Industry Disrupted: The Impact of Automation on Healthcare

Artificial intelligence (AI) has officially segued from the stuff of Arthur C. Clarke novels into reality’s mainstream, and self-driving cars are just the tip of the iceberg. Sometimes AI’s impact is quiet, like an algorithm that helps you find information quicker. Other times it has the potential to completely revolutionize an industry. Such is the case for healthcare where, according to an analysis from from the Office of the National Coordinator for Health IT and JASON, a group of independent scientists, the rapid evolution of AI is significant.

There are several factors experts say make the industry “ripe” for AI technology, beginning with mounting frustration with existing medical systems that are both tedious and error-prone. The rise of big data, cloud networking, and the smart devices is another significant factor, as is consumers’ growing comfort level with at-home services like online buying and distance-based degrees. Let’s look at how the technology is shaping global healthcare for practitioners and patients.

Artificial intelligence is no small part of the American healthcare system: Accenture predicts the market is expected to grow by 40 percent each year through at least 2021. Artificial intelligence can play a role in everything from medical records management to heart surgery, and with notable success. According to Intel, for instance, researchers found its Saffron AI system could diagnose heart conditions faster and more accurately than doctors. Such technologies could improve health outcomes and save time, improving care accessibility by reducing costs and letting doctors dedicate more time to their patients. Here are just some of the roles AI plays in various stages of healthcare, as reported by PwC:

• Preventative care: Improving wellness through patient care portals, AI chatbots, and health apps.
• Diagnosis: Early disease detection using patient health data, scientific literature, and diagnostic images.
• Treatment: Better decision-making using predictive analyses to identify high-risk patients and prioritize health tasks.
• Drug therapy: Drug interaction and overdose prevention applications.
• Long-term care: Coordinating care plans and boosting patients’ adherence to them.
• End-of-life care: Helping people remain independent longer, thereby reducing hospitalizations and assisted care terms.
• Research: Streamlining medical discoveries and trials, reducing the time and cost required to bring new therapies to market.
• Training: Providing more natural, simulatory medical training and making it accessible anywhere.

The ways AI supports the above stages of healthcare are varied, and, thanks to rapid technological advancement, become more so with each passing day.

Healthcare AI is nothing if not tenacious: things that seemed impossible last year may be old news by the next. For example, according to Wired, smartphone apps can now not only help you eat or exercise better, but they also can track changes in your talking or texting patterns that could indicate depression or suicide risk. Those patients can then turn to chatbots programmed in cognitive behavioral therapy while waiting to see a human professional. So, which types of AI technology are growing the fastest? An Accenture market analysis indicates some of the largest investments in health AI dollars are in the following applications:

• Robot-assisted surgery
• Virtual nursing assistants
• Administrative workflow assistance
• Dosage error reduction
• Clinical trial participate identifying
• Preliminary diagnoses
• Automated image diagnoses

This list only underscores the notion that AI has a place in all levels of healthcare. What does that mean for its practitioners?

The idea that robots are coming to take our jobs is certainly not new, and with so many medical tasks ripe for automation, one might worry for healthcare jobs. The news from Accenture’s analysis is generally positive. The firm suggests, for instance, that AI can fill the gaps in rising job shortages by allowing existing workers to do their jobs better and faster. Improved portability and accessibility of such technologies can also make it easier to expand the patient base to include those living in traditionally remote or underserved areas. Artificial intelligence will also create new jobs, if of a different nature: demand for programmers and IT professionals knowledgeable of health AI and machine learning is on the rise.

Artificial intelligence’s impact on healthcare professionals is not limited to hiring needs. As Accenture points out, medical practitioners can only enjoy the fruits of AI’s labor if they know how to use it properly. The industry needs an AI-savvy workforce, which requires not just an overhaul in training and education, but of the entire healthcare culture: it only works well if everyone is on board. Even the most experienced workers must be open and committed to change.

This is not to say AI is perfect. Like all technology, it has its limitations.

As dramatic as its impact might be, healthcare AI has not taken over the industry and, according to some experts, probably never will. Here are a few reasons why.

Data and privacy concerns

“Big data” is now a household term, and with good reason: wearables, phones, and other smart devices constantly monitor and store information about our preferences and behaviors. The same can be said of health information, but only to a point. Privacy laws and HIPPA regulations limit the type of data applications collect and how it can be used. According to Wired, the way this information is stored also makes it difficult to use and share across platforms, making it inaccessible to AI tech.

Regulatory difficulties

The U.S. Food and Drug Administration (FDA) evaluates the efficacy and safety of medical devices and therapies before they hit the market. This is a straightforward (if lengthy) process when it comes to drug therapies; not so with AI-enabled tech. For starters, AI is relatively new, leaving regulators to play catch-up when it comes to understanding its impact and potential. Then, according to another Wired report, the ever-changing landscape of tools makes an already complicated regulatory process even more so, prompting the FDA to establish a taskforce dedicated exclusively to digital health.

The human factor

Perhaps one of the biggest limitations to even broader adoption of AI healthcare technology is that fact that it simply is not human. Even the best programmed AI cannot approach medical problems as creatively as human researchers, nor can it understand how solutions might affect patients on a personal or emotional level. Also, chatbots aside, many patients are more comfortable discussing health concerns with a trained clinician than with a machine no matter how efficient the latter might be. In the end, bedside manner is nearly as important as outcomes.

It is impossible to say if (and how) AI will overcome these hurdles, or how patients will respond to these changes. At least for now, expect a human doctor—not a robot—to attend your next medical appointment, even if some of the scheduling, diagnostic, and record-keeping tasks received an assist from AI’s growing capabilities.

Aimee Hosler

Writer

Aimee Hosler is a long-time journalist specializing in education and technology. She is an advocate for experiential learning among all ages and serves as the director of communications for a non-profit community makerspace. She holds a degree in journalism from California Polytechnic State University in San Luis Obispo.