At two o’clock in the morning, on August 21, in a dark hospital room on R1 at Maine Medical Center, Kirsten suffered a seizure. This seizure caused her to fall and fracture her C5 vertebrae. Since then, she’s been taking Levetiracetam twice a day, had seizure precautions during her entire hospital stay, and endured a forty-eight hour electroencephalogram (EEG). Every decision in the hospital considered the seizure risk, and every medication had to be checked for interactions with the seizure-prevention drug.
However, Kirsten likely did not have a seizure that lead to a fall and neck fracture.
No empirical evidence exists that would lead a doctor to objectively conclude Kirsten had a seizure. In fact, the only evidence anybody has on which to base a diagnoses is the report of a single nurse/CNA, who said they witnessed upper-body movements for sixty to ninety seconds after her fall. The problems with this report are several. Allow me to enumerate:
- The fall occurred at 02:00 in a dark hospital room. The witness was not in the room at the time, and frequency of room visits on that floor at that early hour is very low. It is not possible to determine the duration of seizure activity if the start time is unknown. This should be obvious.
- Reports are conflicting. One person told me on the morning of Kirsten’s fall that she was sitting up in bed working on her computer when she suffered the seizure, which caused her to fall out of bed. The computer is not damaged. Another person said that when she got out of bed to use the bathroom, her blood pressure dropped, precipitating the seizure.
- Kirsten’s physiological condition that morning was not unique, and yet there is no history of seizures.
While I do not easily subscribe to conspiracy theories of any kind, as it is nearly impossible to get so many people to stick to a single explanation of events, the motivation of the hospital to shield themselves from a malpractice lawsuit cannot be ignored.
The explanation of consensus by the medical professionals is more likely attributable to incompetence, lack of diligence, over confidence in the medical records, over reliance on eyewitness testimony, and the awkwardness related to challenging the diagnoses of a colleague.
A seizure simply doesn’t make sense. Kirsten has no history of epilepsy or seizure activity of any kind. Nor has she had any since. What she did have, which was documented in her medical records, is a history of confusion and weakness, including falls caused by a simple lack of muscle strength. An appeal to Occam’s Razor should have won the argument. It did not.
Why should I care? The problem with the diagnosis for me was that it complicated everything else, diverted medical attention from real problems, and likely prolonged recovery.
I discussed my concerns with every doctor who would listen, any many who would rather have not. I was told, time and time again, that there was a witness in the room, a witness who, apparently, described in detail the seizure activity. This person, I was told, had been trained to recognize a seizure. I was then told that the risk of keeping Kirsten on Levetiracetam (Keppra) was far less than the catastrophic damage that would be done if Kirsten had a seizure, considering her freshly broken neck and extremely soft and brittle bones. This argument made sense, and I decided to mostly give up the fight; I didn’t want Kirsten to suffer a catastrophic injury because of my hunch.
Kirsten was in the emergency department a month prior. She scared her co-workers on the ninth floor when she became extremely confused. It was discovered at the time that her potassium level was life-threateningly low. A similar state of confusion is what prompted her doctor to call an ambulance on August 17th. The most obvious explanation for these states of confusion was endocrine imbalance, and remained the explanation until the fall. After the neck fracture, the explanation for these episodes of delirium, which became increasingly frequent and more intense, was that she was experiencing “absence seizures.” Not flailing around like a fish, as Dr. Dolan describes a normal seizure, but a less severe type. The problem is, of course, that absence seizures last for ten to twenty seconds, and Kirsten was experiencing these episodes from between two to fifteen minutes. Endocrine explanations were no longer considered, and were dismissed when I suggested them. Non-seizure explanations were off the table. Neurologists were consulted and ran their tests. Vitals were recorded and examinations administered. Nothing.
One morning in rounds (the doctors meet outside the patient’s room to discuss the history and plan daily), when the resident responsible for Kirsten’s care authoritatively said that Kirsten suffered a seizure that lead to the neck fracture, I spoke up in front of the entire team, which included the attending physician. I said that the evidence simply doesn’t add up. A young doctor recently graduated from medical school and beginning his residency posited that perhaps what Kirsten suffered was instead a “mechanical fall.” The room was dark, there are various cords and stands and chairs and tables and carts, Kirsten was extremely weak from muscle loss, and her blood pressure was very low. Finally, a doctor was suggesting what I had been for so long and to so many. Later that day I pulled the attending physician aside to more completely express my ideas. After laying out my case, point by point, she suggested that perhaps the episodes of delirium were nothing more than her body withdrawing from the high levels of cortisol in which her brain had been bathing for so long. Finally, my persistence and persuasiveness was beginning to pay off.
At this point I didn’t much care about the fall and neck fracture, I just wanted the team to look for explanations of the delirium to causes besides seizure activity. She was on Levetiracetam and the episodes weren’t reduced. Furthermore, if they assumed the problem was correctly diagnosed and the appropriate treatment administered, they wouldn’t continue to look for the true cause. They seemed all too willing to simply let the problem go until a neurologist could do a more thorough exam after Kirsten left the hospital.
I repeated and reiterated my concerns to Kirsten’s doctor at the New England Rehabilitation Hospital the day she was transferred there. Again it was argued that Kirsten should remain on Levetiracetam to reduce the risk of a catastrophic seizure. And again I agreed. However, the side effects of Levetiracetam meant that Kirsten would be more tired and confused that she would otherwise, which would prolong her recovery. It was decided, after discussing it with Kirsten, that the dosage would be reduced by a third. It was a small victory.
Today we visited Dr. Dolan, her neurologist. After asking her some basic questions to determine neurological health, he told us that not only did he not think Kirsten had a seizure, but that putting her on Levetiracetam would not have been the preferred solution if she had. That drug, he said, only works 50% of the time for 50% of the patients who take it. The likely cause of her episodes of delirium, he said, was high levels of cortisol, and the best way to prevent another seizure, if there was any risk of it at all, was to get the endocrine imbalances under control and normalized. As evidence, he said, the episodes decreased in frequency and intensity as her levels of cortisol were normalized. He said Kirsten should stay on Levetiracetam to keep her neurosurgeon happy, but to call him when the neck brace comes off so he could advise us on a titration plan. Then follow up with him in six months.
Kirsten’s brain is just fine. She is not, nor ever has been, an epileptic. It feels very good to finally be validated. It should have been obvious. There is an important lesson here for everybody, but especially those people who work in a field of science (I count medical professionals among these): go with the explanation that is most consistent with the evidence, and don’t rely heavily on one person’s description, regardless of how much training they’ve had. When it comes to evidence, eye-witness testimony is notoriously weak and unreliable. Occam’s razor should not be casually dismissed. The most likely explanation is usually correct.
Kirsten was an epileptic for less time than she was a diabetic. I am thankful for that.
She’s getting around now without her walker, and walked with assistance down and up our stairs without resting in the middle. I wash her hair and change the pads on her neck collar a couple times each week. She pushes harder than I think she should, but her dedication to normalcy is paying off.