CHPR: Tell us a little about yourself.

Dr. Kelly: I’m a neurologist, and I’ve spent much of my career focused on traumatic brain injury (TBI). Early on, neurologists were not very involved in TBI care. Severe cases tended to move from neurosurgery to rehabilitation medicine, and neurologists often stayed on the sidelines. I saw that as a major gap in the practice of neurology. Over time, though, neurology became more engaged, especially through work related to sports concussions. I was the neurologist for the Chicago Bears for 10 years, and our work helped shape the concussion protocols that are now used in professional sports. Later, during the wars in Iraq and Afghanistan, the sports concussion community was asked to help the military understand and treat blast-related concussion and TBI. That led to my work at Walter Reed, where I directed an interdisciplinary program for service members with TBI and related conditions. After that, I helped establish a similar program at the University of Colorado for veterans.

CHPR: How did those experiences shape the way you think about TBI care?

Dr. Kelly: One of the clearest lessons was that TBI care works best when it is interdisciplinary and individualized. Before you can individualize care, though, you need to understand the injury itself. In trauma cases, brain injuries can get pushed into the background because the team often is dealing with orthopedic injuries, abdominal surgeries, or other emergencies. But for recovery from the brain injury, the details matter enormously. Was the person knocked unconscious? If so, for how long? If not, were they dazed, confused, or amnestic? What is the gap in memory? How did their neurologic recovery evolve afterward? Was there a fall, a motor vehicle crash, a blast exposure, an assault? I sometimes say you should approach the history like a legal case, because you want a high level of rigor. The history is the foundation for diagnosis, prognosis, and treatment planning.

“Sleep is not just symptom control after TBI. Restorative sleep is part of brain recovery. Before escalating medications for agitation, I will always ask: Is this patient sleeping?” - James P. Kelly, MD

CHPR: Once you have that history, how do you use it clinically?

Dr. Kelly: You compare who the person is now with who they were before the injury. Since you usually don’t meet these patients until after the catastrophe, you need input from family members, close friends, or colleagues. What was their baseline personality and level of independence? How did they function at work, at home, and socially? Were they already irritable, anxious, impulsive, depressed, or cognitively limited? Then you ask: What has changed? Are you seeing new deficits, an exaggeration of prior traits, or both?

CHPR: Do capacity questions come up often after TBI?

Dr. Kelly: Yes, they come up quite often. The standard is the same as with any other patient: Can the person understand the relevant information, appreciate the risks and benefits, reason through the options, and communicate a choice? But after TBI, you have to be especially careful because patients may sound articulate while still having impaired insight, poor judgment, impulsivity, or trouble anticipating consequences.

CHPR: What behavioral or psychiatric symptoms do you commonly see after TBI?

Dr. Kelly: One common pattern is that the patient becomes a caricature of the person they were before. Salient personality features become exaggerated. This is sometimes benign (eg, a cantankerous uncle becoming more cantankerous), but not always (eg, a person who was mildly impulsive becoming much more impulsive). The most commonly injured part of the brain in TBI is the frontal lobe. The frontal lobes help us plan, organize, and execute behavior—but they also help us inhibit behavior. When the frontal lobes are injured, the patient may have less inhibitory control, less organization, and more unpredictability. That’s why hospital teams often see irritability, impulsivity, aggression, poor judgment, and poor frustration tolerance. When you see those symptoms, it helps to ask whether they are truly new or whether the injury has amplified traits that were already there.

CHPR: Are there any clues that help distinguish TBI-related symptoms from primary psychiatric symptoms?

Dr. Kelly: I would start by looking for signs that point to a neurologic injury. Headaches are common and nonspecific, but persistent headaches after TBI are important. Eye movement abnormalities, double vision, balance problems, dizziness, and cranial nerve findings are much harder to attribute to a primary psychiatric disorder. For example, a patient with double vision, impaired eye movements, or severe disequilibrium may look frightened, avoidant, or uncooperative, but the problem may be neurologic. A detailed neurological evaluation can help identify what is neurologic, what is psychiatric, and what is overlapping. Many symptoms, like sleep disturbance, agitation, poor concentration, anxiety, and depression, can come from more than one source.

CHPR: Speaking about sleep disturbance, I’ve noticed that patients with TBI frequently experience problems sleeping. Is this a common finding?

Dr. Kelly: It’s very common and enormously important. One of the simplest and most useful interventions is to regulate the sleep-wake cycle. When someone is recovering from TBI, day and night may not make sense to the injured brain. The patient may be awake at night and unavailable for therapies during the day because they are exhausted. If we can help the patient sleep at night, they’ll be more available for rehabilitation and medical care during the day. I learned early from psychiatry colleagues that low-dose trazodone at bedtime (eg, 50 mg at night) can be very helpful for sleep maintenance and deeper, more restful sleep. That can be useful across the spectrum of TBI, including mild TBI, where patients often fall asleep from exhaustion and then wake repeatedly through the night. Sometimes a patient also needs a short-acting sedating medication to initiate sleep, like a low dose of temazepam. We’re not looking to sedate the patient indefinitely. The goal is to re-establish a normal sleep-wake cycle, use the mildest effective medication, and taper as soon as feasible. Restorative sleep is part of brain recovery after TBI. Before escalating medications for agitation, I will always ask: Is this patient sleeping?

CHPR: If the patient is sleeping reasonably well but remains impulsive or aggressive, what do you do next?

Dr. Kelly: Valproic acid and carbamazepine can be useful. These medications were not originally developed for post-TBI impulsivity, but clinicians noticed that they could blunt the jumpiness, irritability, and aggressive outbursts seen after TBI (Plantier D and Luauté J, Ann Phys Rehabil Med 2016;59(1):42–57). Patients sometimes describe these effects very clearly. They say, “I still feel upset. I still have the same emotions. I just don’t act on them.” That can be very helpful for families, because the patient has a little more control. But these are not benign medications. With valproic acid, watch platelet counts and liver function. With carbamazepine, watch for hyponatremia. There is some evidence that patients with TBI and alcohol misuse may benefit particularly from valproic acid, but the key is to use these medications thoughtfully (Beresford TP et al, Neuropsychiatry Clin Neurosci 2022;34(3):224–232). They can be especially helpful when the patients’ impulsivity is disrupting their family life or safety. But I would avoid using medication as the first response to every behavioral problem, especially if the environment can be structured and the behavior can be managed without them.

CHPR: Sometimes patients appear to develop new-onset psychosis after a TBI. How do you approach these cases?

Dr. Kelly: It can happen, and in fact it’s been described in the literature as psychotic disorder due to traumatic brain injury (PD-TBI) (Fujii D and Fujii DC, J Neuropsychiatry Clin Neurosci 2012;24(3):278–289). But I‘d be cautious about making that diagnosis too quickly. In reported cases, post-TBI psychosis tends to look more like positive symptoms, such as persecutory delusions and auditory hallucinations, rather than the negative symptoms we often associate with schizophrenia. It’s also often associated with frontal or temporal abnormalities, cognitive impairment, or seizure history. So, when psychosis appears after TBI, I want to know the timing, the injury pattern, the neurologic findings, and whether the person had any premorbid vulnerability. In many of these cases, I suspect the person already had some vulnerability. The injury may unmask or precipitate symptoms, especially in the age range (late teens, early 20s) when a first psychotic episode might have occurred anyway. Frontal lobe injury reduces the brain’s ability to inhibit and organize behavior, so limbic or psychiatric-looking symptoms can become more visible.

CHPR: How do you manage patients who are paranoid, frightened, or behaving unsafely after TBI?

Dr. Kelly: First, try to step back and ask what is driving the behavior. Is this delirium? Post-traumatic amnesia? Sleep deprivation? Pain? Frontal disinhibition? A medication effect? Substance withdrawal? Or is this a true psychotic disorder emerging after the injury? These are very different problems, and if you treat them all as “psychosis,” you can miss things that are reversible. The first steps are often environmental: Reduce overstimulation, keep the room calm, reorient the patient, and bring in familiar people when possible. Nursing and rehabilitation staff are often essential because they can tell you what triggers the behavior and what helps settle the patient. Also, normalize sleep, treat pain, and review medications that may be contributing to confusion, agitation, or disinhibition, like benzodiazepines and anticholinergic medications. Many behavioral problems improve when those issues are addressed.

CHPR: And if the patient is still dangerous or severely distressed?

Dr. Kelly: Then medication may be necessary. In severe TBI, I use antipsychotics carefully because some evidence suggests they can prolong post-traumatic amnesia and delay cognitive recovery (Williamson D et al, BMJ Open 2019;9(7):e029604). Clinically, I also worry that they can interfere with rehabilitation by causing extrapyramidal symptoms, reducing arousal, and making it harder for the patient to participate in therapy. When an antipsychotic is needed for safety, I use the lowest effective dose and reassess frequently. It’s important to avoid letting the medication become a default long-term treatment simply because the patient was difficult early in recovery.

CHPR: What about depression after TBI? Is this something we should watch for?

Dr. Kelly: Depression is one of the most common psychiatric conditions across the TBI spectrum. Neurologically, patients may have headaches, dizziness, eye movement problems, and cognitive deficits. Psychiatrically, depression is at the top of the list. Sometimes depression is a reaction to deficits, but sometimes it looks like a true mood disorder. Either way, it can worsen recovery and increase suicide risk. In military populations, depression appears to be a major driver of suicide after TBI (Brenner LA et al, JAMA Network Open 2023;6(7):e2326296), sometimes even more than PTSD. Depression can be easy to miss because fatigue, poor concentration, sleep disruption, and irritability may be attributed to the brain injury itself. I also ask families what they are seeing at home. Is the person withdrawing? Giving up on rehab? Saying that they’re a burden or that they’ll never get better? Those comments deserve direct suicide screening.

CHPR: Once you’ve identified depression after TBI, what’s your approach?

Dr. Kelly: The good news is that depression after TBI is treatable. Standard antidepressant approaches can help. But I need to re-emphasize the importance of addressing sleep. If sleep is poor, mood will often be worse. Improving restorative sleep may lift a significant part of the mood syndrome. Pain, cognitive frustration, and functional losses may also be sustaining the depression. So, you have to treat more than the mood symptoms: Address pain, help the person function around cognitive deficits, bring in rehabilitation therapies, and help the family understand what recovery may realistically look like. I warn families to watch for worsening depression, suicidal thoughts, escalating irritability or aggression, substance use, giving away possessions, unsafe behavior, or a sudden refusal to participate in care. I frame it as part of recovery monitoring, not as something they are expected to manage alone.

CHPR: What do you say to patients and families about the recovery process?

Dr. Kelly: I tell families that there is no standard timeline for recovery after TBI, nor is recovery usually a straight line. It depends on the injury, of course, but also on things like sleep, pain, baseline functioning, substance use, and ability to consistently engage in rehabilitation. Families often ask, “When will they be back to normal?” I usually say, “We’re looking for gradual gains over weeks to months. But setbacks are common, especially if the person is not sleeping, is in pain, or is getting overstimulated.” I also give families concrete things to watch for at home: worsening depression, suicidal thoughts, escalating agitation or aggression, new confusion, seizures, substance use, unsafe driving, or trouble managing basic care. I encourage them to provide structure, reduce overstimulation, and help the person practice small daily skills, without making every interaction feel like a test.

CHPR: The cognitive deficits must be very frustrating for patients. Are there any treatments you find helpful?

Dr. Kelly: Mental fog is a major source of distress. Patients become frustrated because they cannot concentrate or think clearly, and that frustration can worsen behavior. Some patients also struggle with daytime sleepiness or fatigue, especially when nighttime sleep is fragmented. Stimulants like methylphenidate can help with attention and processing speed (Chien YJ et al, Brain Sci 2019;9(11):291), while wakefulness-promoting agents like modafinil may be more useful when excessive daytime sleepiness is prominent (Kaiser PR et al, Neurology 2010;75(20):1780–1785). I tend to be cautious. I may continue a stimulant if I inherit a patient who is clearly benefiting from its use, but I rarely start one. These medications can create problems if the patient runs out, misuses substances, or begins using alcohol or other sedatives to counterbalance the effects.

CHPR: Has the prognosis for TBI changed over the course of your career?

Dr. Kelly: It has improved tremendously because we have learned what to do and what not to do. Neurosurgical care, intracranial pressure monitoring, trauma systems, rehabilitation, and severe TBI guidelines have all improved. We’re also better at avoiding interventions that delay recovery. For example, patients used to be treated as though they were all going to have seizures and were placed on anticonvulsants unnecessarily, with all of those adverse effects layered on top of the brain injury. Now we are more thoughtful. Patients who once would have been dismissed as having a terrible prognosis can sometimes make dramatic recoveries.

CHPR: Any final takeaways for hospital psychiatrists?

Dr. Kelly: Start with the injury. Find out what happened, find out who the person was before, pay close attention to sleep, and use medications carefully. The goal is not simply to suppress behavior; it’s to understand the injured brain, support recovery, and help the person regain as much of their previous life as possible.

CHPR: Thank you for your time, Dr. Kelly.