Traumatic brain injury (TBI) is a pressing global public health and socioeconomic concern, affecting approximately 70 million individuals annually (1, 2). While the majority of TBI cases are classified as mild, about 20% involve severe injuries (1, 3, 4). Severe TBI (sTBI) is distinguished by extensive intracranial damage, significant disruption of neurological function, and a high mortality rate, ranging between 7 and 39% (3, 5, 6).

The management of sTBI is a formidable challenge in neurocritical care, necessitating prompt stabilization of hemodynamic and ventilatory parameters, continuous intracranial pressure (ICP) monitoring, and a systematic, tier-based therapeutic approach to mitigate intracranial hypertension and prevent secondary brain injury (7). Elevated ICP is a central therapeutic focus in sTBI management, as unchecked pressure increases within the rigid cranial vault can lead to the reduction of cerebral perfusion pressure, where sustained malignant intracranial hypertension can result in cerebral ischemia, brain herniation, and potentially death (8).

Despite ongoing controversy, decompressive craniectomy (DC) remains a pivotal surgical intervention in sTBI, demonstrating clear benefits in reducing intracranial pressure, preventing brain herniation, and improving survival in selected patients (9, 10). By removing a portion of the skull, DC creates space for the swollen brain to expand, alleviating the pressure within the rigid cranial vault and reducing the risk of secondary complications (11, 12). Although standard recommendations exist regarding the extent of decompression, the size of the initial craniectomy is often individualized in clinical practice, based on patient-specific anatomy, intraoperative findings, and surgeon judgment. However, in some cases, initial DC may fail to achieve sufficient ICP control, leading to ongoing intracranial hypertension, brain herniation, and potentially irreversible neurological damage (13). These scenarios pose major therapeutic challenges, as conventional medical measures—including osmotic agents (e.g., mannitol), sedation and analgesia, controlled hyperventilation, and mild hypothermia—are often insufficient to reverse the effects of sustained ICP elevation. If unrelieved, elevated ICP can compromise cerebral perfusion, result in ischemia, and markedly increase the risk of fatal brain herniation (14).

To bridge this therapeutic gap, secondary decompressive craniectomy (secondary DC) has been employed as a salvage intervention in patients with refractory intracranial hypertension after initial surgery, which creates additional space within the rigid cranial vault to accommodate brain swelling and prevent secondary complications. By enlarging the craniectomy area, this approach enhances the effectiveness of ICP reduction, ultimately improving cerebral perfusion and reducing the likelihood of adverse outcomes. Despite its theoretical advantages, the efficacy and safety of secondary DC remains uncertain, and the procedure is not without significant risks. The expanded surgical scope introduces greater technical challenges, which may increase the likelihood of complications such as infection, cerebrospinal fluid leaks, or long-term neurological sequelae. Here, this study aims to explore the clinical profiles and outcomes of patients undergoing secondary DC, providing valuable insights into its effectiveness and safety in the context of refractory ICP.

This retrospective cohort study provides a comprehensive understanding of the clinical characteristics, outcomes, and risks associated with secondary decompressive craniectomy in severe TBI. Our findings highlight significant differences in baseline characteristics between patients who underwent secondary DC and those who received only primary DC. Specifically, patients in the secondary DC group presented with more severe initial conditions, including significantly lower GCS scores at admission, higher preoperative ICP, and more advanced Marshall CT stages. Logistic regression analysis further identified higher preoperative ICP, and smaller decompression areas during primary DC were significant predictors associated with the need for secondary DC. Notably, no significant differences were observed between the two groups in terms of complications or prognosis.

This study underscores the critical relationship between higher preoperative intracranial pressure and the need for secondary DC, as well as the influence of smaller decompression areas during the primary DC procedure. ICP is primarily regulated by the volume of three components: brain tissue, cerebrospinal fluid. Following traumatic brain injury, the intracranial volume may increase due to hemorrhage, cerebral edema, or hydrocephalus, leading to a dangerous rise in ICP. The rigid structure of the skull exacerbates this problem, as the brain has limited space to accommodate swelling (18). Accordingly, if ICP remains high despite initial medical management, decompressive craniectomy is typically performed to alleviate the pressure. DC has long been used as a surgical intervention to manage refractory ICP by removing a portion of the skull, thereby providing additional space for the swollen brain and reducing pressure. The physiological goal of DC is to reduce ICP to a level that allows for improved cerebral blood flow and prevents brain herniation. DC can be particularly effective in cases of elevated ICP not responding to medical management, but the timing and extent of decompression remain key factors in determining its success (18, 19).

Guidelines from major TBI management consortia, such as the Brain Trauma Foundation, recommend early and adequate decompression to prevent irreversible brain injury (20, 21). However, there is still no universal agreement on the optimal timing for primary DC. Most guidelines suggest that DC should be performed when ICP exceeds 20–25 mmHg for extended periods, despite maximal medical management. By removing a substantial portion of the skull, this procedure effectively reduces ICP and mitigates the risk of secondary brain injury. However, despite its benefits, excessive decompression may occur in some patients, leading to a range of postoperative complications. These complications can include syndrome of the trephined, paradoxical herniation, cerebrospinal fluid disturbances, and delayed brain shifts, all of which may negatively impact neurological recovery. Therefore, the optimal extent and location of decompression plays a critical role in surgical outcomes. Typical decompression areas include the frontal, temporal, and parietal regions, with the extent of the craniectomy depending on the severity of the injury and the patient’s clinical status. That is to say, there is no universally defined “ideal” decompression area, which can lead to situations where ICP remains elevated even after the initial DC (22, 23). This insufficiency in primary decompression may subsequently result in the need for secondary decompressive craniectomy to further alleviate intracranial pressure. This highlights the importance of ensuring an appropriately sized decompression area during the primary procedure to reduce the likelihood of requiring further interventions. Future research should focus on defining the optimal decompression area based on injury severity and individual patient factors to minimize the need for secondary surgical interventions.

However, it is noteworthy that after rigorous adjustment, we found no significant differences in prognosis or complications between patients who underwent only primary decompressive craniectomy and those who required secondary DC. This may suggest that secondary DC does not result in an increased risk of complications or secondary injury compared to primary DC. Therefore, when primary decompression fails to adequately control intracranial pressure, performing secondary DC may not exacerbate risks and could be a reasonable intervention to prevent further neurological deterioration (24). However, repeated surgeries can increase both the economic and physiological burden on patients. The need for additional procedures not only prolongs the hospital stay and escalates medical costs but also exposes patients to extended recovery times (25). Therefore, while secondary DC may be necessary in some cases to prevent further neurological deterioration, it emphasizes the importance of ensuring that the initial decompressive surgery is adequately performed. Optimizing surgical decision-making to reduce the need for subsequent interventions could help mitigate the physical and financial burdens on patients and the healthcare system. Further research should focus on refining guidelines for primary DC (26).

Beyond its acute role in ICP control, our findings also relate to the ongoing debate on optimal craniectomy size and its long-term implications. Our observation that smaller initial craniectomy areas were associated with a higher likelihood of requiring secondary decompression underscores the importance of ensuring adequate decompressive size. This finding dovetails with a broader debate in neurotrauma: how large should a decompressive craniectomy be? Too small a bone flap may fail to adequately control ICP, necessitating revision as seen in our study, whereas excessively large defects may carry their own risks. For example, large craniectomy defects have been linked to the syndrome of the trephined, a delayed complication characterized by neurological deterioration due to loss of cranial integrity, which typically improves after cranioplasty (27). Furthermore, craniectomy size can influence reconstructive outcomes; large or bilateral defects may increase the complexity and complication rates of cranioplasty. A recent multicenter study identified defect size as a factor in predicting cranioplasty complications (28). Taken together, these considerations highlight the need to balance acute ICP control with long-term risks when determining the extent of decompression.

According to the BTF guidelines, DC is recommended as a second-tier therapy for refractory intracranial hypertension in severe TBI, with the primary goal of improving survival, although its impact on long-term functional outcome remains complex. Our findings support these recommendations in several respects. Secondary DC effectively controlled intracranial pressure without significantly worsening functional outcomes, consistent with the notion that timely decompression can be life-saving without invariably resulting in poor neurological recovery, as also reflected in the RESCUEicp trial. In addition, our observation that inadequate initial decompression often necessitated secondary DC underscores the guideline emphasis on performing a sufficiently large bone flap removal (e.g., a wide fronto-temporo-parietal craniectomy of at least 12 cm diameter reaching the skull base and midline when appropriate).

Building on these guideline principles, our study also provides actionable recommendations for neurosurgeons. First, ensuring adequate initial decompression is critical; maximizing the extent of the bone flap and performing dural expansion when indicated may help reduce the need for subsequent surgery. Second, patients presenting with markedly elevated ICP or diffuse/bilateral injury patterns should be recognized as high risk and considered for a more aggressive primary approach, including larger or bilateral DC where appropriate. Third, developing institutional protocols that standardize minimum craniectomy size and indications for bilateral decompression may reduce variability among surgeons and improve outcomes (29). Finally, even after adequate DC, vigilant postoperative ICP monitoring remains essential, with timely consideration of secondary DC in cases of refractory hypertension before irreversible injury occurs.

Taken together, these findings align with BTF guideline recommendations while also adding nuance: in our cohort, secondary DC did not significantly increase the risk of severe disability compared with primary DC alone, in contrast to earlier concerns raised by DECRA and other trials. This suggests that, with careful patient selection and appropriate surgical technique, secondary DC can be considered a viable salvage strategy consistent with current guidelines, while emphasizing that the best opportunity to prevent secondary intervention lies in performing an extensive and technically adequate primary decompression.

Several limitations of this study must be acknowledged. Firstly, as a retrospective cohort study, it is inherently limited by biases associated with observational data, such as selection bias and information bias. The decision to perform secondary DC was not randomized but based on clinical judgment, which introduces the potential for unmeasured confounders that may affect both the surgical decision and the outcomes. Therefore, prospective studies are needed to mitigate these limitations and provide stronger evidence for the effectiveness of secondary DC. Secondly, the retrospective single-center design and relatively small sample size, particularly the secondary DC group of 16 patients, limit the statistical power of our analysis and raise the possibility of Type II errors, that is, failing to detect true differences or associations due to insufficient sample size. For example, although no significant differences in functional outcomes were observed between groups after IPTW adjustment, small to moderate effects may have gone undetected. The wide confidence intervals for some outcome measures further reflect this uncertainty. Accordingly, we have interpreted the negative findings with caution and recommend that future studies with larger multicenter cohorts be conducted to validate our results. Additionally, this study was conducted at a single center, which may limit the generalizability of the findings to other institutions with different patient populations, surgical practices, and postoperative care protocols. Multi-center studies would be invaluable in improving the external validity of the results and offering a broader perspective on the outcomes associated with secondary DC. Furthermore, due to the retrospective nature of this study, surgeon-specific data on craniectomy size were not consistently documented, preventing a formal stratified analysis of inter-surgeon variability. Although our review did not indicate that smaller craniectomy sizes or secondary decompressions were concentrated within specific surgical teams (Supplementary Table 2), we acknowledge this as a limitation. Future prospective studies with standardized documentation of surgical decision-making would better clarify the influence of individual surgical practice on outcomes. While the study assessed functional outcomes at 6 months post-surgery, a longer follow-up period would be beneficial to fully evaluate the long-term cognitive, psychological, and quality-of-life outcomes in patients who undergo secondary DC. These additional insights would be crucial for understanding the lasting effects of this intervention and improving long-term patient care. Lastly, although baseline characteristics were generally comparable between groups after weighting, some imbalances remained. For example, except for CT, all baseline variables showed p values > 0.05, suggesting no statistically significant differences. However, the SMD for CT exceeded the conventional threshold, indicating residual imbalance. This may be explained by the clinical context and the nature of our dataset: CT is closely linked to treatment selection and disease status in practice, which makes it inherently more difficult to achieve balance through weighting. Moreover, given the limited sample size and heterogeneity of the study population, subtle imbalances are expected and may persist despite statistical adjustment.

Conclusively, this study provides a comprehensive analysis of the clinical characteristics, outcomes, and risk factors associated with extended decompressive craniectomy in patients with severe traumatic brain injury. The findings indicate that higher ICP and smaller decompression areas during the initial surgery may contribute to the need for secondary DC. These factors underscore the critical importance of achieving adequate decompression in the early stages of treatment to prevent the need for further surgical interventions. Additionally, the study found no significant differences in complications or prognosis between the secondary DC and primary DC groups, suggesting that the potential of secondary DC as a life-saving procedure for managing refractory ICP, without an additional risk of postoperative complications compared to primary DC.