1
|
Li B, Eisenberg N, Beaton D, Lee DS, Al-Omran L, Wijeysundera DN, Hussain MA, Rotstein OD, de Mestral C, Mamdani M, Roche-Nagle G, Al-Omran M. Predicting inferior vena cava filter complications using machine learning. J Vasc Surg Venous Lymphat Disord 2024:101943. [PMID: 39084408 DOI: 10.1016/j.jvsv.2024.101943] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 04/28/2024] [Revised: 06/03/2024] [Accepted: 06/26/2024] [Indexed: 08/02/2024]
Abstract
OBJECTIVE Inferior vena cava (IVC) filter placement is associated with important long-term complications. Predictive models for filter-related complications may help guide clinical decision-making but remain limited. We developed machine learning (ML) algorithms that predict 1-year IVC filter complications using preoperative data. METHODS The Vascular Quality Initiative database was used to identify patients who underwent IVC filter placement between 2013 and 2024. We identified 77 preoperative demographic and clinical features from the index hospitalization when the filter was placed. The primary outcome was 1-year filter-related complications (composite of filter thrombosis, migration, angulation, fracture, and embolization or fragmentation, vein perforation, new caval or iliac vein thrombosis, new pulmonary embolism, access site thrombosis, or failed retrieval). The data were divided into training (70%) and test (30%) sets. Six ML models were trained using preoperative features with 10-fold cross-validation (Extreme Gradient Boosting, random forest, Naïve Bayes classifier, support vector machine, artificial neural network, and logistic regression). The primary model evaluation metric was area under the receiver operating characteristic curve (AUROC). Model robustness was assessed using calibration plot and Brier score. Performance was evaluated across subgroups based on age, sex, race, ethnicity, rurality, median Area Deprivation Index, planned duration of filter, landing site of filter, and presence of prior IVC filter placement. RESULTS Overall, 14,476 patients underwent IVC filter placement and 584 (4.0%) experienced 1-year filter-related complications. Patients with a primary outcome were younger (59.3 ± 16.7 years vs 63.8 ± 16.0 years; P < .001) and more likely to have thrombotic risk factors including thrombophilia, prior venous thromboembolism (VTE), and family history of VTE. The best prediction model was Extreme Gradient Boosting, achieving an AUROC of 0.93 (95% confidence interval, 0.92-0.94). In comparison, logistic regression had an AUROC of 0.63 (95% confidence interval, 0.61-0.65). Calibration plot showed good agreement between predicted/observed event probabilities with a Brier score of 0.07. The top 10 predictors of 1-year filter-related complications were (1) thrombophilia, (2) prior VTE, (3) antiphospholipid antibodies, (4) factor V Leiden mutation, (5) family history of VTE, (6) planned duration of IVC filter (temporary), (7) unable to maintain therapeutic anticoagulation, (8) malignancy, (9) recent or active bleeding, and (10) age. Model performance remained robust across all subgroups. CONCLUSIONS We developed ML models that can accurately predict 1-year IVC filter complications, performing better than logistic regression. These algorithms have potential to guide patient selection for filter placement, counselling, perioperative management, and follow-up to mitigate filter-related complications and improve outcomes.
Collapse
Affiliation(s)
- Ben Li
- Department of Surgery, University of Toronto, Toronto, Canada; Division of Vascular Surgery, St. Michael's Hospital, Unity Health Toronto, Toronto, Canada; Institute of Medical Science, University of Toronto, Toronto, Canada; Temerty Centre for Artificial Intelligence Research and Education in Medicine (T-CAIREM), University of Toronto, Toronto, Canada
| | - Naomi Eisenberg
- Division of Vascular Surgery, Peter Munk Cardiac Centre, University Health Network, Toronto, Canada
| | - Derek Beaton
- Data Science & Advanced Analytics, Unity Health Toronto, University of Toronto, Toronto, Canada
| | - Douglas S Lee
- Division of Cardiology, Peter Munk Cardiac Centre, University Health Network, Toronto, Canada; Institute of Health Policy, Management and Evaluation, University of Toronto, Toronto, Canada; ICES, University of Toronto, Toronto, Canada
| | - Leen Al-Omran
- School of Medicine, Alfaisal University, Riyadh, Saudi Arabia
| | - Duminda N Wijeysundera
- Institute of Health Policy, Management and Evaluation, University of Toronto, Toronto, Canada; ICES, University of Toronto, Toronto, Canada; Department of Anesthesia, St. Michael's Hospital, Unity Health Toronto, Toronto, Canada; Li Ka Shing Knowledge Institute, St. Michael's Hospital, Unity Health Toronto, Toronto, Canada
| | - Mohamad A Hussain
- Division of Vascular and Endovascular Surgery and the Center for Surgery and Public Health, Brigham and Women's Hospital, Harvard Medical School, Boston, MA
| | - Ori D Rotstein
- Department of Surgery, University of Toronto, Toronto, Canada; Institute of Medical Science, University of Toronto, Toronto, Canada; Li Ka Shing Knowledge Institute, St. Michael's Hospital, Unity Health Toronto, Toronto, Canada; Division of General Surgery, St. Michael's Hospital, Unity Health Toronto, Toronto, Canada
| | - Charles de Mestral
- Department of Surgery, University of Toronto, Toronto, Canada; Division of Vascular Surgery, St. Michael's Hospital, Unity Health Toronto, Toronto, Canada; Institute of Health Policy, Management and Evaluation, University of Toronto, Toronto, Canada; ICES, University of Toronto, Toronto, Canada; Li Ka Shing Knowledge Institute, St. Michael's Hospital, Unity Health Toronto, Toronto, Canada
| | - Muhammad Mamdani
- Institute of Medical Science, University of Toronto, Toronto, Canada; Temerty Centre for Artificial Intelligence Research and Education in Medicine (T-CAIREM), University of Toronto, Toronto, Canada; Data Science & Advanced Analytics, Unity Health Toronto, University of Toronto, Toronto, Canada; Institute of Health Policy, Management and Evaluation, University of Toronto, Toronto, Canada; ICES, University of Toronto, Toronto, Canada; Li Ka Shing Knowledge Institute, St. Michael's Hospital, Unity Health Toronto, Toronto, Canada; Leslie Dan Faculty of Pharmacy, University of Toronto, Toronto, Canada
| | - Graham Roche-Nagle
- Department of Surgery, University of Toronto, Toronto, Canada; Division of Vascular Surgery, Peter Munk Cardiac Centre, University Health Network, Toronto, Canada; Division of Vascular and Interventional Radiology, University Health Network, Toronto, Canada
| | - Mohammed Al-Omran
- Department of Surgery, University of Toronto, Toronto, Canada; Division of Vascular Surgery, St. Michael's Hospital, Unity Health Toronto, Toronto, Canada; Institute of Medical Science, University of Toronto, Toronto, Canada; Temerty Centre for Artificial Intelligence Research and Education in Medicine (T-CAIREM), University of Toronto, Toronto, Canada; Li Ka Shing Knowledge Institute, St. Michael's Hospital, Unity Health Toronto, Toronto, Canada; Department of Surgery, King Faisal Specialist Hospital and Research Center, Riyadh, Saudi Arabia.
| |
Collapse
|
2
|
Wang C, Liu Q, Xiao D, Chen X, Liu R, Wang S. Laparoscopic-assisted retrieval of inferior vena cava filter: A case report and literature review. Vascular 2024:17085381241241108. [PMID: 38498983 DOI: 10.1177/17085381241241108] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [Key Words] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Indexed: 03/20/2024]
Abstract
BACKGROUND The retrieval of inferior vena cava filters beyond the retrieval window poses challenges, requiring alternative techniques. OBJECTIVES To discuss the laparoscopy-assisted retrieval approach for difficult inferior vena cava filters. RESEARCH DESIGN Case report. SUBJECTS A 57-year-old male with a retrievable inferior vena cava filter placed 8 months prior. MEASURES Laparoscopy-assisted retrieval technique utilized after unsuccessful interventional attempts. RESULTS Successful retrieval of the filter despite thickened intimal tissue involvement, with no postoperative complications. CONCLUSIONS Laparoscopy-assisted retrieval offers a direct visual approach for challenging filter removal, proving minimally invasive, safe, and effective.
Collapse
Affiliation(s)
| | | | - Dalei Xiao
- Luohe Central Hospital, Luohe City, China
| | | | - Rong Liu
- Luohe Central Hospital, Luohe City, China
| | | |
Collapse
|
3
|
Sridharan N, Williams AO, Rojanasarot S, Anderson N, Wifler W, Jaff MR, Chaer R. Cost Burden and Cost Influencers of Inferior Vena Cava Filter Placement and Retrieval among Medicare Beneficiaries with Acute Venous Thromboembolism. J Vasc Interv Radiol 2023; 34:164-172.e2. [PMID: 36265817 DOI: 10.1016/j.jvir.2022.10.012] [Citation(s) in RCA: 0] [Impact Index Per Article: 0] [Reference Citation Analysis] [Abstract] [MESH Headings] [Track Full Text] [Journal Information] [Subscribe] [Scholar Register] [Received: 01/04/2022] [Revised: 09/22/2022] [Accepted: 10/06/2022] [Indexed: 12/03/2022] Open
Abstract
PURPOSE To examine the frequency, costs, and cost influencers of inferior vena cava filters (IVCFs) placements and retrievals among a national sample of patients using Medicare data. MATERIALS AND METHODS This retrospective cohort study used the U.S. Medicare 100% database, a nationally representative sample of all U.S. patients aged ≥65 years, from 2014 through 2020. Procedures and clinical characteristics were identified from the diagnosis and procedure codes on Medicare claims. Beneficiaries aged ≥65 years with newly diagnosed venous thromboembolism (VTE) were identified and followed to obtain data on IVCF placements and retrievals. Data on the costs of the index IVCF procedures and any subsequent IVCF placements and retrievals were obtained. Multivariate models were used to estimate the impact of patient and clinical characteristics on costs. RESULTS Among 501,216 patients with newly diagnosed VTE, 4,995 (1%) received an IVCF placement; of these, 1,215 (24.3%) had a retrieval procedure. Beneficiaries with IVCF placements and retrievals differed from a demographic and clinical perspective than from those without. Costs varied by the site of service, VTE acuity, and VTE type. Cost influencers included age, race, census region, service location, and VTE type. CONCLUSIONS IVCF placement costs were driven by baseline patient characteristics (age, race, geographic residence, acute VTE diagnosis, and inpatient site of service), whereas retrieval costs were driven by age and deep vein thrombosis diagnosis. Strategies to mitigate the retrieval costs or the need to retrieve IVCFs may reduce the overall cost burden of IVCFs.
Collapse
Affiliation(s)
| | | | | | | | | | | | - Rabih Chaer
- UPMC Presbyterian Hospital, Pittsburgh, Pennsylvania
| |
Collapse
|
6
|
Han K, Kim MD, Kim GM, Kwon JH, Lee J, Won JY, Moon S, An H. Comparison of Retrievability and Indwelling Complications of Celect and Denali Infrarenal Vena Cava Filters: A Randomized, Controlled Trial. Cardiovasc Intervent Radiol 2021; 44:1536-1542. [PMID: 34312689 DOI: 10.1007/s00270-021-02928-9] [Citation(s) in RCA: 2] [Impact Index Per Article: 0.7] [Reference Citation Analysis] [Abstract] [Key Words] [MESH Headings] [Track Full Text] [Journal Information] [Submit a Manuscript] [Subscribe] [Scholar Register] [Received: 01/07/2021] [Accepted: 07/15/2021] [Indexed: 11/27/2022]
Abstract
PURPOSE To compare the Celect and Denali filters in terms of complex filter retrieval and indwelling complications after a 2-month indwelling time. MATERIALS AND METHODS In this prospective, randomized trial, 153 subjects were assessed for eligibility between May 2016 and July 2018. A total of 136 participants were randomly assigned to receive either Celect (n = 68) or Denali (n = 68) filter placement in the infrarenal inferior vena cava. Tilt angles at placement and retrieval and rates of overall filter retrieval, indwelling complication, and complex retrieval were compared. RESULTS Of 136 participants (mean age, 62 ± 12.8 years, 55 male), 24 (17.6%) were lost to follow-up. The mean indwelling time of filter was 60.4 ± 7 days and there was no significant difference in the baseline characteristics between the two groups. Filter retrieval was successful in all participants (112/112, 100%). Significantly higher rates of filter tilt > 15° (n = 8) and strut penetration (n = 14) were found with the Celect filter than with the Denali filter (1 significant tilt and 1 penetration) (P = 0.033 and 0.001, respectively). No filter fractures were observed and there was no significant difference in tip embedment, filter fracture, filter migration, or mean fluoroscopy times. There were 3 cases of complex retrieval (1 for Denali vs. 2 for Celect, P = 0.500), for which the loop-snare technique was used. CONCLUSION Denali filters demonstrated significantly lower rates of tilt angle > 15° and strut penetration. However, there was no significant difference in the complex filter retrieval rate between the Celect and Denali filters.
Collapse
Affiliation(s)
- Kichang Han
- Department of Radiology, Severance Hospital, Research Institute of Radiological Science, Yonsei University College of Medicine, 50-1 Yonsei-ro Seodaemun-gu, Seoul, 03722, Korea
| | - Man-Deuk Kim
- Department of Radiology, Severance Hospital, Research Institute of Radiological Science, Yonsei University College of Medicine, 50-1 Yonsei-ro Seodaemun-gu, Seoul, 03722, Korea.
| | - Gyoung Min Kim
- Department of Radiology, Severance Hospital, Research Institute of Radiological Science, Yonsei University College of Medicine, 50-1 Yonsei-ro Seodaemun-gu, Seoul, 03722, Korea
| | - Joon Ho Kwon
- Department of Radiology, Severance Hospital, Research Institute of Radiological Science, Yonsei University College of Medicine, 50-1 Yonsei-ro Seodaemun-gu, Seoul, 03722, Korea
| | - Junhyung Lee
- Department of Radiology, Severance Hospital, Research Institute of Radiological Science, Yonsei University College of Medicine, 50-1 Yonsei-ro Seodaemun-gu, Seoul, 03722, Korea
| | - Jong Yun Won
- Department of Radiology, Severance Hospital, Research Institute of Radiological Science, Yonsei University College of Medicine, 50-1 Yonsei-ro Seodaemun-gu, Seoul, 03722, Korea
| | - Sungmo Moon
- Department of Radiology, Severance Hospital, Research Institute of Radiological Science, Yonsei University College of Medicine, 50-1 Yonsei-ro Seodaemun-gu, Seoul, 03722, Korea
| | - Hyeongsu An
- Department of Radiology, Severance Hospital, Research Institute of Radiological Science, Yonsei University College of Medicine, 50-1 Yonsei-ro Seodaemun-gu, Seoul, 03722, Korea
| |
Collapse
|