Now we can keep tabs
YOUR phone is buzzing: it is time for your medication. You point the camera at your face so that an app can confirm your identity. It then approves the dose of the pill you show it, and watches as you put the pill on your tongue and drink a glass of water. Algorithms listen in and inspect your face and throat for any signs that you might be faking. The app dings. You may now go about the rest of your day.
The app, AiCure, is being used in US drug trials to ensure participants stick to the rules. It could also have a role to play in stopping the spread of antibiotic resistance, its creators say.
The New York start-up that built AiCure has won $19.5 million in funding – nearly a third of which is government backing via the US National Institutes of Health – to address the problem of people not taking their medication properly. Some of these are individuals who enrol in multiple trials in order to make money.
Researchers have blamed so-called “professional patients” for the rise in placebo effects in clinical trials of antidepressants, which have seen placebo response rates as high as 50 per cent. Some estimate that professional patients now make up 1 in 20 trial participants. Many fail as a result.
Amir Kalali at Quintiles, one of the world’s largest contract research organisations, says that AiCure is part of a new trend to devise monitoring techniques that are hard to deceive. There are smart pills with embedded microchips that broadcast a signal only when they have reached the acid of the stomach, for example. AiCure has the advantage that it can be installed on any smartphone with a video camera, says the company’s chief scientist Alejandro Jaimes.
Crucially, if a trial participant is not taking their medication as planned, the app doesn’t tell them. “If you let them know they’ve made a mistake, they can learn from that how to better defeat the system,” says Jaimes. Instead AiCure sends details of all irregular activity to the trial organiser, who can either call the person to ask if they need help, or boot them off the trial.
David Mischoulon at Massachussetts General Hospital in Boston envisages the app being useful outside of trials. Many people get confused over which drug to take and when, he says. “The app is a great idea for the motivated patient who needs help, particularly if their regimen is complicated.”
“The software can be easily trained to monitor doses of new medication, or even injections“
Often, patients fail to keep up doses of immunosuppressants after transplant operations, says Kalali. And many people stop taking antibiotics when they feel better, despite having a two-week prescription. Non-compliance plays a significant role in bacteria becoming resistant to multiple antibiotics.
The software behind the app can help prevent this as it can be trained to monitor doses of new types of medication, says Jaimes – or even injections. “If it’s visible, we can monitor it.”
But such techniques raise some ethical concerns. “There’s nothing inherently wrong with monitoring adherence,” says Matt Lamkin at the University of Tulsa College of Law in Oklahoma. But he wonders whether people will be able to opt out of using the app. “It turns on what the consequences are for refusing to be monitored,” he says.
Some companies already track their employees’ health, so it’s not hard to foresee a future in which insurance premiums depend on whether you have taken all your medication. “If you lose access to healthcare, that’s a problem,” says Lamkin.
This article appeared in print under the headline “An app full of AI helps the medicine go down”