Brain injury survivors are a growing clinical group, and most return to the community. Rehabilitation teams rarely include psychiatrists, and a wide range of behavioral and mood dyscontrol symptoms simply are labeled "agitation." Although behavioral interventions may be used, medication management is prevalent, often with benzodiazepines or neuroleptics. Benzodiazepine use has significant drawbacks, such as sedation, disinhibition, and impaired coordination. Neuroleptic use, especially with typical agents, can present risk of extrapyramidal symptoms, restlessness, and tardive dyskinesia in vulnerable individuals.
Effective medication strategies are necessary for achieving the goal of an alert but calm patient.1 Prior case series described propranolol2 and carbamazepine3 for combativeness following brain injury. Preceded by a short case series,4 the present report is the largest case series detailing the use of divalproex for agitation symptoms following brain injury.
A retrospective chart review (by P.E.S., a psychiatrist) was undertaken of all acute brain injury rehabilitation unit inpatients who received divalproex from December 1996 through September 1998, selecting those who received divalproex for "agitation" symptoms. Excluded were those who took divalproex for any indication prior to injury or received divalproex only for seizure prophylaxis after injury. All resulting charts were reviewed (N=29). Abstracted chart data included demographics, mechanism of injury, initial head computed tomography (CT) results, comments on specific agitation symptoms contained in clinical notes, psychotropic medications administered during acute and rehabilitation care, length of stay, discharge site, and any psychiatric consultation.
The patient cohort (N=29) had been on the inpatient brain injury rehabilitation unit, which had one attending physiatrist (D.N.K.). All patients had received benzodiazepines with or without a neuroleptic, resulting in unsuccessful control of agitation symptoms prior to starting a divalproex titration. Age ranged from 13 to 89 years (mean 48.2). Patients had been transferred from acute care from 4 to 180 days (mean 28.6) after brain injury, with the subsequent inpatient rehabilitation unit length of stay ranging from 11 to 151 days (mean 46.4). For the 8 patients for whom this measure was used, Agitated Behavior Scale5 scores ranged in the 30s and 40s (indicating severe agitation) prior to starting divalproex and the 20s (indicating minimal agitation) when symptoms abated.
The initial brain CT scans of 28 of the 29 patients showed significant injuries, often in multiple sites. Of the motor vehicle event patients (n=14), 3 showed bifrontal hemorrhagic contusions, 6 unilateral frontal hemorrhagic contusions or subdural hematomas, 3 nonfrontal hemorrhagic contusions, 1 multiple bilateral orbit and facial fractures, and 1 unspecified brain injury. Of the fall patients (n=7), 3 showed bifrontal and 4 unilateral frontal hemorrhagic contusion or acute subdural hematoma. Intracranial hemorrhage patients (n=5) showed 2 frontal and 3 other site parenchymal hemorrhages. Scans of the 1 blunt trauma patient showed a left temporal hematoma; the 1 vasculitis patient showed global patchy lucencies; and the 1 gunshot wound patient showed right temporal damage. Only the 1 anoxia patient showed no CT scan brain abnormalities.
Chart review revealed a wide range of symptoms used interchangeably with "agitation": easily aggravated, argumentative, escalating temper, insistent, frustrated, labile, biting, combative, profane, punching, pushing, striking out, yelling; disinhibited, distractible, impulsive, pacing, perseverating, restless; pulling at tubes, removing braces, resisting nursing care; insomnia; sexually preoccupied. All patients had at least three of these symptoms commented upon throughout progress and therapy notes while treated with other psychotropics prior to divalproex. In the retrospective review, overall clinical descriptions of each member of this cohort were compatible with a diagnosis of mood or personality syndrome due to brain injury. Five patients had received a formal psychiatric consultation (by P.E.S. and others) as part of their admission and had formally received one of these diagnoses.
For a subgroup of 18 patients (62%), progress notes included comments on decreased, significantly improved, or resolved agitation symptoms within 7 days after achieving a mean daily dose of 1,257 mg. Final doses in this subgroup were distributed as follows: 250 mg for 1 patient, 750 mg for 1 patient, 1,000 mg for 5 patients, 1,125 mg for 1 patient, 1,250 mg for 2 patients, 1,500 mg for 6 patients, and 2,000 mg for 2 patients. Doses were titrated to effectiveness and adjusted for side effects of sedation or tremor rather than blood level. None of these 18 patients were noted to appear sedated at their final dose. No excessive weight gain received comment. Although all 18 had received a benzodiazepine and/or a neuroleptic before starting divalproex, after achieving the above divalproex doses only 2 of these 18 patients received any subsequent lorazepam for anxiety, and 4 patients still required a low-dose neuroleptic for paranoia or sundowning delirium. After, and in addition to, divalproex, 3 patients began an antidepressant for new depressive symptoms. No mechanism of injury or brain injury site correlated with this improvement subgroup of 18 patients.
The response of a separate subgroup of 8 patients (28%) showed a rapid resolution of agitation symptoms to near total recovery with a mean daily dose of 714 mg and no other psychotropic medication. Because of such a rapid recovery at a low dose, their divalproex was discontinued, and their agitation symptoms did not recur. No mechanism of injury or brain injury site correlated with this improvement group.
Three patients did not fit into the two improvement subgroups. One patient's agitation showed no response to several weeks at the maximum tolerated dose of 1,500 mg. Divalproex worsened the existing lethargy of 2 patients (6.9%) and was discontinued.
Most patients went to their homes (n=23) or to other community sites (n=4). For all of these patients, their brain injury history was not apparent in casual conversation, although deficiencies were apparent in mental status examinations, cognitive testing, and neuropsychological batteries. The subgroup of agitated but lethargic patients (n=2) were intolerant of all medication attempts and eventually moved to a skilled nursing facility.
Although this report has all the limitations associated with a retrospective chart review of naturalistic clinical practice, its large cohort gives important guidance for subsequent clinical and research pursuits. There are no medications indicated for agitation symptoms related to brain injury or other neurologic conditions. For agitation associated with geriatric dementia, treatment guidelines have emerged from case series and consensus reports6 and from studies of carbamazepine,7 valproate,8—10 propranolol titration, and others.11 In contrast to carbamazepine and propranolol, divalproex does have a U.S. Food and Drug Administration (FDA) indication for bipolar mania, which appears similar to the agitation symptoms in this cohort with brain injury. Divalproex also has FDA indications for seizure disorder and migraine prophylaxis, both commonly encountered in the brain injury population.
After acute injury, unlike mania or dementia, brain neuronal healing can occur; however, the cohort in this report required treatment for agitation symptoms in a rehabilitation setting approximately two months after injury, when the rate of recovery has slowed. Divalproex may provide valuable symptomatic relief during a specific period of brain healing (as it apparently did for 28% of this cohort), or it may be maintenance therapy (as in up to 62% of this cohort).
This case series of 29 patients with agitation symptoms following relatively severe brain injury responded to divalproex doses similar to those in conventional psychiatric practice.12 Until open-label or controlled studies are published, this study provides guidance for treatment of patients with symptoms compatible with a diagnosis of mood (mania) or personality disorder due to brain injury.
Abbott Laboratories provided financial support only after the retrospective record review was designed. Only the retrospective chart review was funded by Abbott Laboratories.
This work was presented in part at the 1999 annual meeting of the American Psychiatric Association and the 1999 annual meeting of the Association of Academic Physiatrists.