ABSTRACT
Objectives:
This study aimed to assess the impact of distinct treatment modalities on outcomes in patients with moderate-to-severe traumatic brain injury (TBI).
Materials and Methods:
This retrospective single-center study analyzed 228 patients with moderate-to-severe TBI (Glasgow Coma Scale [GCS] 3–12) admitted to Hualien Tzu Chi Hospital from 2017 to 2019. Patients were grouped by age (<65 or ≥65) and GCS (3–8 or 9–12). Four treatment strategies were assessed: decompressive craniectomy (DC), craniotomy, intracranial pressure (ICP) monitoring alone, and conservative management. Outcomes included hospital length of stay (LOS), ICU duration, and mortality. Statistical analyses included Kaplan–Meier survival curves, Log-rank tests, Gehan–Breslow Wilcoxon tests, and Kruskal–Wallis tests.
Results:
In Group 1 (n = 58, <65, GCS: 9–12), DC was associated with the longest median hospital (43.0 days) and ICU stay (7.0 days). In Group 2 (n = 44, ≥65, GCS: 9–12), significant differences in ICU stay (P < 0.05) were observed. In Group 3 (n = 77, <65, GCS 3–8), ICP monitoring alone resulted in the longest hospital LOS (33.5 days), whereas DC was associated with the longest ICU stay (13 days). In Group 4 (≥65 years, GCS: 3–8, n = 49), DC correlated with the longest hospital LOS (39.0 days) and ICU stay (10.0 days). In severe TBI patients, the ICP monitoring group displayed better survival over time. Moreover, DC and ICP monitoring only achieved the lowest mortality rates (25%) in Group 4.
Conclusion:
Early DC and ICP monitoring substantially improve survival and aggravation in elderly patients with severe TBI, underscoring their critical role in outcome optimization.
KEYWORDS: Decompressive craniectomy, Delayed intracerebral hemorrhage, Early surgical decompression, Elderly patients, Traumatic brain injury
INTRODUCTION
Traumatic brain injury (TBI) is a common cause for visits to the emergency department (ED). The data of the Ministry of Health and Welfare of Taiwan showed that severe TBI is among the most severe injuries common in children and young adults and mostly caused by traffic accidents [1]. The World Health Organization (WHO) even estimates over 10 million cases associated with TBI-related mortality or hospitalization annually [2]. One of the main characteristics of TBI is that patients with severe TBI can experience subsequent mental and medical problems. Moreover, in those aged over 75, fall is the leading cause of TBI and often relates to hospitalization and even death. Therefore, efficiently predicting outcomes and providing effective care among different age groups was essential to facilitate better patient outcomes.
Current management of severe TBI consists of surgical and nonsurgical modalities, and the treatment varies in individuals [3]. Typically, patients treated surgically with severe TBI are defined by Glasgow Coma Scale (GCS) <9 or an Abbreviated Injury Score (AIS) of head >3 [4]. Intracranial pressure (ICP) is commonly applied to monitor cerebral blood flow and oxygen delivery, optimizing a healthy metabolic environment, especially in the secondary TBI prevention [5]. Patients with ICP >30 mmHg would typically undergo craniotomy, and those with ICP below 20 mmHg would be managed conservatively [6].
Primary brain injury involves traumatic shearing and tearing of axons, resulting in diffuse axonal injury and focal damage. Secondary brain injury arises from processes such as ischemia, cerebral edema, seizures, and oxidative stress, which contribute to further neuronal, axonal, and glial damage [7]. In the case of severe TBI, primary brain injury can quickly turn into a secondary brain injury if the common conservative medical approach (wait-and-see approach) fails to intervene the elevation of ICP [8]. Our study aims to evaluate the impact of decompressive craniectomy (DC), craniotomy, ICP monitoring only, and conservative treatment on clinical outcomes in elderly patients with moderate-to-severe TBI.
MATERIALS AND METHODS
This retrospective cohort study analyzed patients with moderate-to-severe TBI who were admitted to the ED of Hualien Tzu Chi Hospital within 24 h of injury between June 1, 2017, and May 31, 2019. Among the 787 TBI patients who presented to the ED during the study period, a total of 228 met the eligibility criteria. This study was conducted in accordance with the Declaration of Helsinki. The study was approved by the Research Ethics Committee (REC) of Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation (IRB 108-267B). All personal information of the patients was deidentified before analysis. The need for informed consent was waived by the REC due to the nature of the retrospective chart review study.
The study compiled data on gender, age, date of birth, hospital stays, ICU stays, GCS scores at admission and discharge, history of prior brain injury, medication usage, and diagnoses classified according to the International Statistical Classification of Diseases and Related Health Problems, 10th Revision, Clinical Modification. In addition, it included procedural codes related to DC, craniotomy, ICP monitoring only, and conservative treatment. The treatment modalities were classified based on the documented medical records, which specify the primary intervention and any secondary treatments administered. DC is defined as a neurosurgical intervention involving the permanent removal of a skull segment to alleviate ICP, without subsequent bone replacement. In contrast, craniotomy entails the removal of skull bone, which is immediately repositioned postsurgery, with burr hole procedures also categorized under craniotomy in this study. Patients who require ICP monitoring following craniotomy, due to sustained instability in vital signs, are classified within the ICP monitoring group, as their treatment hinges on continuous ICP management. Conservative management, conversely, refers to nonsurgical treatment reliant exclusively on medical therapies, such as the administration of mannitol or other pharmacological agents to mitigate cerebral edema.
Early surgical intervention was defined as any procedure performed within 24 h of admission. The primary outcome assessed was in-hospital mortality, whereas secondary outcomes included death, increased intracranial pressure (IICP), and the need for additional surgical interventions. The hospital length of stay (LOS) was standardized to 30 days for survival curve analysis. The aggravation included seizure, delayed intracranial hematoma (ICH), and death occurred after the initial intervention.
To compare surgical and nonsurgical treatments for moderate-to-severe TBI, the 228 patients were stratified into four groups based on age and GCS scores upon hospital arrival [Figure 1]. According to the WHO definition, individuals aged 65 years and older were classified as elderly, whereas those younger than 65 years were categorized as young. GCS scores were interpreted based on the Parkland Trauma Handbook (Third Edition, 2009), where scores of 13–15 indicate mild head injury, 9–12 correspond to moderate head injury, and scores of 8 or lower represent severe head injury. Patients were grouped into elderly (≥65 years) and young (<65 years) categories, as well as moderate (GCS 9–12) and severe (GCS 3–8) TBI groups. Group 1 consisted of 58 patients aged <65 with GCS 9–12, Group 2 included 44 patients aged >65 with GCS 9–12, Group 3 comprised 77 patients aged <65 with GCS 3–8, and Group 4 included 49 patients aged >65 with GCS 3–8. The outcomes of DC, craniotomy, ICP monitoring only, and conservative treatment were then assessed within these classifications.
Figure 1.
The flow chart shows how the patients were divided into four groups by Glasgow Coma Scale and age. Each group was subdivided into four kinds of treatments: craniectomy, craniotomy, intracranial pressure monitor, and conservative treatment
Statistical analysis was performed using GraphPad Prism 8. Normally distributed data were summarized as mean (standard deviation), nonnormally distributed data as median (interquartile range), and categorical data as proportions (percentages). A median and standard deviation graph was used to illustrate the characteristics of the four groups across different ages and GCS scores.
Since the standard deviation values were close to the mean, the median was selected as the primary measure of central tendency. The Kruskal–Wallis test was applied to assess statistical significance among the four treatment modalities, whereas hospital LOS comparisons were conducted using Kaplan–Meier survival curves and logistic regression. P < 0.05 was considered statistically significant, whereas P < 0.1 was regarded as marginally significant for all analyses.
RESULTS
In this retrospective cohort study, 787 patients were admitted to the emergency room within 24 h of head injuries. Of these, 559 patients with a GCS score of 13–15 were excluded from the analysis [Figure 1]. The remaining 228 patients were divided into four groups: Group 1 consisted of 58 patients aged <65 with GCS 9–12, Group 2 included 44 patients aged >65 with GCS 9–12, Group 3 comprised 77 patients aged <65 with GCS 3–8, and Group 4 included 49 patients aged >65 with GCS 3–8. Among Group 1 patients, 7 underwent craniectomy, 6 underwent craniotomy, 9 received ICP monitoring only, and 36 were treated conservatively. In Group 2, 4 underwent craniectomy, 11 underwent craniotomy, 7 received ICP monitoring only, and 22 were treated conservatively. For Group 3, 22 underwent craniectomy, 10 underwent craniotomy, 20 received ICP monitoring only, and 25 received conservative treatment. Finally, in Group 4, 8 underwent craniectomy, 9 underwent craniotomy, 8 received ICP monitoring only, and 24 were managed conservatively. According to established TBI guidelines in 2020, patients in Group 3 (GCS 3–8, <65 years) and Group 4 (GCS 3–8, ≥65 years) ultimately required ICP monitoring. However, among patients younger than 65 years, only two individuals underwent craniectomy following an initial craniotomy.
In Group 1, the median hospital LOS and ICU stay varied based on treatments administered. Among these patients, those who underwent DC had the longest hospital LOS (43.0 days), followed by those with ICP monitoring only (24.0 days), craniotomy (16.5 days), and conservative treatment (9.5 days). The medians of ICU stay also exhibited a similar pattern, but with ICP monitoring only ranking the longest (9.0 days), followed by DC (7.0 days), then craniotomy (3.5 days), and finally, conservative treatment (3.0 days). Kruskal–Wallis test revealed significant differences (P < 0.05) in the distributions of both hospital LOS and ICU stay. The survival curves for craniotomy, craniectomy, and ICP monitoring only overlapped substantially due to similar outcomes and no mortality events, with no significant differences detected, resulting in superimposed lines in Figure 2a.
Figure 2.
A 30-day observation of progression curves for patients with Glasgow Coma Scale 9–12 showed no significant difference in treatments: (a) <65 years old (b) ≥65 years old (c) No age differences
However, in Group 2, the Kruskal-Wallis test revealed a significant difference in the ICU stay (P < 0.05), with no significant difference observed in hospital LOS. Both tests yielded comparable results, with DC ranking the longest (39 days and 10 days), followed by craniotomy (19.5 days and 9.5 days), ICP monitoring only (17 days and 8 days), and conservative treatment (15 days and 2.5 days). Notably, there were no significant differences in age distribution among treatment groups across the <65, ≥65 subgroups, or the overall cohort [Figure 2a-c].
In Group 3, the hospital LOS and ICU stay varied depending on treatments administered as well. Patients subjected to ICP monitoring only had the longest hospital LOS (33.5 days), followed by those who underwent DC (16 days), conservative treatment (10.0 days), and craniotomy (4.5 days). Nonetheless, the ICU stay displayed a different pattern, with DC resulting in the longest (11.5 days), followed by ICP monitoring only (6.5 days), craniotomy (4.5 days), and conservative treatment (2.0 days). The progression curve also showed a similar trend, with ICP monitoring only resulting in the longest stay, followed by DC, conservative treatment, and craniotomy [Figure 3a]. In addition, the aggravation rate following different treatment modalities was highest in patients who underwent DC (82%), followed by those who received craniotomy (80%), conservative management (44%), and ICP monitoring alone (40%).
Figure 3.
A 30-day observation of progression curves for patients with Glasgow Coma Scale 3–8. (a) <65 years old (b) ≥65 years old (c) No age differences
In Group 4, patients undergoing DC had a hospital LOS of 23 days, followed by those receiving ICP monitoring only (20 days), craniotomy (19 days), and conservative management (6 days). ICU stay revealed a different pattern for those with ICP monitoring only and craniotomy, ranking the longest (9 days), followed by DC (7 days) and then conservative management (4 days). The Kruskal–Wallis test identified a significant association between ICU stay and treatment modality (P < 0.05), whereas no such correlation emerged for hospital LOS. Nevertheless, log-rank and Gehan–Breslow Wilcoxon tests discerned significant differences in progression curves (P < 0.05), ranking treatments in descending order as craniotomy, ICP monitoring only, DC, and conservative management [Figure 3b]. A 30-day analysis of progression curves for patients with GCS scores of 3–8 further corroborated the efficacy of ICP monitoring, yielding statistically significant results through the Log-rank and Gehan–Breslow Wilcoxon tests [Figure 3c]. Clinical aggravation rates were highest in the conservative treatment group (79.2%), followed by ICP monitoring only (75%), craniotomy (55.6%), and DC (37.5%), while mortality rates aligned as follows: DC and ICP monitoring only both demonstrated the lowest mortality (25%), underscoring the pivotal role of early DC in patient stabilization, followed by craniotomy (33.3%) and conservative treatment (75%) [Supplementary Table 1a]. Progression curves [Figure 3c] revealed that the DC and craniotomy groups exhibited initially stable survival trajectories with modest declines, whereas patients managed conservatively experienced a sharp survival decrement, with a significant proportion succumbing within the first 10 days. In contrast, the ICP monitoring group displayed a less pronounced decline in survival compared to other groups.
DISCUSSION
Age is an important factor influencing the final outcome after severe TBI, especially after the age of 40 [9]. In addition, the mortality rate of geriatric trauma increased by 6.8% with each year increase in age after the age of 60 [10]. In a study from 2015, 50% (n = 97) of the elderly with severe TBI expired within 14 days [11]. In a meta-analysis, 22 studies were included to calculate an overall mortality rate (38.3%) of TBI. Pooled mortality rate for severe head injuries was 65.3% [12]. The way to prevent unfavorable outcomes of the elderly with severe TBI is still controversial.
Wan et al.’s study showed that the elderly with operative treatment were more likely to have a better outcome (52.9% vs. 95.2%) or reduced mortality (32.9% vs. 88.1%) at 6 months after injury compared with patients with conservative treatment. Moreover, surgical treatment can reduce mortality and improve outcomes among the elderly with TBI [13]. Other studies suggested that patients with DC showed no significant improvement in their neurological outcomes and even aggravated their functional outcomes [14,15]. Kinoshita et al. directly classified the patients according to their treatment, but this study also considered the severity and the age of the patient, which led to the different outcomes between the two studies [15]. In Group 4 of our study, the aggravation rate of DC (37.5%) was much lower than the other treatments (55.6%, 75%, and 79.2%). Although there was a longer hospital LOS, patients with DC had lower mortality. The result suggested that early DC could decrease the need for additional operation among patients with advanced age and lower GCS scores, which is similar to Wan et al.’s study [13]. Unfortunately, most patients in Group 4 who received conservative treatment, rather than DC, experienced poorer outcomes (mortality rate = 78.26%) [Supplementary Table 1a]. Notably, two cases in Group 4 underwent DC 24 h later after admission, and both resulted in death. We speculated that most of the initial clinical management within 24 h in Group 4 was appropriate and correct.
It is worth mentioning that there was the longest hospital LOS in patients with DC in Group 1 (43.0 days). We speculated that patients with DC tend to keep hospitalization until undergoing cranioplasty (57.1%) due to the relatively inadequate transport system in Hualian [Supplementary Table 1b]. There was a similar trend in Group 2 (39 days). Nevertheless, only 25% of patients with DC underwent cranioplasty [Supplementary Table 1b]. This could have been that most of the patients in Group 2 lacked physiological reserve due to their age, resulting in long-stay hospitalization.
ICP monitoring enabled precise measurement of ICP, allowing surgeons to assess cerebral perfusion pressure. Furthermore, elevated ICP could indicate the risk of secondary brain injury or re-bleeding. Real-time data monitoring facilitated timely interventions, enabling proactive management to maintain cerebral homeostasis and improve patient outcomes. According to the Brain Trauma Foundation, ICP monitoring is regarded as a recommended approach to severe TBI. The relationship between ICP monitoring and mortality rate had also been studied in previous studies. Shen et al.’s meta-analysis showed a better survival with ICP monitoring, including overall, 2 weeks, 6 months, and hospital mortality [16]. Moreover, Dawes et al.’s study which included 14 trauma centers over 2 years had similar results, especially for patients with high ISS (median = 26) and low GCS (median = 3) [17].
This study was consistent with Dawes’ study [17]. In Group 3, the patients treated with ICP monitoring only and conservative treatment had similar ISS and multitrauma ratio. Nevertheless, the hospital LOS of conservative treatment was much lower than that of ICP monitoring only (10 days vs. 33.5 days). It is also found that the mortality of those patients with conservative treatment was at least two times higher than that of patients with ICP monitoring [Supplementary Table 1c]. The decrease in the hospital LOS represented the high mortality in Group 3. In conclusion, ICP monitoring enabled surgeons to follow up patients’ condition, leading to the decrease in the mortality rate. Moreover, device selection for ICP monitoring is prognostically significant, with external ventricular drains (EVDs) offering superior control of refractory intracranial hypertension; therefore, routine EVD placement is recommended in TBI patients unless only parenchymal monitoring is available [18].
Furthermore, patients in Group 1 with ICP monitoring only had the longest ICU LOS (9 days), even though their AIS (head) scores were identical to others. Prominently, we observed the highest ISS (20.22) and multitrauma ratio (67%) of these patients [Supplementary Table 1b]. Multitrauma could increase severity and prolong the days of recovery. Thus, the longest ICU LOS in patients with ICP monitoring only was for the high multitrauma ratio, which was at least two times higher than the other three procedures. Comprehensive data on patients over 65 years with GCS 9–12 are presented in Supplementary Table 1d.
Although attending physicians typically manage patients optimally, some patients deteriorate and may require secondary DC after the initial 24 h of admission; therefore, we focus on these atypical cases [Supplementary Table 2]. Epidural hematoma (EDH) is a serious complication of TBI, with 1%–4% in all TBI cases [19]. Approximately 85% of EDH cases are associated with arterial injury, especially tearing of middle meningeal artery, with the following hematoma over the cerebral convexity in the middle cranial fossa, while the rests are due to injury of dural sinuses and the confluence of sinus [19,20]. However, as surgical procedures are expected to cause hemorrhage, postoperative EDH might ensue after evacuation of the mass lesion [21,22,23]. Risk factors include postoperation hypotension, anticoagulation, and extensive surgical excision. In addition, Akpinar et al. [24] reported EDH occurred after craniotomy was done in a 41-year-old male with chronic subdural hematoma (SDH) and further proposed that rapid dynamic intracranial changes, such as the sudden restoration of normal perfusion pressure in the injured area, might precipitate vascular damage and epidural leakages.
In our study, there was 7% of our patients suffered from EDH, and Group 1 rank the most cases and the highest ratio of EDH. In addition, 4.4% of our patients presented with the complication of late EDH [Supplementary Table 2]. However, only one patient in Group 4 had late EDH after craniotomy. Initial loss of consciousness occurred after falling. After admission, computed tomography (CT) showed the right SDH, and emergent craniotomy was done to remove the hematoma. Seven hours later, the patient was observed with signs of IICP, and CT indicated the right EDH with midline shift. As the injured site was on the right temporal, it is possible that the right middle meningeal artery of our patient was slightly injured in the beginning, then necrotized owing to the post-surgical hypotension of craniotomy.
Occasionally, TBI patients arrive at ER with minimal ICH, and either conservative treatment or ICP monitor placement is applied for further observations. Most of them recover and discharge, while unfortunately some patients increased ICP with complications such as delayed ICH and seizure. In this study, within subgroups of ICP monitoring only and conservative treatment, there were 64% of patients recover without any complications, while ratio of patients with seizure and delayed ICH were 5.3% and 6%, respectively [Table 1]. In addition, Group 2 had the highest ratio of patients with seizure and Group 3 ranked first in ratio of patients with delayed ICH. Regarding patients with late DC, there is one patient with delayed ICH in Group 4, two patients with delayed ICH and two patients with brain swelling in Group 3, and two patients with delayed ICH in Group 2.
Table 1.
A comparison table showed the number and ratio of patients recover without complications, death, or having complications such as seizure and delayed intracranial hematoma in subgroups of intracranial pressure monitor and conservative treatment in each group
| Group 1 (45) | Group 2 (28) | Group 3 (45) | Group 4 (32) | Total (150) | |
|---|---|---|---|---|---|
| Recover without complication | 38 | 21 | 25 | 12 | 96 |
| Seizure | 3 | 2 | 2 | 1 | 8 |
| Delayed ICH | 1 | 2 | 5 | 1 | 9 |
| Death | 2 | 4 | 14 | 20 | 40 |
| Ratio of recovery without complication (%) | 84.4 | 75 | 55.6 | 37.5 | 64 |
| Ratio of seizure (%) | 6.7 | 7.1 | 4.4 | 3.1 | 5.3 |
| Ratio of delayed ICH (%) | 2.2 | 7.1 | 11.1 | 3.1 | 6 |
| Ratio of death (%) | 4.4 | 14.3 | 31.1 | 62.5 | 26.7 |
ICH: Intracranial hematoma
The incidence of delayed ICH varies among studies and was affected by the criteria that attending doctors apply for rehabilitation and follow-up CT scan [25]. Fukamachi et al. classified posttraumatic ICHs into four types, and further regarded Types 3 and 4 as true delayed ICH, which were hematomas developing from normal areas in the initial CT scan as Type 3 and hematomas developing from contusion areas that were of mixed density or isodensity as Type 4 [26]. Our analysis identified several factors associated with delayed ICH, including older age [Figure 4a and b], acute brain hematomas [Figure 4c and d], multiple ICHs [Figure 4e and f], smaller initial volume, coagulopathy, falls, and prior operations such as DC. These findings align with previously reported risk factors [27]. In Group 3, a patient experienced a head contusion after falling down the stairs, with CT imaging revealing acute SDH and bilateral multifocal ICH. DC was done 8 h after ICP monitor placement, and the patient finally recovered and discharged.
Figure 4.
Progressions of delayed intracranial hematoma (ICH) with different variants. (a and b) An elderly patient with a sparse brain gyrus and small amount of hematoma. After 3 days of intracranial pressure monitoring, the patient developed a large amount of hematoma, and brain gyrus became relatively dense due to mass effect. (c and d) A drunk driver with acute SAH and IVH. A day later, sites and areas of hematoma increased. (e and f) A congenital muscular dystrophy patient with a small volume of multiple ICH. Eight hours later, a delayed ICH resulted in IICP
Mechanisms for the progression of delayed ICH in multifocal ICH patients are still unclear; however, Cepeda et al. [28] proposed that the gradual accumulation of small lesions results in a great volume of ICH. The GCS is a practical tool to evaluate the conscious level, depth, and duration of coma and to identify the development of complications and the possibility of ultimate recovery [27]. In several studies, a low GCS score is suggested to be associated with poor TBI conditions, progression of intracranial lesions, and unfavorable outcomes [29,30]. However, since GCS is composed of eye opening, verbal response, and motor response, initial assessment of GCS is affected by certain scenarios such as alcohol intake, psychoactive drug use, mass hemorrhage of trauma, and tracheal intubation. If there are no such influences, we suggest that the severity of GCS might be a way to distinguish patients with either singular or multiple TBI. Further studies are required to validate this hypothesis.
Advanced age is a well-established predictor of poorer clinical outcomes [31,32]. According to the US Centers for Disease Control and Prevention database, the mortality rate of patients over 65 is twice more than any younger groups [33]. In addition, advanced age may be connected with delayed ICH progression and refractory IICP [28]. While the relationship between advanced age and poor TBI progression remains unclear, a potential explanation is the age-related decline in recovery capacity, resulting in prolonged brain tissue and vascular repair, along with delayed resolution of ICH. Moreover, studies have shown that brain volume decreases with age, providing more space for hematoma expansion, which may contribute to the development of delayed ICH [34,35,36]. The compromised condition also challenged the surgical intervention, contributing to higher mortality rates in older individuals. In Group 4, a 66-year-old male was brought to the ER following a traffic accident caused by drunk driving. Upon admission, a CT scan revealed hematoma in the left frontal region and sparse brain gyrus [Figure 4a and b]. 72 h after the placement of an ICP monitor, IICP developed, and a follow-up CT scan showed a relatively dense gyrus with a large area of hematoma. Despite undergoing DC to evacuate the hematoma, the patient’s condition continued to deteriorate, ultimately resulting in death.
According to the 2020 updates in the Brain Trauma Foundation’s Guidelines for the surgical management of Traumatic Brain injury, secondary DC is recommended to improve mortality, favorable outcomes, and reduce ICP as well as duration of intensive care for late refractory ICP elevation [37]. In addition, two trials suggested similar guidance for the treatment of late ICP elevation [38,39]. In Group 3, a patient with congenital muscular dystrophy sustained a head contusion from a fall, with initial CT scan showing small multiple intracranial hemorrhages and an ICP of 20 mmHg [Figure 4e and f]. Eight hours later, ICP increased and delayed ICH occurred. After immediate DC, the patient’s condition stabilized, and recovery followed. Another Group 3 patient, a drunk driver, presented with subarachnoid hemorrhage and intraventricular hemorrhage, prompting ICP monitoring [Figure 4c and d]. Following hematoma expansion observed after 24 h, DC was indicated; however, surgical intervention was delayed due to family refusal. Despite subsequent DC, the patient did not achieve recovery. These cases underscore the necessity of ICP monitoring in patients with multifocal hematomas and minimal mass effect, advocating for early DC upon signs of IICP, which aligns with the 2020 Brain Trauma Foundation guidelines on the surgical management of TBI, emphasizing the critical role of timely surgical intervention in improving patient outcomes.
This study has certain limitations. As a retrospective, nonrandomized, and one single-institute study from Hualian Tzu Chi Hospital, generalizability may be limited. Treatment decisions were based on physicians’ clinical judgment rather than randomization, which, although made with careful consideration of each patient’s best interests, may introduce potential confounding. Small sample sizes in certain subgroups, such as elderly patients with specific GCS scores, may have reduced statistical power. In addition, outcomes were limited to in-hospital measures, including 60-day mortality, without long-term follow-up to assess functional recovery. Patients were grouped based on treatments administered within the first 24 h after admission to enhance comparability among the four treatment modalities; however, complete standardization remains challenging due to the inherent variability and complexity of clinical practice. Variability in both the timing and indications for treatments, particularly for interventions such as DC and ICP monitoring, should therefore be considered when interpreting comparisons across groups, especially in delayed interventions. Nonetheless, our findings provide meaningful insights and underscore the need for future prospective studies to define the optimal timing of DC in elderly patients and to examine management strategies for specific aggravation events and their impact on patient prognosis.
CONCLUSION
Early DC in elderly patients with severe TBI is associated with enhanced survival and improved clinical outcomes. Moreover, ICP monitoring in patients with severe TBI and polytrauma has been shown to reduce mortality and mitigate disease progression. In cases of multifocal hematomas with minimal mass effect, ICP monitoring remains crucial for guiding timely neurosurgical intervention. Further research is warranted to determine the optimal timing of DC in elderly patients to maximize therapeutic benefits.
Data availability statement
The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.
Conflicts of interest
There are no conflicts of interest.
Acknowledgments
The authors thank the statistical consultant, Dr. Jen-Hung Wang (Department of Medical Research, Hualien Tzu Chi Hospital, Buddhist Tzu Chi Medical Foundation, Hualien, Taiwan) and Dr. Tsung-Cheng Hsieh (Institute of Medical Sciences and Research Center for Big Data Teaching, Tzu Chi University, Hualien, Taiwan) for advice on statistical analysis.
SUPPLEMENTARY MATERIAL
Supplementary Table 1a-d: Comparison tables among four types of treatment in patients of four groups
Supplementary Table 1a.
Group 4: Patients over 65 with Glasgow Coma Scale 3–8
| Group 4 | Craniectomy | Craniotomy | ICP monitoring | Conservative treatment |
|---|---|---|---|---|
| Hospital LOS | 23 | 19 | 20 | 6 |
| ICU LOS | 7 | 9 | 9 | 4 |
| Initial GCS | 5 | 5.44 | 4.63 | 4.17 |
| AIS - head | 4.75 | 4.56 | 4 | 4.43 |
| ISS | 24 | 22.44 | 19.63 | 26.79 |
| Multitrauma (ratio) (%) | 12.5 | 11.1 | 12.5 | 29.2 |
| Aggravation rate (%) | 37.5 | 55.6 | 75 | 79.2 |
| Mortality (%) | 25 | 33.3 | 25 | 75 |
AIS: Abbreviated Injury Score, GCS: Glasgow Coma Scale, ICP: Intracranial pressure, ICU: Intensive care unit, ISS: Injury Severity Score, LOS: Length of stay
Supplementary Table 1b.
Group 1: Patients under 65 with Glasgow Coma Scale 9–12
| Group 1 | Craniectomy | Craniotomy | ICP monitoring | Conservative treatment |
|---|---|---|---|---|
| Hospital LOS | 43.0 | 16.5 | 24.0 | 9.5 |
| ICU LOS | 7 | 3.5 | 9 | 3 |
| Initial GCS | 10.14 | 10.6 | 9.1 | 10.78 |
| AIS - head | 4.14 | 3 | 3.55 | 3.13 |
| ISS | 18.57 | 13.16 | 20.22 | 14.16 |
| Multitrauma (ratio) (%) | 0 | 16.7 | 67 | 33 |
| Aggravation rate (%) | 57 | 33 | 22 | 6 |
| Mortality (%) | 0 | 0 | 0 | 5.6 |
AIS: Abbreviated Injury Score, GCS: Glasgow Coma Scale, ICP: Intracranial pressure, ICU: Intensive care unit, ISS: Injury Severity Score, LOS: Length of stay
Supplementary Table 1c.
Group 3: Patients under 65 with Glasgow Coma Scale 3–8
| Group 3 | Craniectomy | Craniotomy | ICP monitoring | Conservative treatment |
|---|---|---|---|---|
| Hospital LOS | 16.0 | 4.5 | 33.5 | 10.0 |
| ICU LOS | 11.5 | 4.5 | 6.5 | 2.0 |
| Initial GCS | 4.05 | 5.2 | 5.9 | 5.1 |
| AIS - head | 4.32 | 4.6 | 4.05 | 4.08 |
| ISS | 26.77 | 25.9 | 26.3 | 24.6 |
| Multitrauma (ratio) (%) | 46 | 30 | 65 | 60 |
| Aggravation rate (%) | 82 | 80 | 40 | 44 |
| Mortality (%) | 50 | 70 | 20 | 44 |
AIS: Abbreviated Injury Score, GCS: Glasgow Coma Scale, ICP: Intracranial pressure, ICU: Intensive care unit, ISS: Injury Severity Score, LOS: Length of stay
Supplementary Table 1d.
Group 2: Patients over 65 with Glasgow Coma Scale 9–12
| Group 2 | Craniectomy | Craniotomy | ICP monitoring | Conservative treatment |
|---|---|---|---|---|
| Hospital LOS | 39 | 19.5 | 17 | 15 |
| ICU LOS | 10 | 9.5 | 8 | 2.5 |
| Initial GCS | 10.75 | 10.55 | 10.86 | 10.82 |
| AIS - head | 4.5 | 3.91 | 3.43 | 3.091 |
| ISS | 22 | 18.91 | 14.14 | 11.86 |
| Multitrauma (ratio) (%) | 25 | 27 | 0 | 23 |
| Aggravation rate (%) | 50 | 36 | 57 | 18 |
| Mortality (%) | 0 | 27 | 29 | 9.1 |
AIS: Abbreviated Injury Score, GCS: Glasgow Coma Scale, ICP: Intracranial pressure, ICU: Intensive care unit, ISS: Injury Severity Score, LOS: Length of stay
Supplementary Table 2.
A comparison table showed the number and ratio of odd cases, epidural hematoma, and late epidural hematoma in each group
| Group 1 (58) | Group 2 (44) | Group 3 (77) | Group 4 (49) | Total (228) | |
|---|---|---|---|---|---|
| Number of odd cases | 0 | 2 | 4 | 2 | 8 |
| Ratio of odds cases in group (%) | 0 | 4.5 | 5.2 | 4.1 | 3.5 |
| EDH | 8 | 3 | 2 | 3 | 16 |
| EDH ratio (%) | 13.8 | 6.8 | 2.6 | 6.12 | 7 |
| Late EDH | 5 | 0 | 4 | 1 | 10 |
| Late EDH ratio (%) | 8.6 | 0 | 5.2 | 2 | 4.4 |
EDH: Epidural hematoma
Funding Statement
Nil.
REFERENCES
- 1.Kuo CY, Chiou HY, Lin JW, Tsai SH, Chiang YH, Lin CM, et al. Seatbelt use and traumatic brain injury in Taiwan: A 16 Year Study. Iran J Public Health. 2015;44:470–8. [PMC free article] [PubMed] [Google Scholar]
- 2.Hyder AA, Wunderlich CA, Puvanachandra P, Gururaj G, Kobusingye OC. The impact of traumatic brain injuries: A global perspective. NeuroRehabilitation. 2007;22:341–53. [PubMed] [Google Scholar]
- 3.Galgano M, Toshkezi G, Qiu X, Russell T, Chin L, Zhao LR. Traumatic brain injury: Current treatment strategies and future endeavors. Cell Transplant. 2017;26:1118–30. doi: 10.1177/0963689717714102. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Pélieu I, Kull C, Walder B. Prehospital and emergency care in adult patients with acute traumatic brain injury. Med Sci (Basel) 2019;7:12. doi: 10.3390/medsci7010012. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Bouzat P, Sala N, Payen JF, Oddo M. Beyond intracranial pressure: Optimization of cerebral blood flow, oxygen, and substrate delivery after traumatic brain injury. Ann Intensive Care. 2013;3:23. doi: 10.1186/2110-5820-3-23. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Gallbraith S, Teasdale G. Predicting the need for operation in the patient with an occult traumatic intracranial hematoma. J Neurosurg. 1981;55:75–81. doi: 10.3171/jns.1981.55.1.0075. [DOI] [PubMed] [Google Scholar]
- 7.Alvis Miranda H, Castellar Leones SM, Moscote Salazar LR. Decompressive craniectomy and traumatic brain injury: A review. Bull Emerg Trauma. 2013;1:60–8. [PMC free article] [PubMed] [Google Scholar]
- 8.Stocker RA. Intensive care in traumatic brain injury including multi-modal monitoring and neuroprotection. Med Sci (Basel) 2019;7:37. doi: 10.3390/medsci7030037. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 9.MRC CRASH Trial Collaborators. Perel P, Arango M, Clayton T, Edwards P, Komolafe E, et al. Predicting outcome after traumatic brain injury: Practical prognostic models based on large cohort of international patients. BMJ. 2008;336:425–9. doi: 10.1136/bmj.39461.643438.25. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Solomon D, Kaminski O, Schrier I, Kashtan H, Stein M. Isolated traumatic brain injury in the very old. Isr Med Assoc J. 2019;21:779–84. [PubMed] [Google Scholar]
- 11.Røe C, Skandsen T, Manskow U, Ader T, Anke A. Mortality and one-year functional outcome in elderly and very old patients with severe traumatic brain injuries: Observed and predicted. Behav Neurol 2015. 2015:845491. doi: 10.1155/2015/845491. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 12.McIntyre A, Mehta S, Aubut J, Dijkers M, Teasell RW. Mortality among older adults after a traumatic brain injury: A meta-analysis. Brain Inj. 2013;27:31–40. doi: 10.3109/02699052.2012.700086. [DOI] [PubMed] [Google Scholar]
- 13.Wan X, Liu S, Wang S, Zhang S, Yang H, Ou Y, et al. Elderly patients with severe traumatic brain injury could benefit from surgical treatment. World Neurosurg. 2016;89:147–52. doi: 10.1016/j.wneu.2016.01.084. [DOI] [PubMed] [Google Scholar]
- 14.Pompucci A, De Bonis P, Pettorini B, Petrella G, Di Chirico A, Anile C. Decompressive craniectomy for traumatic brain injury: Patient age and outcome. J Neurotrauma. 2007;24:1182–8. doi: 10.1089/neu.2006.0244. [DOI] [PubMed] [Google Scholar]
- 15.Kinoshita T, Yoshiya K, Fujimoto Y, Kajikawa R, Kiguchi T, Hara M, et al. Decompressive craniectomy in conjunction with evacuation of intracranial hemorrhagic lesions is associated with worse outcomes in elderly patients with traumatic brain injury: A propensity score analysis. World Neurosurg. 2016;89:187–92. doi: 10.1016/j.wneu.2016.01.071. [DOI] [PubMed] [Google Scholar]
- 16.Shen L, Wang Z, Su Z, Qiu S, Xu J, Zhou Y, et al. Effects of intracranial pressure monitoring on mortality in patients with severe traumatic brain injury: A meta-analysis. PLoS One. 2016;11:e0168901. doi: 10.1371/journal.pone.0168901. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17.Dawes AJ, Sacks GD, Cryer HG, Gruen JP, Preston C, Gorospe D, et al. Intracranial pressure monitoring and inpatient mortality in severe traumatic brain injury: A propensity score-matched analysis. J Trauma Acute Care Surg. 2015;78:492–501. doi: 10.1097/TA.0000000000000559. [DOI] [PubMed] [Google Scholar]
- 18.Liu H, Wang W, Cheng F, Yuan Q, Yang J, Hu J, et al. External ventricular drains versus intraparenchymal intracranial pressure monitors in traumatic brain injury: A prospective observational study. World Neurosurg. 2015;83:794–800. doi: 10.1016/j.wneu.2014.12.040. [DOI] [PubMed] [Google Scholar]
- 19.Louis ED, Mayer SA, Rowland LP. Traumatic Brain Injury. In: Merritt's Neurology, 13th ed. Lippincott Williams &Wilkins, a Wolters Kluwer business. 2016:364–79. [Google Scholar]
- 20.Matsumoto K, Akagi K, Abekura M, Tasaki O. Vertex epidural hematoma associated with traumatic arteriovenous fistula of the middle meningeal artery: A case report. Surg Neurol. 2001;55:302–4. doi: 10.1016/s0090-3019(01)00328-7. [DOI] [PubMed] [Google Scholar]
- 21.Fukamachi A, Koizumi H, Nagaseki Y, Nukui H. Postoperative extradural hematomas: Computed tomographic survey of 1105 intracranial operations. Neurosurgery. 1986;19:589–93. doi: 10.1227/00006123-198610000-00013. [DOI] [PubMed] [Google Scholar]
- 22.Seifman MA, Lewis PM, Rosenfeld JV, Hwang PY. Postoperative intracranial haemorrhage: A review. Neurosurg Rev. 2011;34:393–407. doi: 10.1007/s10143-010-0304-3. [DOI] [PubMed] [Google Scholar]
- 23.Chughtai KA, Nemer OP, Kessler AT, Bhatt AA. Post operative complications of craniotomy and craniectomy. Emerg Radiol. 2019;26:99–107. doi: 10.1007/s10140-018-1647-2. [DOI] [PubMed] [Google Scholar]
- 24.Akpinar A, Ucler N, Erdogan U, Yucetas CS. Epidural Hematoma Complication after Rapid Chronic Subdural Hematoma Evacuation: A Case Report. Am J Case Rep. 2015;16:430–3. doi: 10.12659/AJCR.894539. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.Cooper PR. Delayed traumatic intracerebral hemorrhage. Neurosurg Clin N Am. 1992;3:659–65. [PubMed] [Google Scholar]
- 26.Fukamachi A, Nagaseki Y, Kohno K, Wakao T. The incidence and developmental process of delayed traumatic intracerebral haematomas. Acta Neurochir (Wien) 1985;74:35–9. doi: 10.1007/BF01413274. [DOI] [PubMed] [Google Scholar]
- 27.Middleton PM. Practical use of the Glasgow Coma Scale;a comprehensive narrative review of GCS methodology. Australas Emerg Nurs J. 2012;15:170–83. doi: 10.1016/j.aenj.2012.06.002. [DOI] [PubMed] [Google Scholar]
- 28.Cepeda S, Gómez PA, Castaño Leon AM, Martínez Pérez R, Munarriz PM, Lagares A. Traumatic intracerebral hemorrhage: Risk factors associated with progression. J Neurotrauma. 2015;32:1246–53. doi: 10.1089/neu.2014.3808. [DOI] [PubMed] [Google Scholar]
- 29.Bullock R, Golek J, Blake G. Traumatic intracerebral hematoma – Which patients should undergo surgical evacuation?CT scan features and ICP monitoring as a basis for decision making. Surg Neurol. 1989;32:181–7. doi: 10.1016/0090-3019(89)90176-6. [DOI] [PubMed] [Google Scholar]
- 30.Stein SC, Spettell C, Young G, Ross SE. Delayed and progressive brain injury in closed head trauma: Radiological demonstration. Neurosurgery. 1993;32:25–30. doi: 10.1227/00006123-199301000-00004. [DOI] [PubMed] [Google Scholar]
- 31.Hukkelhoven CW, Steyerberg EW, Rampen AJ, Farace E, Habbema JD, Marshall LF, et al. Patient age and outcome following severe traumatic brain injury: An analysis of 5600 patients. J Neurosurg. 2003;99:666–73. doi: 10.3171/jns.2003.99.4.0666. [DOI] [PubMed] [Google Scholar]
- 32.Czosnyka M, Balestreri M, Steiner L, Smielewski P, Hutchinson PJ, Matta B, et al. Age, intracranial pressure, autoregulation, and outcome after brain trauma. J Neurosurg. 2005;102:450–4. doi: 10.3171/jns.2005.102.3.0450. [DOI] [PubMed] [Google Scholar]
- 33.Centers for Disease Control and Prevention, US Department of Health and Human Services. [[Last accessed on 16 Sep 2025]];Surveillance Report of Traumatic Brain Injury-related Hospitalizations and Deaths by Age Group, Sex, and Mechanism of Injury—United States, 2016 and 2017. [Google Scholar]
- 34.Sowell ER, Peterson BS, Thompson PM, Welcome SE, Henkenius AL, Toga AW. Mapping cortical change across the human life span. Nat Neurosci. 2003;6:309–15. doi: 10.1038/nn1008. [DOI] [PubMed] [Google Scholar]
- 35.Raz N, Rodrigue KM. Differential aging of the brain: Patterns, cognitive correlates and modifiers. Neurosci Biobehav Rev. 2006;30:730–48. doi: 10.1016/j.neubiorev.2006.07.001. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 36.Fotenos AF, Snyder AZ, Girton LE, Morris JC, Buckner RL. Normative estimates of cross sectional and longitudinal brain volume decline in aging and AD. Neurology. 2005;64:1032–9. doi: 10.1212/01.WNL.0000154530.72969.11. [DOI] [PubMed] [Google Scholar]
- 37.Hawryluk GW, Rubiano AM, Totten AM, O’Reilly C, Ullman JS, Bratton SL, et al. Guidelines for the management of severe traumatic brain injury:2020 update of the decompressive craniectomy recommendations. neurosurgery. 2020;87:427–34. doi: 10.1093/neuros/nyaa278. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 38.Cooper DJ, Rosenfeld JV, Murray L, Arabi YM, Davies AR, Ponsford J, et al. Patient outcomes at twelve months after early decompressive craniectomy for diffuse traumatic brain injury in the randomized DECRA clinical trial. J Neurotrauma. 2020;37:810–6. doi: 10.1089/neu.2019.6869. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 39.Hutchinson PJ, Kolias AG, Timofeev IS, Corteen EA, Czosnyka M, Timothy J, et al. Trial of Decompressive craniectomy for traumatic intracranial hypertension. N Engl J Med. 2016;375:1119–30. doi: 10.1056/NEJMoa1605215. [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The datasets generated and/or analyzed during the current study are available from the corresponding author on reasonable request.