Neurocrit Care DOI 10.1007/s12028-014-0003-y
ORIGINAL ARTICLE
Association Between Optic Nerve Sheath Diameter and Mortality in Patients with Severe Traumatic Brain Injury Mypinder S. Sekhon • Paul McBeth • Jie Zou • Lu Qiao • Leif Kolmodin • William R. Henderson Steve Reynolds • Donald E. G. Griesdale
•
Ó Springer Science+Business Media New York 2014
Abstract Purpose Increased intracranial pressure (ICP) is associated with worse outcomes following traumatic brain injury (TBI). Studies have confirmed that ICP is correlated with optic nerve sheath diameter (ONSD) on ultrasound. The aim of our study was to assess the independent relationship between ONSD measured using CT and mortality in a population of patients admitted with severe TBI. Methods We conducted a retrospective cohort study of patients with a TBI requiring ICP monitoring admitted to the ICU between April 2006 and May 2012 to two neurotrauma centers. ONSD was independently measured by two physicians blinded to patient outcomes. Multivariable logistic regression modeling was used to assess an association between ONSD and hospital mortality. M. S. Sekhon (&) P. McBeth J. Zou L. Qiao L. Kolmodin W. R. Henderson D. E. G. Griesdale Division of Critical Care Medicine, Department of Medicine, Vancouver General Hospital, University of British Columbia, Room 2438, Jim Pattison Pavilion, 2nd Floor, 855 West 12th Avenue, Vancouver, BC V5Z 1M9, Canada e-mail: [email protected] S. Reynolds Department of Critical Care Medicine, Royal Columbian Hospital, University of British Columbia, New Westminster, BC, Canada D. E. G. Griesdale Department of Anesthesiology, Pharmacology and Therapeutics, Vancouver General Hospital, University of British Columbia, West 12th Avenue, Vancouver, BC V5Z 1M9, Canada D. E. G. Griesdale Centre for Clinical Epidemiology and Evaluation, Vancouver Coastal Health Research Institute, University of British Columbia, 899 West 12th Avenue, Vancouver, BC V5Z 1M9, Canada
Results A total of 220 patients were included in the analysis. Overall, the cohort had a mean age of 35 (SD 17) years and 171 of 220 (79 %) were male. The median admission GCS was 6 (IQR 3–8). Intra-class correlation coefficient between raters for ONSD measurements was 0.92 (95 % CI 0.90–0.94, P < 0.0001). On multivariable analysis, each 1 mm increase in ONSD was associated with a twofold increase in hospital mortality (OR 2.0, 95 % CI 1.2–3.2, P = 0.007). Using linear regression, ONSD was independently associated with increased ICP in the first 48 h after admission (b = 4.4, 95 % CI 2.5–6.3, P < 0.0001). Conclusions In patients with TBI, ONSD measured on CT scanning was independently associated with ICP and mortality. Keywords Optic nerve sheath diameter Computed tomography Traumatic brain injury Intracranial pressure Cerebral spinal fluid
Introduction The burden of traumatic brain injury (TBI) on an individual and societal level is substantial. Each year in the United States, 1.4 million patients suffer a TBI. Of these, 235,000 patients are hospitalized, and 50,000 die [1]. Importantly, less than half of patients with a severe TBI have a favorable neurologic outcome at 1 year [2] and 40 % of survivors will experience long-term disabilities [3]. There are multiple clinical variables associated with poor outcome, including: advanced age, low admission Glasgow Coma Scale (GCS) score, pre-hospital hypotension, hypoxemia, and pupillary abnormalities [4–6]. However, even when they are combined into multivariable models, they remain
123
Neurocrit Care
only moderately discriminative [7]. Thus, investigation into additional variables to improve prognostication is warranted. Elevated intracranial pressure (ICP) is an important modifiable secondary injury that has been associated with worse outcomes in patients with severe TBI [8]. Following TBI, intracranial compliance can decrease due to cerebral edema or space occupying lesions, which ultimately results in increased ICP [5]. Subsequently, cerebral perfusion pressure decreases resulting in inadequate cerebral oxygen delivery and secondary cerebral ischemia [5, 6]. As such, invasive ICP monitoring is currently a level II recommendation from the Brain Trauma Foundation in patients who present with a low GCS [9] but can be associated with complications. Recently, non-invasive determination of ICP has garnered greater interest, in particular estimation of the optic nerve sheath diameter (ONSD). As an extension of the central nervous system, the optic nerve is surrounded by the subarachnoid membrane and cerebrospinal fluid (CSF). Increasing CSF pressure results in increased ONSD [10]. Multiple studies have demonstrated that ONSD correlates with ICP using ultrasound in patients with TBI [11–14]. ONSD measurements using ultrasound are reliable [15] and readily taught to novice operators [16], although it requires technical expertise. As an alternative, ONSD based on CT scans provides a readily available measure based on a test that is ubiquitous in hospitalized patients with TBI. Recently, ONSD calculated on CT scanning revealed an association with mortality on an unadjusted analysis in patients with TBI [17]. This study further demonstrated that ONSD was associated with CT scan findings suggestive of elevated ICP (midline shift >5 mm, basal cistern, and sulcal effacement); however, an association between elevated ONSD and ICP could not be established as only 7 of 77 (9 %) patients underwent invasive ICP monitoring. Therefore, we set out to determine if there is an association between ONSD measured on CT and hospital mortality using a multivariable analysis. Additionally, we evaluated a potential association between ONSD on CT and increased ICP, measured by invasive neuromonitoring. Finally, we aimed to quantify the reliability of ONSD measurements between 2 independent assessors and assess the relationship between ONSD and within the first 48 h following TBI.
British Columbia, Fraser Health, and Vancouver Coastal Health Authority approved the protocol and waived the requirement for written informed consent.
Methods
Measurement of ONSD on CT
This article reports on our 2 centers retrospective cohort study in accordance with the strengthening the reporting of observational studies in epidemiology (STROBE) guidelines [18]. The Research Ethics Boards at the University of
Independently, two investigators (MS, PM), not involved with data collection and blinded to the patient’s condition, measured bilateral ONSD on from the initial admission CT head. Each observer scrolled through the CT head slices to
123
Patient Inclusion and Data Collection We used an administrative database of all patients admitted to the intensive care units (ICUs) at Vancouver General Hospital (VGH) and Royal Columbian Hospital (RCH) between April 2006 and May 2012. Patients with a closed TBI were included if they had an ICP monitor inserted during their stay. Affiliated with the University of British Columbia, VGH (27 beds) and RCH (18 beds) are both closed, mixed medical-surgical ICUs which operate on an approximately 1:1.2 nurse-to-patient ratio. Both ICUs are staffed by fellowship-trained intensivists with medical specialty trainees in attendance. Data were collected using standardized electronic case report forms using Microsoft Access 2013Ò database (Redmond, WA). We collected data on the location, mechanism, and severity of TBI and ICU management. The location and time of injury as well as intervening transfers were collected. We collected the following baseline characteristics: admission Glasgow Coma Score (GCS), pupillary reactivity, pre-hospital or admission hypoxemia (arterial oxygen saturation
ORIGINAL ARTICLE
Association Between Optic Nerve Sheath Diameter and Mortality in Patients with Severe Traumatic Brain Injury Mypinder S. Sekhon • Paul McBeth • Jie Zou • Lu Qiao • Leif Kolmodin • William R. Henderson Steve Reynolds • Donald E. G. Griesdale
•
Ó Springer Science+Business Media New York 2014
Abstract Purpose Increased intracranial pressure (ICP) is associated with worse outcomes following traumatic brain injury (TBI). Studies have confirmed that ICP is correlated with optic nerve sheath diameter (ONSD) on ultrasound. The aim of our study was to assess the independent relationship between ONSD measured using CT and mortality in a population of patients admitted with severe TBI. Methods We conducted a retrospective cohort study of patients with a TBI requiring ICP monitoring admitted to the ICU between April 2006 and May 2012 to two neurotrauma centers. ONSD was independently measured by two physicians blinded to patient outcomes. Multivariable logistic regression modeling was used to assess an association between ONSD and hospital mortality. M. S. Sekhon (&) P. McBeth J. Zou L. Qiao L. Kolmodin W. R. Henderson D. E. G. Griesdale Division of Critical Care Medicine, Department of Medicine, Vancouver General Hospital, University of British Columbia, Room 2438, Jim Pattison Pavilion, 2nd Floor, 855 West 12th Avenue, Vancouver, BC V5Z 1M9, Canada e-mail: [email protected] S. Reynolds Department of Critical Care Medicine, Royal Columbian Hospital, University of British Columbia, New Westminster, BC, Canada D. E. G. Griesdale Department of Anesthesiology, Pharmacology and Therapeutics, Vancouver General Hospital, University of British Columbia, West 12th Avenue, Vancouver, BC V5Z 1M9, Canada D. E. G. Griesdale Centre for Clinical Epidemiology and Evaluation, Vancouver Coastal Health Research Institute, University of British Columbia, 899 West 12th Avenue, Vancouver, BC V5Z 1M9, Canada
Results A total of 220 patients were included in the analysis. Overall, the cohort had a mean age of 35 (SD 17) years and 171 of 220 (79 %) were male. The median admission GCS was 6 (IQR 3–8). Intra-class correlation coefficient between raters for ONSD measurements was 0.92 (95 % CI 0.90–0.94, P < 0.0001). On multivariable analysis, each 1 mm increase in ONSD was associated with a twofold increase in hospital mortality (OR 2.0, 95 % CI 1.2–3.2, P = 0.007). Using linear regression, ONSD was independently associated with increased ICP in the first 48 h after admission (b = 4.4, 95 % CI 2.5–6.3, P < 0.0001). Conclusions In patients with TBI, ONSD measured on CT scanning was independently associated with ICP and mortality. Keywords Optic nerve sheath diameter Computed tomography Traumatic brain injury Intracranial pressure Cerebral spinal fluid
Introduction The burden of traumatic brain injury (TBI) on an individual and societal level is substantial. Each year in the United States, 1.4 million patients suffer a TBI. Of these, 235,000 patients are hospitalized, and 50,000 die [1]. Importantly, less than half of patients with a severe TBI have a favorable neurologic outcome at 1 year [2] and 40 % of survivors will experience long-term disabilities [3]. There are multiple clinical variables associated with poor outcome, including: advanced age, low admission Glasgow Coma Scale (GCS) score, pre-hospital hypotension, hypoxemia, and pupillary abnormalities [4–6]. However, even when they are combined into multivariable models, they remain
123
Neurocrit Care
only moderately discriminative [7]. Thus, investigation into additional variables to improve prognostication is warranted. Elevated intracranial pressure (ICP) is an important modifiable secondary injury that has been associated with worse outcomes in patients with severe TBI [8]. Following TBI, intracranial compliance can decrease due to cerebral edema or space occupying lesions, which ultimately results in increased ICP [5]. Subsequently, cerebral perfusion pressure decreases resulting in inadequate cerebral oxygen delivery and secondary cerebral ischemia [5, 6]. As such, invasive ICP monitoring is currently a level II recommendation from the Brain Trauma Foundation in patients who present with a low GCS [9] but can be associated with complications. Recently, non-invasive determination of ICP has garnered greater interest, in particular estimation of the optic nerve sheath diameter (ONSD). As an extension of the central nervous system, the optic nerve is surrounded by the subarachnoid membrane and cerebrospinal fluid (CSF). Increasing CSF pressure results in increased ONSD [10]. Multiple studies have demonstrated that ONSD correlates with ICP using ultrasound in patients with TBI [11–14]. ONSD measurements using ultrasound are reliable [15] and readily taught to novice operators [16], although it requires technical expertise. As an alternative, ONSD based on CT scans provides a readily available measure based on a test that is ubiquitous in hospitalized patients with TBI. Recently, ONSD calculated on CT scanning revealed an association with mortality on an unadjusted analysis in patients with TBI [17]. This study further demonstrated that ONSD was associated with CT scan findings suggestive of elevated ICP (midline shift >5 mm, basal cistern, and sulcal effacement); however, an association between elevated ONSD and ICP could not be established as only 7 of 77 (9 %) patients underwent invasive ICP monitoring. Therefore, we set out to determine if there is an association between ONSD measured on CT and hospital mortality using a multivariable analysis. Additionally, we evaluated a potential association between ONSD on CT and increased ICP, measured by invasive neuromonitoring. Finally, we aimed to quantify the reliability of ONSD measurements between 2 independent assessors and assess the relationship between ONSD and within the first 48 h following TBI.
British Columbia, Fraser Health, and Vancouver Coastal Health Authority approved the protocol and waived the requirement for written informed consent.
Methods
Measurement of ONSD on CT
This article reports on our 2 centers retrospective cohort study in accordance with the strengthening the reporting of observational studies in epidemiology (STROBE) guidelines [18]. The Research Ethics Boards at the University of
Independently, two investigators (MS, PM), not involved with data collection and blinded to the patient’s condition, measured bilateral ONSD on from the initial admission CT head. Each observer scrolled through the CT head slices to
123
Patient Inclusion and Data Collection We used an administrative database of all patients admitted to the intensive care units (ICUs) at Vancouver General Hospital (VGH) and Royal Columbian Hospital (RCH) between April 2006 and May 2012. Patients with a closed TBI were included if they had an ICP monitor inserted during their stay. Affiliated with the University of British Columbia, VGH (27 beds) and RCH (18 beds) are both closed, mixed medical-surgical ICUs which operate on an approximately 1:1.2 nurse-to-patient ratio. Both ICUs are staffed by fellowship-trained intensivists with medical specialty trainees in attendance. Data were collected using standardized electronic case report forms using Microsoft Access 2013Ò database (Redmond, WA). We collected data on the location, mechanism, and severity of TBI and ICU management. The location and time of injury as well as intervening transfers were collected. We collected the following baseline characteristics: admission Glasgow Coma Score (GCS), pupillary reactivity, pre-hospital or admission hypoxemia (arterial oxygen saturation
