Factors Affecting Outcome in Treatment of Chronic Subdural Hematoma in ICU Patients: Impact of Anticoagulation

Synthetic minority over-sampling technique-enhanced machine learning models for predicting recurrence of postoperative chronic subdural hematoma

Objective

Chronic subdural hematoma (CSDH) is a neurological condition with high recurrence rates, primarily observed in the elderly population. Although several risk factors have been identified, predicting CSDH recurrence remains a challenge. Given the potential of machine learning (ML) to extract meaningful insights from complex data sets, our study aims to develop and validate ML models capable of accurately predicting postoperative CSDH recurrence.

Methods

Data from 447 CSDH patients treated with consecutive burr-hole irrigations at Wenzhou Medical University's First Affiliated Hospital (December 2014-April 2019) were studied. 312 patients formed the development cohort, while 135 comprised the test cohort. The Least Absolute Shrinkage and Selection Operator (LASSO) method was employed to select crucial features associated with recurrence. Eight machine learning algorithms were used to construct prediction models for hematoma recurrence, using demographic, laboratory, and radiological features. The Border-line Synthetic Minority Over-sampling Technique (SMOTE) was applied to address data imbalance, and Shapley Additive Explanation (SHAP) analysis was utilized to improve model visualization and interpretability. Model performance was assessed using metrics such as AUROC, sensitivity, specificity, F1 score, calibration plots, and decision curve analysis (DCA).

Results

Our optimized ML models exhibited prediction accuracies ranging from 61.0% to 86.2% for hematoma recurrence in the validation set. Notably, the Random Forest (RF) model surpassed other algorithms, achieving an accuracy of 86.2%. SHAP analysis confirmed these results, highlighting key clinical predictors for CSDH recurrence risk, including age, alanine aminotransferase level, fibrinogen level, thrombin time, and maximum hematoma diameter. The RF model yielded an accuracy of 92.6% with an AUC value of 0.834 in the test dataset.

Conclusion

Our findings underscore the efficacy of machine learning algorithms, notably the integration of the RF model with SMOTE, in forecasting the recurrence of postoperative chronic subdural hematoma. Leveraging the RF model, we devised an online calculator that may serve as a pivotal instrument in tailoring therapeutic strategies and implementing timely preventive interventions for high-risk patients.

Effect of long-term anticoagulant therapy on the outcome of chronic subdural hematoma: a propensity score-matched analysis

OBJECTIVE

Chronic subdural hematomas (cSDHs) are particularly common in older adults who have increased risk of falls and the conditions that require anticoagulants (ACs). In such cases, clinicians are often left with the dilemma of co-managing the cSDH and the ongoing need for ACs.

METHODS

Patients who underwent surgical management for cSDH at the authors' institution between January 2006 and June 2022 were identified. Propensity score-matched analysis was used to obtain a balance in patients who were on ACs before the procedure versus those who were not, and in patients who were on ACs postprocedure versus those who were not. Length of hospitalization, periprocedural complications, reintervention rate during the same admission, rebleeding risk, and reintervention rates after discharge were compared.

RESULTS

In total, 104 patients were on long-term ACs before the procedure, whereas 372 were not. After matching, 55 pairs were included in the analysis. Postprocedure, 74 patients were started on long-term ACs; the rest were not. A total of 49 patients in each group were then included in the analysis after matching. Comparing the preprocedure AC group with the non-AC group, no significant differences were found in length of hospitalization (8.5 ± 6.7 days vs 8.1 ± 7.7 days, p = 0.75), periprocedural complications (7.3% vs 7.3%, p > 0.99), or reintervention during the same admission (1.8% vs 5.5%, p = 0.31). In the comparison of postprocedure AC and non-AC groups, no significant differences were seen in recurrence rate (8.2% vs 14.3%, p = 0.52), reintervention rate after discharge (4.1% vs 14.3%, p = 0.16), or disability (i.e., mRS ≤ 2; 83.7% vs 89.8%, p = 0.55).

CONCLUSIONS

Being treated with long-term ACs before cSDH procedures does not affect length of hospitalization, periprocedural complications, or reintervention during the same admission. Similarly, administration of long-term ACs after a procedure for cSDH does not increase rebleeding risk or reintervention rate. Patients who are on long-term ACs can have similar interventions to those who are not on ACs. In addition, it is safe to restart patients on AC agents in a 7- to 14-day window after admission for cSDH with or without acute/subacute components.

Intracranial hemorrhage following drainage of chronic subdural effusion and hematoma: A case report and review of the literature

Acute intracranial hemorrhage (AIH) after drainage of chronic subdural hematoma is a rare but serious complication. An 86-year-old man with bilateral frontotemporal subdural effusion, hematoma, and cerebral hernia was admitted to our department and treated with bilateral burr hole surgery and closed-system drainage under local anesthesia. After the operation, computed tomography (CT) showed AIH in the left temporal and occipital lobe, and then a series of head CT showed that the hematoma gradually increased day by day. This patient had a medical history of hypertension, diabetes, atrial fibrillation, and taking warfarin. He was treated conservatively, but had not recovered at discharge after 1 month. We reviewed the relevant literature and analyzed the operation opportunity, causes of cerebral hemorrhage, and preventive measures in similar patients. We suppose that the coagulation abnormality and rapid fluctuations of intracranial pressure were the main causes of development of AIH in our patient. Several possible reasons such as brain shift and impaired vascular autoregulation are also associated with postoperative AIH. We must be aware of this complication and keep some preventive measures in our mind.

A Grading System For The Prediction Of Unilateral Chronic Subdural Hematoma Recurrence After Initial Single Burr Hole Evacuation

Background and purpose

Previous studies have identified many risk factors related to the recurrence of chronic subdural hematomas (CSDHs). Among these factors, there may be deviations in measuring the midline shift, preoperative hematoma volume (PreHV), postoperative hematoma residual volume, and postoperative pneumocephalus in bilateral CSDHs. The aims of this study were to eliminate the impact of complicated situations on parameter measurement and to identify actual predictors for CSDH recurrence, and finally, to develop a grading system to predict unilateral CSDH (uCSDH) recurrence.

Patients and methods

A total of 342 patients with uCSDH were identified. Predictors of uCSDH recurrence were obtained from univariable and multivariable logistic regression models. A prognostic grading system was developed based on the results of multivariable logistic regression and receiver operating characteristic (ROC) analyses. All patients were scored according to the grading system, and differences in the recurrence rate were reanalyzed according to the scores.

Results

Age, antiplatelet or anticoagulant use, midline shift, severe brain atrophy, drainage volume, and the ratio of the postoperative pneumocephalus volume (PostPV) to the postoperative hematoma cavity volume (PostHCV) were identified as independent risk factors for predicting the recurrence of uCSDH. The cut-off values of age, drainage volume, midline shift, and the ratio of the PostPV to the PostHCV were 67 years, 101 mL, 11.2 mm, and 31.61%, respectively. The recurrence rates were 1.7%, 12.4%, 19.4%, 53.3%, and 58.3% for scores of 0–1, 2, 3, 4, and 5–6, respectively, which significantly increased as the score increased (P<0. 001).

Conclusion

The prognostic grading system for uCSDH on the basis of multivariable logistic regression and ROC analyses is applicable and reliable.

Role of Regulatory T cells in Atorvastatin Induced Absorption of Chronic Subdural Hematoma in Rats

Chronic subdural hematoma (CSDH) is a neurological disorder with a substantial recurrence rate. Atorvastatin is an effective drug for treating hyperlipidemia and known to improve neurological outcome after intracerebral hemorrhage. Previous studies have reported that atorvastatin treatment promotes hematoma absorption in CSDH, while the underlying mechanisms remain unclear. In this study, we investigated whether the anti-inflammatory effects of atorvastatin mediate absorption of CSDH. 144 male, Wistar rats (6 months old) were randomly divided into the following groups: 1) sham surgery control, 2) treatment: CSDH + atorvastatin, and 3) vehicle control: CSDH + saline. Atorvastatin or saline was orally administered daily for 19 days after CSDH procedure. A T2WI MRI was used to evaluate CSDH volume changes during the time course of the study. Flow cytometry and immunohistochemical staining were used to measure the number of regulatory T cells (Treg). ELISA was used to measure cytokine level in the hematoma border. Neurological function and cognitive outcome were evaluated using Foot-Fault test and Morris Water Maze test, respectively. When compared to saline treatment, atorvastatin treatment accelerated the absorption of CSDH as indicated by decreased hematoma volume in T2WI MRI data on 14^th and 21^st day after CSDH (P<0.05). Atorvastatin treatment significantly increased the number of Treg in circulation and hematoma border from 3^rd to 21^st day after CSDH. Atorvastatin treatment significantly decreased the levels of interleukins (IL-6 and IL-8) and tumor necrosis factor-α (TNF-α), but increased IL-10 level in the hematoma border. Atorvastatin treatment also improved neurological function and cognitive outcome compared to vehicle treated group. Atorvastatin induced anti-inflammatory responses and increased Treg in circulation and brain which may contribute to the accelerated CSDH absorption in rats.

Acute parenchymal hemorrhage of three cases report after burr hole drainage of chronic subdural hematoma

Chronic subdural hematoma (CSDH) is one of the most common neurological diseases, which mainly occurs among elderly people and usually develop after minor head injuries. Over the years, a simple burr hole evacuation of the hematoma has been accepted as the widespread method for most cases of CSDH, but acute parenchymal hemorrhage is a rare and deadly complication after surgery. We report three elderly cases of post-operative parenchymal hemorrhage and analyse the underlying factors and formulate relevant strategies in this article. Three advanced age patients had been admitted to our department with gradually increasing headache and limb activity disorder urgently and underwent an emergency operation of burr hole drainage of CSDH in frontal-temporal region after preoperative evaluations and examinations. Unfortunately, acute post-operative parenchymal hemorrhage occurred in three advanced age patients. Ultimately, the patients achieved satisfying outcome with no significant neurological deficit through conservative treatment. The exact mechanism of such uncommon complications are difficult to explain and remain poorly understood. Advanced age, hypertension, amyloidosis, high perfusion triggered by rapid hematoma release, cerebrospinal fluid (CSF) loss, oral anticoagulant, primary disease aggravation were the main mechanisms which were speculated in our report. Simultaneously, positive measures could be adopt to prevent this rare complication.

Development of the Subdural Hematoma in the Elderly (SHE) score to predict mortality

OBJECTIVE

The purpose of this study was to describe the development of a novel prognostic score, the Subdural Hematoma in the Elderly (SHE) score. The SHE score is intended to predict 30-day mortality in elderly patients (those > 65 years of age) with an acute, chronic, or mixed-density subdural hematoma (SDH) after minor, or no, prior trauma.

METHODS

The authors used the Prognosis Research Strategy group methods to develop the clinical prediction model. The training data set included patients with acute, chronic, and mixed-density SDH. Based on multivariate analyses from a large data set, in addition to review of the extant literature, 3 components to the score were selected: age, admission Glasgow Coma Scale (GCS) score, and SDH volume. Patients are given 1 point if they are over 80 years old, 1 point for an admission GCS score of 5-12, 2 points for an admission GCS score of 3-4, and 1 point for SDH volume > 50 ml. The sum of points across all categories determines the SHE score.

RESULTS

The 30-day mortality rate steadily increased as the SHE score increased for all SDH acuities. For patients with an acute SDH, the 30-day mortality rate was 3.2% for SHE score of 0, and the rate increased to 13.1%, 32.7%, 95.7%, and 100% for SHE scores of 1, 2, 3, and 4, respectively. The model was most accurate for acute SDH (area under the curve [AUC] = 0.94), although it still performed well for chronic (AUC = 0.80) and mixed-density (AUC = 0.87) SDH.

CONCLUSIONS

The SHE score is a simple clinical grading scale that accurately stratifies patients' risk of mortality based on age, admission GCS score, and SDH volume. Use of the SHE score could improve counseling of patients and their families, allow for standardization of clinical treatment protocols, and facilitate clinical research studies in SDH.

Effects of Pre-Existing Comorbidities on Outcomes in Patients with Chronic Subdural Hematoma

OBJECTIVE

The number of patients with chronic subdural hematoma (CSDH) showing comorbidities and/or impaired activities of daily living (ADL) before the onset of CSDH has increased with the recent aging of society. The purposes of this study were to evaluate ADL worsening by comparing premorbid ADL and ADL at discharge and to investigate the effects of pre-existing comorbidity-related clinical factors on the outcomes.

METHODS

A total of 570 patients with CSDH admitted from 2006 to 2016 were studied retrospectively. Clinical factors, including pre-existing comorbidities, related to outcomes were identified by multivariate analysis. A variation of the modified Rankin Scale (mRS) using 5 united categories of mRS scores 0/1, 2, 3/4, 5, and dead was used for evaluation of ADL.

RESULTS

Of 570 patients, 390 (68.4%) had pre-existing comorbidities and 120 (21.1%) showed premorbid impaired ADL (mRS scores 2 and worse). Considering pre-existing impaired ADL, ADL deteriorated after CSDH in 92 patients (16.1%), whereas ADL impairment at discharge was found in 173 patients (30.4%). Comorbidities related to ADL deterioration on multivariate analysis were hemodialysis and chronic heart failure. Antithrombotic use for cardiovascular diseases was a predictor of acute-on-chronic subdural hematoma, which was the sole common predictor for ADL deterioration and the occurrence of surgical complications.

CONCLUSIONS

In patients with CSDH, pre-existing comorbidity-linked factors related to outcomes were hemodialysis, chronic heart failure, and antithrombotic use. Patients with acute-on-chronic subdural hematoma with these factors should be regarded as a high-risk group.

Flow, Liver, Flow: A Retrospective Analysis of the Interplay of Liver Disease and Coagulopathy in Chronic Subdural Hematoma

BACKGROUND

Chronic subdural hematoma (cSDH) is a common neurosurgical ailment, particularly in elderly patients. A recent study uncovered an association between liver disease and recurrence in patients with cSDH. Here, we explored that relationship to identify recurrence predictors in at-risk patients.

OBJECTIVE

We hypothesized that the association between liver disease and recurrence was attributable to coagulopathy secondary to liver disease.

METHODS

We retrospectively reviewed all patients with cSDH treated with burr-hole drainage by 2 surgeons between 2007 and 2015. Comorbidities and laboratory findings for each patient were examined by Pearson χ² analysis or Mann-Whitney U tests.

RESULTS

We identified 261 cSDH in 215 patients. Patients were a mean age of 65.6 years, and 72% were male. Sixteen patients with cSDH required repeat surgery (6.1%). There were 123 coagulopathic patients (47.1%), and 14 with liver disease (5.4%), all of whom were coagulopathic (P < 0.001). Coagulopathic patients with liver disease were more likely to experience recurrence than patients with coagulopathy alone (relative risk = 4.09, P = 0.019). Patients with liver disease had significantly elevated prothrombin time (P = 0.013) and reduced platelet counts (P < 0.001). Platelets also were reduced in coagulopathic patients with liver disease, as compared with those with coagulopathy alone (P = 0.002). Thrombocytopenia remained significant in a multivariate analysis (P < 0.001).

CONCLUSIONS

Liver disease is significantly associated with the recurrence of cSDH. Although coagulopathy alone does not predict recurrence, patients with coagulopathy and liver disease are at greater risk for recurrence than those with coagulopathy alone. Liver disease effects are reflected in certain hematologic laboratory values.

Optimal perioperative management of antithrombotic agents in patients with chronic subdural hematoma

OBJECTIVE

The use of antithrombotic agents such as anticoagulants and antiplatelet agents is widespread, and the opportunities to treat patients with chronic subdural hematoma (CSDH) under antithrombotic therapy are growing. However, whether antithrombotic therapy contributes to postoperative complications and recurrences of CSDH and how these agents should be managed in the surgical treatment of CSDH remains unclear.

METHODS

We retrospectively analyzed 150 consecutive patients with CSDH who underwent neurosurgical interventions at Kyushu Rosai Hospital from 2011 to 2015 and followed them for more than 3 months.

RESULTS

Of the 150 study patients, 44 received antithrombotic therapy. All anticoagulants and 76% of the antiplatelet agents were discontinued before surgical treatment of CSDH and resumed within 1 week except in 4 patients whose treatment was terminated and 7 patients who developed postoperative complications or underwent reoperations before resumption of these agents. Postoperative hemorrhagic complications associated with surgical treatment of CSDH occurred in 8 patients (5.3%), and there was no significant difference in the incidence of these complications between patients with and without antithrombotic therapy (6.8% vs. 4.7%, respectively; p=0.90). Postoperative thromboembolic complications occurred in 5 patients (5.4%), including 4 patients with antithrombotic therapy; these complications developed before resumption of antithrombotic agents in 2 patients. There was a significant difference in the incidence of postoperative thromboembolic complications between patients with and without antithrombotic therapy (9.1% vs. 0.9%, respectively; p=0.04). There were no significant differences in the incidence of radiographic deterioration or reoperation of ipsilateral or contralateral hematomas between patients with and without antithrombotic therapy after surgical treatment of unilateral CSDH.

CONCLUSION

A history of antithrombotic therapy was significantly correlated with the incidence of postoperative thromboembolic complications in patients with CSDH. Antithrombotic agents should be resumed as soon as possible when no hemorrhagic complication is confirmed after neurosurgical intervention for CSDH.