Babylon, a UK-based telehealth company, is answering the worldwide need for remote care in the setting of social isolation. But have their services matured enough to actually help in the fight against the coronavirus?
Babylon was big before COVID-19
Founded by Ali Parsa in 2013, Babylon offers instant 24/7 medical assistance via smartphone. An AI chatbot takes patients’ history and performs triage. Very simple cases it can handle on its own. Anything more complicated is quickly contacted by a Babylon GP via the app. The doctors treat most cases remotely, even issuing prescriptions to patients’ pharmacies. Complex cases are requested to visit physical locations for more detailed follow-up, while the AI-chatbot instructs patients to visit emergency services if a life-threatening condition cannot be ruled out. The company’s big break came in November 2017, when an official partnership with the NHS was launched. With GP at Hand, NHS patients can actually make Babylon their primary care provider, letting them use their services for free. Within just one year, 36555 patients had signed on.
Babylon in the COVID-19 era
Since then, Babylon expanded in the US and Rwanda. The company’s linear growth became exponential due to the COVID-19, as the company upgraded their systems to be able to handle 250 thousand simultaneous AI-consultations, up almost 90-fold from their pre-COVID-19 capacity. In order to provide targeted care for the virus, Babylon launched a separate app for monitoring COVID-19 symptoms, while anyone can get their COVID-like symptoms examined by Babylon’s AI for free.
Is Babylon AI really ready for primetime?
Babylon’s track record isn’t exactly pristine. At the time of the GP at Home launch, their AI was caught missing diagnoses of fatal diseases, like pulmonary embolism. This, however, was not due to the lack of caution; the AI sent patients to emergency services at almost twice the rate actual GPs would. Was Babylon’s AI not trained well enough at the time? Has it gotten better? Nobody outside Babylon knows, as they have no intention of measuring their bot’s competence in clinical studies, dismissing them as way too static to capture AI’s evolving potential. Nonetheless, with all the vocal opposition to the company, if any patient was actually harmed because of mistreatment, we would know. Thus, the fact that criticism is still pure theoretical is a good sign.
The Fantastrial team actually experimented with Babylon’s COVID-19 symptom checker, and we all thought that it was definitely on the cautious side, suggesting trips to emergency rooms to hypothetical patients without particularly alarming symptoms.
In the end, time will tell if Babylon’s service will contribute to the fight against COVID-19 – let’s just hope patients (and their privacy) aren’t hurt during this process of trial and error.