Filipe, W.F. (Wills Floris) (2018) Postoperative venous thromboembolisms in high-grade gliomas: a prediction model. thesis, Medicine.
Full text available on request.Abstract
Introduction: high-grade glioma (HGG) patients have a 20%-30% risk of developing postoperative venous thromboembolisms (VTE), despite postoperative administration of low molecular weight heparin (LMWH) during admission. Prolonged postoperative LMWH may be beneficial for this patient population, but the risk of intracranial haemorrhage may increase. This highlights the importance of risk-stratification to identify high-risk patients in which the prophylactic effect of LMWH outweighs that potential increased risk of ICH. The aim of this study is to develop a prediction model of postoperative VTE in HGG patients to identify the high-risk patients. Methods: Data of patients that underwent HGG surgery in the UMCG between 2012 and 2016 was collected retrospectively. Exclusion criteria were pregnancy and age under 18. Possible predictors were assessed. The prediction model was created using logistic regression analysis by forward selection and backwards elimination methods. Model performance was assessed using the area under the receiver operator curve (AUROC) and superior method was used to create a final weighted risk model. Results: 273 patients were included, of which 8.8% suffered a VTE. The regression analysis by backwards elimination method had the superior AUROC of 0.701 and was used for weighted risk scores. Variables included in the final weighted risk score were gender, recurrent disease, medical history of VTE and a BMI above 25 and all had wide confidence intervals. Tumour necrosis volume was found to be an independent predictor of VTE but was not included in the regression analysis due to numerous missing values. Conclusion: The prediction model has a fair performance in the study population and is made with retrospective dichotomised risk factors, thus making it a clinically user-friendly model. However, the model is likely overfitting to the study population due to a low VTE incidence and low power.
| Item Type: | Thesis (Thesis) |
|---|---|
| Supervisor name: | Department of neurosurgery and University Medical Centre Groningen and Dijk, Prof. dr. J.M.C. van and and and Wagemakers, Dr. M. |
| Faculty: | Medical Sciences |
| Date Deposited: | 25 Jun 2020 10:41 |
| Last Modified: | 25 Jun 2020 10:41 |
| URI: | https://umcg.studenttheses.ub.rug.nl/id/eprint/269 |
Actions (login required)
 |
View Item |