The Night Is Where Epilepsy Care Falls Apart

Over 65 million people worldwide live with epilepsy. Roughly 80 percent of them get no warning before a seizure begins. And for the subgroup of patients with uncontrolled seizures specifically, the annual mortality risk runs close to 1 in 150.

Those three numbers don't sit comfortably next to each other. They describe a condition that is common, unpredictable, and — for a meaningful slice of patients — genuinely dangerous. And yet, if you read the public-facing materials of most companies operating in this space, you'd think epilepsy was a clinic-hours problem.

It isn't. It never was.

The night is where the danger lives

Sudden Unexpected Death in Epilepsy — SUDEP — is the single most studied and most feared outcome in epilepsy care. The strongest predictors are uncontrolled tonic-clonic seizures and unwitnessed nocturnal events. SUDEP cases overwhelmingly occur during sleep. The patient is usually found face-down in bed in the morning. The mechanism is still debated — respiratory, cardiac, autonomic — but the timing isn't. The night is when it happens.

This isn't a niche fact. It's the load-bearing data point behind every conversation about seizure monitoring, alert devices, and home safety protocols. And it's the reason families spend years sleeping with one eye open.

Pharma has the daytime covered

There is no shortage of investment in seizure pharmacology. The pipeline of anti-seizure medications has expanded steadily for decades. There are companies — well-funded, scientifically serious companies — building third-generation cannabinoid therapies, KCNQ2/3 openers, vigabatrin alternatives for infantile spasms, and selective targets for Dravet and Lennox-Gastaut.

These drugs do real work. They reduce seizure frequency, sometimes dramatically. They give families more clear days than they had before.

But "more clear days" is a daytime metric. Trial endpoints are usually monthly seizure counts, percent reduction, and time-to-first-seizure. The patient or caregiver fills out a seizure diary. That diary is by definition a record of what was observed. Nocturnal seizures that nobody witnessed — and nocturnal events that families heard but couldn't classify — go uncounted. The very events most associated with SUDEP are the events least likely to make it into the data that drives drug development.

This isn't a criticism of the science. It's a structural blind spot. You can't measure what you can't see.

The sleep companies don't connect the dots either

Here's the part that surprised us most while researching this space.

There are companies whose entire identity is sleep and wake biology — pitolisant, orexin antagonists, melatonin receptor agonists, narcolepsy and idiopathic hypersomnia franchises. These companies understand the architecture of sleep at a level most epileptologists don't have time for.

But when those same companies publish materials about their rare-epilepsy programs, sleep gets listed as one bullet among many — appetite, balance, gait, behavior, sleep disorders. One line. No discussion of nocturnal seizures. No mention of SUDEP. No connective tissue at all between their daytime franchise and the nighttime reality of the patients they're trying to serve.

That gap isn't malicious. It's just what happens when product organizations are siloed. The sleep team and the epilepsy team don't share the same office, and the consequence is that the most important nighttime population in neurology gets treated like every other comorbidity bullet.

What contactless monitoring actually adds

Komori Care doesn't claim to detect seizures. We're a wellness-pathway device and our framing is deliberate: movement awareness, presence awareness, environmental context. Not diagnosis. Not treatment. Not prevention.

What we do build is the layer that has been structurally missing — continuous, contactless, multimodal visibility into what is happening overnight in a person's actual bed. Headboard radar for movement and bed presence. Thermal for skin-temperature trends. WiFi CSI for whole-room context. Acoustic event classification for sounds. Ambient sensors for temperature, humidity, CO2, and light.

None of those signals alone tell you whether a seizure happened. Used together, over weeks and months in real homes, they tell you something most clinical workflows have never had access to: what normally happens at night, and when something is different. That's a useful piece of information for a family. It's also, eventually, a useful piece of information for the clinicians and researchers trying to understand why nocturnal events keep killing people.

Why we wrote this

We have a feasibility amendment in motion with UVA Neurology under Dr. Mark Quigg, MD. We're presenting at NIH Curing the Epilepsies 2026 and the Epilepsy Pipeline Conference in June. If you're a clinician, researcher, foundation officer, or family member who has felt the absence of overnight visibility in epilepsy care, we'd like to hear from you.

The drugs are getting better. The night is still the gap. We're trying to close it — carefully, honestly, and without overpromising.

Sources

• World Health Organization, Epilepsy fact sheet (most recent edition) — 65 million prevalence figure
• Epilepsy Foundation and ILAE consensus statements on SUDEP risk in poorly-controlled tonic-clonic epilepsy
• Devinsky et al., Lancet Neurology — long-term mortality and SUDEP risk stratification
• General no-warning prevalence figures vary 50–80 percent across studies depending on syndrome and aura definition; conservative reporting in patient-facing materials cites the upper end