Transcranial ultrasonography to detect early intracranial complications post-decompressive craniectomy for stroke: A feasibility study

INTRODUCTION

Decompressive craniectomy is a critical surgical intervention aimed at reducing elevated intracranial pressure (ICP) by removing a portion of the skull to allow for brain expansion. This procedure is often performed in response to severe traumatic brain injury, stroke, or other conditions that lead to increased ICP and a risk for herniation.^[1,2] The post-operative management of patients who have undergone decompressive craniectomy is crucial, given the high risk of complications such as intracranial bleeding, infections, cerebral edema, and hydrocephalus.^[3] In patients with stroke, the risk of hemorrhagic transformation and increase in bleed size following initiation of antiplatelet and anticoagulation therapy needs to be monitored closely.

Conventionally, computed tomography (CT) has been the imaging modality of choice for monitoring these patients. CT offers high-resolution images that can detect various complications such as bleeding and swelling. However, the need for patient transfer to the radiology suite introduces logistical challenges and risks, particularly for critically ill patients who may be unstable or at risk for further deterioration during transport.^[4] Furthermore, repeated CT scans contribute to cumulative radiation exposure and cost, which may be particularly concerning for patients requiring frequent imaging.^[5]

In contrast, transcranial ultrasonography (TCS) presents a promising alternative. TCS is a non-invasive imaging technique that can be performed at the bedside, eliminating the need for patient transfer and avoiding radiation exposure.^[6] It is also relatively cost-effective and can provide real-time imaging, which is advantageous for rapid decision-making in the intensive care setting.^[7] Despite these potential benefits, the comparative efficacy of TCS versus CT in detecting post-operative complications following decompressive craniectomy has not been extensively studied.

This study aimed to evaluate the feasibility and diagnostic accuracy of TCS in comparison to CT for monitoring post-operative stroke patients following decompressive craniectomy. Specifically, we assessed the sensitivity, specificity, and operational efficiency of TCS compared to CT.

MATERIALS AND METHODS

This prospective observational study was conducted at a tertiary care center with a specialized neurocritical care unit. Adult patients (18–75 years) undergoing decompressive craniectomy for ischemic or hemorrhagic stroke were included if they had a suitable transcranial acoustic window. Patients with severe hemodynamic instability, redo surgeries, or cranial fractures that could interfere with TCS application were excluded from the study. We used the STROBE checklist when writing our report [Supplementary file 1].

A convenience sample of 10 patients was studied over 3 months. Baseline assessments included demographic and clinical data such as age, gender, stroke subtype, and pre-operative imaging findings. TCS was performed at predefined intervals using a Sonosite M-Turbo ultrasound system with a phased-array probe (1–5 MHz frequency) by an experienced intensivist with over 5 years of expertise in transcranial ultrasound. The procedure was conducted through the decompressive craniectomy site, allowing direct insonation of brain structures. The probe was positioned perpendicular to the exposed brain surface. Key anatomical structures, including the midline, ventricular system, and any post-surgical collections, were evaluated in multiple planes. Special attention was given to detecting midline shift, intracranial hemorrhage, hydrocephalus, and any collection to facilitate early identification of complications. The same operator performed the TCS at specific intervals: 6, 12, 24, 36, 48, 60, and 72 h post-surgery. The images stored were later interpreted by two independent intensivists who had more than 10 years of experience. An image was deemed interpretable if it clearly demonstrated at least three of the following structures: (1) Midline shift or falx cerebri, (2) ventricular system, (3) contralateral skull or midline echo, (4) post-operative collection or hematoma, and (5) basal cisterns. These criteria were applied consistently, and images not meeting them despite optimal technique were labeled non-interpretable. The final assessment was performed independently by two experienced intensivists using these predefined criteria.

The primary measured parameters included midline shift, abnormal findings such as any collection, ventricle size, and herniation noted. Clinical deterioration, such as altered consciousness, worsening motor function, or new focal deficits, was assessed alongside TCS findings. A routine post-operative CT was done at 24-h post-surgery. Additional imaging with CT or magnetic resonance imaging was performed as needed based on clinical indications. The primary outcome measured was feasibility, defined as the ability to acquire interpretable TCS images at each designated time point. Secondary outcomes included correlations between TCS findings and radiological imaging results.

RESULTS

The study included 10 patients with an average age of 67.5 ± 5.2 years. The majority had ischemic strokes (80%), while 20% had intraventricular hemorrhage (IVH) due to hypertension. Patient demographics are given in Table 1. The feasibility of acquiring interpretable TCS images was highest at 6 h post-surgery with a 90% success rate. This decreased to 70% at 12 h followed by a rate improved to 80% at 24 h but dropped again to 70% at 36 h, reflecting clinical variability. Full feasibility (100%) was achieved at 72 h, and Figure 1 shows the variation in TCS acquisition success over time. A root cause analysis identified key factors affecting the feasibility of acquiring interpretable TCS images. Patient-related issues included agitation and restlessness, while operator-dependent factors involved reliance on a single trained sonographer and shift-related constraints. Resource limitations, such as restricted ultrasound machine availability and high workload, also contributed.

Close

Figure 1:

Transcranial ultrasonography (TUS) acquisition success rate over various time points.

Export to PPT

The total number of CT scans was a median of 3 (2.25–3.75) performed over 9.6 ± 3.2 days in the intensive care unit (ICU). TCS demonstrated a sensitivity of 85.7% and a specificity of 100%, with an area under the curve of 0.93 on ROC analysis, indicating high diagnostic accuracy for identifying postoperative complications. The complications studied were midline shift, hemorrhage, hydrocephalus, or any collection. The kappa coefficient of 0.783 reflected substantial agreement with CT findings [Figure 2].

Close

Figure 2:

Receiver operating characteristic (ROC) curve of transcranial ultrasonography to diagnose intracranial complications. AUC: Area under the curve.

Export to PPT

Complications were observed in 40% of patients, including hemorrhagic transformation in two cases, midline shift due to subdural hemorrhage in one case, and hydrocephalus in another [Figure 3]. Notably, TCS detected the midline shift at 36 h. Hemorrhagic transformations were identified at 48 and 60 h, while hydrocephalus was detected at 72 h. Both patients with IVH showed a reduction after surgery. One patient required surgical exploration, and antiplatelet therapy was discontinued in two patients. TCS was significantly faster, with a median time of 3 (2–5) min compared to 20 (17–28) min for CT, including patient transfer time (P < 0.001).

Close

Figure 3:

(a) Transcranial ultrasound showing hemorrhage (red line tracing), (b) Computed tomography showing hemorrhage (red line tracing), (c) Transcranial ultrasound showing midline shift (red arrow), (d) Intraoperative image showing subdural collection, (e) Computed tomography showing midline shift (red arrow), (f) Computed tomography showing hemorrhage in ventricle (red line tracing), (g) Transcranial ultrasound showing hemorrhage in ventricle (red line tracing), (h) Hemorrhage in ventricle (red line tracing, white arrow), (i) Computed tomography showing hemorrhage in ventricle (red line tracing, white arrow), (j) Transcranial ultrasound showing dilated ventricle, (k) Computed tomography showing dilated ventricle, (l) Computed tomography showing hemorrhage in ventricle (red line tracing, white arrow), (m) Transcranial ultrasound showing hemorrhage in ventricle (white arrow: ventricular hemorrhage, red line tracing: ventricle).

Export to PPT

DISCUSSION

This study demonstrates that TCS is a feasible and effective alternative to CT for post-operative monitoring of patients who have undergone decompressive craniectomy.

The management of stroke patients after decompressive craniectomy is challenging due to the high incidence of post-operative complications, including hemorrhagic transformation, cerebral edema, and hydrocephalus. Early detection of these complications is crucial for improving patient outcomes, especially when considering antiplatelet or anticoagulant therapy initiation. In the present study, 40% of patients experienced post-operative intracranial complications detected by either TCS or CT. Similar studies have reported that up to 43% of patients may experience intracranial hematomas or midline shifts post-craniectomy, necessitating close monitoring.^[8] TCS’s high sensitivity makes it a valuable tool for early detection of these complications, supporting timely interventions.

The decision to initiate antiplatelet or anticoagulant therapy in stroke patients after decompressive craniectomy remains complex due to the risk of intracranial hemorrhage. In the current study, patients were closely monitored using TCS every 12 h for the first 72 h postoperatively. This approach allowed early detection of any evolving intracranial pathology, potentially guiding the safe initiation of antithrombotic therapy. The current guidelines recommend delaying therapy for at least 24–48 h post-craniectomy until hemostasis is confirmed.^[9,10] TCS can facilitate individualized therapy decisions by providing continuous intracranial monitoring without radiation exposure.

Hemorrhagic transformation is a major concern when starting antiplatelet or anticoagulant therapy. In the present study, hemorrhagic transformations were identified at 48 and 60 h in two patients, respectively, which needed stoppage of antithrombotic therapy after antiplatelet initiation, emphasizing the importance of individualized monitoring. This demonstrated the possible utility beyond the usual period of complication. Reported bleeding rates in the literature range from 6% to 25%, depending on clinical and surgical factors.^[11] Using TCS, ICU teams can monitor for early signs of bleeding, allowing prompt adjustments in therapy.

The current study demonstrated that TCS provided accurate, real-time monitoring, detecting intracranial pathologies within a median time of 3 min compared to 20 min for CT. This substantial reduction in imaging time highlights TCS’s practicality in neurocritical care. TCS detected all midline shifts, hematomas, and signs of hydrocephalus, confirming its high diagnostic accuracy, which has been similarly reported in prior research.^[12] Its use allows more frequent evaluations, enabling dynamic management in unstable patients.

The acquisition of TCS was successfully completed in more than 50% of the patients at all time points with the lowest success rates at 12, 36, and 60 h. The root cause analysis revealed patient-related, operator-dependent, and system-related factors. Addressing these factors through expanded training, optimized duty schedules, and dedicated machine access could improve TCS imaging consistency.

Transporting critically ill patients for CT scans poses significant risks, including hemodynamic instability, hypoxemia, and increased ICP.^[13] The present study demonstrated the possibility of the use of TCS to minimize the need for patient transport, reducing associated risks and resource utilization. Its portability, low operational cost, and elimination of transport expenses make TCS a financially sustainable imaging solution in resource-constrained settings.^[14]

Several limitations must be acknowledged in this study. The small sample size focused only on stroke and single-center design may limit the generalizability of the findings. In addition, the study focused primarily on the technical feasibility and accuracy of TCS without evaluating long-term clinical outcomes or cost-effectiveness in a broader healthcare context. A skilled operator with training in TCS is required to interpret the findings. Future research with a larger sample size, multicenter trials, and longitudinal follow-up is needed to validate these findings and assess the broader impact of TCS on patient outcomes and healthcare costs.

Further research should explore the integration of TCS into standard post-operative care protocols and assess its impact on clinical outcomes and healthcare resource utilization. Comparative studies examining the cost-effectiveness of TCS versus CT in different clinical settings could provide valuable insights for optimizing post-operative management strategies.

CONCLUSION

TCS offers a rapid, cost-effective, and non-invasive alternative to CT for monitoring post-operative complications following decompressive craniectomy in stroke. The high sensitivity and specificity of TCS, combined with its operational advantages, make it a valuable tool in the intensive care setting. Incorporating TCS into routine postoperative care could enhance patient safety, improve diagnostic efficiency, and reduce the need for radiation exposure.