‍How it works: 

‍From Alcor: 

“The service supports live daily calls, at a predetermined time, from an Emergency Medical Technician (EMT) to confirm you are well and inquire about any changes to your health status. If you do not answer the call after several tries, the EMT will alert a designated contact, such as a friend or neighbor, to check on you! In addition to live daily calls from an EMT, the service comes with a smartphone app!

The app supports several features, including: Member electronic check-in to skip a call from an EMT for the day; Mobile Emergency Response Service to immediately connect you to an EMT 24/7; Geolocation enabled Pin Drop feature so you can be found in case of an emergency; And other features!"

‍High-level summary of issues: Alcor Check-In App: There aren't any complicated details to highlight here. The system has a mechanism in place to cap the worst case scenario at one day. We would like the system to be faster, but given that this kind of solution requires active user interaction, there's some upper bound on how frequently to set a check-in before it becomes too cumbersome to use. Also, if you miss the check-in for non-emergency reasons, such as being away from the phone, the situation escalates to your contacts, which, if it occurs too frequently, could cause the user to stop using the system.

How does the system stack up against other current options?

On paper, in terms of the best case, worst case, and average case, the system scores the same as the Cryonics Institute Check-In app. However, it should be pointed out that the CI app does not have EMT (medical call center) involvement in the system, which may or may not play a critical role in some situations.

Summary: This is a solid stop-gap design for preventing patients from being unattended for days, with the average case lasting approximately 12 hours (assuming death occurs with the same frequency throughout the day - perhaps not). In any case, at the very least, we'd typically be dealing on the order of multiple hours or more, and so while this kind of monitoring can be helpful as a backup to some of the less reliable sensor systems, it is fundamentally not quick enough for the response times we'd like to see.