“im rly depressed and im feeling insecure and its just so exhausting why is life like this.”
“Feeling low but too lazy to get out of bed to check my BG [blood glucose]. I’ll go back to sleep instead.”
“I wouldn’t say I’m suicidal. I don’t want to die. I just really don’t want to wake up anymore. . .”
I recently overheard people making remarks like these. But they weren’t telling me because I’m a behavioral psychologist in the Department of Family Medicine at UCLA. They were telling the entire world on Twitter.
Social media, it turns out, is a new vital sign, and it’s one that health providers are learning how to read.
People don’t just use social media to talk about celebrity sightings or joke about why TSA agents take so long to search them. They’re also using it to publicly tell the world about health issues that they often don’t even tell their doctors. They’ll share openly on Facebook and elsewhere about their feelings, their plans to do healthy things like exercise, and their intentions to do unhealthy things like use drugs.
Users post videos showing whether they’re anxious and how fast their hearts are beating. They don’t always share this information explicitly, but it can be inferred–these days, with ever greater accuracy. Together, psychologists and computer scientists can now read between the lines of social media to interpret and predict how people feel and what they’ll do. They can even tell from people’s posts whether they’re telling the truth, whether they’re gay, and whether they’re stressed out.
Health providers will soon be able to use social media as a powerful tool to monitor patients outside of healthcare settings. In the clinic, most health providers measure physiological vital signs like blood pressure and heart rate, but the best doctors also notice psychological vital signs, like observing whether a patient is anxious or sad, or reluctant to take medication.
But how do health providers observe patients when they can’t see them? Smartwatches and self-monitoring devices can track vital signs like heart rate and blood pressure, but they don’t monitor psychological vital signs. If health providers could detect when their patients were sad, or why they were reluctant to take medications, for example, they could provide real-time interventions to improve health care. That’s where social media comes in.
Take depression, the world’s top cause of disability. It affects approximately 8% of Americans every year, costs the U.S. health care system approximately $200 billion annually, and leads to other negative health outcomes like reduced medication adherence and suicide. Although real-time monitoring of depression could save money and lives, it has been virtually impossible to track depression outcomes when patients are outside of a clinic. Until now.
Teams of psychologists and computer scientists are working together to analyze the words and images that people organically share on social media sites like Twitter, Facebook, and Instagram. They’re developing technologies that learn to think like a large team of psychologists. Working side-by-side with psychologists and public health experts, these machines can quickly learn to identify psychological patterns from millions of social media texts and images and use that information to predict people’s emotions and behaviors. The tools incorporate advanced computer science approaches like “sentiment analysis indicators,” and “behavioral insights on big data.”
For example, in our research, we’ve used these methods to learn to monitor stress, anxiety, and depression levels among UCLA students, and have even been able to intervene and provide psychological services to a student we found expressing suicidal intentions. Health systems will be able to use social media data to monitor and predict a broad range of health issues, like diabetes self-management and medication adherence, and public health departments can use them to predict and address regional trends, like planning the number of vaccinations needed for an upcoming flu season.
The practical implications of this approach are clear, actionable, and stand to significantly improve health care. Apps and health monitoring tools that incorporate social media vital signs will soon be available for health systems nationwide. Armed with them, health providers will be able to gain real-time knowledge of when patients show unhealthy psychological behavior.
They’ll be able to know whether patients are experiencing prolonged sadness and when they’re not taking medications. They’ll be able to use this information to diagnose clinical disorders and request patients to come for follow-up appointments. If health providers receive an alert that a patient’s social media posts suggest she’s self-medicating her depression, for instance, they could immediately contact her and intervene.
Implementing these tools won’t require additional time from doctors, but we will need to rethink our current medical approach, as few health systems are equipped to deal with remote patient monitoring. That leaves to big questions unanswered: First, will doctors want to have access to more information about their patients? And second, will they be required to act on it? Today, when doctors receive information about patients’ vital signs, they have a medical responsibility to respond based on that information if need be. If health providers would be required to act on social media vital-sign information, then some might prefer to not receive it at all.
But the answer to whether we use social media as a vital sign shouldn’t be based on whether it’s convenient for health providers; it should be based on whether it improves patient health. If social media is as valuable a resource as research is suggesting, then we need to start talking about how to restructure our health systems to incorporate this new approach.
Tools for analyzing social data are already successfully being used in other fields like consumer behavior, education, and crime prediction. Incorporating social media as a medical vital sign is an investment in the future well-being of our society. Big Brotherish as it might sound, studies nevertheless show that patients will willingly share this publicly available information with the world if it can be used to improve their health.
What’s more, we won’t need expensive infrastructure to implement this approach. A growing number of tools like Google’s Deep Mind and IBM’s Watson are already being applied to health care and will soon be ready for wide-scale use by medical professionals. It’s time that health care providers and health technologists start discussing how to incorporate social media into clinical care. Our lives depend on it.
Sean Young, PhD, is a UCLA medical school professor in the Department of Family Medicine and the executive director of the University of California Institute for Prediction Technology (UCIPT).