Could Big Data Beat Our Opioid Crisis?
Experts in the AI and Big Data sphere consider October 2021 to be a dark month. Their pessimism isn’t fueled by rapidly shortening days or chilly weather in much of the country—but rather by the grim news from Facebook on the effectiveness of AI in content moderation.
This is unexpected. The social media behemoth has long touted tech tools such as machine learning and Big Data as answers to its moderation woes. As CEO Mark Zuckerberg explained for CBS News, “The long-term promise of AI is that in addition to identifying risks more quickly and accurately than would have already happened, it may also identify risks that nobody would have flagged at all—including terrorists planning attacks using private channels, people bullying someone too afraid to report it themselves, and other issues both local and global.”
In theory, AI systems can detect rule-breaking posts and remove them long before human staffers can judge the content and act. There’s just one problem. The systems don’t seem capable of performing what Zuckerberg described. According to internal communications leaked to The Wall Street Journal and published as part of its “Facebook Files” series, Facebook’s own engineers doubt its AI systems will ever match management’s vision. In fact, one senior engineer stated the company catches just 2% of rule-breaking posts, adding: “Recent estimates suggest that unless there is a major change in strategy, it will be very difficult to improve this beyond 10%-20% in the short-medium term.”This revelation raises questions for other groups investigating how AI and Big Data can improve the world. After all, if an organization as vast and as well-funded as Facebook is stumbling with AI implementation, what hope do others have? It’s worth noting, the company has the posts, personal data, and site activity of billions of users worldwide to base its analysis on. And yet it still hasn’t mastered application of this incredible data source.
Despite all its advantages, AI isn’t displacing radiologists because, as Harvard Business Review explains, interpreting images is just part of what a human specialist does. Instead, AI is being utilized in medical imaging to fill other important roles, such as serving as a “safety net” to catch potential cancers humans could miss while working with specialists to make the best use of their busy day.
What’s more, as AI and Big Data mature in the medicine field, they are being harnessed for a new and surprising battle: the fight against the opioid epidemic ravaging America. Overdose deaths have tragically soared nearly 30% during the pandemic. But contrary to popular opinion, Opioid Use Disorder (OUD) and overdoses are not a problem exclusive to the homeless, working poor, or any race. In fact, opioid addiction and overdose crosses all racial, social, and economic lines.
As the above makes clear, the problem continues to plague the U.S. However, AI and Big Data are emerging as potent new weapons in the fight. Let’s get into the specifics. Just as an ER doctor must stop a car crash victim’s bleeding before damaged organs can be repaired, the first step to fighting OUD is to keep the patient safe while they use drugs. How? By preventing overdoses.
Cognizant of this truth, an upstart from Canada named Brave already leverages tech to accomplish this mission. Brave’s app can connect an opioid user to a volunteer when they are about to abuse drugs. Both parties remain anonymous to each other, but the app records the user’s emergency plan, and the volunteer communicates with them before, during, and after to ensure they are safe. Although these volunteers are human today, there is a strong use case for AI with the Brave app. (In a drastic situation, does it matter if it is a human or AI that messages you to ensure you’re safe?)
While Brave is pioneering the use of tech to keep drug users from overdosing, another company is revolutionizing addiction treatment. QuickMD was blazing the telemedicine trail more than a year before Americans embraced it en masse at the outset of the pandemic. The company developed an innovative approach to OUD treatment using “teleMAT,” the combination of telemedicine and MAT (medication-assisted treatment).
MAT is a newer alternative to methadone, the commonly known legacy OUD treatment available since the 1970s. The “medicine” portion of MAT is Suboxone, the most popular brand name of buprenorphine + naloxone. Suboxone is a key component of treating OUD because it eliminates withdrawal symptoms commonly associated with opioids while also removing the cravings a patient feels when abstaining from them.
TeleMAT raises treatment to the next level by both making it available to those in need and helping patients feel more comfortable with treatment. Why is this so critical? Before QuickMD, an opioid user ready to kick the habit might only take advantage of the MAT approach if they were lucky enough to live close to a treatment center. In rural areas where help was often most needed, up to half of opioid users did not have access to MAT programs.
TeleMAT changes this situation completely. QuickMD has licensed physicians throughout the United States, eliminating the plight of patients who can’t find a nearby doctor. Another benefit of the “tele” portion of teleMAT is that it eliminates social stigma concerns. Especially in a small town, a visit to an addiction center can inform a patient’s friends and loved ones of matters they might prefer to remain private. TeleMAT patients can instead reach their doctor from the car or any other private location—the addiction center has been replaced by the smartphone.
And like any smartphone application, patients may be concerned about the security of their medical data. But as QuickMD COO Jared Sheehan explains, “Security is a major focus for us. Both our patients and physicians deserve the highest levels of privacy and protection, so we are constantly focusing on security practices and are launching two-factor authentication to further raise our security level. On top of cybersecurity, we follow all HIPAA rules to protect our patients’ privacy, and I believe our implementation of security protocol is better than any physical addiction center.” This dedication to cybersecurity is extremely important for any organization harnessing the power of telemedicine, especially in sensitive areas of treatment such as addiction.
TeleMAT is a unique innovation to defeat opioid addiction, a major societal problem, but those interested in state-of-the-art tech may wonder, “What does this have to do with AI and Big Data?” The answer is nothing short of a gamechanger. Traditional data on opioids comes in the form of overdoses, hospitalizations, and deaths. QuickMD’s dataset, drawn from tens of thousands of patients but anonymized and devoid of identifying information, creates a completely different picture, one revealing hot spots of OUD patients who need help and treatment before they become a statistic.
The most interesting thing about QuickMD’s approach to Big Data is how it drives the company’s efforts to bring new doctors to the program. Dr. Talib Omer, an assistant professor in emergency medicine at Keck Medicine of USC and the founder of QuickMD, explains: “One of the fascinating outcomes of the dataset we’re building is the ability to locate those areas most needed to effectively treat OUD. We know in real time where the demand for teleMAT is, and we use the information to rapidly deploy doctors where they can be beneficial. Providing such specialized help in real time would be unthinkable just a few years ago.”
QuickMD uses this approach for other aspects of its practice. Anonymized patient data can provide more critical information, such as what component of teleMAT treatment is cumbersome, and what other aspects of treatment are successful with patients. As Dr. Omer explains, “We are a piece of the puzzle, and data helps show what other pieces fit well with ours.” This information helps guide the company, including its recent decision to partner with Pear Therapeutics, a leading provider of digital CBT (cognitive behavioral therapy).
Beyond such developments, QuickMD offers an ambitious vision of OUD treatment in five to 10 years. Patient successes will create a virtuous cycle in which sound treatments become more effective as physicians and data experts more closely collaborate. At the same time, AI will begin to play a larger role in treatment, especially for at-risk patients. Meanwhile, wearables incorporating the monitoring of vitals will ensure overdoses are avoided based on evaluating key metrics such as sleep intervals and brain function. Likewise, for patients already in teleMAT treatment, AI chatbots could perform check-ins, flagging nurses and physicians if their immediate attention is needed.
This autumn needn’t be such a dismal season after all. Hope is on the horizon—brought to us by tech. As it turns out, Facebook’s digital army of “fact-checkers” won’t be displacing us anytime soon. Even better, thrilling advances in OUD treatment show what’s possible when mankind and machines collaborate for the good of all.
Post a Comment