Keita Shibahashi, Hidenori Hoda, Takuto Ishida, Takayuki Motoshima, Kazuhiro Sugiyama, and Yuichi Hamabe
The objective of this study was to derive and validate a new screening model using a nomogram that allows clinicians to quantify the risk of blunt cerebrovascular injury (BCVI).
In this study, the authors examined 258,935 patients from a prospectively collected nationwide Japanese database (January 2009–December 2018) who experienced blunt injury. Patients were randomly divided into training (n = 129,468) and validation (n = 129,467) cohorts. First, the authors investigated the prevalence of BCVI, which was defined as blunt injury to any intracranial vessel, the extracranial vertebral artery, the extracranial carotid (common, internal) artery, or the internal jugular vein. Then, a new arterial BCVI screening model using a nomogram was derived, based on multivariate logistic regression analysis through quantifying the association of potential predictive factors with BCVI in the training cohort. The model’s discriminatory ability was validated using the area under the receiver operating characteristic curve (AUC) in the validation cohort.
Multivariate analysis in the training cohort showed that 13 factors were significantly associated with arterial BCVI and were included in our model. These factors were 1) male sex; 2) high-energy impact; 3) hypotension on hospital arrival; 4) Glasgow Coma Scale score < 9; 5) injury to the face; 6) injury to the neck; 7) injury to the spine; 8) skull base fracture; 9) cervical spine fracture or subluxation; and those with negative associations, i.e., 10) injury to the lower-extremity region; 11) supratentorial subdural hemorrhage; 12) lumbar spine fracture or subluxation; and 13) soft tissue injury of the face. In the validation cohort, the model had an AUC of 0.83 (95% CI 0.81–0.86). When the definition of BCVI was narrowed to include only carotid (common, internal) and vertebral artery injuries, the AUC of the model in predicting these injuries was 0.89 (95% CI 0.87–0.91).
A new screening model that incorporates an easy-to-use nomogram to quantify the risk of BCVI and assist clinicians in identifying patients who warrant additional evaluation was established.
Keita Shibahashi, Kazuhiro Sugiyama, Jun Tomio, Hidenori Hoda, and Akio Morita
The optimal surgical treatment for acute subdural hemorrhage (ASDH) remains controversial. The purpose of this study was to compare outcomes in patients who underwent craniotomy with those in patients who underwent decompressive craniectomy for the treatment of ASDH.
Using the Japan Trauma Data Bank, a nationwide trauma registry, the authors identified patients aged ≥ 18 years with ASDH who underwent surgical evacuation after blunt head trauma between 2004 and 2015. Logistic regression analysis was used to estimate a propensity score to predict decompressive craniectomy use. They then used propensity score–matched analysis to compare patients who underwent craniotomy with those who underwent decompressive craniectomy. To identify the potential benefits and disadvantages of decompressive craniectomy among different subgroups, they estimated the interactions between treatment and the subgroups using logistic regression analysis.
Of 236,698 patients who were registered in the database, 1788 were eligible for propensity score–matched analysis. The final analysis included 514 patients who underwent craniotomy and 514 patients who underwent decompressive craniectomy. The in-hospital mortality did not differ significantly between the groups (41.6% for the craniotomy group vs 39.1% for the decompressive craniectomy group; absolute difference −2.5%; 95% CI −8.5% to 3.5%). The length of hospital stay was significantly longer in patients who underwent decompressive craniectomy (median 23 days [IQR 4–52 days] vs 30 days [IQR 7–60 days], p = 0.005). Subgroup analyses demonstrated qualitative interactions between decompressive craniectomy and the patient subgroups, suggesting that patients who were more severely injured (Glasgow Coma Scale score < 9 and probability of survival < 0.64) and those involved in high-energy injuries may be good candidates for decompressive craniectomy.
The results of this study showed that overall, decompressive craniectomy did not appear to be superior to craniotomy in ASDH treatment in terms of in-hospital mortality. In contrast, there were significant differences in the effectiveness of decompressive craniectomy between the subgroups. Thus, future studies should prioritize the identification of a subset of patients who will possibly benefit from the performance of each of the procedures.