Statistics from Altmetric.com
Intracranial pressure (ICP) monitoring is now a widely accepted tool in the management of patients with head injuries.1,2 However, intracranial haemorrhage is a recognised as a possible complication following placement of an ICP device.3,4 The purpose of this study was to investigate the incidence of haemorrhage after ICP monitor insertion through a thorough review of post-insertion computed tomography scans, and to classify them in a clinically relevant manner.
Materials and methods
Over 5 months, the Neurosurgery Service at Harborview Medical Center treated 314 patients with traumatic head injury. There were 247 males and 67 females with a median (SD) age of 35.16 (22.9) years (range 0.4 to 102 years), and all were admitted to the hospital. Placement of an ICP monitor (Camino™, intraparenchymal) was undertaken in 130 of these patients. We retrospectively analysed the patient’s hospital charts and all available radiological studies, with particular attention paid to our own interpretation of CT scans before and after ICP monitor insertion. The final numbers in the study were 101 males and 29 females with a median (SD) age of 36.6 (21.9) years (range 1.8 to 102 years).
Indications for ICP monitoring followed the head injuries (ICP monitoring) guidelines:5 (a) patients with severe head injury, GCS⩽8 with an abnormal head CT; (b) patients with severe head injury, GCS⩽8 with a normal head CT, and having two or more of the following: age >40 years, systolic blood pressure <90, or posturing; (c) patients with GCS 9−12 and abnormal head CT, if undergoing therapies for other injuries with possible deleterious effects on ICP; and (d) subsequent to removal of intracranial mass.
The fibre optic device was placed at the bedside (intensive care unit, emergency room) or at completion of the surgery in the operating theatre. Some patients needed replacement of an ICP device because of technical problems with the device. The right side was preferred for the insertion of the ICP monitor.
The institutional protocol was to perform CT scanning during the first 24 hours after the insertion of ICP device. We were not able to obtain CT in this time frame in four cases. A grading system for haemorrhages after ICP monitor insertion from our institution was used (fig 1). Grade 0 was used to report patients with no complications on post-placement studies. Grade 1 is a small punctuate haemorrhage or localised subarachnoid haemorrhage (SAH). Grade 2 haemorrhage is an intracranial bleed, diffuse SAH or extra-axial haematoma without a new neurological deficit and does not require operative intervention. In a case of grade 3 complication, revision craniotomy is required or there is a new neurological deficit, even a death.
Of the 314 patients with traumatic head injury, ICP monitor insertion was performed in 130 (41%). Nineteen patients had more than one ICP monitor inserted; altogether, 155 procedures were carried out. Right sided procedures prevailed (n = 102, 66%). The majority of the patients in this study were admitted with the diagnosis of a closed head injury (n = 116, 89.2%), 10 patients had open head injury, and four suffered a gunshot wound to the head. One hundred and six procedures (68%) were performed at the bedside, and 49 insertions (32%) took place in the operating theatre.
We retrospectively analysed the patient’s hospital charts and all available radiological studies. There were 140 procedures performed without any haemorrhagic complications on follow up radiological studies (grade 0). After 10 insertions (6.5%), a small punctuate haemorrhage or localised subarachnoid haemorrhage occurred. These complications were classified as grade 1 haemorrhages. Five patients (3.2%) sustained an intracerebral haematoma that did not necessitate evacuation or manifest as a new neurological deficit (grade 2). There were no haemorrhagic complications requiring evacuation or resulting in a noticeable change in the patient’s clinical condition (grade 3). Altogether, the complication rate was 9.7% for this study. More haemorrhagic complications occurred after ICP monitor placement in the operating theatre (n = 8/49, 16.3%), compare the bedside procedures (n = 7/106, 6.6%). This distribution did not reach statistical significance (p = 0.057).
There is a wide range (0−15.3%) in the literature of reported incidences of intracranial haemorrhages following placement of an ICP monitoring device.3,4 However, most studies had multiple targets such as outcome, different treatment options or indication criteria and these published reports failed to distinguish between large haematomas requiring surgical evacuation and small punctuate haemorrhages picked up incidentally only on imaging. In head trauma, there are multiple lesions on radiological examinations, and without detailed knowledge of the patient’s surgical procedures, a punctuate haemorrhage can be counted as an evolving contusion or go unnoticed. Due to previous metal artefacts from tip of the ICP monitor catheter, some small lesions were detected only after its removal.
In our traumatic brain injury group, we found a complication rate of 9.7% with no grade 3 haemorrhage. Although the most common grade 1 haemorrhage seems to be unimportant, we do not know its long term consequences and it may cause a false reading of a high ICP with subsequent unnecessary therapeutic interventions. The incidence of grade 3 haemorrhage was 0.15% (1 in 684 procedures) in our institution for the general neurosurgical population (trauma, tumors, cerebrovascular),4 and there is a similar complication rate and stratification for the paediatric subpopulation.4
Although intracranial pressure monitoring plays an indispensable role in the management of head injuries, the indications for this invasive neurosurgical procedure should always be carefully considered. Even with the utmost precautions, haemoragic complications may occur. Classification of the complications in the clinically relevant scheme may help to compare results of future studies.
Competing interests: none declared