Myths and Realities of AI-Healthcare
Editor’s Note: AI has had a transformative effect on many industries, the healthcare industry included. In this insightful piece, Kerrie Holley, SVP and Technology Fellow at Optum, and Dr. Siupo Becker, VP of Health Care Strategies at UnitedHealthcare, explore some of the ways that AI and healthcare providers have been able to work together, while dispelling some popular myths about the role of AI in healthcare. We’d love to hear from you about what you think about this piece.
AI Healthcare Myths
There is so much excitement involving AI in healthcare, but what exactly is AI in healthcare supposed to fix? People look to AI to predict future disease, prevent disease, enhance disease treatment, overcome obstacles to health care access, solve the burden of overworked and burnt out clinicians, and overall improve the health of people while decreasing the cost of healthcare. While some of this is achievable, AI is not a miracle panacea to all health and healthcare related problems.
Roy Amara, a previous head of the Institute of the Future, coined Amara’s Law, which states, “We tend to overestimate the effect of a technology in the short run and underestimate the effect in the long run.” One of the major myths regarding AI is that AI will replace physicians and other healthcare providers. AI relies upon the knowledge base provided by trained and experienced clinicians. AI cannot replace the “care” aspect of human interaction and its associated documented therapeutic effect. AI does not have the capability to determine the best solution when a holistic review of a patient would recommend an approach that relies on human creativity, judgment and insight. For example, take an otherwise healthy ninety year old patient who develops an imminently treatable cancer. Logic and current medicine would support aggressive treatment to destroy the cancer. The human aspect comes into play when this same patient lets their clinician know that they are widowed and alone, and although not depressed, feel they have lived a full life and decline treatment. AI and most physicians would argue for treatment. Patient autonomy and holistic review of an individual’s wishes and autonomy in their healthcare decisions takes priority here and would have been missed by an autonomous AI agency operating without human oversight.
AI can apply counterintuitive strategies to health management, but the steps from raw data to decision are complex and need human perceptions and insights. The process is a progression, starting with clinical data, obtained from innumerable sources that is built and developed to become relevant information, which is then used and applied to populations and/or individuals. The transformation from raw data to insights to intelligence is a process that is guided by clinicians working with data scientists using AI. The clinical interpretation of data is dependent upon humans and their understanding of disease processes and its effect on the timeline of progression of disease that molds this early knowledge. Algorithms for disease management, identification of risk factors predicting the probability of development of disease, all this is based on human understanding and interpretation of the disease process and the human state. The use of AI and clinicians activities are intertwined and together the potential for improving health is remarkable.
AI has inherent benefit and broad application, but it is AI in collaboration with the human interface which allows AI tools to be so impactful. AI will not replace health care providers, but is a powerful tool to augment disease identification, and management along with the physician. Let’s explore some of the ways that AI and healthcare providers have been able to work together, while also dispelling the myths that AI can do it all on its own.
AI Will Cure Disease
AI is not a replacement for a medicinal cure that may one day end diseases (e.g., coronary artery disease or cancer); however, advances in AI, the massive accumulation of data (i.e., Big Data), and data sharing in health care could lead to what does. Some people believe that if AI can be used to predict who is at risk for disease, then we can intervene and change behaviors or start treatment that would circumvent the disease from ever becoming present. Of course helping people avoid getting a disease is not the same as curing a disease. Defining what we mean by a cure can be confusing, and is never more evident than in certain diseases, such as Human Immunodeficiency Virus (HIV). Magic Johnson, NBA All star proclaimed he was cured from HIV, because doctors were unable to detect virus in his body after, and with ongoing treatment for HIV. Without the anti-retroviral medications, HIV would have increased in number and once again been found in his body. Was he ever truly cured? For certain diseases what defines a cure is not well defined. However, preventing a disease for an individual is better than trying to cure that disease.
The norm today is that we do this also, often without AI, working to prevent disease. Routinely, healthcare companies take in data from Electronic Health Records (EHRs), health care claims, prescriptions, biometrics and numerous data sources to create proprietary models for the identification of “at risk” patients. Healthcare constituents need to use artificial intelligence to support decisions and make recommendations based on the assessment and findings of clinicians providing healthcare to all patients for prevention to be effective.
The healthcare ecosystem comprises consumers in need of health care services, clinicians and providers who deliver health care services, the government who regulates, insurance and payers who pay for services and the various agencies who administer and coordinate services. In the ideal state each of these ecosystem constituents are in sync optimizing patient care. A simple example is medical coding where the medical jargon does not always sync with the coding terminology resulting in gaps in care appearing in identifying the true disease process of an individual patient. The current system is heavily dependent on coding of diagnoses by hospitals, providers, medical coders and billing agents. This coding process, CAC, discussed previously, will improve as we see greater AI adoption increasing the opportunity to prevent diseases.
AI can provide clinicians with more and better tools, augment a clinician’s diagnostic capabilities by analyzing a holistic picture of the individual patient with broader data streams and technological understanding of the disease process and who is both at risk and will be most greatly impacted. AI has become a more accurate tool for identifying disease diagnosis in images and with the rise of intelligent spaces (e.g. hospitals, homes, clinicians’ work spaces) AI triggers a more viable source of diagnostic information. The volume of data streams makes this unmanageable for a human but highly possible for intelligent machines powered by AI. By identifying and stratifying individuals most at risk AI can intervene to alert physicians and healthcare companies to intervene and address modifiable risk factors to prevent disease.
AI algorithms, personalized medicine and predictive patient outcomes can be used to study different diseases and identify the best practice treatments and outcomes leading to potential increase in cure of a specific disease. AI can further analyze whether and why a specific population may not respond to certain treatments versus another.
One area where AI has made significant impact already, is in identification of cancer in radiologic studies. University of Southern California Keck School of Medicine published a study in 2019 showing improved cancer detection using AI. Specifically, “a blinded retrospective study was performed with a panel of seven radiologists using a cancer-enriched data set from 122 patients that included 90 false-negative mammograms.” Findings showed that all radiologists experienced a significant improvement in their cancer detection rate. The mean rate of cancer detection by radiologists of 51% increased to a mean of 62% with use of AI, while false positives (detection of cancer when cancer is not actually present) remained essentially the same.¹ Early detection of disease states makes an enormous positive impact on treatment and cure in cancer outcomes not just with radiologists but similarly with dermatologists. Computer scientists at Stanford created an AI skin cancer detection algorithm using deep learning for detection of malignant melanoma (a form of skin cancer) and found that AI identified cancers as accurately as dermatologists. China has used AI in the analysis of brain tumors. Previously, neurosurgeons performed tumor segmentation (used in diagnosis of brain cancer) manually. With AI the results were accurate, reliable and created greater efficiency. Where early and accurate detection can lead to cure AI has a proven place in the diagnosis of cancers.
Identifying and diagnosing cancer in its early stages coupled with proper care treatment increases the opportunity to cure patients of cancer. AI can significantly contribute to the diagnosis of patients based on signals or symptoms missed by human detection. Gathering data from large populations pools allows an understanding, and awareness of the most effective plans for treatment. This leads to better treatment for individual patients based on what has been proven as the most effective treatment plans for other patients who have had similar types of cancers. This is where AI and specifically deep learning algorithms using large data sets containing data of millions of patients diagnosed with cancer all over the world, plays a role. AI can provide evidence based recommendations to clinicians on the treatment plans with the greatest opportunity for positive outcomes.
Human error can be life or death for cancer patients, detecting it early makes all the difference. AI helps with early detection and diagnosis by augmenting current diagnostic tools for clinicians. An important part of detecting lung cancer is finding if there are small lesions on the lungs from Computed Tomography (CT) scans. There is some chance of human error and this is where Artificial Intelligence plays a role. Clinicians using AI increase the likelihood of early detection and diagnosis which is a matter of life and death for these patients. With the big data available in relation to cancer and its treatment, AI has the potential to assist in creating structure out of these databases and pulling relevant information to guide patient decision making and treatment in the near future.
Giving patients tools which help them, also helps their doctors. Tools which rapidly analyze large amounts of patient data to provide signals that may otherwise go undetected. Machine learning algorithms that can learn about a patient continuously should increasingly play a role in healthcare. The shift of largely using computers to using devices that patients wear for early detection of disease states is on the rise. In spite of the plethora of good news seemingly almost daily of new machine learning algorithms or new devices or AI products improving healthcare, AI alone will not fix the healthcare problems facing societies. There are many challenges needed to be solved to make our healthcare system work better and AI will help but let’s dispel the next myth about AI alone fixing the problems facing healthcare.
AI Will Replace Doctors
AI will not replace doctors now or in the near future. There is a lot of discussion suggesting that AI will replace doctors. In 2012, entrepreneur and Sun Microsystems co-founder, Vinod Khosla, articulated computers will replace 80 percent of what doctors do. Vinod sees a future where machines will replace 80 percent of doctors in a healthcare future that will be driven by entrepreneurs, not medical professionals. He wrote an article with the provocative title, “Do We Need Doctors Or Algorithms?”²
We can look at AI as a doctor replacement thing or a clinician augmentation. AI can perform a double check and see patterns from millions of patients that one doctor cannot possibly see. One doctor can’t possibly see a million patients across their lifetime but AI can. The diagnostic work done by a doctor arguably focuses heavily on pattern recognition. So augmenting diagnoses with AI makes sense.
Key arguments for AI replacing doctors include:
- AI can get more accurate every day, every year, every decade at a rate and scale not possible for human doctors
- AI will be able to explain possibilities and results with confidence scores
- AI can improve the knowledge set, raise the insight of a doctor (perhaps not trained in a specific specialty)
- AI may be the only way to provide access to best-in-class health care to millions of people who don’t have access to healthcare services or can’t afford healthcare services
Arguments for not replacing doctors include:
- Doctors are better at decision making
- Doctors have empathy which may be critical to clinical care
- Doctors have a human connection which may directly influence how a patient feels and facilitate a patient adhering to a treatment plan
- American Medical Association (AMA) Journal of Ethics states “a patients’ desire for emotional connection, reassurance, and a healing touch from their caregivers is well documented”
- Doctors may observe or see critical signals because of human senses
- AI cannot converse with patients like a human doctor
- Subsconscious factors that may influence a doctor’s ability to treat, if not explicitly identified for AI, will be missed
There are many tasks that AI can do better than a doctor but rarely if at all will AI replace entire business processes, operations, occupation or profession. The most likely scenario is doctors in the foreseeable future will transition to a doctor who wields AI, understanding how to use AI tools for delivering more efficient and better clinical care. AI today provides a lot of point solutions and its opportunity to improve diagnostics is significant. Treatment pathways and even many diagnostics today require decision making something AI is not good at.
A practical problem exists in that AI must live in our current brownfield world where several barriers must be overcome for AI replacing doctors. Today we have a proliferation of systems that do not integrate well with each other. For example, a patient who is cared for at a hospital, urgent care center or provider office may have his/her data spread across several different systems with varying degrees of integration and today a doctor’s ability to navigate the healthcare system is critical to patient care.
The reality is there is not going to be a computer, machine or AI that solves health care, just like there isn’t one solution to all banking, retail, or manufacturing. The path to digitization differs based on clinical speciality and most likely will occur one process at a time within domain or speciality. AI systems are here and on the horizon for assessing mental health, diagnosing disease states, identifying abnormalities and more.
AI Will Decrease Healthcare Costs
US health expenditure projections for 2018–2027 from CMS [cms.gov] show that the projected average growth rate on health care spend is 5.5% with expectations of meeting 6 trillion in spend by 2027. Looking at these numbers, it’s clear that healthcare spending will outstrip economic growth. All components of healthcare are projected to increase at exceedingly high annual rates over the next decade. For example, inpatient hospital care, which is the largest component of national health expenditures, is expected to grow at an annual average rate of 5.6%, which is above its recent five year average growth rate of 5%. AI alone won’t fix these problems but it can help with cost containment and cost reduction.
The myth about AI and healthcare costs is thinking that AI will reinvent or overturn the existing healthcare or medical models in practice today. Or that AI will transform or revolutionize the healthcare industry triggering enormous savings. The reality is that AI as a general purpose technology will be transformative. Tremendous evidence abounds showing that the big technology companies and startups will transform how healthcare is done. AI will be the tool of trade making many of these transformations possible. You might see this as splitting hairs but the point is that the problems facing healthcare lie with resistance to change, historical inefficiencies and inertia, lack of cooperation for the greater good by companies designed to compete with each other and the lack of a game changing technology. Now we have the game changing technology, AI.
Given this background, how and where will AI have an impact? One major area of spend is management of chronic disease. When people with a chronic condition are not adherent to treatment plans, then complications related to the underlying disease occur which can result in expensive hospitalizations and/or need to institute high cost specialty pharmacy therapies. For example, Diabetic Retinopathy (or DR — an eye disease of diabetes) is responsible for causing 24,000 Americans to go blind, annually, as reported by the Centers for Disease Control (CDC). This is a preventable problem and routine exams with early diagnosis and treatment can prevent blindness in up to 95% of diabetics. And yet, greater than 50% of diabetics don’t get their eyes examined, or are too late for effective treatment. Diabetes related disease and blindness is estimated to total more than $500 million per year.³
AI can address cost spending in chronic disease through increasing efficacy and ease of obtaining eye exams in diabetics. AI using machine learning and deep learning have been adopted by various groups to develop automated DR detection algorithms, some of which are commercially available. Although binocular slit-lamp ophthalmoscopy remains the standard against which other DR screening approaches are compared, AI applications with fundus photography are more cost-effective and do not require an ophthalmologist consultation. The scarcity of specialists, ophthalmologists, necessitates the use of non physicians who can do the DR screening using AI algorithms (deep learning) embedded in various tools like a mobile device. Such solutions are already used in developing nations with a scarcity of ophthalmologists, that is, a highly trained doctor is not required for disease detection where AI can be trained to do the same.
Cost savings using AI and fundus photography have been reported at 16–17% (due to fewer unnecessary referrals). A cost effectiveness study from China showed that though the screening cost per patient increased by 35%, the cost per quality-adjusted life year was reduced by 45%.⁴ Taking this a step further, to help diabetics with existing eye related disease, AT&T partnered with Aira combining smart glasses with AI algorithms to improve the patient’s quality of life and increase medication adherence by medication recognition technology. The technologies for medication adherence are varied such as intelligent pill bottle tracking when patients open their bottle using sensors in the cap or sensors in the bottle showing weight decline. Or mobile apps or smart speakers which alert patients to take their medications and these applications can also be armed with machine learning to teach the patients habits, and instead of sending annoying alerts learns the optimal time of day or scenario to send a reminder.
All of these varied strategies using AI in chronic condition management lead to ongoing waterfall health effects and cost savings. In this case, the Diabetic is now screened through ease and efficacy of AI for Diabetic eye related disease, if they are found to have significant disease then the AT&T and Aira “pill bottle reader project” can facilitate medication adherence and issues such as falls related to poor vision, fractures/other musculoskeletal trauma, hospital admissions for poor blood sugar control, infections related to poor blood sugar control, etc can all be avoided at significant cost savings and quality of life improvement for patients.
AI can assist in standardizing and identification of best practice management of chronic disease. It has always been known in the medical community that there is a wide variation in practices, and economists have pointed out that treatment variability results in wasteful health care spending. As an example of how treatment management can impact health care spend, let’s examine low back pain. Over 80% of Americans experience low back pain at some point in their lives. Of those people with low back pain 1.2% account for about 30% of expenditures. When treatment guidelines were adhered to, costs were less. The pattern for increasing costs to both the overall healthcare system and patients is reflected in Figure 1–4 illustrating the impact on costs when consumers, that is, patients fail to adhere to treatment guidelines.
AI can reduce treatment variability by applying AI to the myriad of siloed data sources to identify optimal care pathways leading to updating of current guidelines and improvement in cost expenditure.
Hospitalization and administrative costs are also a major driver for US health care. “More than $1 trillion is wasted each year on costly administration and avoidable hospital readmissions,” per Orion Health CEO Ian McCrae.⁵ Orion Health used AI to predict patient costs and readmission risks, while analyzing clinical and financial outliers to improve the treatment and practice management at the point of care. AI is being applied to the large data pools coming in through numerous inputs including socio-economic data, behavioral data, biometric data, demographics, geographic location, etc, all in an effort to more accurately predict who would benefit from more aggressive treatment management with the goal of avoiding hospitalizations.
Another area of high health care spend is medication research and discovery. Historically, new medications and vaccines took a long time to develop, and demonstrating safety and efficacy was a tedious time-consuming process. Now AI could facilitate this process by increasing the speed of the process in expediting the analysis and research components.
Accenture performed a study in 2017 showing a potential savings of $150 billion in annual health care cost savings by 2026 with the application of AI to health care in the areas of: robot-assisted surgery ($40 B), virtual nursing assistants ($20 B), administrative workflow assistance ($18 B), fraud detection ($17 B), dosage error reduction ($16 B), connected machines ($14 B), clinical trial participant identifier ($13 B), preliminary diagnosis ($5 B), automated image diagnosis ($3 B), and cybersecurity ($2 B).10 These statistics show the potential, value and transformative effect in adopting AI to reduce costs. But these are potential future costs reduction and the real rub lies in translating these costs to reduced patient costs.
Translational challenges continue to exist and include issues such as costs of implementing this technology, and cultural acceptance of AI in healthcare. All of which continue to act as barriers to AI use in helping to control healthcare spend. Investments in AI can take away from short term financial goals. Legacy systems usually exist as silos and the time and expense of creating interoperability can be cost or time prohibitive. Providers themselves may be unaware of the benefits of AI and its applications to their practices in improving the lives of their patients. Safely and respectfully controlling the confidentiality of patient information through these systems and use of AI is another potential barrier. That said, the future of AI in healthcare in both controlling health care costs and improving the health of our population is clear. AI has opened a world of potential opportunities through various means to work with clinicians and health care systems to better the lives and health of our population. The future is full of potential for AI use in healthcare, but limitless, it is not.
¹ https://link.springer.com/journal/10278/32/4
² https://techcrunch.com/2012/01/10/doctors-or-algorithms/
³ https://www.cdc.gov/visionhealth/pdf/factsheet.pdf
Learn faster. Dig deeper. See farther.
Join the O’Reilly online learning platform. Get a free trial today and find answers on the fly, or master something new and useful.
Kerrie Holley is focused on advancing UnitedHealth Group (UHG) in the adoption of emerging technologies — including machine learning, graph databases, deep learning, natural language processing, IoT, virtual reality, genomics, blockchain, and virtual assistants — with a heavy focus on Artificial Intelligence. Kerrie leads an engineering team committed to applying and incubating emerging technologies to make the health care system work better and help people live healthier lives. Dr. Siupo Becker joined Optum/UnitedHealthcare in 2016, focusing on population health management and data analysis focusing on application to improving health care and quality outcomes. She has since been named as Vice President of Health Care Strategies within UnitedHealthcare and drives nationwide health care initiatives that impact the quality and control of health care costs.
