http://informahealthcare.com/bij ISSN: 0269-9052 (print), 1362-301X (electronic) Brain Inj, 2014; 28(3): 336–340 ! 2014 Informa UK Ltd. DOI: 10.3109/02699052.2013.873142
ORIGINAL ARTICLE
Agitation after traumatic brain injury and predictors of outcome Rajiv Singh, Guru Venkateshwara, Krishnan P. S. Nair, Muhammed Khan, & Rafat Saad Osborn Neurorehabilitation Unit, Northern General Hospital, Sheffield, UK
Abstract
Keywords
Objectives: To measure the incidence of agitation after traumatic brain injury (TBI) in an inpatient population and to identify any features associated with an adverse outcome Design: Prospective study of TBI admissions over 30 months in consecutive admissions with TBI to a regional neurorehabilitation unit. Outcome of agitation was compared to patient, injury and treatment features and any associations were sought. The presence of agitation was measured by the Agitation Behaviour Score. A good outcome for agitation was defined as a return to independent living with minimal care requirement at 6 months using an Extended Glasgow Outcome Scale (GOSE) score 44. Results: Over 30 months, there were 146 TBI admissions, of whom 53 cases had agitation (36.3%). Achieving 100% follow-up, 27 (51%) had a good outcome. On a multivariable logistic regression analysis, a good outcome was associated with the type of lesions seen on CT scan, the severity of agitation and the duration of the behaviour. Alcohol excess and type of treatment used for the behaviour were initially significant on univariate testing but dropped out of the logistic regression model. Conclusions: Over a third of TBI admissions, developed agitation and poor functional outcome was associated with CT scan findings, severity and duration of agitation.
Agitation, CT scans, neuropsychology, outcome, traumatic brain injury
Agitation is a serious sequela of traumatic brain injury (TBI) and often interferes with recovery, rehabilitation and community re-integration [1–4]. It can endanger the safety of patients, family and carers [5] and there is a paucity of good evidence on which the management of this condition can be based [6]. Defining agitation has proved to be difficult. A consensus statement from the American Academy of Physical Medicine and Rehabilitation defined agitation as ‘excessive behaviour occurring within the context of an altered state of consciousness and diminished cognitive functioning’ [7] (p. 803). While this provides a good working definition, it is clear that there is a wide range of behaviours that may constitute aggression, including verbal abuse, rage, impulsivity, distractibility, sudden mood change or lack of co-operation. It is, therefore, perhaps best to consider agitation not as a single behaviour, but rather as a global construct and to classify the extent of those behaviours as mild, moderate or severe [8, 9]. The variation in description of agitation affects the interpretation and comparison of study results as well as
Correspondence: Rajiv Singh, Osborn Neurorehabilitation Unit, Northern General Hospital, Sheffield S5 7AU, UK. Tel: 0044 114 2715651. Fax: 0044 114 2715649. E-mail: [email protected]
Received 15 October 2012 Accepted 4 December 2013 Published online 27 December 2013
estimates of the frequency of agitation. As a result, these range widely from 10–96% [8, 10–19]. Furthermore, there is some uncertainty whether short- or long-term outcomes are affected in those with agitation, e.g. achievement of rehabilitation goals or discharge destination [8, 10, 20]. It is also unclear whether agitation is associated with particular demographic or treatment features and no clear picture emerges from previous studies. The goal was to provide a systematic investigation of agitation in a population with TBI (traumatic brain injury). At the same time, the authors wished to investigate whether any particular features were associated with agitation outcome. The aim was to make the study directly relevant to normal neurorehabilitation practice for most clinicians and not overly specialized with highly complex assessments or major resource requirements, e.g. behavioural units or very specialized centres. Hence, a prospective study was designed of all admissions with TBI to a regional rehabilitation centre treated with standard behaviour modification and pharmacological intervention if required. A literature search on MEDLINE and PsycINFO found that all previous studies were from the US and one from Australia; there were no prospective studies from the UK. While there is no particular reason to think that the incidence would differ markedly from other countries, it was thought that it would be interesting to document incidence in the context of a UK population as well as investigate whether any patient or clinical features were associated with the long-term outcome.
20 14
Introduction
History
Agitation after traumatic brain injury
DOI: 10.3109/02699052.2013.873142
Methods Consecutive inpatients admitted to a Neurorehabilitation unit at a University Teaching Hospital with traumatic brain injury (TBI) from January 2007 to June 2009, were assessed for agitation during their stay. Only those with a primary diagnosis of TBI were included and all patients admitted to the unit were over 18. Demographic and injury characteristics were recorded on admission. Agitation was identified and graded by using the Agitation Behavioural Scale [9, 21]. Cases were identified if two or more of the behaviours on this scale were exhibited by an individual over any 24-hour period. If such behaviours were observed by staff, then subsequent daily assessment by neuropsychologists over a 30-minute session was used to assess the severity of agitation by scoring the behaviours observed. This was made by the same observer as often as possible but required nursing or medical staff to make the assessments at weekends after appropriate training. Daily observation ensured ongoing recording of all agitation behaviours rather than depending on retrospective notes reading. Mild agitation is diagnosed between scores of 21–28, moderate from 29–35 and severe agitation above 35. Agitation was resolved when the score was less than 21 on two consecutive days. Severity of brain injury was graded by Glasgow Coma Score in the first 24 hours, but, as this was an inpatient study, almost all admissions to the unit fall into the group of severe injury. This study also examined the predominant type of brain lesion caused by the injury. Using CT scans, reported at a weekly neuroradiology department meeting, the authors were able to classify injuries into three groups; those whose predominant lesion was diffuse axonal injury, those with predominantly brain contusion injuries and those predominantly with intra-cranial haemorrhage. While it is accepted that a certain overlap will exist in some patients, it was, in fact, quite straightforward for the neuroradiology meeting to decide which pathology was most prominent on the scans and no major disagreements took place. It was felt that this might give a more useful item to study rather than simply the severity of injury. The pharmacological treatment received by patients was also examined. There was no attempt made to modify pharmacological treatment for the study, which was left to the treating team if they judged appropriate. All patients received behavioural approaches aimed at identifying antecedents and triggers, environmental modification where appropriate, constant support including family and use of positive and negative reinforcements aimed at modifying the resultant behaviours. Pharmacological management was divided into three groups, with the first group consisting of patients who received no drugs. The second group received only one agent, e.g. beta-blockers, benzodiazepines or mood stabilizers such as carbamazepine. The third group required change of the first medication to another agent as the first was not considered effective. There may have been a period of overlap as one drug was ‘washed out’, but no patient received more than one behavioural drug at any time. This study compared these three groups. There was no attempt to compare different drugs.
337
Rehabilitation outcome was assessed by the Extended Glasgow Outcome Scale (GOSE) at 6 months post-injury. In those instances where the patient could not complete the questions, a carer was used. A cut-off above 4 (moderate lower or better) was considered a good outcome and represents those who are capable of independent living [22]. Outcome was, therefore, a binary variable. Most patients attended clinic follow-up; a few required a visit to their home or nursing home. Local ethics committee approval was taken. Initial univariate analysis was carried out by 2 test and t-test to examine the relationship between outcome and the dependant variables, with statistically significant difference set at p50.05. These features were; age, severity of agitation, duration of agitation, CT scan findings, gender, alcohol excess and pharmacological treatment given. Subsequent multivariable logistic regression with backward step entry was applied; starting with a full model of all variables, a stepwise analysis was used to remove non-significant variables one at a time to determine the independent predictors of outcome. The criteria for entry or removal of a variable were 5% for inclusion and 5% for exclusion. All analyses were performed using SPSS version 16.0.
Results Over a 30-month period, there was a total of 146 traumatic brain injury (TBI) inpatient admissions, of whom 53 cases manifested some degree of agitation. This corresponds to a frequency of 36.3%. Severe TBI made up 79% of agitation cases and 88% of all admissions to the unit. Average age of the agitation group was 41.2 years (SD ¼ 15.8) and of the whole group 45.4 years (SD ¼ 21.0). Average duration of agitation was 39.0 days (SD ¼ 27.7). In terms of type of injury seen on CT scan, 13 (24.5%) individuals had diffuse axonal injury, 17 (32.1%) had contusional injury and 23 (43.4%) had intracranial haemorrhage as their predominant pathology. In 26 (49%) cases there was a history of alcohol excess. Using the ABS, mild agitation made up 10 (18.9%) of the group, moderate was 21 (39.6%) and severe was 22 (41.5%) of the group. This study looked at how many patients had received medication and found that 13 (24.5%) received no behavioural medication, 17 (32.1%) received one agent only and 23 (43.4%) required change of medication to an alternative agent. Follow-up was achieved for all 53 cases. Outcome was a binary variable. Using a score of GOSE44 to determine good outcome, patients divided almost equally, with 27 (50.9%) having a poor outcome. Most of these patients required significant daily care input or discharge to a nursing home. In terms of the pathology resulting from TBI, those with predominantly a contusional injury were far more likely to have a good outcome (71%) compared to those with intracranial haemorrhage (56%) or diffuse axonal injury (8%). The initial univariate analysis found an association between outcome and the type of pathology (p50.002), with excessive alcohol intake (p ¼ 0.03), with the severity of the agitational behaviour as graded on ABS (p50.001) and with the type of treatment used (p50.001). This study also found that outcome was related to the duration for which agitation was present (p50.001). These results are shown
338
R. Singh et al.
Brain Inj, 2014; 28(3): 336–340
Table I. Univariate analyses of outcome and clinical features. Outcome n Type of injury DAI Contusions Intracranial bleed Alcohol excess Behaviour level Mild Moderate Severe Drug treatment Nil One drug Change of drug Age in years; mean (SD) Duration of behaviour days; mean (SD)
Poor (27)
2 or t-test value
Good (26)
12.6, df ¼ 2, p ¼ 0.002* 13 17 23 26
12 5 10 17
1 12 13 9
10 21 22
0 9 18
10 12 4
13 17 23
1 9 17 44.8 (14.5) 58.9 (29.3)
12 8 6 38.1 (16.3) 20.4 (13.9)
4.26, df ¼ 1, p ¼ 0.04* 19.3, df ¼ 2, p ¼ 0.0001*
14.6, df ¼ 2, p ¼ 0.001*
1.58, df ¼ 51, p ¼ 0.12 5.50, df ¼ 51, p ¼ 0.0001
*Significant p value for p50.05.
Table II. Logistic regression technique (backward) of clinical features. 95% CI for Exp (B) B
SE
df
Sig
Exp (B)
Lower
– 3.046 4.330
– 1.174 1.447
– 1 1
– 0.009 0.003
– 21.03 75.92
– 2.108 4.453
Level of behaviour Mild agitation Moderate agitation Severe agitation
– 1.589 2.878
– 1.177 1.311
– 1 1
– 0.177 0.028
– 0.204 0.056
– 0.020 1.360
– 2.049 238.1
Duration of agitation Length in days
3.477
1.393
1
0.013
0.031
2.109
515.5
Type of injury Contusions Intracranial bleed Diffuse axonal injury
Upper – 209.8 1294.4
Variables entered on step 1: age, injury type, alcohol excess, severity of behaviour, drug rx, duration.
in Table I. While there was a 6-year difference in the age of individuals with good and poor outcomes, this was not statistically significant on univariate analysis (p ¼ 0.12). Using a logistic regression technique (backward step), the factors associated with outcome were further analysed, as shown in Table II. On running the analysis, the variables that remained in the equation were type of lesion on CT scan, severity of agitation and duration of behaviour. These, therefore, can be said to be the independent predictors for outcome after agitation in TBI. Factors that were significant on univariate analysis, but dropped out on the logistic regression, were alcohol excess and type of drug treatment given.
Discussion Behavioural problems are among the most disabling sequelae of TBI and are often a barrier to successful rehabilitation. The widely different criteria used to define and diagnose agitation means that reported incidence varies considerably and some of these studies do not have a clear description of how the study population was derived or exclusions made. In an inpatient population with TBI, it was found that 36.3%
manifested agitation. This compares to previous studies, almost all from the US, of rates between 10–96% [8, 10–19]. Interestingly, this rate is similar to the two other studies that were identified which used the Agitation Behaviour Score to measure agitation [8, 17]. This scale is validated in TBI and it was found to be easy to use, with accepted cut-offs for distinguishing mild, moderate and severe behavioural disturbance [21]. This study has also found that certain features, such as CT scan findings and severity of the behaviour, were associated with the outcome of agitation. This is the first study, that the authors know of, to find an association between agitation and the CT findings. While there will often be a degree of overlap in types of lesions seen on CT after TBI, it was not found difficult to attribute a predominant lesion type to each case at a weekly neuroradiology meeting and there was rarely any disagreement between neuroradiologists. Those with contusions had a better outcome than those with intracranial haemorrhage and a markedly better outcome than those with diffuse axonal injury. It is not known why this was the case, but it is speculated that the latter usually reflects more widespread brain damage, affecting more areas of the brain. It is known that CT findings can be used to classify
Agitation after traumatic brain injury
DOI: 10.3109/02699052.2013.873142
severity of TBI [23], but the finding that there is a link between the outcome of agitation and the CT findings requires further investigation. What is it in the type of lesion that determines worse outcome? Is it related to the extent of brain damage incurred or does the resultant cognitive deficit between these injuries differ in some way. The observation that functional outcome is determined by the total score on ABS is perhaps not a surprise and may have been expected intuitively. More severe scores are likely to reflect on a reduced prospect for recovery. Unfortunately the authors were not able to study the effect of severity of the brain injury itself as almost all of these patients had a severe TBI. However, one previous study has found no association to the severity of brain injury [24]. This study found no association with age of the individual, although others have demonstrated an association with younger age [9, 25]. One suggestion that may explain this is that the younger patients have more severe injury, with a wider spectrum of deficits, but this study has not looked at this. Alcohol excess, which was common in this group, was initially significant on univariate testing, but dropped out of the logistic model. Many other studies have found an association to alcohol excess [18, 26–28] or to substance abuse [28, 29]. Worse outcome was associated with a longer duration of agitation, as has been identified previously in other studies [8, 20]. This study also found that individuals treated with more than one medication had a worse outcome. This is not surprising as more severe behaviour will warrant more aggressive treatment and so, while this was significant on univariate analysis, it dropped out on logistic regression. Drug changes are simply a surrogate marker for more severe cases. It is controversial whether medication is beneficial in the treatment of agitation and a well-designed randomized control trial is required to answer this. It is important to clarify that this study was an observation of agitation and its outcome; it was not designed to compare interventions such as drugs. The question of which drugs are most effective in agitation is a controversial field and there are very few well designed RCTs to guide one. A number of other features have been assessed in other papers that were not looked at in this study. Associations have been found with epilepsy [30], depression [10], age at time of injury and frontal lobe pathology [5, 18]. Damage to the paralimbic area of the temporal lobe was also associated with agitation [31]. A higher level of later psychiatric problems has been identified in individuals with agitation [14] and there was no association to gender [17]. The main strength of the study is that it was prospective and looked at all admissions with TBI over 30 months using a standardized measure and in a systematic manner. This study also managed a 100% follow-up rate; it is known that followup in TBI populations is poor with a meta-analysis showing a rate of 42% [8, 32]. Therefore, this represents quite a substantial achievement and required active management including home visits. Patients were managed with a normal behavioural approach supplemented with medication in a typical neurorehabilitation setting. In other words, this study is relevant to normal inpatient rehabilitation practice, not just
339
highly specialized and resourced centres such as behavioural units. Furthermore, many published studies suffer from inadequate description of the population studied or exclude individuals with certain types of behaviour or decreased cognition or comprehension. Retrospective studies suffer from recollection bias or dependency on accurate documentation, e.g. of agitation behaviours and symptoms. This study has tried it’s best to overcome these deficiencies. The main weaknesses of this study are that, because this has been an inpatient population, almost all the patients were at the severe end of the TBI spectrum. The authors, therefore, have no idea of the problems faced by those with mild and moderate TBI who are usually discharged within 24 hours of attendance at Accident and Emergency. Furthermore, because of the relatively small numbers, no relevant sub-group analyses can be done. This study has also not looked in detail at different treatment approaches and has not been able to measure the length of post-traumatic amnesia or degree of cognitive impairment. The authors are starting a new study that incorporates these elements in its design, as well as the localization of intracranial pathology, but will recruit over 5 years in order to have sufficient numbers for valid conclusions. It is also hoped to determine the extent to which problems with agitation persist. It has been suggested that the incidence of such behaviours remains high at 5 years [8], although almost all of these patients had resolved their behaviour by the time of discharge. In conclusion, this study has shown that agitation is common in a TBI inpatient population and that worse outcome is associated with severity of agitation behaviour, duration of the behaviour and CT scan findings. This is the first study, that the authors know of, to have found that CT findings are related to agitation outcome.
Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References 1. Denny-Brown D. Disability arising from closed head injury. Journal of the American Medical Association 1945;127:429–436. 2. Morton MV, Wehman P. Psychosocial and emotional sequelae of individuals with traumatic brain injury: A literature review and recommendations. Brain Injury 1995;9:81–92. 3. Miller L. Major syndromes of aggressive behaviour following head injury: An introduction to evaluation and treatment. Cognitive Rehabilitation 1990;8:14–19. 4. Sandel ME, Mysiw WJ. The agitated brain injured patient. I: Definitions, differential diagnosis, and assessment. Archives of Physical Medicine & Rehabilitation 1996;77:617–623. 5. Grafman J, Schwab K, Warden D, Pridgen A, Brown HR, Salazar AM. Frontal lobe injuries, violence, and aggression: A report of the Vietnam head injury study. Neurology 1996;46:1231–1238. 6. Lombard LA, Zafonte RD. Agitation after traumatic brain injury: Considerations and treatment options. American Journal of Physical Medicine and Rehabilitation 2005;84:797–812. 7. Fugate LP, Spacek LA, Kresty LA, Levy CE, Johnson JC, Mysiw WJ. Definition of agitation following traumatic brain injury: I. A survey of the Brain Injury Special Interest Group of the American Academy of Physical Medicine and Rehabilitation. Archives of Physical Medicine & Rehabilitation 1997;78:917–923. 8. Bogner JA, Corrigan JD, Fugate L, Mysiw WJ, Clinchot D. Role of agitation in prediction of outcomes after traumatic brain injury.
340
9. 10. 11. 12. 13.
14.
15. 16. 17.
18. 19. 20.
R. Singh et al. American Journal of Physical Medicine and Rehabilitation 2001; 80:636–644. Corrigan JD, Bogner JA. Factor structure of the Agitated Behavior Scale. Journal of Clinical & Experimental Neuropsychology 1994; 16:386–392. Baguley IJ, Cooper J, Felmingham K. Aggressive behavior following traumatic brain injury: How common is common? Journal of Head Trauma Rehabilitation 2006;21:45–56. Kim SH, Manes F, Kosier T, Baruah S, Robinson RG. Irritability following traumatic brain injury. Journal of Nervous & Mental Disorders 1999;187:327–335. Rao N, Jellinek HM, Woolston DC. Agitation in closed head injury: Haloperidol effects on rehabilitation outcome. Archives of Physical Medicine & Rehabilitation 1985;66:30–34. Brooke MM, Questad KA, Patterson DR, Bashak KJ. Agitation and restlessness after closed head injury: A prospective study of 100 consecutive admissions. Archives of Physical Medicine & Rehabilitation 1992;73:320–323. Reyes RL, Bhattacharyya AK, Heller D. Traumatic head injury: Restlessness and agitation as prognosticators of physical and psychologic improvement in patients. Archives of Physical Medicine & Rehabilitation 1981;62:20–23. Maryniak O, Manchanda R, Velani A. Methotrimeprazine in the treatment of agitation in acquired brain injury patients. Brain Injury 2001;15:167–174. Levin HS, Grossman RG. Behavioral sequelae of closed head injury. Archives of Neurology 1978;35:720–727. Kadyan V, Mysiw WJ, Bogner JA, Corrigan JD, Fugate LP, Clinchot DM. Gender differences in agitation after traumatic brain injury. American Journal of Physical Medicine and Rehabilitation 2004;83:747–752. Tateno A, Jorge RE, Robinson RG. Clinical correlates of aggressive behavior after traumatic brain injury. Journal of Neuropsychiatry & Clinical Neurosciences 2003;15:155–160. Corrigan JD, Mysiw WJ, Gribble MW, Chock SKL. Agitation, cognition and attention during post-traumatic amnesia. Brain Injury 1992;6:155–160. Nott M, Chapparo C, Baguley IJ. Agitation following traumatic brain injury: An Australian sample. Brain Injury 2006;20: 1175–1182.
Brain Inj, 2014; 28(3): 336–340
21. Corrigan JD. Development of a scale for assessment of agitation following TBI. Journal of Clinical & Experimental Neuropsychology 1989;11:262–277. 22. Wilson J, Pettigrew L, Teasdale G. Structured interviews for the Glasgow Outcome Scale and the Extended Glasgow Outcome Scale: Guidance for their use. Journal of Neurotrauma 1998;15: 573–585. 23. Marshall LF, Bowers SA, Klauber MR, Van Berkum Clark M, Eisenberg HM, Jane JA. A new classification of head injury based on computerized tomography. Journal of Neurosurgery 1991;75: S14–S20. 24. Johnson R, Balleny H. Behaviour problems after brain injury: Incidence and need for treatment. Clinical Rehabilitation 1996;10: 173–181. 25. Evans RW. The postconcussion syndrome and the sequelae of mild head injury. Neurologic Clinics 1992;10:815–847. 26. Sparadeo FR, Gill D. Effects of prior alcohol use on head injury recovery. Journal of Head Trauma Rehabilitation 1989;4: 75–82. 27. Swanson JW, Holzer III CE, Ganju VK, Jono RT. Violence and psychiatric disorder in the community: Evidence from the epidemiologic catchment area surveys. Hospital & Community Psychiatry 1990;41:761–770. 28. Cassidy JW. Neuropharmacological management of destructive behaviour after TBI. Journal of Head Trauma Rehabilitation 1994; 9:43–60. 29. Fleminger S. Managing agitation and aggression after head injury: Minimum use of drugs and early care in rehabilitation units are recommended. British Medical Journal 2003;327: 4–5. 30. Galski T, Palasz J, Bruno RL, Walker JE. Predicting physical and verbal aggression on a brain trauma unit. Archives of Physical Medicine & Rehabilitation 1994;75:380–383. 31. Tonkonogy JM. Violence and temporal lobe lesion: Head CT and MRI data. Journal of Neuropsychiatry & Clinical Neurosciences 1991;3:189–196. 32. Corrigan JD, Bogner J, Mysiw WJ, Clinchot D, Fygate L. Systematic bias outcome studies of persons with traumatic brain injury. Archives of Physical Medicine & Rehabilitation 1997;78: 132–138.
Copyright of Brain Injury is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
ORIGINAL ARTICLE
Agitation after traumatic brain injury and predictors of outcome Rajiv Singh, Guru Venkateshwara, Krishnan P. S. Nair, Muhammed Khan, & Rafat Saad Osborn Neurorehabilitation Unit, Northern General Hospital, Sheffield, UK
Abstract
Keywords
Objectives: To measure the incidence of agitation after traumatic brain injury (TBI) in an inpatient population and to identify any features associated with an adverse outcome Design: Prospective study of TBI admissions over 30 months in consecutive admissions with TBI to a regional neurorehabilitation unit. Outcome of agitation was compared to patient, injury and treatment features and any associations were sought. The presence of agitation was measured by the Agitation Behaviour Score. A good outcome for agitation was defined as a return to independent living with minimal care requirement at 6 months using an Extended Glasgow Outcome Scale (GOSE) score 44. Results: Over 30 months, there were 146 TBI admissions, of whom 53 cases had agitation (36.3%). Achieving 100% follow-up, 27 (51%) had a good outcome. On a multivariable logistic regression analysis, a good outcome was associated with the type of lesions seen on CT scan, the severity of agitation and the duration of the behaviour. Alcohol excess and type of treatment used for the behaviour were initially significant on univariate testing but dropped out of the logistic regression model. Conclusions: Over a third of TBI admissions, developed agitation and poor functional outcome was associated with CT scan findings, severity and duration of agitation.
Agitation, CT scans, neuropsychology, outcome, traumatic brain injury
Agitation is a serious sequela of traumatic brain injury (TBI) and often interferes with recovery, rehabilitation and community re-integration [1–4]. It can endanger the safety of patients, family and carers [5] and there is a paucity of good evidence on which the management of this condition can be based [6]. Defining agitation has proved to be difficult. A consensus statement from the American Academy of Physical Medicine and Rehabilitation defined agitation as ‘excessive behaviour occurring within the context of an altered state of consciousness and diminished cognitive functioning’ [7] (p. 803). While this provides a good working definition, it is clear that there is a wide range of behaviours that may constitute aggression, including verbal abuse, rage, impulsivity, distractibility, sudden mood change or lack of co-operation. It is, therefore, perhaps best to consider agitation not as a single behaviour, but rather as a global construct and to classify the extent of those behaviours as mild, moderate or severe [8, 9]. The variation in description of agitation affects the interpretation and comparison of study results as well as
Correspondence: Rajiv Singh, Osborn Neurorehabilitation Unit, Northern General Hospital, Sheffield S5 7AU, UK. Tel: 0044 114 2715651. Fax: 0044 114 2715649. E-mail: [email protected]
Received 15 October 2012 Accepted 4 December 2013 Published online 27 December 2013
estimates of the frequency of agitation. As a result, these range widely from 10–96% [8, 10–19]. Furthermore, there is some uncertainty whether short- or long-term outcomes are affected in those with agitation, e.g. achievement of rehabilitation goals or discharge destination [8, 10, 20]. It is also unclear whether agitation is associated with particular demographic or treatment features and no clear picture emerges from previous studies. The goal was to provide a systematic investigation of agitation in a population with TBI (traumatic brain injury). At the same time, the authors wished to investigate whether any particular features were associated with agitation outcome. The aim was to make the study directly relevant to normal neurorehabilitation practice for most clinicians and not overly specialized with highly complex assessments or major resource requirements, e.g. behavioural units or very specialized centres. Hence, a prospective study was designed of all admissions with TBI to a regional rehabilitation centre treated with standard behaviour modification and pharmacological intervention if required. A literature search on MEDLINE and PsycINFO found that all previous studies were from the US and one from Australia; there were no prospective studies from the UK. While there is no particular reason to think that the incidence would differ markedly from other countries, it was thought that it would be interesting to document incidence in the context of a UK population as well as investigate whether any patient or clinical features were associated with the long-term outcome.
20 14
Introduction
History
Agitation after traumatic brain injury
DOI: 10.3109/02699052.2013.873142
Methods Consecutive inpatients admitted to a Neurorehabilitation unit at a University Teaching Hospital with traumatic brain injury (TBI) from January 2007 to June 2009, were assessed for agitation during their stay. Only those with a primary diagnosis of TBI were included and all patients admitted to the unit were over 18. Demographic and injury characteristics were recorded on admission. Agitation was identified and graded by using the Agitation Behavioural Scale [9, 21]. Cases were identified if two or more of the behaviours on this scale were exhibited by an individual over any 24-hour period. If such behaviours were observed by staff, then subsequent daily assessment by neuropsychologists over a 30-minute session was used to assess the severity of agitation by scoring the behaviours observed. This was made by the same observer as often as possible but required nursing or medical staff to make the assessments at weekends after appropriate training. Daily observation ensured ongoing recording of all agitation behaviours rather than depending on retrospective notes reading. Mild agitation is diagnosed between scores of 21–28, moderate from 29–35 and severe agitation above 35. Agitation was resolved when the score was less than 21 on two consecutive days. Severity of brain injury was graded by Glasgow Coma Score in the first 24 hours, but, as this was an inpatient study, almost all admissions to the unit fall into the group of severe injury. This study also examined the predominant type of brain lesion caused by the injury. Using CT scans, reported at a weekly neuroradiology department meeting, the authors were able to classify injuries into three groups; those whose predominant lesion was diffuse axonal injury, those with predominantly brain contusion injuries and those predominantly with intra-cranial haemorrhage. While it is accepted that a certain overlap will exist in some patients, it was, in fact, quite straightforward for the neuroradiology meeting to decide which pathology was most prominent on the scans and no major disagreements took place. It was felt that this might give a more useful item to study rather than simply the severity of injury. The pharmacological treatment received by patients was also examined. There was no attempt made to modify pharmacological treatment for the study, which was left to the treating team if they judged appropriate. All patients received behavioural approaches aimed at identifying antecedents and triggers, environmental modification where appropriate, constant support including family and use of positive and negative reinforcements aimed at modifying the resultant behaviours. Pharmacological management was divided into three groups, with the first group consisting of patients who received no drugs. The second group received only one agent, e.g. beta-blockers, benzodiazepines or mood stabilizers such as carbamazepine. The third group required change of the first medication to another agent as the first was not considered effective. There may have been a period of overlap as one drug was ‘washed out’, but no patient received more than one behavioural drug at any time. This study compared these three groups. There was no attempt to compare different drugs.
337
Rehabilitation outcome was assessed by the Extended Glasgow Outcome Scale (GOSE) at 6 months post-injury. In those instances where the patient could not complete the questions, a carer was used. A cut-off above 4 (moderate lower or better) was considered a good outcome and represents those who are capable of independent living [22]. Outcome was, therefore, a binary variable. Most patients attended clinic follow-up; a few required a visit to their home or nursing home. Local ethics committee approval was taken. Initial univariate analysis was carried out by 2 test and t-test to examine the relationship between outcome and the dependant variables, with statistically significant difference set at p50.05. These features were; age, severity of agitation, duration of agitation, CT scan findings, gender, alcohol excess and pharmacological treatment given. Subsequent multivariable logistic regression with backward step entry was applied; starting with a full model of all variables, a stepwise analysis was used to remove non-significant variables one at a time to determine the independent predictors of outcome. The criteria for entry or removal of a variable were 5% for inclusion and 5% for exclusion. All analyses were performed using SPSS version 16.0.
Results Over a 30-month period, there was a total of 146 traumatic brain injury (TBI) inpatient admissions, of whom 53 cases manifested some degree of agitation. This corresponds to a frequency of 36.3%. Severe TBI made up 79% of agitation cases and 88% of all admissions to the unit. Average age of the agitation group was 41.2 years (SD ¼ 15.8) and of the whole group 45.4 years (SD ¼ 21.0). Average duration of agitation was 39.0 days (SD ¼ 27.7). In terms of type of injury seen on CT scan, 13 (24.5%) individuals had diffuse axonal injury, 17 (32.1%) had contusional injury and 23 (43.4%) had intracranial haemorrhage as their predominant pathology. In 26 (49%) cases there was a history of alcohol excess. Using the ABS, mild agitation made up 10 (18.9%) of the group, moderate was 21 (39.6%) and severe was 22 (41.5%) of the group. This study looked at how many patients had received medication and found that 13 (24.5%) received no behavioural medication, 17 (32.1%) received one agent only and 23 (43.4%) required change of medication to an alternative agent. Follow-up was achieved for all 53 cases. Outcome was a binary variable. Using a score of GOSE44 to determine good outcome, patients divided almost equally, with 27 (50.9%) having a poor outcome. Most of these patients required significant daily care input or discharge to a nursing home. In terms of the pathology resulting from TBI, those with predominantly a contusional injury were far more likely to have a good outcome (71%) compared to those with intracranial haemorrhage (56%) or diffuse axonal injury (8%). The initial univariate analysis found an association between outcome and the type of pathology (p50.002), with excessive alcohol intake (p ¼ 0.03), with the severity of the agitational behaviour as graded on ABS (p50.001) and with the type of treatment used (p50.001). This study also found that outcome was related to the duration for which agitation was present (p50.001). These results are shown
338
R. Singh et al.
Brain Inj, 2014; 28(3): 336–340
Table I. Univariate analyses of outcome and clinical features. Outcome n Type of injury DAI Contusions Intracranial bleed Alcohol excess Behaviour level Mild Moderate Severe Drug treatment Nil One drug Change of drug Age in years; mean (SD) Duration of behaviour days; mean (SD)
Poor (27)
2 or t-test value
Good (26)
12.6, df ¼ 2, p ¼ 0.002* 13 17 23 26
12 5 10 17
1 12 13 9
10 21 22
0 9 18
10 12 4
13 17 23
1 9 17 44.8 (14.5) 58.9 (29.3)
12 8 6 38.1 (16.3) 20.4 (13.9)
4.26, df ¼ 1, p ¼ 0.04* 19.3, df ¼ 2, p ¼ 0.0001*
14.6, df ¼ 2, p ¼ 0.001*
1.58, df ¼ 51, p ¼ 0.12 5.50, df ¼ 51, p ¼ 0.0001
*Significant p value for p50.05.
Table II. Logistic regression technique (backward) of clinical features. 95% CI for Exp (B) B
SE
df
Sig
Exp (B)
Lower
– 3.046 4.330
– 1.174 1.447
– 1 1
– 0.009 0.003
– 21.03 75.92
– 2.108 4.453
Level of behaviour Mild agitation Moderate agitation Severe agitation
– 1.589 2.878
– 1.177 1.311
– 1 1
– 0.177 0.028
– 0.204 0.056
– 0.020 1.360
– 2.049 238.1
Duration of agitation Length in days
3.477
1.393
1
0.013
0.031
2.109
515.5
Type of injury Contusions Intracranial bleed Diffuse axonal injury
Upper – 209.8 1294.4
Variables entered on step 1: age, injury type, alcohol excess, severity of behaviour, drug rx, duration.
in Table I. While there was a 6-year difference in the age of individuals with good and poor outcomes, this was not statistically significant on univariate analysis (p ¼ 0.12). Using a logistic regression technique (backward step), the factors associated with outcome were further analysed, as shown in Table II. On running the analysis, the variables that remained in the equation were type of lesion on CT scan, severity of agitation and duration of behaviour. These, therefore, can be said to be the independent predictors for outcome after agitation in TBI. Factors that were significant on univariate analysis, but dropped out on the logistic regression, were alcohol excess and type of drug treatment given.
Discussion Behavioural problems are among the most disabling sequelae of TBI and are often a barrier to successful rehabilitation. The widely different criteria used to define and diagnose agitation means that reported incidence varies considerably and some of these studies do not have a clear description of how the study population was derived or exclusions made. In an inpatient population with TBI, it was found that 36.3%
manifested agitation. This compares to previous studies, almost all from the US, of rates between 10–96% [8, 10–19]. Interestingly, this rate is similar to the two other studies that were identified which used the Agitation Behaviour Score to measure agitation [8, 17]. This scale is validated in TBI and it was found to be easy to use, with accepted cut-offs for distinguishing mild, moderate and severe behavioural disturbance [21]. This study has also found that certain features, such as CT scan findings and severity of the behaviour, were associated with the outcome of agitation. This is the first study, that the authors know of, to find an association between agitation and the CT findings. While there will often be a degree of overlap in types of lesions seen on CT after TBI, it was not found difficult to attribute a predominant lesion type to each case at a weekly neuroradiology meeting and there was rarely any disagreement between neuroradiologists. Those with contusions had a better outcome than those with intracranial haemorrhage and a markedly better outcome than those with diffuse axonal injury. It is not known why this was the case, but it is speculated that the latter usually reflects more widespread brain damage, affecting more areas of the brain. It is known that CT findings can be used to classify
Agitation after traumatic brain injury
DOI: 10.3109/02699052.2013.873142
severity of TBI [23], but the finding that there is a link between the outcome of agitation and the CT findings requires further investigation. What is it in the type of lesion that determines worse outcome? Is it related to the extent of brain damage incurred or does the resultant cognitive deficit between these injuries differ in some way. The observation that functional outcome is determined by the total score on ABS is perhaps not a surprise and may have been expected intuitively. More severe scores are likely to reflect on a reduced prospect for recovery. Unfortunately the authors were not able to study the effect of severity of the brain injury itself as almost all of these patients had a severe TBI. However, one previous study has found no association to the severity of brain injury [24]. This study found no association with age of the individual, although others have demonstrated an association with younger age [9, 25]. One suggestion that may explain this is that the younger patients have more severe injury, with a wider spectrum of deficits, but this study has not looked at this. Alcohol excess, which was common in this group, was initially significant on univariate testing, but dropped out of the logistic model. Many other studies have found an association to alcohol excess [18, 26–28] or to substance abuse [28, 29]. Worse outcome was associated with a longer duration of agitation, as has been identified previously in other studies [8, 20]. This study also found that individuals treated with more than one medication had a worse outcome. This is not surprising as more severe behaviour will warrant more aggressive treatment and so, while this was significant on univariate analysis, it dropped out on logistic regression. Drug changes are simply a surrogate marker for more severe cases. It is controversial whether medication is beneficial in the treatment of agitation and a well-designed randomized control trial is required to answer this. It is important to clarify that this study was an observation of agitation and its outcome; it was not designed to compare interventions such as drugs. The question of which drugs are most effective in agitation is a controversial field and there are very few well designed RCTs to guide one. A number of other features have been assessed in other papers that were not looked at in this study. Associations have been found with epilepsy [30], depression [10], age at time of injury and frontal lobe pathology [5, 18]. Damage to the paralimbic area of the temporal lobe was also associated with agitation [31]. A higher level of later psychiatric problems has been identified in individuals with agitation [14] and there was no association to gender [17]. The main strength of the study is that it was prospective and looked at all admissions with TBI over 30 months using a standardized measure and in a systematic manner. This study also managed a 100% follow-up rate; it is known that followup in TBI populations is poor with a meta-analysis showing a rate of 42% [8, 32]. Therefore, this represents quite a substantial achievement and required active management including home visits. Patients were managed with a normal behavioural approach supplemented with medication in a typical neurorehabilitation setting. In other words, this study is relevant to normal inpatient rehabilitation practice, not just
339
highly specialized and resourced centres such as behavioural units. Furthermore, many published studies suffer from inadequate description of the population studied or exclude individuals with certain types of behaviour or decreased cognition or comprehension. Retrospective studies suffer from recollection bias or dependency on accurate documentation, e.g. of agitation behaviours and symptoms. This study has tried it’s best to overcome these deficiencies. The main weaknesses of this study are that, because this has been an inpatient population, almost all the patients were at the severe end of the TBI spectrum. The authors, therefore, have no idea of the problems faced by those with mild and moderate TBI who are usually discharged within 24 hours of attendance at Accident and Emergency. Furthermore, because of the relatively small numbers, no relevant sub-group analyses can be done. This study has also not looked in detail at different treatment approaches and has not been able to measure the length of post-traumatic amnesia or degree of cognitive impairment. The authors are starting a new study that incorporates these elements in its design, as well as the localization of intracranial pathology, but will recruit over 5 years in order to have sufficient numbers for valid conclusions. It is also hoped to determine the extent to which problems with agitation persist. It has been suggested that the incidence of such behaviours remains high at 5 years [8], although almost all of these patients had resolved their behaviour by the time of discharge. In conclusion, this study has shown that agitation is common in a TBI inpatient population and that worse outcome is associated with severity of agitation behaviour, duration of the behaviour and CT scan findings. This is the first study, that the authors know of, to have found that CT findings are related to agitation outcome.
Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of the paper.
References 1. Denny-Brown D. Disability arising from closed head injury. Journal of the American Medical Association 1945;127:429–436. 2. Morton MV, Wehman P. Psychosocial and emotional sequelae of individuals with traumatic brain injury: A literature review and recommendations. Brain Injury 1995;9:81–92. 3. Miller L. Major syndromes of aggressive behaviour following head injury: An introduction to evaluation and treatment. Cognitive Rehabilitation 1990;8:14–19. 4. Sandel ME, Mysiw WJ. The agitated brain injured patient. I: Definitions, differential diagnosis, and assessment. Archives of Physical Medicine & Rehabilitation 1996;77:617–623. 5. Grafman J, Schwab K, Warden D, Pridgen A, Brown HR, Salazar AM. Frontal lobe injuries, violence, and aggression: A report of the Vietnam head injury study. Neurology 1996;46:1231–1238. 6. Lombard LA, Zafonte RD. Agitation after traumatic brain injury: Considerations and treatment options. American Journal of Physical Medicine and Rehabilitation 2005;84:797–812. 7. Fugate LP, Spacek LA, Kresty LA, Levy CE, Johnson JC, Mysiw WJ. Definition of agitation following traumatic brain injury: I. A survey of the Brain Injury Special Interest Group of the American Academy of Physical Medicine and Rehabilitation. Archives of Physical Medicine & Rehabilitation 1997;78:917–923. 8. Bogner JA, Corrigan JD, Fugate L, Mysiw WJ, Clinchot D. Role of agitation in prediction of outcomes after traumatic brain injury.
340
9. 10. 11. 12. 13.
14.
15. 16. 17.
18. 19. 20.
R. Singh et al. American Journal of Physical Medicine and Rehabilitation 2001; 80:636–644. Corrigan JD, Bogner JA. Factor structure of the Agitated Behavior Scale. Journal of Clinical & Experimental Neuropsychology 1994; 16:386–392. Baguley IJ, Cooper J, Felmingham K. Aggressive behavior following traumatic brain injury: How common is common? Journal of Head Trauma Rehabilitation 2006;21:45–56. Kim SH, Manes F, Kosier T, Baruah S, Robinson RG. Irritability following traumatic brain injury. Journal of Nervous & Mental Disorders 1999;187:327–335. Rao N, Jellinek HM, Woolston DC. Agitation in closed head injury: Haloperidol effects on rehabilitation outcome. Archives of Physical Medicine & Rehabilitation 1985;66:30–34. Brooke MM, Questad KA, Patterson DR, Bashak KJ. Agitation and restlessness after closed head injury: A prospective study of 100 consecutive admissions. Archives of Physical Medicine & Rehabilitation 1992;73:320–323. Reyes RL, Bhattacharyya AK, Heller D. Traumatic head injury: Restlessness and agitation as prognosticators of physical and psychologic improvement in patients. Archives of Physical Medicine & Rehabilitation 1981;62:20–23. Maryniak O, Manchanda R, Velani A. Methotrimeprazine in the treatment of agitation in acquired brain injury patients. Brain Injury 2001;15:167–174. Levin HS, Grossman RG. Behavioral sequelae of closed head injury. Archives of Neurology 1978;35:720–727. Kadyan V, Mysiw WJ, Bogner JA, Corrigan JD, Fugate LP, Clinchot DM. Gender differences in agitation after traumatic brain injury. American Journal of Physical Medicine and Rehabilitation 2004;83:747–752. Tateno A, Jorge RE, Robinson RG. Clinical correlates of aggressive behavior after traumatic brain injury. Journal of Neuropsychiatry & Clinical Neurosciences 2003;15:155–160. Corrigan JD, Mysiw WJ, Gribble MW, Chock SKL. Agitation, cognition and attention during post-traumatic amnesia. Brain Injury 1992;6:155–160. Nott M, Chapparo C, Baguley IJ. Agitation following traumatic brain injury: An Australian sample. Brain Injury 2006;20: 1175–1182.
Brain Inj, 2014; 28(3): 336–340
21. Corrigan JD. Development of a scale for assessment of agitation following TBI. Journal of Clinical & Experimental Neuropsychology 1989;11:262–277. 22. Wilson J, Pettigrew L, Teasdale G. Structured interviews for the Glasgow Outcome Scale and the Extended Glasgow Outcome Scale: Guidance for their use. Journal of Neurotrauma 1998;15: 573–585. 23. Marshall LF, Bowers SA, Klauber MR, Van Berkum Clark M, Eisenberg HM, Jane JA. A new classification of head injury based on computerized tomography. Journal of Neurosurgery 1991;75: S14–S20. 24. Johnson R, Balleny H. Behaviour problems after brain injury: Incidence and need for treatment. Clinical Rehabilitation 1996;10: 173–181. 25. Evans RW. The postconcussion syndrome and the sequelae of mild head injury. Neurologic Clinics 1992;10:815–847. 26. Sparadeo FR, Gill D. Effects of prior alcohol use on head injury recovery. Journal of Head Trauma Rehabilitation 1989;4: 75–82. 27. Swanson JW, Holzer III CE, Ganju VK, Jono RT. Violence and psychiatric disorder in the community: Evidence from the epidemiologic catchment area surveys. Hospital & Community Psychiatry 1990;41:761–770. 28. Cassidy JW. Neuropharmacological management of destructive behaviour after TBI. Journal of Head Trauma Rehabilitation 1994; 9:43–60. 29. Fleminger S. Managing agitation and aggression after head injury: Minimum use of drugs and early care in rehabilitation units are recommended. British Medical Journal 2003;327: 4–5. 30. Galski T, Palasz J, Bruno RL, Walker JE. Predicting physical and verbal aggression on a brain trauma unit. Archives of Physical Medicine & Rehabilitation 1994;75:380–383. 31. Tonkonogy JM. Violence and temporal lobe lesion: Head CT and MRI data. Journal of Neuropsychiatry & Clinical Neurosciences 1991;3:189–196. 32. Corrigan JD, Bogner J, Mysiw WJ, Clinchot D, Fygate L. Systematic bias outcome studies of persons with traumatic brain injury. Archives of Physical Medicine & Rehabilitation 1997;78: 132–138.
Copyright of Brain Injury is the property of Taylor & Francis Ltd and its content may not be copied or emailed to multiple sites or posted to a listserv without the copyright holder's express written permission. However, users may print, download, or email articles for individual use.
