The 685 Japanese Children Who Were Hospitalised in One Evening by a Pokémon Episode

The Pokémon television series had been on the air in Japan for eight months when the episode now known to Pokémon fandom as “Electric Soldier Porygon” was broadcast on the TV Tokyo network at 6:30 PM Tokyo time on 16 December 1997. The episode was the 38th of the series and aired in the standard early-evening children’s slot to an audience of approximately 4.6 million Japanese households.

At approximately 6:51 PM, twenty-one minutes into the broadcast, the episode contained a sequence in which the digital Pokémon Porygon was attacked inside a virtual computer environment by anti-virus countermeasures. The animators had used a 4-second alternating red-and-blue flash sequence at 12 hertz to represent the electronic combat. The flash had not been pre-tested for photosensitive triggering. By 7:30 PM that evening approximately 685 Japanese children had been hospitalised — most with the symptoms of seizures, convulsions, vomiting, dizziness, and breathing difficulty.

It was the largest single mass-emergency event in the history of Japanese television broadcasting. The series was suspended for four months. The Japanese broadcasting authorities subsequently issued new flash-rate guidelines (since adopted internationally) that have prevented any comparable subsequent event from recurring.

The medical investigation

The Japanese Ministry of Health investigation through the following months identified two distinct medical phenomena that had been combined in the public emergency.

The first was a real but small population of children with previously-undiagnosed photosensitive epilepsy — a recognised neurological condition in which the visual cortex responds to specific frequencies of flashing light (typically between 5 and 30 hertz) with electrical discharges that propagate into epileptic seizure activity. The 12-hertz red-blue alternation in the Pokémon sequence sat precisely in the most-provocative band of the photosensitive trigger frequencies. Among the 4.6 million Japanese households watching the broadcast, the substantial minority of children with previously-undetected photosensitive epilepsy who experienced actual seizures from the flash sequence is estimated at approximately 80–100 individuals.

The second was a much larger population of children who had experienced no actual seizure but had developed psychogenic symptoms — dizziness, nausea, headaches, fainting, breathing difficulty — after either watching the flash sequence themselves or learning about other children who had reportedly fainted. This second population produced the substantial majority of the 685 hospitalisations and is the textbook modern case of mass psychogenic illness spread through a single broadcast television event. The Japanese government investigation, the subsequent academic literature on the case (particularly Furusho’s 1998 Brain & Development paper), and the international convergent medical consensus all place approximately 500–600 of the 685 hospitalisations in this second category.

The MPI propagation

The case is significant for the medical-history literature on mass psychogenic illness because of its unusual propagation mechanism. Previous MPI episodes — the Strasbourg dancing plague of 1518, the Aachen dancing mania of 1374, the Tanganyika laughter epidemic of 1962, the various 20th-century school-faint episodes — had all required direct face-to-face propagation between affected individuals. The Pokémon Shock was the first major documented MPI episode that propagated primarily through broadcast television, with the secondary wave of cases triggered by the news coverage of the initial cases rather than by direct contact.

The pattern is straightforward in retrospect. The Japanese national news bulletins on the evening of 16 December 1997 carried emergency-response coverage of the hospital admissions; the public-health agencies issued substantial advisories warning parents to check their children for delayed symptoms; the medical-television commentary speculated about possible causes through the late-evening hours. The children who had been watching Pokémon and had not experienced any acute physical reaction were now under sustained adult attention and sustained media-environmental anxiety. A measurable fraction developed delayed psychogenic symptoms and were brought to hospital over the following 12 hours.

The wider medical-emergency-response community has substantially absorbed the lesson. Modern Japanese broadcasting industry guidelines limit flash sequences to a maximum of three flashes per second; the British equivalent (the Ofcom code) and the international ITU broadcasting standard both adopted comparable post-Pokémon-Shock limits in 1998–2000. The international guidelines have prevented any subsequent broadcast event of comparable medical magnitude.

Aftermath

The Pokémon television series was suspended from 16 December 1997 until 16 April 1998. The TV Tokyo network and the production studio (OLM Inc., working under the brand name Game Freak) issued a coordinated public apology, restructured the post-production approval process to include a photosensitivity-review step (still in operation), and proceeded with a relaunch that was substantially more cautious about visual-effect sequences. The episode “Electric Soldier Porygon” has not been rebroadcast in any market since 16 December 1997.

The Pokémon franchise itself was unaffected commercially. The global expansion of the series through the late 1990s and early 2000s — the trading-card game, the international television syndication, the Pokémon video-game sequels — proceeded substantially on schedule. The franchise’s lifetime commercial revenue is now estimated at approximately $90 billion, making it the highest-grossing media franchise in history.

The Pokémon Shock remains the largest single-event broadcast-media medical emergency on record and the only major MPI episode with confirmed propagation entirely through television. It is now a standard reference case in the medical-broadcasting literature, the photosensitive-epilepsy literature, and the mass psychogenic illness literature. The substantial majority of the 685 children admitted to Japanese hospitals on the evening of 16 December 1997 had not, in any organic-medical sense, been made ill by the broadcast. They had been made ill by the news coverage of the broadcast. The distinction matters scientifically; it did not matter to the parents in the hospital waiting rooms that night.