Abstract:
Physicians historically have embraced technology to enhance the diagnosis and treatment of medical conditions. We have not shied away from CT scans, bone marrow transplants, or immunotherapy. However, we have been slow to implement technology in our office settings. Other industries, such as airlines, hospitality, and finance, have used technology to streamline the customer experience. The future of medicine is going to require that we make use of the technology that is readily available to provide our patients with a better experience.
Because of the use of technology by Amazon, Facebook, and Uber, a new term has entered our lexicon. The Amazon effect describes the expectations of customers or patients who have become accustomed to the level of convenience provided by Amazon and seek the same experience with every online—and now off-line—service.
Patients are less willing than before to spend time on mundane procedures, such as repeatedly providing their demographic and insurance information or using telephones to schedule an appointment. They expect instant access to medical care, outstanding customer service, immediate test results, and the ability to access their health records from every electronic device they have. And exactly as with Amazon, they expect transparent prices and extensive data that will help them make informed decisions related to their health. Artificial intelligence (AI) just may be the solution to the expectations of patients who have experienced and expect the same response from healthcare as they have received from Amazon.
A multitude of opportunities are available to leverage technology to deploy more precise, efficient, and impactful interventions at exactly the right moment in a patient’s care, beginning with scheduling appointments, online payments, obtaining results of imaging and lab tests, and providing current and accurate educational materials for various medical conditions. Now is the time for the healthcare profession to make use of AI, which is poised to be the engine that drives improvements across the continuum of care from the first telephone call to the office to interacting with the practice between appointments.
AI offers a number of advantages over traditional analytics and clinical decision-making techniques. Learning algorithms can become more precise and accurate as they interact with training data, allowing doctors and their patients to gain unprecedented insights into diagnostics, care processes, treatment variability, and patient outcomes.
At the 2018 World Medical Innovation Forum on AI presented by Partners Healthcare, leading researchers and clinical faculty members showcased the 12 technologies and areas of the healthcare industry that are most likely to see a major impact from AI within the next decade.(1)
Using computers as a vehicle for communication is not a new idea, but creating direct interfaces between technology and the human mind without the need for keyboards, mice, and monitors is a cutting-edge area of research that has significant applications for some patients. Neurological diseases and trauma to the nervous system can take away some patients’ abilities to speak, move, and interact meaningfully with people and their environments. Brain–computer interfaces (BCIs) backed by AI could restore those fundamental experiences to those who feared them lost forever.
BCIs could drastically improve quality of life for patients with ALS, strokes, or locked-in syndrome, as well as the 500,000 people worldwide who experience spinal cord injuries every year.(2)
New Radiology Tools
Radiologic images obtained by MRI machines, CT scanners, and conventional x-rays offer noninvasive visibility into the inner workings of the human body. But many diagnostic processes still rely on physical tissue samples obtained through biopsies, which often include significant risks.
AI will enable the next generation of radiology tools to be accurate and detailed enough to replace the need for tissue samples in some cases, experts predict. AI is an opportunity to bring together the diagnostic imaging team with the surgeon or interventional radiologist and the pathologist. There is no doubt that putting together different teams and aligning goals is going to be a big challenge.
Succeeding in this quest may allow clinicians to develop a more accurate understanding of how tumors behave as a whole instead of basing treatment decisions on the properties of a small segment of the malignancy. Providers also may be able to better define the aggressiveness of cancers and target treatments more appropriately. AI is helping to enable “virtual biopsies” and advance the innovative field of radiomics, which harnesses image-based algorithms to characterize the phenotypes and genetic properties of tumors.
Shortages of trained healthcare providers, including ultrasound technicians and radiologists, can significantly limit access to life-saving care in underserved rural areas of the United States. AI could help mitigate the impacts of this deficit of qualified clinical staff by taking over some of the diagnostic duties typically allocated to humans. For example, AI imaging tools can screen chest x-rays for signs of tuberculosis or lung cancer, often achieving a level of accuracy comparable to that achieved by humans. This capability could be deployed through an app available to providers in low-resource areas, reducing the need to have a trained diagnostic radiologist on site.
EHRs have played an instrumental role in the healthcare industry’s journey toward digitalization and data capture, but the switch has brought myriad problems associated with cognitive overload, endless documentation, and user burnout. Perhaps one of the main reasons that middle- aged and older physicians are leaving medicine is the requirement to use the EHR.(3) Users spend the majority of their time on three tasks: clinical documentation, order entry, and sorting through the in-basket. EHR developers are now using AI to create more intuitive interfaces and automate some of the routine processes that consume so much of a physician’s time. Voice recognition and dictation are helping to improve the clinical documentation process, but natural language processing tools might not be going far enough. Changes such as video recording a clinical encounter may enable AI to document accurately the doctor–patient encounter. Then AI and machine learning can be used to index those videos for future information retrieval.
The virtual assistant will soon be available in the home and in independent living communities. Siri and Alexa will be able to remind patients to take their medications, refill their prescriptions, and exercise regularly. AI also may help to process routine requests from the inbox, such as medication refills and lab results notifications. It also may help to prioritize tasks that truly require the clinician’s attention, making it easier for users to work through their to-do lists. For example, if a chest x-ray is abnormal or there has been a significant change since a previous x-ray, this information can be prioritized in the physician’s inbox and allow that physician to take action on the findings very quickly.
Controlling Antibiotic Resistance
Antibiotic resistance, perhaps as a result of giving antibiotics to cows and pigs, is a growing threat to populations around the world, because overuse of these critical drugs fosters the evolution of superbugs that no longer respond to treatments. Multi–drug-resistant organisms can wreak havoc in the hospital setting and claim thousands of lives every year. Clostridium difficile alone accounts for approximately $5 billion in annual costs for the U.S. healthcare system and claims more than 30,000 lives each year.(4)
One of the issues and obstacles will be the reluctance of the public to share their healthcare data.
Data from the EHR can help to identify infection patterns and highlight patients at risk before they begin to show symptoms. Leveraging machine learning and AI tools to drive these analytics can enhance their accuracy and create faster, more accurate alerts for healthcare providers and decrease the risk of these potentially lethal infections.
In 2018, the FDA took a historic step with the approval of a new AI-based system for the detection of diabetic retinopathy, a diabetes-related eye disorder that can result in permanent blindness. This marks the first time that a fully automated, AI-based diagnostic tool has reached the market in the United States—one that does not require additional expert review and is therefore suited for use in primary care and low-resource settings. This AI tool represents a potential advancement in care for a condition that affects a large swath of people worldwide. In the United States alone, over 30 million people live with diabetes, and roughly 24,000 a year experience vision loss due to diabetic retinopathy. And more than half of patients with diabetes do not seek regular eye exams, highlighting the potential role that primary care providers can play in recognizing the early signs of diabetic eye disease.(5)
Data Management
AI tools offer the healthcare profession the opportunity to harness the voluminous amount of data that EHRs generate. Every day in our digital world, we generate more than 2.5 million terabytes of data. (That’s 2.5 followed by 12 zeros.) We are sitting on mountains of data that can be retrieved, analyzed, and then utilized to create smarter, faster clinical trials and drug development that can reduce the cost of product development as well as bring the new drugs and products to market much faster than the current method and systems.
Pathologists provide one of the most significant sources of diagnostic data for providers across the spectrum of care delivery. Seventy percent of all decisions in healthcare are based on a pathology result.(6) Somewhere between 70% and 75% of all the data in an EHR are from a pathology result.(7) AI has the opportunity to deliver a digital pathology report, which will make the diagnosis of disease much faster and avoid delays in treatment that are not only costly but also anxiety-producing for patients.
Analytics that can drill down to the pixel level on extremely large digital images can allow providers to identify nuances that may escape the naked human eye. We are getting to the point where we can do a better job of assessing whether a cancer is going to progress rapidly or slowly and how that might change how patients will be treated based on an algorithm rather than clinical staging or the histopathologic grade noted on a pathology slide that may take several days to produce a report. AI also can improve productivity by identifying features of interest in slides before a human clinician reviews the data.
In the medical environment, smart devices are critical for monitoring patients in the ICU and elsewhere. Using AI to enhance the ability to identify deterioration, suggest that sepsis is taking hold, or sense the development of complications can significantly improve outcomes and may reduce costs related to nosocomial situations.
Currently, humans cannot aggregate large amounts of data. However, with the use of AI, we can identify conditions before they become clinically relevant, and, as a result, we can intervene earlier and decrease morbidity and even mortality. Inserting intelligent algorithms into monitoring devices such as those used in the operating room or in the ICU can reduce cognitive burdens for physicians while ensuring that patients receive care in as timely a manner as possible.
Immunotherapy is one of the most promising avenues for treating cancer. By using the body’s own immune system to attack malignancies, patients may be able to control and even cure certain tumors. However, only a small number of patients respond to current immunotherapy options, and oncologists still do not have a precise and reliable method for identifying which patients will benefit from the use of immunotherapy. Machine learning algorithms and their ability to synthesize highly complex datasets may be able to identify new options for targeting therapies to an individual’s unique genetic makeup. This will be precision medicine at its finest.(8)
Recently, the most exciting development has been checkpoint inhibitors, which are a type of immunotherapy that blocks proteins that stop the immune system from attacking the cancer cells.(9)
EHRs are a goldmine of patient data, but extracting and analyzing that wealth of information in an accurate, timely, and reliable manner has been a continual challenge for providers and developers. One of the issues and obstacles will be the reluctance of the public to share their healthcare data. Healthcare breaches are not a rarity; they occur more frequently than we would like to admit. We don’t have to go back too far in our collective consciousness to recall Cambridge Analytica and Facebook collecting our data and perhaps even foreign countries impacting our elections. A recent report from the Department of Health and Human Services’ Office for Civil Rights showed that in 2016, in the United States alone, there were 329 reported data breaches of more than 500 records.(10) A report from the Ponemon Institute estimated that 90% of healthcare organizations have experienced a data breach in the past two years.(11) Every mishap of that sort is not only extremely dangerous but also expensive—the same report from the Ponemon Institute calculated that the average cost of a breach is $8 million!(11) As a result, patients will become cautious about sharing their healthcare data. However, patients tend to trust their physicians more than they might trust a company like Facebook, which may help to ease any discomfort with contributing data to large-scale research initiatives.
The selfie just may be the doctor of tomorrow.
The selfie just may be the doctor of tomorrow. The quality of cell phone cameras is increasing every year, and these cameras can produce images that are clear enough to enable analysis by AI algorithms. Dermatology and ophthalmology are early beneficiaries of this trend. Using smartphones to collect images of eyes, skin lesions, wounds, infections, medications, or other categories may help underserved areas cope with a shortage of specialists while reducing the time-to-diagnosis for certain complaints.
Researchers in the United Kingdom have even developed a tool that identifies developmental diseases by analyzing images of a child’s face. The algorithm can detect discrete features, such as a child’s jaw line, eye and nose placement, and other attributes that might indicate a craniofacial abnormality. Currently, the tool can match the ordinary images to more than 90 disorders to provide clinical decision support.(12)
As the healthcare industry shifts away from fee-for-service to value, so too is it moving further and further from reactive care. Getting ahead of chronic diseases, costly acute events, and sudden deterioration is the goal of every doctor—and reimbursement structures are finally allowing them to develop the processes that will enable proactive, predictive interventions. AI will provide much of the bedrock for that evolution by powering predictive analytics and clinical decision support tools that clue providers in to problems long before they might otherwise recognize the need to act. AI can provide earlier warnings for conditions such as seizures or hypoglycemia, which often require intensive analysis of highly complex datasets. Today, there are dogs that can predict glucose levels and the risk of impending seizures.(13) Tomorrow, those dogs may be replaced by AI.
Machine learning also can help support decisions around whether or not to continue care for critically ill patients, such as those who have entered a coma after cardiac arrest. Currently, doctors must visually inspect EEG data from these patients. The process is time-consuming and subjective, and the results may vary with the skill and experience of the individual clinician. With an AI algorithm and lots of data from many patients, it’s easier to match up what you’re seeing to long-term patterns and maybe detect subtle improvements that would affect decisions around care. Leveraging AI for clinical decision support, risk scoring, and early alerting is one of the most promising areas of development for this revolutionary approach to data analysis.
Bottom Line: By powering a new generation of tools and systems that make clinicians more aware of nuances, more efficient when delivering care, and more likely to get ahead of developing problems, AI will usher in a new era of clinical quality and exciting breakthroughs in patient care.
References
2019 disruptive dozen. Partners Healthcare. https://worldmedicalinnovation.org/wp-content/uploads/2019/04/Partners-FORUM-2019-BROCHURE-D12-AI-190326_0832-F-FOR-WEB-X3-SM.pdf .
Spinal cord injury—facts and statistics. SCI Info Pages. www.sci-info-pages.com/spinal-cord-injury-facts-and-statistics/ .
Bae J, Encinosa WE. National estimates of the impact of electronic health records on the workload of primary care physicians. BMC Health Serv Res. 2016;16:172.
Nearly half a million Americans suffered from Clostridium difficile infections in a single year. CDC. February 25, 2015. www.cdc.gov/media/releases/2015/p0225-clostridium-difficile.html
Watch out for diabetic retinopathy. Centers for Disease Control and Prevention; 2018. https://www.cdc.gov/features/diabetic-retinopathy/index.html
Aston G. Hospital labs go under microscope. Hosp Health Netw. 2014;88(5):36-40.
Bresnick J. Top 12 ways artificial intelligence will impact healthcare. Health IT Informatics. https://healthitanalytics.com/news/top-12-ways-artificial-intelligence-will-impact-healthcare .
Murphy B. Precision medicine should have its place in new pay models. American Medical Association. June 13, 2018. www.ama-assn.org/delivering-care/precision-medicine/precision-medicine-should-have-its-place-new-pay-models .
Immune checkpoint inhibitors and their side effects. American Cancer Society. www.cancer.org/treatment/treatments-and-side-effects/treatment-types/immunotherapy/immune-checkpoint-inhibitors.html
Largest healthcare data breaches. HIPAA Journal. January 4, 2017. www.hipaajournal.com/largest-healthcare-data-breaches-of-2016-8631/
Data breach study: impact of business continuity management. IBM. www.ibm.com/services/business-continuity/cyber-resilience
Ferry Q, Steinberg J, Webber C, et al. Diagnostically relevant facial gestalt information from ordinary photos. eLife. 2014;3:e02020. DOI: 10.7554/eLife.02020
Willingham E. Dogs detect the scent of seizures. Scientific American. March 29, 2019.
Topics
Environmental Influences
Adaptability
Technology Integration
Related
Case Study: Are the Right People in the Right Seats?Core Requirements on Every CallThe Vital Role of the Outgoing CEO