Conversion Disorder: Does Any Medication Treat the Seizures Themselves?
A single patient with video-EEG-confirmed nonepileptic seizures and a depression screen that crossed threshold three months after her father's death. The disagreement isn't about the seizures — no voice thinks a pill treats those — it's about whether her PHQ-9 describes grief or a second, independently treatable condition.
T.H., a 34-year-old woman, was three weeks from starting a new job in hospital billing when her father died suddenly of a stroke, a loss that arrived with no warning and no chance to say anything to him beforehand. Six weeks later, at her father's house sorting through his belongings, she had her first episode — her sister found her on the floor, limbs shaking, unresponsive for almost four minutes, and called an ambulance certain she was watching a seizure. The emergency department agreed and started her on levetiracetam. She has had eleven more episodes since, always during moments of intense emotional strain, and three days of inpatient video-EEG monitoring last month captured two of them on camera with no epileptiform activity accompanying either — the basis for stopping the anticonvulsant and diagnosing psychogenic nonepileptic seizures instead of epilepsy. She started the new job on schedule and has kept it, though two episodes have happened at work, and her manager has been patient so far in a way she doesn't expect to last indefinitely.
Her PHQ-9 today comes back at 13, moderate range, three months out from her father's death — a number that could describe two genuinely different things, and the distinction matters for what happens next. LaFrance and colleagues' randomized pilot trial in exactly this diagnosis found CBT-informed psychotherapy reduced seizure frequency significantly against treatment as usual, while sertraline alone did not separate from treatment as usual on that same outcome — so nothing in the literature argues for a medication aimed at the seizures themselves. What the trial doesn't resolve is her mood: whether a PHQ-9 of 13 this soon after losing her father describes ordinary, expectable grief, or a second, independently treatable depressive episode that happens to have arrived at the same time as an already difficult diagnosis.
After the video-EEG results
Whatever else this team decides, I want to be clear that no medication has trial support for the seizures themselves. LaFrance and colleagues randomized patients with this exact diagnosis to CBT-informed psychotherapy, sertraline, both, or treatment as usual — CBT-ip significantly reduced seizure frequency, sertraline alone did not separate from usual care on that outcome. Levetiracetam is already stopped, correctly, and nothing should replace it under the banner of treating her episodes.
I'm not proposing anything treat her seizures — I agree completely with that read of LaFrance's data. What I'd treat is the depression: a PHQ-9 of 13 clears the threshold for a real, independently significant depressive burden, and the fact that it arrived alongside a harder diagnosis to treat doesn't make it less real or less deserving of its own treatment.
Grief and depression overlap heavily at three months, I'll grant that much going in — but overlap isn't identity. A score this elevated, with functional consequences already showing at work, is worth treating on its own criteria rather than assumed away as normal mourning until it resolves on its own.
Three months after losing a parent suddenly, with no warning and no chance to say goodbye, a PHQ-9 of 13 is well inside what an expectable grief trajectory can look like — this isn't a score that reliably distinguishes grief from a depressive episode at this timepoint, and I don't think we can read it as clearly one or the other yet.
My concern is less about whether an SSRI could plausibly help and more about what starting one on top of a seizure disorder she's already been told, accurately, has no pill for teaches her about her own experience — that the answer to a hard year is medication, precisely the narrative the psychotherapy-based PNES treatment model is trying to help her move away from. I'd rather reassess her PHQ-9 after four to six weeks of grief-focused and CBT-ip work before deciding whether what remains looks more like depression than mourning.
Agreed: no medication started today for the seizures themselves, and none started today for mood either — CBT-ip referral placed immediately, with a structured PHQ-9 recheck in four to six weeks specifically to answer the grief-versus-depression question the group couldn't resolve from a single number.
Not agreed, and stated plainly rather than papered over:
The psychiatrist's case for an independent depressive episode strengthens considerably, and sertraline would be started for that reason alone — not for the seizures.
The psychologist's read is favored, and the case for medication may never arise — consistent with a normal, if difficult, grief trajectory rather than a second diagnosis.