About a decade ago EL was working with his mentor EK, a leading healthcare data scientist, on a project for an AI tool in healthcare. EL had been a startup founder and had worked for Microsoft for many years. Both were born in the USSR, emigrated to Israel, then to the US. So had some friends. Their friend RK had learned IT and encryption working for the Soviet rocket science program. Their doctor was also born in the USSR, became an internist, and emigrated to the US, learning English driving a cab while doing a second internal medicine residency, required to become a board-certified US physician.
So the two took their idea to their doctor. “Guys, I don’t need any artificial intelligence! I can’t get real intelligence! I can’t get records from the hospital, I can’t get the echo results from the cardiologist, I can’t figure out what is the blue pill my patient says he’s now taking….I need REAL intelligence!”
The guys were stunned. They had no idea of the problems with information communication in healthcare, and decided to get together and do something for their doctor. “Why don’t the patients have the records? We could put a whole records file on a credit care chip!’ “Yes, great. What about security?” “Each card could have it’s own security code, and one for the company, and one for the doctor. Triple protection!” “OK, but there are 600 electronic record companies in the US. How can the patient’s records connect to ALL of them?” This required some thought. “Print! Every computer has a print function. We can build a mini-computer, plug it in with a USB cable, and emulate a printer so records can be printed to and from the system to any electronic records system.”
So they made those ideas a reality over about 18 months. By then I was the Medical Director of a Clinical Integration Network with the hospital with over 600 physicians. The doc sent me an email: “I have couple of software guys I like you to meet.” One line. Verbatim. I almost ignored it, but he was in my network, and I was his Medical Director, so I invited him to have coffee on Friday at 0930 in the hospital cafeteria. They all showed up, and explained the above story. “Wow. It looks to me like you guys have invented the 21st Century solution to medical records!” “Oh, so you do understand!” said the data scientist. “Yes, but why are you talking to me?” “We think you can help us get a pilot at the hospital.” “You know, I think I can!” So I took the idea to one of the quickest minds I know, the VP and Chief Information Officer. He listened for at least 90 seconds and decided. “Yes, we can do that.”
We set up a pilot with doctors in primary care, 2 here, 2 there, an orthopedist, a cardiologist, a gastroenterologist, and recruited 300 patients. NONE of my patients turned me down. EVERYONE wanted to have their own personal health records on their own chip card. Many patients thought the records systems should have been built this way in the first place! We got the hospital CEO to be a patient, get referred to a specialist, have a procedure, and have the pre-procedure records delivered on the card, and the procedure delivered back on the card. No faxes!! No paper!! It worked!!
Next, a bigger pilot with the Cancer Center, 6000 patients. Then we understood the roadblocks: the health information management felt threatened that jobs printing, scanning, filing faxes and paper would be eliminated. I.T. did NOT want to help us. We converted the USB connected minicomputer with 8 months and 20 developers into a mobile app: syncMD. Debugged and working, it went to Apple’s Play store and Google’s online to patients for free. Then we were referred by that CIO VP to the vendor copying records for the hospital. They could NOT believe that we could train their staff in 10 minutes, but we did. Instead of burning a DVD for patients, we helped them send the files to enrolled patients. Folks could download the app, register, and receive electronic copy of their requested records on their smartphone while they were in the records department, maybe 15 minutes altogether.
Why there are only some thousands of users and not billions is another story….