Role of the “In-Window” Craniectomy in the Management of Serious Head Trauma: 12 Case Series, Technical Note, Literature Review, and Comparison With Other Surgical Managements Authors: Hugues Bulteau, Philippe Poidevin, Marc Baroncini Published: 2026 ABSTRACT BACKGROUND AND OBJECTIVES:Decompressive bone flap remains the standard surgical for refractory intracranial hypertension in connection with serious head trauma. However, this surgical procedure is not free from risk in itself (infectious, hydraulic troubles) and the cranioplasty procedure, which will be performed remotely. In certain cases, an alternative “hinge” or “semifloating” bone flap technique can be discussed.METHODS:We aimed to present 12 cases of “semifloating” bone flaps made between January 2022 and September 2025 in our institution, describing our surgical technique (and alternatives published in the literature), the outcomes, and the possible complications. We also provide key points identified after a literature review of this procedure.RESULTS:In our short series, 2 of the 12 semifloating craniectomies were insufficient to control intracranial pressure and required a decompressive bone flap at J1. None of our patients presented with a cerebrospinal fluid-related complication; none presented with an infectious complication or autologous bone resorption; moreover, all the craniectomies returned to their physiological position in the following weeks. No major technical difficulties were encountered, and the procedure was associated with a short operative time.CONCLUSION:Given the possibility of controlling intracranial hypertension in most cases, the lower number of complications caused, and the possibility of avoiding one (or even more) invasive surgery(ies) with a high risk, particularly infection, hinge craniectomy should be considered but only in carefully selected traumatic brain injury patients with focal mass lesions and should not be extrapolated to cases of diffuse malignant cerebral edema. ABBREVIATIONS: acute subdural hematoma decompressive craniectomy epidural hematoma Glasgow Coma Scale hinge craniectomy intracranial pressure traumatic brain injury. Decompressive craniectomy (DC) is a well-known neurosurgical procedure, particularly in traumatic and vascular contexts. However, it has significant consequences, whether related to skin, infection, neurological, and even more so, functional and public health implications. This led us to question the role of “(semi)floating” or “hinged” or “in-window” bone flaps and their indications. In their scoping review, Layard Horsfall et al1 define the “in-window” craniectomy as a partial decompression to relieve raised intracranial pressure (ICP) with subsequent immediate replacement of the bone flap, over the surgical cranial window, in a “loose” or “hinged” fashion in an attempt to accommodate cerebral edema and subsequent swelling, without further neurosurgical intervention.1 We report all cases (n = 12) of traumatic cranial surgery for which we performed a hinge craniectomy (HC) at our center between January 2022 and September 2025, as well as to conduct a literature review to identify the indications, benefits, and risks and to outline the various possible surgical techniques. We will try to mirror these 12 HC, with cases that in our opinion would have required a HC which unfortunately was not conducted (either a “too aggressive” DC was performed, or conversely, a standard craniotomy was performed but was not effective enough) and the complications resulting from it. The aim of this study was to attract the attention of the neurosurgical community on this intermediate technique between craniotomy and DC so that every neurosurgeon can keep it in mind when considering the most appropriate surgical strategy in the specific context of each patient. METHODS This case series has been reported in line with the Preferred Reporting Of Case Series in Surgery Guidelines.2This study was conducted in accordance with the principles of the Declaration of Helsinki. Ethical approval was waived by the Institutional Review Board due to the retrospective nature of the study. Written informed consent was obtained from all patients (or their legal guardians) for inclusion in the study and publication of anonymized data and images. The data supporting the findings of this study are available from the corresponding author upon reasonable request. Case Series Hinged Craniectomies Case 1 A 40-year-old man was found on the ground under unknown circumstances, with a Glasgow Coma Scale (GCS) score of 9/15. Owing to persistent consciousness disturbances, the patient was intubated and neurosedated. The initial computed tomography (CT) scan showed a right frontotemporal contusion. ICP monitoring was initiated on arrival at the intensive care unit. ICP became refractory to maximal neurosedation 2 days after trauma. A repeat CT scan showed worsening of the contusion with subfalcine and uncal herniation with exclusion of the left lateral ventricle (Figure 1A and 1B). This warranted surgery on day 2 posttraumatic brain injury (TBI), involving evacuation of an acute subdural hematoma (ASDH), temporal contusectomy to disengage the temporal uncus at the tentorial notch, enlargement plasty, and HC (to avoid further complicating the procedure by adding frontal contusectomy) (Figure 1C and 1D). Postoperative scans showed bone flap lifting in the days after surgery (Figure 1E-1G), thus demonstrating the intended hinge effect adjusted according to ICP. One month postoperatively, the flap returned to its original position (Figure 1H). At 3 months postoperatively, the patient regained satisfactory autonomy, with a modified Rankin Scale (mRS) score of 1. Describes the scannographic evolution of the case 1. A and B, Show the scan performed 2 days after the trauma on which a left frontotemporal contusion causes a deviation of the midline and an increase of the intracranial pressure. C and D, Are the early postoperative scan, 1 day after surgery. E-G, Show the expected hinge effect 4 days after surgery. H, At 1 month, the bone flap has returned to physiological position, without diastasis. Case 2 This case involved a 41-year-old woman, who fell down the stairs with an initial severe clinical presentation and a GCS of 3, necessitating immediate endotracheal intubation. A CT scan revealed predominant bifrontal contusions, and despite maximal medical management, uncontrolled ICP was observed on day 6, prompting surgical intervention. A Cairns-Hinterberg bifrontal scalp flap was performed, with evacuation of a small ASDH, enlargement plasty, and HC fixed at the anteroinferior aspect. Postoperative scans clearly showed elevation of the bone flap; then, follow-up imaging showed progressive return of the bone flap to its physiological position. Others Surgical Strategies The following cases are presented for illustrative purposes only and aim to highlight potential advantages and limitations of different surgical strategies. These retrospective considerations should not be interpreted as definitive comparative evidence. Decompressive Craniectomy A 25-year-old man was a victim of a motorcycle accident with an immediate coma and bilateral unreactive mydriasis, requiring intubation ventilation sedation immediately on site. The CT scan showed multiple intracranial hemorrhagic lesions, especially ASDH on the right convexity and a right basifrontal contusion (Figure 2A). During surgery, moderate cerebral edema pushed the surgeon to decide on a gesture of DC. Three weeks later, contralateral subdural hygroma required a subduroperitoneal shunt (Figure 2B). Autologous cranioplasty was performed 1 month after DC, which required ventricular puncture because of unsatisfactory brain relaxation even with anesthetic measures. Currently, he is progressing well in rehabilitation, and his sooner CT scan is satisfactory, whether it is the rested right hemicraniectomy bone flap, the position of the subdural drain, or the collection it drains, although we still noted a voluminous subcutaneous collection next to the right hemicranium (Figure 2C). Show the iterative complications mentioned in the decompressive craniectomy case: A, Initial computed tomography scan, with the right acute subdural hematoma; B, Voluminous contralateral convexity subdural hygroma; C, Latest scan, with the left subdural shunt. With an initial HC, it might have been avoiding the 2 supplementary surgeries (subduroperitoneal drain and autologous cranioplasty), the necessity of a ventricular puncture to rest the bone flap, the long-term carrying of intracranial and intraperitoneal foreign material, and the days spent with a transcalvarial hernia and major midline shift when the bone flap was not rested, and the contralateral convexity hygroma was present. Moreover, if an initial HC is performed and ICP is not controlled, it does not prevent the realization of a DC. Craniotomy A 49-year-old man was found in a coma, GCS 6/15, with an anisocoria right greater than left, after which a large tree branch fell on his head. Intubation, ventilation, and neurosedation were performed. The CT scan showed an ASDH of the right convexity and a right temporopolar epidural hematoma (EDH) (Figure 3A). The first surgery is performed to evacuate the ASDH and EDH. Postoperatively, the anisocoria did not regress, and above all, ICP monitoring showed peak values around 30 mm Hg despite maximum neurosedation including pentothal and salt, in connection with a large frontal contusion unmasked by the evacuation of the ASDH. A second surgery was performed 2 days later to re-extend the bone flap and perform DC. Subsequently, the ICP was controlled. After long-term medical support, autologous cranioplasty was performed 5 months later. The patient was transferred to rehabilitation where the evolution was very slowly favorable, brain imaging showed a posttraumatic hydrocephalus (Figure 3B), and a ventriculoperitoneal shunt was performed (Figure 3C). Currently, the patient is still in rehabilitation, where we observe minimal progress and a posttraumatic epilepsy. Illustrates the course of the case of the insufficient craniotomy: A, Initial computed tomography scan with the bulky right subdural hematoma; B and C, Posttraumatic hydrocephalus requiring a ventriculoperitoneal shunt. In this case, if during the first surgery, the bone flap had been placed on a hinge, many benefits could be expected: immediate ICP control, no need for DC on day 2, and no need for autologous cranioplasty; and because we know that the absence of a bone flap causes numerous hydraulic and comitial complications as mentioned by Omerhodzic et al,3 we could expect the absence of chronic posttraumatic hydrocephalus, and so the necessity of a ventriculoperitoneal shunt, and the absence of a crisis. Failures of HC: Lessons Learned Case 1 A 22-year-old woman was admitted after a severe TBI after being struck by a vehicle while running. On initial assessment, her GCS score was 5. She was intubated, mechanically ventilated, and sedated at the scene. Initial head CT revealed a left temporoparietal cerebral contusion associated with left parietal subarachnoid hemorrhage, a right temporal EDH, and a left tentorial ASDH extending to the cerebellar falx. ICP monitoring was initiated, with an initial ICP of 55 mm Hg, which decreased to 18 mm Hg after optimized management of secondary cerebral insults, including hyperosmolar therapy with hypertonic saline. A repeat CT scan demonstrated radiological worsening, with progression of the left temporopolar intracerebral hemorrhage and enlargement of the right temporal EDH (Figure 4A). An emergency HC was therefore performed, without evacuation of the left temporal contusion. At the end of the procedure, ICP had decreased to 1 mm Hg. However, a secondary rise in ICP occurred on the following evening, and a scan showed a lifting of the temporal portion of the flap but insufficient (Figure 4B-4D), prompting urgent conversion to a DC (Figure 4E). The patient was successfully extubated 15 days later. Neurological recovery was rapidly favorable. Autologous cranioplasty was performed 2 months later. At 3 months after the initial injury, the patient had already resumed her university studies. Show the scannographic evolution of the first case of failure of hinge craniectomy. A, Initial computed tomography scan with left temporal contusion. B-D, The scan after ICP reascent shows a lifting of the flap, but it is insufficient because it is still associated with a mass effect and high ICP. E, Postoperative scan of the necessary decompressive craniectomy. ICP, intracranial pressure. Case 2 A 35-year-old man was admitted after a severe TBI sustained in a motorcycle accident without helmet use, with a GCS score at 6. He was intubated, and sedated. Initial head CT demonstrated a frontal cerebral contusion associated with a left ASDH (Figure 5A and 5B). Emergency surgical management consisted of evacuation of the contusion and ASDH, associated with a HC. ICP monitoring was initiated at the end of the procedure, with an initial ICP of 13 mm Hg. A postoperative CT scan demonstrated insufficient outward displacement of the bone flap, with intracranial hypertension signs, indicating inadequate decompression (Figure 5C and 5D). The patient subsequently underwent urgent conversion to a DC (Figure 5E). Show the scannographic evolution of the second case of failure of hinge craniectomy. A and B, Initial computed tomography scan with the left frontal contusion and the left acute subdural hematoma. C and D, Insufficient hinge craniectomy, without effective lifting of the bone flap. E, Postoperative scan of the required decompressive craniectomy. Data Availability The data that support the findings of this study are available from the corresponding author upon reasonable request. RESULTS This was an ambispective monocentric descriptive series of 12 cases, spanning from January 2022 to September 2025 (see Table for patient characteristics). These 12 young patients, with no medical history or treatment (particularly no anticoagulant or antiplatelet therapy), were in the context of severe TBI, with altered consciousness. Preoperatively, they were all intubated, ventilated, and neurosedated, 4 had anisocoria, and one had bilateral mydriasis. Six of the 12 patients underwent ICP monitoring. TABLE.Summary of Clinical ResultsInitial contextSurgeryFollow-upAmbispective12 casesJanuary 2022–September 2025All in context of severe TBIPredominant radiological lesion Contusion: 9 ASDH: 3Surgery timing: H0–J7ICP monitoring: 6/12Bone flap 10 hemicraniectomies 2 bifrontal11 accompanied by an intradural evacuation procedure (ASDH/contusion)/1 without any other surgical procedureSystematic expansion of the dural-meningeal plastyPost-operative bone flap: 6/12 with significant elevation2/12 require subsequent DCAll bone flaps have returned to their physiological positionNo hydrocephalusNo infectionNo autologous resorptionSatisfactory mRS score: 0-1ASDH, acute subdural hematoma; DC, decompressive craniectomy; ICP, intracranial pressure; mRS, modified Rankin Scale; TBI, traumatic brain injury. Summary of Clinical Results ASDH, acute subdural hematoma; DC, decompressive craniectomy; ICP, intracranial pressure; mRS, modified Rankin Scale; TBI, traumatic brain injury. These 12 cases of HC were performed between H0 and Day 7 after TBI. The main hemorrhagic lesion leading to increased ICP was contusion in 9 of 12 cases and ASDH in 3 cases. Ten hemicraniectomy flaps and 2 bifrontal flaps were performed: 11 with an associated intradural evacuation procedure (contusion or ASDH), one without, and each with dural-meningeal expansion plasty. Postoperatively, 6 of the 12 flaps showed significant elevation. No complications occurred, during or after the procedure; however, 2 of 12 patients (of which, the only one who did not undergo evacuation of an intracranial hemorrhagic lesion) required a second standard DC because of uncontrolled and increasing ICP. During follow-up consultations, all flaps had returned to their physiological position (without bone diastasis), most of them quickly. To confirm that the flap had returned to its physiological position, we considered that both bone edges should be in apposition without significant diastasis, allowing for ossification. No hydrocephalus, infections, or autologous resorption was observed. Each patient achieved satisfactory autonomy, with a mRS score between 0 and 1. These follow-up observations had a minimum follow-up of 3 months postoperatively. DISCUSSION Technical Note At our center, we performed HC with only fixation with separate suture stitches of Ethibond® 2.0 (see Figure 6 for CT scans, and Figure 7 for surgical photographs); peripheral and central dural suspensions were also performed as in a traditional craniotomy to prevent postoperative EDH. But in the literature, other techniques are described, notably this one by Schmidt et al,4 Goettler and Tucci,5 Kenning et al,6 which consists in fixing 3 titanium plates on the bone flap: one anterosuperior which is screwed at the bone flap and at the rest of the cranium to allow the bone flap to be lifted while leaving it attached to the rest of the skull, and 2 others posteroinferior which are screwed only to the bone flap to prevent the invagination in the bony depression one the parenchymal flare has resolved. Conversely, in the technique described by Ko and Segan,7 the inferior plates act as the hinge, the superior edges are left unattached, and the future reconstitution of the cranial vault can be performed under local anesthesia in a minor procedure of ambulatory surgery. 3-dimensional scannographic illustrations of the technical note: A and B, Hemicraniectomy bone flap; C and D, Bicoronal bone flap. Per-operative illustrations of the technical note. Valença et al8 have also published a technique that involves dividing the bone flap in 2 using an angled craniotome to allow it to open gradually as the brain parenchyma herniates transcalvarially and to prevent subsequent invagination once the parenchymal herniation is resolved. In addition, the flap was fixed to the dura mater to prevent epidural bleeding and to the skin to prevent flap invagination. Valença et al8 estimate that only a 15% reduction is achieved with craniectomy alone and that duraplasty can further decrease ICP by an additional 55%. At our center, we routinely perform dural enlargement surgery, which is usually nonsealed. Literature Review A systematic literature search was conducted using the PubMed database. Keywords were combined, using Boolean operators “and” and “or” and included: “severe TBI,” “hinged craniotomy,” “floating craniotomy” and “in-window craniotomy.” In the series by Ko and Segan,7 which included 16 patients with increased ICP (because of EDH, spontaneous intracranial hemorrhages, middle cerebral artery infarcts, and ASDH), ICP was controlled for each patient, and no patient required further cranial decompression. In the series of Schmidt et al,4 comprising 25 patients, one had 1 subsequent cranioplasty (because of infection), 1 cranial deformity, 1 esthetic issue because of ongoing elevation of the flap, but no instances of bone flap resorption. We can also cite the work of Rocque et al,9 which discusses the following issues that are more common in children: infection, autologous resorption, growth-related luxations (especially in infants), as well as preservation problems (because the abdominal subcutaneous space is limited in younger children, and there are many issues with resorption during cryopreservation); according to the authors of this article, all these risks are reduced with the HC. In addition, Figaji et al10 add that a thinner bone flap in children may not cause a significant mass effect under elastic scalp tissue. In all cases, it is advisable to avoid the use of standard plates and screws in adults and instead opt for absorbable plates. Traxler et al11 evaluated the projected increase in cranial volume based on the virtual “hinged” elevation of the frontal bone in cranial CT scans, with the gain in cranial volume determined by computer-generated measurements, and the results showed a cranial capacity enlarged by 6% or 78 cm3 with a 10-mm elevation. Furthermore, Søndergaard et al12 presented a study on the correlation between increased ICP and increased intracranial volume: before reaching 20 mm Hg of ICP, the HC allows for tolerance of an increase in intracranial volume to 190 mL, which is sufficient in many patients. Possible consequences resulting from local pressure when the swollen brain pushes against the overlying bone, such as regional reduction in cerebral blood flow or focal cerebral metabolic abnormalities (such as altered glucose metabolism), may not be detected by a global monitoring technique, such as ICP measurement, but the sensitivity of this monitoring may be enhanced by adding PtiO2 monitoring and cerebral microdialysis. Kenning et al6 suggested that DC resulted in a greater degree of postoperative brain deformation than HC, and they found that there was no significant difference between HC and DC in operative time, need for reoperation, duration of mechanical ventilation, or intensive care unit stay. Moreover, Valença et al8 believe that the absence of the bone flap causes a major disturbance to the ICP regulatory mechanisms. Because of the lower ICP, cerebrospinal fluid absorption may be deficient because a pressure gradient between the cerebrospinal fluid and venous blood is required. They supposed that with the in-window craniectomy, a “normal” intracranial biomechanical condition is reestablished after 1 to 2 months; thus, they expect a lower incidence of hydrocephalus. In their series of 5 patients, only one developed shunt-dependent hydrocephalus. According to Deng et al,13 the incidence of shunt-dependent hydrocephalus after severe TBI was approximately 10%, but it increased to 21% when DC was performed. Layard Horsfall et al1 in their scoping review included 15 studies with a total of 283 patients: only 3.2% required subsequent formal DC; 6 studies compared HC with DC after TBI and stroke, finding at least equivalent control of ICP, and also reported reduced rates of complications, including infection, in HC compared with DC. Limitations The main limits of our study are the small number of cases, and the absence of precise ICP figure before and after the surgery. Postoperative ICP monitoring was not systematically available, which limits any definitive conclusion regarding ICP control and represents a major limitation of the study. The unusually high rate of favorable functional outcome in our series likely reflects a strong selection bias. HC was preferentially performed in young patients with focal lesions, preserved brainstem reflexes, and absence of diffuse malignant cerebral edema. Therefore, HC should not be considered a replacement for DC but rather a complementary option in carefully selected patients. CONCLUSION This consideration arises from the fact that in traumatic cranial surgery for the evacuation of a hemorrhagic lesion (eg, ASDH, contusion) causing increased ICP, performing a craniotomy and therefore replacing the bone flap may be insufficient, especially in the days after surgery because of edema. It may even require a second, more aggressive DC, with subsequent associated issues. Therefore, we investigated the role of HC, as an intermediate procedure between an insufficient evacuation surgery of an expanding hemorrhagic lesion with craniotomy and highly aggressive DC. Compared with DC, HC likely reduces, edema and brain expansion exaggerated by rapid cerebral decompression in the DC, compared with a HC based on the extracerebral herniation index6,8; the rate of hydrocephalus; the rate of hygromas. In addition, it avoids clinical signs and metabolic problems associated with the absence of a bone flap (sinking skin flap syndrome/trephine syndrome), which is known to have unfavorable effects on functional prognosis. In most cases, it avoids the need for subsequent cranioplasty. Therefore, it is an anatomic solution that progressively adapts to herniated brain tissue and changes in ICP. This helps reduce public health costs, particularly those related to cranioplasty surgery itself, its complications (notably cutaneous and infectious), and the associated hospitalization. In the current literature, HC do not seem to be inferior to DC for controlling ICP refractory to well-conducted medical treatment, although they often allow for the postponement of a second cranioplasty surgery. Although their use remains less common, they remain useful in selected indications of less severe refractory intracranial hypertension (for example, they may be insufficient in cases of malignant middle cerebral artery infarction) than in those requiring a standard DC. Therefore, it is a surgical alternative in cases of intermediate ICP elevation.