Article Text
PDF
Abstract
Background To assess if ultrasound measurement of the optic nerve sheath diameter (ONSD) can accurately predict the presence of raised intracranial pressure (ICP) and acute pathology in patients in the emergency department.
Methods This 3-month prospective observational study used ultrasound to measure the ONSD in adult patients who required CT from the emergency department. The mean ONSD from both eyes was measured using a 7.5 MHz ultrasound probe on closed eyelids. A mean ONSD value of >0.5 cm was taken as positive. Two radiologists independently assessed CT scans from patients in the study population for signs of raised ICP and signs of acute pathology (cerebrovascular accident, subarachnoid, subdural or extradural haemorrhage and tumour). Specificity, sensitivity and κ values, for interobserver variability between reporting radiologists, were generated for the study data.
Results In all, 26 patients were enrolled into the study. The ONSD measurement was 100% specific (95% CI 79% to 100%) and 86% sensitive (95% CI 42% to 99%) for raised ICP. For any acute intracranial abnormality the value of ONSD was 100% specific (95% CI 76% to 100%) and 60% sensitive (95% CI 27% to 86%). κ Values were 0.91 (95% CIs 0.73 to 1) for identification of raised ICP on CT and 0.84 (95% CIs 0.62 to 1) for any acute pathology on CT, between the radiologists.
Conclusions This study shows that ultrasound measurement of ONSD is sensitive and specific for raised ICP in the emergency department. Further observational studies are needed but this emerging technique could be used to focus treatment in unstable patients.
- Optic nerve ultrasound intracranial pressure
- emergency care systems
- emergency departments
- imaging
- ultrasound
- research
- trauma
- head
Statistics from Altmetric.com
- Optic nerve ultrasound intracranial pressure
- emergency care systems
- emergency departments
- imaging
- ultrasound
- research
- trauma
- head
Introduction
Traumatic head injury and acute neurological signs and symptoms are common presentations to emergency departments. It is important in all these patients to identify raised intracranial pressure (ICP) and focal pathology as they are associated with significant morbidity and mortality if not treated. Accurate measurement of ICP, within an emergency department with no access to neurosurgical input, is challenging. The current accepted standard for identifying raised ICP and focal intracranial pathology is an urgent CT scan organised from the emergency department.1 This CT scanning is relatively quick and non-invasive.1 2 The reported accuracy of CT varies depending on pathology, but all subsequent decision making about definitive patient management is based on initial CT findings.2 Any signs of raised ICP evident on CT may be inaccurate and are subjective. CT scanning is also associated with x-ray exposure and access to CT in National Health Service (NHS) hospitals is variable, particularly out of hours.3 Access in the developing world is less good.1
Several recent studies suggest that the optic nerve sheath diameter (ONSD) in adults increases almost instantaneously to greater than 5 mm if significantly raised ICP is present.4 5 A study measuring ONSD in a population of 30 patients in whom emergency doctors had high clinical suspicion of raised ICP found a sensitivity of 100% and a specificity of 95% for raised ICP compared to CT results.4 Another study on 59 patients with head injuries reported a sensitivity of 100% and specificity of 63% for raised ICP compared to CT.4 This study also showed a sensitivity of 84% for any traumatic intracranial pathology identifiable on CT.5 Both of these studies showed measuring ONSD was feasible in an emergency department, and did not interrupt patient treatment. The measurement of ONSD is easy to teach, easily reproducible, can be performed in seconds and is very well tolerated by patients.6 7 It also has a high interobserver reliability.5 6
Cranial CT scanning was used in this study as the best standard available and the only method of identifying raised ICP acutely in the emergency department.
Aims
The aim of this study was to assess if ONSD measurement by ultrasound can accurately predict the presence, or absence, of raised ICP and acute pathology (cerebrovascular accident, subarachnoid, subdural or extradural haemorrhage, or tumours). This was in a patient population containing patients who were traumatic and patients who were non-traumatic, referred for urgent CT from the emergency department.
Methods
This study was carried out in the emergency department of the Norfolk and Norwich University Hospital, Norwich, UK, which is a large urban teaching hospital whose Emergency Department sees approximately 80 000 patients per annum. Patients were enrolled over a 3-month period from November 2007 to February 2008. A group of emergency department middle grades conducted the study measurements after a period of training in ONSD ultrasound. This training consisted of each middle grade successfully performing five ONSD measurements prior to the start of the study. This number of ONSD scans has been shown to be reproducible with a high interobserver reliability in previous studies.6 7
During the study period, any adult patient presenting to the emergency department who, after initial assessment and emergency treatment, was referred for a CT scan was considered eligible for enrolment. Patients under 18, or those with significant ocular trauma, were excluded. If a trained emergency department middle grade was present on shift and the patient met the eligibility criteria they were enrolled in the study and underwent ONSD measurement. Consent was obtained where possible, and in patients who were comatose consent was gained from the next of kin. The ONSD measurement was taken by an emergency department middle grade who was not the patient's primary clinician in an attempt to avoid selection and measurement biases. The measurements were taken with a 7.5 MHz ultrasound probe, applied directly on an intravenous cannula dressing that had been applied to each of the participants closed eyes. Each subject had the ONSD from both eyes measured and an average taken. The study participant was then referred for urgent CT as normal.
A positive ONSD measurement was taken as an average ONSD from both eyes greater than 5 mm. At a later date the CT scan images from study participants were reviewed by two radiologists independently, one radiologist specialising in neurological imaging. Criteria for the diagnosis of raised ICP on CT scanning were one or more of: mass effect with midline shift 3 mm or more, collapsed third ventricle, hydrocephalus, effacement of sulci with evidence of significant oedema or abnormal mesencephalic cisterns. The radiologists examined the CT scans independently and were blinded to patient presentation, ONSD measurement and discharge location, again to avoid bias.
The primary outcome measure of the study was the sensitivity and specificity of ONSD measurement in identifying raised ICP evident on CT scan. The secondary outcome measure was the sensitivity and specificity of ONSD in identifying the presence of any acute intracranial pathology evident on CT scan. Analysis of interobserver variability between reporting radiologists was calculated using κ values.
Full NHS research and ethical approval was gained for the study from the York Research and Ethics committee, as a specialist research ethics committee governing research involving patients who were unconscious.
Results
During the 3-month study period 26 patients were enrolled into the study. Patient demographics and presentation are listed in table 1. The patient path is outlined in the flowchart (figure 1).
Patient characteristics
Patient flow diagram.
Of the 26 patients enrolled in the study, 6 had raised ICP on ONSD measurement (table 2). CT scanning confirmed all six of these patients had signs of raised ICP. A single patient had a positive CT but no signs of raised ICP on ONSD measurement.
Patient ICP levels
Analysis showed, for the primary outcome measure, ONSD measurement had a specificity of 100% (95% CI 79% to 100%) and sensitivity of 86% (95% CI 42% to 99%) for raised ICP. For the secondary outcome measure, of any acute intracranial abnormality evident on CT, ONSD had a specificity of 100% (95% CI 76% to 100%) and sensitivity of 60% (95% CI 27% to 86%).
κ Values were 0.91 (95% CIs 0.73 to 1) for identification of raised ICP on CT and 0.84 (95% CIs 0.62 to 1) for any acute pathology on CT, between radiologists.
Of the seven patients who had signs of raised ICP on CT scanning, one died in the department, three were referred for neurosurgical care and three were referred to a ward for medical treatment only.
Discussion
For the primary outcome measure of raised ICP, ultrasound ONSD measurement was 100% specific and 86% sensitive. This translates to a positive predicted value of 100% and a negative predicted value of 95% for raised ICP identification with ONSD measurement. For the secondary outcome measure of identifying any acute intracranial pathology, ONSD was 100% specific and 60% sensitive.
These figures agree with previous studies showing high specificity and sensitivity for ICP identification.4 There was also very low interobserver variability between reporting radiologists, showing that CT diagnosis of raised ICP and acute intracranial pathology, with the study diagnostic criteria, is reproducible.
This study was small due to limited availability of trained ultrasound operatives, and the observational nature of the study means some patients referred for CT were not identified or enrolled. All eligible patients who were referred for CT scan for any clinical reason were enrolled to limit selection bias. Blinding of the radiologists to ONSD measurement was achieved, but it was impossible to blind the ultrasound operator from the study participant's clinical state. This may have introduced some measurement bias, although the ONSD measurement was not taken by the patient's primary treating clinician to limit the effect of this bias as much as possible.
There were differences in mean presentation Glasgow Coma Score (GCS) of trauma and non-trauma groups. This is explained by the National Institute for Health and Clinical Excellence (NICE) guidelines for CT scanning in traumatic head injury. In the trauma group many patients had a GCS of 15 on admission to the emergency department but were scanned due to history of significant loss of consciousness, persistent vomiting or use of anticoagulants. There was also a discrepancy on mean age between trauma and non-trauma groups. This is explained by all of the patients in the non-trauma group being diagnosed as having intracerebral bleeds or cerebrovascular accidents (CVAs), with higher incidences in older patients. This accounted for the higher mean age in this group.
Conclusions
This small study has shown that ultrasound measurement of ONSD, with minimal training, is a specific and sensitive measure of raised ICP in any patient undergoing urgent CT from the emergency department. It may play a role in the evaluation of raised ICP in medical and trauma cases. This test is quick to perform, reproducible and, in the unstable or quickly deteriorating patient, could be used to vary treatment for raised ICP. Larger observational trials are needed to evaluate the use of this emerging diagnostic tool further.
Footnotes
Competing interests None.
Ethics approval This study was conducted with the approval of the York Research and Ethics committee.
Provenance and peer review Not commissioned; externally peer reviewed.
Linked Articles
- Primary survey