Significance of D-Dimer in the Development of Postoperative Venous Thrombosis in Patients with Lung Tumor. Cancer Manag Res 2020;12:5169-79. [Crossref] [PubMed]
19. Di Minno MND, Calcaterra I, Papa A, et al. Diagnostic accuracy of D-Dimer testing for recurrent venous thromboembolism: A systematic review with meta-analysis.: VTE recurrence and D-dimer. Eur J Intern Med 2021;89:39-47. [Crossref] [PubMed]
20. Chi G, Lee JJ, Memar Montazerin S, et al. Association of D-dimer with short-term risk of venous thromboembolism in acutely ill medical patients: A systematic review and meta-analysis. Vasc Med 2022;27:478-86. [Crossref] [PubMed]
21. Doubre H, Monnet I, Azarian R, et al. Plasma tissue factor activity in lung cancer patients predicts venous thromboembolism and poor overall survival. Res Pract Thromb Haemost 2024;8:102359. [Crossref] [PubMed]
22. Wang Y, Li Q, Zhou Y, et al. A systematic review of risk prediction model of venous thromboembolism for patients with lung cancer. Thorac Cancer 2024;15:277-85. [Crossref] [PubMed]
23. Weitz JI, Fredenburgh JC, Eikelboom JW. A Test in Context: D-Dimer. J Am Coll Cardiol 2017;70:2411-20. [Crossref] [PubMed]
24. Li H, Tian Y, Niu H, et al. Derivation, validation and assessment of a novel nomogram-based risk assessment model for venous thromboembolism in hospitalized patients with lung cancer: A retrospective case control study. Front Oncol 2022;12:988287. [Crossref] [PubMed]
25. Cai Y, Dong H, Li X, et al. Development and validation of a nomogram to assess postoperative venous thromboembolism risk in patients with stage IA non-small cell lung cancer. Cancer Med 2023;12:1217-27. [Crossref] [PubMed]
26. Zhang Y, Shi Z, Yi J, et al. Correlation between clinicopathological characteristics of lung adenocarcinoma and the risk of venous thromboembolism. Thorac Cancer 2022;13:247-56. [Crossref] [PubMed]
27. Liang G, Hu Z, Xu Q, et al. Development and validation of a nomogram model for predicting venous thromboembolism risk in lung cancer patients treated with immune checkpoint inhibitors: A cohort study in China. Cancer Med 2024;13:e70115. [Crossref] [PubMed]
28. Nwagha T, Nweke M. Stratification of Risk Factors of Lung Cancer-Associated Venous Thromboembolism and Determining the Critical Point for Preemptive Intervention: A Systematic Review With Meta-analysis. Clin Med Insights Oncol 2023;17:11795549231175221. [Crossref] [PubMed]
29. Qin Y, Zhang L, Liang X, et al. Venous and Arterial Thromboembolism in Patients with Metastatic Lung Cancer. Clin Appl Thromb Hemost 2023;29:10760296231159121. [Crossref] [PubMed]
30. Kuderer NM, Poniewierski MS, Culakova E, et al. Predictors of Venous Thromboembolism and Early Mortality in Lung Cancer: Results from a Global Prospective Study (CANTARISK). Oncologist 2018;23:247-55. [Crossref] [PubMed]
31. Gregson J, Kaptoge S, Bolton T, et al. Cardiovascular Risk Factors Associated With Venous Thromboembolism. JAMA Cardiol 2019;4:163-73. [Crossref] [PubMed]
32. Chen X, Huang J, Liu J, et al. Derivation and External Validation of a Risk Assessment Model of Venous Thromboembolism in Hospitalized Chinese Patients. Clin Appl Thromb Hemost 2023;29:10760296221151164. [Crossref] [PubMed]
33. Smith D, Raices M, Diego C, et al. Incidence of venous thromboembolism and bleeding after pulmonary lobectomy: evaluating the timing for thromboprophylaxis. J Thromb Thrombolysis 2021;51:997-1004. [Crossref] [PubMed]
34. Shargall Y, Wiercioch W, Brunelli A, et al. Joint 2022 European Society of Thoracic Surgeons and The American Association for Thoracic Surgery guidelines for the prevention of cancer-associated venous thromboembolism in thoracic surgery. J Thorac Cardiovasc Surg 2023;165:794-824.e6. [Crossref] [PubMed]
35. Xenos ES, Vargas HD, Davenport DL. Association of blood transfusion and venous thromboembolism after colorectal cancer resection. Thromb Res 2012;129:568-72. [Crossref] [PubMed]
36. Nielsen AW, Helm MC, Kindel T, et al. Perioperative bleeding and blood transfusion are major risk factors for venous thromboembolism following bariatric surgery. Surg Endosc 2018;32:2488-95. [Crossref] [PubMed]
37. Kaida S, Miyake T, Murata S, et al. A Prospective Multicenter Observational Study of Venous Thromboembolism after Gastric Cancer Surgery (SHISA-1601). Eur Surg Res 2021;62:10-7. [Crossref] [PubMed]
38. Raphael J, Feng X, Shotwell MS, et al. Association of Intraoperative Red Blood Cell Transfusions With Venous Thromboembolism and Adverse Outcomes After Cardiac Surgery. Ann Surg 2023;278:e650-60. [Crossref] [PubMed]
39. Lavikainen LI, Guyatt GH, Sallinen VJ, et al. Systematic Reviews and Meta-analyses of the Procedure-specific Risks of Thrombosis and Bleeding in General Abdominal, Colorectal, Upper Gastrointestinal, and Hepatopancreatobiliary Surgery. Ann Surg 2024;279:213-25. [Crossref] [PubMed]
40. Ke L, Cui S, Chen S, et al. Dynamics of D-dimer in non-small cell lung cancer patients receiving radical surgery and its association with postoperative venous thromboembolism. Thorac Cancer 2020;11:2483-92. [Crossref] [PubMed]
41. de Hond AAH, Steyerberg EW, van Calster B. Interpreting area under the receiver operating characteristic curve. Lancet Digit Health 2022;4:e853-5. [Crossref] [PubMed]
42. Parodi S, Verda D, Bagnasco F, et al. The clinical meaning of the area under a receiver operating characteristic curve for the evaluation of the performance of disease markers. Epidemiol Health 2022;44:e2022088. [Crossref] [PubMed]
43. Byun JY, Aiyeolemi A, Qdaisat A, et al. Association between epidermal growth factor receptor-tyrosine kinase inhibitors and venous thromboembolism among older patients with advanced non-small cell lung cancer. Cancer 2024;130:3412-25. [Crossref] [PubMed]