755
NeuroRehabilitation 35 (2014) 755–761 DOI:10.3233/NRE-141160 IOS Press
Hydrocephalus during rehabilitation following severe TBI. Relation to recovery, outcome, and length of stay Mia Linnemanna,b , Maiken Tibæka,b and Lars Peter Kammersgaarda,b,∗ a Research
Unit on Brain Injury Rehabilitation, Copenhagen (RUBRIC), Denmark of Neurorehabilitation, TBI Unit. Copenhagen University Hospital of Glostrup, Denmark
b Department
Abstract. BACKGROUND: Post traumatic hydrocephalus (PTH) is a frequent complication during rehabilitation following severe TBI. However, the diagnosis of PTH is not straightforward and despite shunting recovery may be delayed. OBJECTIVE: To study the influence of PTH on recovery and outcome during rehabilitation. METHODS: We studied 417 patients with severe TBI admitted consecutively to a single hospital-based neurorehabilitation department serving Eastern Denmark between 2000 and 2010. Demographics (age and gender) and clinical characteristics (length of acute treatment, post traumatic amnesia (PTA), level of consciousness, injury severity (ISS), and admission FIMTM ), and PTH were related to recovery (discharge FIMTM ), outcome (GOS), and length of rehabilitation stay. RESULTS: Patients with PTH were older, brain injury more severe, and acute treatment was longer. At discharge they had more disability, longer rehabilitation stays, and unfavorable outcome. However, after adjusted multiple regression analyses PTH was not associated with disability at discharge or outcome. Instead, PTH was associated with longer stay for rehabilitation. CONCLUTIONS: Shunting for PTH does not affect recovery and outcome per se, but prolongs lengths of stay by almost 3 weeks. Therefore, patients treated for PTH are as likely to benefit from rehabilitation as patients without, but require longer rehabilitation stays. Keywords: Brain injury, traumatic, hydrocephalus, rehabilitation, outcome, prediction
1. Introduction Patients with traumatic brain injury (TBI) sometimes develop hydrocephalus leading to increased intracranial pressure that requires permanent diversion of cerebrospinal fluid (CSF) to alleviate the effects of high intracranial pressure on brain func∗ Address for correspondence: Dr. Lars Peter Kammersgaard, Consultant research neurologist, Research Unit on Brain Injury Rehabilitation, Copenhagen (RUBRIC), Department of Neurorehabilitation, TBI Unit, Copenhagen University Hospital, Glostrup, Ketteg˚ard Alle 30, DK-2650 Hvidovre, Denmark. E-mail: [email protected].
tion (Bauer, McGwin, Jr., Melton, George, & Markert, 2011). This condition is referred to as post traumatic hydrocephalus (PTH). PTH is treated by surgical implantation of a shunt that leads away excess CSF to the peritoneum or the atrium of the heart – a ventriculoperitoneal or ventriculoatrial shunt. Studies have reported PTH to emerge in somewhere between 5% and 45% largely depending on severity of TBI and whether the patients studied are from the acute setting or the post-acute phase (Groswasser, Cohen, ReiderGroswasser, & Stern, 1988; Licata, Cristofori, Gambin, Vivenza, & Turazzi, 2001; Marmarou et al., 1996; Mazzini et al., 2003).
1053-8135/14/$27.50 © 2014 – IOS Press and the authors. All rights reserved
756
M. Linnemann et al. / Hydrocephalus during rehabilitation following severe TBI
However, the diagnosis of true PTH is not always simple and the criteria used to select patients for shunt surgery during the post-acute phase are not strict. Some rely on imaging criteria, others on measurements of CSF dynamics, clinical deterioration during rehabilitation, or a combination of these aspects (Groswasser et al., 1988; Licata et al., 2001; Marmarou et al., 1996; Mazzini et al., 2003). Because there are no firm diagnostic criteria for PTH the diagnosis of “true” hydrocephalus may be delayed. Consequently, decision to implant a shunt during rehabilitation might be postponed leading to impeded recovery of brain function. Therefore, PTH as a complication during rehabilitation may lead to prolonged hospital stay, and poor outcome following severe TBI despite treatment with a shunt (Marmarou et al., 1996; Mazzini et al., 2003). In this retrospective study a large number of patients with severe TBI undergoing lengthy in-hospital rehabilitation were included with the aim to investigate if shunting for PTH after the acute stage of TBI extended length of rehabilitation stay in hospital, increased disability at discharge, and worsened global outcome at discharge.
2. Methods 2.1. Participants The study included all 444 patients aged 18 years or older with severe TBI consecutively admitted for rehabilitation to the Department of Neurorehabilitation/Traumatic Brain Injury Unit at Glostrup Hospital, Copenhagen, Denmark, over a decade beginning October 1st 2000 to September 30th 2010. We excluded 27 patients, who were shunted in the acute stage of TBI, from the analyses because a decision to establish permanent CSF diversion in the neurosurgical departments was influenced by somewhat different criteria: Namely conversion from a temporary ventriculostomy drain to a permanent shunt because repeated measurements showed increased intracranial pressure. In the rehabilitation phase clinical signs of PTH were if patients slowed in expected progress, did not improve, or even declined in function. Radiological signs of PTH were enlarged lateral and third ventricles, and effaced cortical sulci present on CT or MRI of the brain under these clinical conditions. Imaging was available around the clock, seven days per week. All patients with clinical and radiological signs of PTH were treated with a shunt; therefore no patient was excluded from
shunting because of anticipated dismal prognosis after TBI. The neurosurgical departments that also treated the patients in the acute stage performed the shunting surgery procedure. All patients with PTH were given ventriculoperitoneal shunts. The rehabilitation unit is part of an acute hospital that also has an intensive care unit, departments of internal medicine, surgery, and a radiological service. Patients were referred from the only two neurosurgical/ neurointensive wards serving the eastern part of Denmark with a total population of 2.5 million inhabitants. Our rehabilitation department takes care of early, centralized neurorehabilitation, for all patients who sustain acute severe TBI in this geographical area. There is a close collaboration between the departments of neurosurgery and the neurorehabilitation unit. Therefore, all patients with the most severe brain injuries are admitted to our rehabilitation unit if they survive the acute stage of TBI. Patients were admitted to the rehabilitation unit if they fulfilled either of the following clinical criteria for severe TBI after cessation of sedation in the intensive care units: 1) Glasgow coma scale (Teasdale & Jennett, 1974) (GCS) score 3 to 9, 2) Glasgow coma scale score 10 to 12 if combined with severe focal neurological deficits or agitated behavior, or 3) severe cognitive dysfunction (continued post traumatic amnesia or lowered consciousness) combined with either hemiparesis or agitated behavior. No patient surviving the acute treatment was denied access to our rehabilitation unit because of futility or pre-existing comorbidity. 2.2. Measurements On the day of admission we prospectively recorded demographics (such as age and sex), time from injury to admission for rehabilitation (a substitute parameter for the extend and complexity of acute treatment), length of sedation during the acute stage, GCS on admission for rehabilitation, and level of consciousness according to the Aspen criteria assessed by the Rancho Los Amigos Scale (Giacino et al., 1997) (RLAS). Patients were considered in vegetative state if they were on level II on the RLAS. Duration of posttraumatic amnesia (PTA) was based on prospective weekly measurements using the Galveston Orientation and Amnesia Test (Levin, O’Donnell, & Grossman, 1979) (GOAT). Emergence from PTA was defined as two consecutive GOAT scores of 76 points or greater. The FIMTM scale (Hamilton, Laughlin, Fiedler, & Granger, 1994) was used to measure disability on admission and at discharge. The FIMTM scale consists of 18 items with score ranges
M. Linnemann et al. / Hydrocephalus during rehabilitation following severe TBI
from 1 to 7 for each item. The scores are eventually summed up to yield a total score between 18 (worst) and 126 points. Finally, we assessed injury severity score (ISS) based on clinical examination and review of case notes from the acute stage (Linn, 1995). ISS was constructed by summing up the squares of the scores derived from the three highest scoring organ systems on theAbbreviated Injury Scale (AIS). We used the 1998 version of the AIS to calculate the ISS (Palmer et al., 2012). Rehabilitation was organized in interdisciplinary teams. A team included neurorehabilitation neurologists, nurses, neuropsychologists, physio- and occupational therapists, and social workers. Speech and language pathologists and nutritionists were included if necessary. All patients received two hours of therapy six days a week by the physio- and occupational therapists and nurses used ADL actively to perform rehabilitation. Nurses and therapists performed evaluations of function by validated scales every 4th week as a part of routine. Measurements of function were performed more frequently if patients did not improve or declined in function. A neuropsychologist evaluated level of consciousness and orientation weekly. All acute patient care, rehabilitation, and treatments in Denmark are free of charge. Therefore, decision about the length of stay for rehabilitation in hospital is a clinical decision based on performance and improvement rather than economy. Patients were not discharged before the rehabilitation team decided that further rehabilitation could be continued in a less intensive rehabilitation facility or at home. The average length of stay for rehabilitation in our department is close to three month. Shunting was performed if the patient needed permanent CSF diversion by surgical placement of a shunt. We scrutinized all patient charts retrospectively to record when (acute phase versus rehabilitation phase) a decision about permanent CSF diversion was made. At discharge length of stay for rehabilitation was recorded. Disability at discharge was measured by the FIMTM scale and global outcome assessed by the Glasgow Outcome Scale (GOS) (Jennett & Bond, 1975). The GOS is a 5 level hierarchical scale, which gives a gross assessment of global outcome categorized into dead (level 1), vegetative (level 2), severe disability (level 3), moderate disability (level 4), and good recovery (level 5) (Jennett & Bond, 1975). The output from the scale was dichtomized into unfavourable outcome (levels 1 through 3) versus favourable outcome (moderate disability and good recovery) (Nichol et al., 2011).
757
2.3. Statistical analyses Statistical analyses were performed with the SPSS package for Windows® version 18.0. The distribution of age for patients with versus without shunt was analyzed with One-Way ANOVA test, because age had a normal distribution. For all other continuous variables comparisons were made by the Mann-Whitney U test and distribution given as median and interquartile range (IQR). For dichotomized variables (sex and number of patients in vegetative state on admission) we performed comparisons by using Fischer’s Exact test. Stepwise linear regression models were used to establish the independent relation between shunting and FIMTM at discharge, and between shunting and length of rehabilitation stay adjusted for demographics and clinical characteristics (age, sex, ISS, length of acute stay, PTA, and FIMTM on admission). Similarly, stepwise logistic regression models were used to determine the independent relation between shunting and favourable outcome adjusted for clinical characteristics and demographics. Multiple regression analyses were based on the Wald statistics. All findings were considered significant if p-values were equal to or less than 0.05. 2.4. Ethical considerations The study was approved by the Danish Data Protection Agency and complied with the Danish Committee System on Biomedical Research Ethics Guidelines. The Danish Ethical Committee approved the study. 3. Results Forty-seven (11.3%) of the 417 patients developed PTH and required shunting during their in-hospital rehabilitation. Table 1 shows a comparison of clinical characteristics between patients with versus without shunt. Shunted patients were older (p = 0.009), had longer time in sedation during the acute treatment phase (p = 0.015), and stayed longer in the acute treatment phase (p = 0.04). At the time of admission for rehabilitation patients who eventually needed shunting had lower GCS (p < 0.001), were more often in vegetative state (p < 0.001), and had lower total FIMTM scores (p < 0.001). Furthermore, they had more severe head injury as measured by the length of PTA. We found no differences between the two groups with regard to gender (p = 0.99) and overall injury severity as measured by the ISS (p = 0.27).
758
M. Linnemann et al. / Hydrocephalus during rehabilitation following severe TBI Table 1 Clinical characteristics for TBI patients with vs. without post traumatic hydrocephalus −PTH
+PTH Number of patients Age, years, mean (SD) Sex, male Sedation, days, median (IQR) Acute phase, days, median (IQR) GCS∗ , median (IQR) Vegetative on admission PTA, median (IQR) FIM admission, median (IQR) ISS, median (IQR)
47 (11.3%) 50.8 (18.0) 36/11 (76.6%) 8 (5 to 12) 19 (14 to 25) 9 (7 to 11) 21/25 (45.7%) 132 (67 to 224) 18 (18 to 18) 27 (25 to 33)
p-value
370 (88.7%) 43.6 (17.8) 282/88 (76.2%) 5 (2 to 10) 17 (11 to 23) 12 (9 to 14) 67/298 (18.4%) 50 (31 to 100) 19 (18 to 34) 25 (25 to 34)
0.009 0.99 0.015 0.04
NeuroRehabilitation 35 (2014) 755–761 DOI:10.3233/NRE-141160 IOS Press
Hydrocephalus during rehabilitation following severe TBI. Relation to recovery, outcome, and length of stay Mia Linnemanna,b , Maiken Tibæka,b and Lars Peter Kammersgaarda,b,∗ a Research
Unit on Brain Injury Rehabilitation, Copenhagen (RUBRIC), Denmark of Neurorehabilitation, TBI Unit. Copenhagen University Hospital of Glostrup, Denmark
b Department
Abstract. BACKGROUND: Post traumatic hydrocephalus (PTH) is a frequent complication during rehabilitation following severe TBI. However, the diagnosis of PTH is not straightforward and despite shunting recovery may be delayed. OBJECTIVE: To study the influence of PTH on recovery and outcome during rehabilitation. METHODS: We studied 417 patients with severe TBI admitted consecutively to a single hospital-based neurorehabilitation department serving Eastern Denmark between 2000 and 2010. Demographics (age and gender) and clinical characteristics (length of acute treatment, post traumatic amnesia (PTA), level of consciousness, injury severity (ISS), and admission FIMTM ), and PTH were related to recovery (discharge FIMTM ), outcome (GOS), and length of rehabilitation stay. RESULTS: Patients with PTH were older, brain injury more severe, and acute treatment was longer. At discharge they had more disability, longer rehabilitation stays, and unfavorable outcome. However, after adjusted multiple regression analyses PTH was not associated with disability at discharge or outcome. Instead, PTH was associated with longer stay for rehabilitation. CONCLUTIONS: Shunting for PTH does not affect recovery and outcome per se, but prolongs lengths of stay by almost 3 weeks. Therefore, patients treated for PTH are as likely to benefit from rehabilitation as patients without, but require longer rehabilitation stays. Keywords: Brain injury, traumatic, hydrocephalus, rehabilitation, outcome, prediction
1. Introduction Patients with traumatic brain injury (TBI) sometimes develop hydrocephalus leading to increased intracranial pressure that requires permanent diversion of cerebrospinal fluid (CSF) to alleviate the effects of high intracranial pressure on brain func∗ Address for correspondence: Dr. Lars Peter Kammersgaard, Consultant research neurologist, Research Unit on Brain Injury Rehabilitation, Copenhagen (RUBRIC), Department of Neurorehabilitation, TBI Unit, Copenhagen University Hospital, Glostrup, Ketteg˚ard Alle 30, DK-2650 Hvidovre, Denmark. E-mail: [email protected].
tion (Bauer, McGwin, Jr., Melton, George, & Markert, 2011). This condition is referred to as post traumatic hydrocephalus (PTH). PTH is treated by surgical implantation of a shunt that leads away excess CSF to the peritoneum or the atrium of the heart – a ventriculoperitoneal or ventriculoatrial shunt. Studies have reported PTH to emerge in somewhere between 5% and 45% largely depending on severity of TBI and whether the patients studied are from the acute setting or the post-acute phase (Groswasser, Cohen, ReiderGroswasser, & Stern, 1988; Licata, Cristofori, Gambin, Vivenza, & Turazzi, 2001; Marmarou et al., 1996; Mazzini et al., 2003).
1053-8135/14/$27.50 © 2014 – IOS Press and the authors. All rights reserved
756
M. Linnemann et al. / Hydrocephalus during rehabilitation following severe TBI
However, the diagnosis of true PTH is not always simple and the criteria used to select patients for shunt surgery during the post-acute phase are not strict. Some rely on imaging criteria, others on measurements of CSF dynamics, clinical deterioration during rehabilitation, or a combination of these aspects (Groswasser et al., 1988; Licata et al., 2001; Marmarou et al., 1996; Mazzini et al., 2003). Because there are no firm diagnostic criteria for PTH the diagnosis of “true” hydrocephalus may be delayed. Consequently, decision to implant a shunt during rehabilitation might be postponed leading to impeded recovery of brain function. Therefore, PTH as a complication during rehabilitation may lead to prolonged hospital stay, and poor outcome following severe TBI despite treatment with a shunt (Marmarou et al., 1996; Mazzini et al., 2003). In this retrospective study a large number of patients with severe TBI undergoing lengthy in-hospital rehabilitation were included with the aim to investigate if shunting for PTH after the acute stage of TBI extended length of rehabilitation stay in hospital, increased disability at discharge, and worsened global outcome at discharge.
2. Methods 2.1. Participants The study included all 444 patients aged 18 years or older with severe TBI consecutively admitted for rehabilitation to the Department of Neurorehabilitation/Traumatic Brain Injury Unit at Glostrup Hospital, Copenhagen, Denmark, over a decade beginning October 1st 2000 to September 30th 2010. We excluded 27 patients, who were shunted in the acute stage of TBI, from the analyses because a decision to establish permanent CSF diversion in the neurosurgical departments was influenced by somewhat different criteria: Namely conversion from a temporary ventriculostomy drain to a permanent shunt because repeated measurements showed increased intracranial pressure. In the rehabilitation phase clinical signs of PTH were if patients slowed in expected progress, did not improve, or even declined in function. Radiological signs of PTH were enlarged lateral and third ventricles, and effaced cortical sulci present on CT or MRI of the brain under these clinical conditions. Imaging was available around the clock, seven days per week. All patients with clinical and radiological signs of PTH were treated with a shunt; therefore no patient was excluded from
shunting because of anticipated dismal prognosis after TBI. The neurosurgical departments that also treated the patients in the acute stage performed the shunting surgery procedure. All patients with PTH were given ventriculoperitoneal shunts. The rehabilitation unit is part of an acute hospital that also has an intensive care unit, departments of internal medicine, surgery, and a radiological service. Patients were referred from the only two neurosurgical/ neurointensive wards serving the eastern part of Denmark with a total population of 2.5 million inhabitants. Our rehabilitation department takes care of early, centralized neurorehabilitation, for all patients who sustain acute severe TBI in this geographical area. There is a close collaboration between the departments of neurosurgery and the neurorehabilitation unit. Therefore, all patients with the most severe brain injuries are admitted to our rehabilitation unit if they survive the acute stage of TBI. Patients were admitted to the rehabilitation unit if they fulfilled either of the following clinical criteria for severe TBI after cessation of sedation in the intensive care units: 1) Glasgow coma scale (Teasdale & Jennett, 1974) (GCS) score 3 to 9, 2) Glasgow coma scale score 10 to 12 if combined with severe focal neurological deficits or agitated behavior, or 3) severe cognitive dysfunction (continued post traumatic amnesia or lowered consciousness) combined with either hemiparesis or agitated behavior. No patient surviving the acute treatment was denied access to our rehabilitation unit because of futility or pre-existing comorbidity. 2.2. Measurements On the day of admission we prospectively recorded demographics (such as age and sex), time from injury to admission for rehabilitation (a substitute parameter for the extend and complexity of acute treatment), length of sedation during the acute stage, GCS on admission for rehabilitation, and level of consciousness according to the Aspen criteria assessed by the Rancho Los Amigos Scale (Giacino et al., 1997) (RLAS). Patients were considered in vegetative state if they were on level II on the RLAS. Duration of posttraumatic amnesia (PTA) was based on prospective weekly measurements using the Galveston Orientation and Amnesia Test (Levin, O’Donnell, & Grossman, 1979) (GOAT). Emergence from PTA was defined as two consecutive GOAT scores of 76 points or greater. The FIMTM scale (Hamilton, Laughlin, Fiedler, & Granger, 1994) was used to measure disability on admission and at discharge. The FIMTM scale consists of 18 items with score ranges
M. Linnemann et al. / Hydrocephalus during rehabilitation following severe TBI
from 1 to 7 for each item. The scores are eventually summed up to yield a total score between 18 (worst) and 126 points. Finally, we assessed injury severity score (ISS) based on clinical examination and review of case notes from the acute stage (Linn, 1995). ISS was constructed by summing up the squares of the scores derived from the three highest scoring organ systems on theAbbreviated Injury Scale (AIS). We used the 1998 version of the AIS to calculate the ISS (Palmer et al., 2012). Rehabilitation was organized in interdisciplinary teams. A team included neurorehabilitation neurologists, nurses, neuropsychologists, physio- and occupational therapists, and social workers. Speech and language pathologists and nutritionists were included if necessary. All patients received two hours of therapy six days a week by the physio- and occupational therapists and nurses used ADL actively to perform rehabilitation. Nurses and therapists performed evaluations of function by validated scales every 4th week as a part of routine. Measurements of function were performed more frequently if patients did not improve or declined in function. A neuropsychologist evaluated level of consciousness and orientation weekly. All acute patient care, rehabilitation, and treatments in Denmark are free of charge. Therefore, decision about the length of stay for rehabilitation in hospital is a clinical decision based on performance and improvement rather than economy. Patients were not discharged before the rehabilitation team decided that further rehabilitation could be continued in a less intensive rehabilitation facility or at home. The average length of stay for rehabilitation in our department is close to three month. Shunting was performed if the patient needed permanent CSF diversion by surgical placement of a shunt. We scrutinized all patient charts retrospectively to record when (acute phase versus rehabilitation phase) a decision about permanent CSF diversion was made. At discharge length of stay for rehabilitation was recorded. Disability at discharge was measured by the FIMTM scale and global outcome assessed by the Glasgow Outcome Scale (GOS) (Jennett & Bond, 1975). The GOS is a 5 level hierarchical scale, which gives a gross assessment of global outcome categorized into dead (level 1), vegetative (level 2), severe disability (level 3), moderate disability (level 4), and good recovery (level 5) (Jennett & Bond, 1975). The output from the scale was dichtomized into unfavourable outcome (levels 1 through 3) versus favourable outcome (moderate disability and good recovery) (Nichol et al., 2011).
757
2.3. Statistical analyses Statistical analyses were performed with the SPSS package for Windows® version 18.0. The distribution of age for patients with versus without shunt was analyzed with One-Way ANOVA test, because age had a normal distribution. For all other continuous variables comparisons were made by the Mann-Whitney U test and distribution given as median and interquartile range (IQR). For dichotomized variables (sex and number of patients in vegetative state on admission) we performed comparisons by using Fischer’s Exact test. Stepwise linear regression models were used to establish the independent relation between shunting and FIMTM at discharge, and between shunting and length of rehabilitation stay adjusted for demographics and clinical characteristics (age, sex, ISS, length of acute stay, PTA, and FIMTM on admission). Similarly, stepwise logistic regression models were used to determine the independent relation between shunting and favourable outcome adjusted for clinical characteristics and demographics. Multiple regression analyses were based on the Wald statistics. All findings were considered significant if p-values were equal to or less than 0.05. 2.4. Ethical considerations The study was approved by the Danish Data Protection Agency and complied with the Danish Committee System on Biomedical Research Ethics Guidelines. The Danish Ethical Committee approved the study. 3. Results Forty-seven (11.3%) of the 417 patients developed PTH and required shunting during their in-hospital rehabilitation. Table 1 shows a comparison of clinical characteristics between patients with versus without shunt. Shunted patients were older (p = 0.009), had longer time in sedation during the acute treatment phase (p = 0.015), and stayed longer in the acute treatment phase (p = 0.04). At the time of admission for rehabilitation patients who eventually needed shunting had lower GCS (p < 0.001), were more often in vegetative state (p < 0.001), and had lower total FIMTM scores (p < 0.001). Furthermore, they had more severe head injury as measured by the length of PTA. We found no differences between the two groups with regard to gender (p = 0.99) and overall injury severity as measured by the ISS (p = 0.27).
758
M. Linnemann et al. / Hydrocephalus during rehabilitation following severe TBI Table 1 Clinical characteristics for TBI patients with vs. without post traumatic hydrocephalus −PTH
+PTH Number of patients Age, years, mean (SD) Sex, male Sedation, days, median (IQR) Acute phase, days, median (IQR) GCS∗ , median (IQR) Vegetative on admission PTA, median (IQR) FIM admission, median (IQR) ISS, median (IQR)
47 (11.3%) 50.8 (18.0) 36/11 (76.6%) 8 (5 to 12) 19 (14 to 25) 9 (7 to 11) 21/25 (45.7%) 132 (67 to 224) 18 (18 to 18) 27 (25 to 33)
p-value
370 (88.7%) 43.6 (17.8) 282/88 (76.2%) 5 (2 to 10) 17 (11 to 23) 12 (9 to 14) 67/298 (18.4%) 50 (31 to 100) 19 (18 to 34) 25 (25 to 34)
0.009 0.99 0.015 0.04
