Artificial intelligence has the potential to revolutionize all fields, and healthcare isn’t exempted.
This technology, which involves machine and deep learning, enables computers to gain the capacity to better understand and process complex forms of data. Essentially, they would have the ability to learn through examples.
When implemented correctly, it’s a development that comes with many possibilities, especially in a data-driven field like healthcare. Machine learning has the potential to improve patient care, provide faster service and diagnoses, and generally provide a better experience for both healthcare providers and patients.
Anyone involved in healthcare (which basically means everyone) can stand to gain from learning more about how AI might affect the industry.
We are quickly moving to a patient-centric world in healthcare where treatment is coming to the patient, the patient is treated more like a customer, and medical facilities of all types must use technology from the business sector. Business sector software designed to improve the customer experience can now be used to improve the patient experience. No technology is driving this shift faster than artificial intelligence (AI). AI is propelling us into an increasingly digital medical experience where patients expect personalized experiences that take into account their individual needs and values, and empower them to get information fast and accurately.
Prescription drugs are ground zero for AI innovation
Although AI has been touted for everything from diagnosis to automating medical imaging to drug discovery, we believe that ground zero for AI innovation in patient-centric healthcare is prescription medicine. Prescribers and patients are suffering in countless ways from the complexity and associated errors in prescriptions.
A single drug has hundreds of factors that must be considered by a doctor or a pharmacist when prescribing or dispensing a drug to a patient. We examined 50 of the most popular drugs and found that the average number of considerations for a single drug is enormous:
It’s an understatement to say that AI and machine learning were among the forefront of conversation drivers recently at HIMSS19 in Orlando. One session actually diving into the ethics of AI while leaders from Microsoft, the American Medical Association and the Cleveland Clinic spoke of the need for organizations to develop the right framework for innovative clinical delivery.
Hemnant Pathak, associate general counsel at Microsoft, led the round table discussion, which included Peter Lee, CVP AI and research at Microsoft; Dr. Susanna Rose, Phd., chief experience officer at the Cleveland Clinic; and Sylvia Trujilo, MPP, JD, senior Washington counsel for the American Medical Association.
Virtually every industry is being sharpened by the emergence of these new technologies, Pathak said during his opening remarks, while pointing out the prevalence of the technology throughout the HIMSS exhibition hall. “In many cases, existing frameworks can be adapted, but as they continue to improve we’re going to need sources from every sector and they need to be given equal weight for their considerations” to the healthcare landscape.
From faster diagnosis to enhancing operations across care management, compliance and accounting, artificial intelligence (AI) is revolutionizing the healthcare landscape. While AI is not a new concept, its adoption in the healthcare industry has lagged behind because of significant changes in technology and the exponential growth of data. AI has tremendous application for healthcare, but leaders will need to balance data management and governance to properly enhance the patience experience and outcomes through AI implementation.
By Dr. Giovanni Colella, CEO and co-founder, OODA Health.
Giovanni Colella, MD
2018 was another year of rapid changes in the healthcare industry. Over the past twelve months, I’ve had fascinating conversations with executives in the provider and payer community and they have described the challenges they face and their visions for the future. One thing has become clear among everyone in the healthcare ecosystem: expectations are changing.
Patients are looking for a retail-like experience for their treatment, and sure enough, a number of new, innovative providers are gaining traction by leading with convenience and patient-centricity. Patients are embracing alternatives to the traditional doctor’s office or the local hospital. Unwilling to wait for an appointment, patients are eager for more responsive, high-quality treatment options.
Healthcare providers are becoming more innovative in an effort to improve the quality and efficiency of patient care while continuing to look for ways to reduce costs. Payers are exploring new approaches to improve their back-office systems and are willing to partner with third-party vendors to accelerate this improvement, rather than trying to develop everything in-house.
I am confident that this year will be full of important change in the industry and that these changes will touch the lives of millions. Of course, there is still a long way to go since the healthcare industry is historically slow to embrace change – it is often difficult to break through a culture that normally stays with the status quo. However, the desire for change is definitely there. If I were to list my top three predictions for trends in healthcare in 2019, they would include the following:
1) Health plans will get much more serious about improving member experience and predictability.
Health plans are eager to improve their member experience, partially because of the emergence of new consumer-centric plan models. As the market becomes more competitive with new entrants, traditional health plans will continue to partner with agile startups to deliver an improved member experience. These partnerships will enable health plans and startups to pilot projects and scale them quickly, leveraging the strengths of both organizations.
2) Employers are focusing on providing access to targeted healthcare solutions that meet specific therapeutic needs.
As of 2018, 83 percent of employees say healthcare is very or extremely important to/for staying in a job or changing a job. Given the importance of healthcare to their employees, and that the average company spends $10,000 on healthcare annually for each worker, employers will recognize the need to invest in specific solutions in areas where traditional networks fall short. These include programs for infertility, women’s health and behavioral health.
Healthcare organizations face unprecedented compliance challenges when it comes to managing business associate agreements (BAAs) amid frequent data breaches, heightened federal scrutiny and anticipated privacy legislation. Actions by the Office for Civil Rights (OCR) have clearly demonstrated stricter enforcement of HIPAA rules in recent years, and the industry has already witnessed a notable uptick in public shaming and fines associated with missing just a single BAA.
In December 2018 alone, OCR announced two notable settlements. Advanced Care Hospitalists (FL) entered into a $500,000 no-fault settlement with OCR, and Pagosa Springs Medical Center (CO) agreed to pay $111,400, both for missing a single BAA.
Simply put, BAAs have become a cornerstone of OCR compliance initiatives. And the outlook is not likely to change as trends point to continued advancement of privacy laws. As of close of 2018, 12 states had already updated their privacy laws regarding notification to patients, shortening the standard 60 days from the federal guidelines to 45 days, and in some states (CO, FL), the breach notification window is down to 30 days.
Breaches involving protected health information (PHI) are typically reported publicly at the Covered Entity (CE) level. When a breach involving a third party, or Business Associate (BA), occurs, one of the first things the federal government investigates is whether a BAA is in place with the CE. If a BAA does not exist, it typically sets off a chain reaction of investigations into other areas of HIPAA compliance.
While most headlines related to BAA compliance relate to CEs, HIPAA experts predict that 2019 will usher in greater focus on BAs and their management of these agreements as well. Many believe that unprepared BAs—especially small and mid-sized companies that lack resources to address HIPAA compliance—will become targets, increasing industry concern over proper BAA compliance.
Healthcare’s BAA management conundrum
Today’s healthcare organizations are feeling the heat, yet most are challenged to effectively manage BAAs due to limited resources for reviewing and managing massive and growing numbers of these agreements—reaching upwards of several thousand in larger organizations and health systems. Exacerbating this challenge is the current consolidation trend, which creates a fragmented landscape for BAA oversight that extends across multiple departments, facilities, affiliations and a multitude of different owners.
Consequently, manual, inconsistent workflows common to BAA management in today’s organizations open the door to significant risk. In truth, the most basic information often eludes the executive suite in most CEs and BAs, including the total number of existing agreements, where they are located and the terms of each.
BAAs are also the subject of intense negotiations between CEs, BAs and other subcontractors that often result in obligations that go beyond HIPAA and HITECH, causing contractual obligations to vary significantly between agreements. Subsequently, when organizations need to know the terms of these agreements, they must manually extract the information one agreement at a time. Within a framework of manual processes, the resources required to conduct this kind of data extraction across hundreds or thousands of BAAs is simply unfeasible for many organizations.
Yet, compliance professionals need quick and easy access to this information to ensure optimal response to breaches, which have become the norm for healthcare organizations as opposed to the exception. Consider the findings of a 2018 Black Book Market Research study: 90 percent of healthcare organizations have experienced a data breach since the third quarter of 2016, and nearly 50 percent have had more than five.
The approval of electrocardiogram’s (EKG) through the FDA that enables atrial fibrillation detection right from a patient’s watch band is just one example of how the digitization of medical devices, a part of the Internet of Things movement, is leading product development and innovation in medicine. However, while medical devices built on a connected services platform include components for data storage, security, accessibility, and mobile applications, along with advanced analytics, successfully implementing artificial intelligence to drive actionable intelligence remains a challenge from an execution perspective. According to Gartner, 85 percent of data science projects fail. Successful integration of data science into medical device development requires a rethinking around the role of data science in product design and life-cycle management.
Viewing data science as a product
While data science is rightly defined as the process of using mathematical algorithms to automate, predict, control or describe an interaction in the physical world, it must be viewed as a product. This distinction is necessary because, like any medical product, data science begins with a need and ends with something that provides clear medical utility for healthcare providers and patients.
It is erroneous to restrict the realm of data science to just the designing of algorithms. While data scientists are good at fitting models, their true value comes from solving real-world problems with fitted data models. A successful algorithm development process in data science includes business leaders, product engineers, medical practitioners, and data scientists collaborating to discover, design and deliver. For instance, a typical data science integration with a medical device product would include many of the following activities:
Identifying the medical need
Identifying proper data variables
Developing the right analytic models
Designing analytic algorithm integrations
Performing testing and verification
Deploying beta versions
Monitoring real-time results
Maintaining and updating algorithms
Considering data science as a product or feature of a product provides organizations with a different paradigm for execution focused on a tangible outcome. Data scientists are trained to develop accurate models that solve a problem, but the challenge many companies face is operationalizing those models and monetizing their outputs. Furthermore, conceptualizing data science as a product will ensure companies focus on its implementation, rather than just its development.
Advanced analytics: Part of the process, not an afterthought
Designing intelligence (even AI) into a connected medical device first depends on whether the data is being used to make a real-time decision or report on the outcome of a series of events. Most companies don’t realize the layers of advanced analytics that create actionable intelligence. By understanding these layers, which range from simple rule- and complex rule-based analytics to asynchronous event rules, complex event processing, and unsupervised learning models, companies can move quickly into developing mature analytics that have an impact from day one. As a company matures its analytics system from descriptive and diagnostic to predictive and prescriptive, it should also evolve to include strategic opportunities to provide business value, including automating decisions that can be delegated to a smart decision-support system.
Successful integration involves viewing advanced analytics as an architecture and not as a single solution to be implemented. The best way to make sure that you are successful in analytic development is to follow a continual process of discovery, design and delivery. For instance, data science architecture may begin with a business question, requiring you to determine if you have the right data and can actually leverage that data in the existing IT system. If you don’t answer this basic question, you will have challenges fully vetting the analytic opportunities available to you.
Recognizing common challenges in data science execution
Data science execution is often impaired by common missteps, like incongruence between customer and business needs and solving technical problems when it’s too late to have a positive impact. Another significant mistake from the business side is treating data science like a one-time accomplishment and not realizing it is a continuous process, or like a software development process with an unwarranted fixation on tools rather than skills and capabilities.
To use a common metaphor, data science is not a single moon shot, but laps around a track. Ultimately your goal is to run progressively faster around the track. An equally major drawback hindering execution is artisan thinking where design is seen as the ultimate end to the data science process. In fact, the most desirable approach is a modular system with emphasis on consistently maintaining and improving what has already been designed. This is particularly true for medical devices where innovation and changes in technology are continuing to better support and enable patients and practitioners.
In today’s world, we have access to more data than ever before. Analysis of big data is allowing us to solve ever more complicated problems at scale in a way that was not previously possible.
It’s a revolution.
Now it’s time to use the revolution of artificial intelligence in medicine. But what does this mean for the way we diagnose and develop new treatments?
Giving researchers access to pools of big data and equipping them with the tool of machine learning allows then a great opportunity to speed up processes that would have previously taken years. The typical timeline for new drug development from concept to market-ready product can range between 12 and 30 years. A significant portion of this time is devoted to research and development.
Patient diagnosis and treatment also experience, in particular for more acute diseases and ailments, a substantial amount of study to perfect. How can we leverage artificial intelligence in medicine, and its application in machine learning to help speed up this process?
We are starting to see the power of this technology to support medical diagnosis and treatment in several recent papers published by medical research institutes. These papers show the breadth at which the technology can be applied to increase the efficiency of the processes of diagnosis and treatment significantly. Thus allowing doctors to spend more time working directly with patients and saving institutions money.
Here we look at how artificial intelligence in medicine is driving forward our understanding of a variety of conditions. From drug development, through to improving the testing process and finally onto diagnosis — it is clear that there are multiple benefits of partnering with the technology.
Step 1: Drug Development
A recent article published in Nature highlighted several companies using machine learning to improve the drug development process. One company highlighted is Berg.
Berg is using machine learning to better understand and map human biology in far greater details. In their own words, “instead of hypothesizing the mechanism of a disease and focusing on only a few related compounds, we profile the entire disease by analyzing various patient biofluids (OMICS) and cell models (bio systems) as well as clinical information (EHRs).” This knowledge is then fed into Berg’s data analysis systems to be applied to drug development. The use of this artificial intelligence in medicine helps us to develop drugs more quickly and efficiently.
Artificial intelligence has the potential to revolutionize all fields, and healthcare isn’t exempted.
