Artificial intelligence (AI) has two faces in healthcare.
One face sings the praises of AI as the tonic that will enable healthcare to
deliver better clinical outcomes at a lower cost and the second face is full of
skepticism and raises barriers to adoption at every turn. It is heartening to
see that a third face is emerging, the thoughtful and appropriate use of AI to
predict adverse health events; to identify and stratify patients in need of
health, social, and human services; and the application of AI in the automation
of tasks, activities, and processes.
To understand the likely evolution of AI-based automation,
it’s important to evaluate the interaction of humans and machines across these
five levels. At each level of automation, the following questions must be asked
Who produces insights? – Does the human
or the machine (AI) analyze data and deliver insights from such analysis? Does
the human or the machine describe what something is, how it trends, why
something is happening, and what might happen next?
Who decides and how? – Once all relevant
analysis has been conducted, does the human or the machine make the decision
based on the derived insights?
Who acts based on the decision? – Finally,
a decision should lead to an action by either a human or a machine? The action
can be in the digital or physical environment.
Based on the responses to these questions, IDC has
identified the following five levels of AI-based automation:
Human Led – At the first level, it is
the human who analyzes the data using limited technology, such as tools for
only descriptive analytics; it is the human who makes the decision based on the
analysis (or experience); and it is the human who acts based on the decision.
Human Led, Machine Supported – At the
second level, the human continues to lead data analysis, decision making, and
action steps but is now more reliant on the machine across these steps.
Machine-led, Human Supported – At the
third level, it is the machine that is using a wide range of analytic and AI
techniques to conduct the analysis and produce insights. These insights are
reviewed by humans. The human still makes the decision based on machine’s
recommendations, and it is the human who acts based on the decision. However,
at this level, the machine acts to provide oversight over human decision making
Machine Led, Human Governed – At the
fourth level, the machine analyzes data and produces insights without the need
for human review. At this level, the machine decides based on the analysis of
all available data and a framework of human-developed governance policies and
procedures. At this stage, it is also the machine that acts based on the
decision under the governance of humans.
Machine Led – At the fifth level, the
world has likely achieved general AI. At this stage, there is a full AI-based
automation without the need for human involvement. At this level, we need to
think of machines that set their own goals and understand all mathematical,
economic, legal, and other external constraints. Most AI academics and experts
in labs of commercial enterprises predict this level of AI to arrive no sooner
than in about 50 years.
recent years, one of the shortcomings in the commercial sphere of AI has been
the misrepresentation of the scope of possible automation. Too often, we hear
claims of AI systems automating end-to-end processes and predictions of massive
labor losses, this does a disservice to organizations trying to plan for the
appropriate level of investment in AI. There is a need for a pragmatic
framework that decision makers across industries can use to assess
opportunities and risks of AI-based automation. The levels of AI-based
automation must also be viewed in the context of the scope of automation. We
define this scope where:
Task is the smallest possible unit of work
performed on behalf of an activity.
Activity is a collection of related tasks to be
completed to achieve the objective.
Process is a series of related activities that
produces a specific output.
System (or an ecosystem) is a set of connected
IDC’s AI automation framework was developed to help wade through the hyperbole associated with AI. Our goal is to help provide a planning tool and key piece of vendor evaluations processes to fully understand the role AI is playing in software and guide strategic decision making.
By Abhinav Shashank, CEO and co-founder, Innovaccer.
Once while I was scrolling through the news feed on my phone, there was one specific line that really made me wonder: “There’s a 40 percent chance of gusty and blustery winds today.” Statements such as this one strongly influence people’s behavior, as they are based on evidence or data findings from years of surveying, studying, and analyzing past trends and occurrences. However, my question is “Why are we not able to make such claims in healthcare- even today?”
Can we predict the vulnerabilities a patient might face in the future or the current health risks a population segment faces?
Is risk scoring the answer we have been looking for?
Almost all kinds of care organizations have some risk scoring methodology to target care interventions. With quality, costs, and patient experience taking the center stage in healthcare, care organizations need to stratify patients based on their need for immediate intervention.
The need of the hour is to address high-risk issues that impact large groups of patients and ensure that these needs are met in a timely fashion. Often, frequent fliers among high-risk patients come into the emergency department as if it’s their second home.
What if we take the method of risk scoring to a whole new level?
Traditionally, providers and health systems have relied on claims-based risk models, such as CMS-HCC, ACG and DxCG, which were built to forecast the risk of populations/sub-populations but not for individual patients. Hence, these models give an accurate prediction of the average risk of the population but exhibit very poor accuracy if used to predict risk for individual patients.
Although risk scoring has turned out to be a key factor in addressing the needs of the patient population, this method cannot provide all the important insights that are needed to drive necessary interventions. Since healthcare already has the right data from sources such as EHRs, claims, labs, pharmacy, social determinants of health (SDoH) and others, can we predict the future cost of care instead of just stating the risk score of the patient?
The right machine learning-driven approach to predict the future cost of care for patients
It all starts with the right data. The first step is to integrate the data from multiple sources- whether it is clinical or non-clinical data, such as SDoH. The data from these sources can allow us to use the comprehensive patient’s data for multiple predictive models to predict future health cost with greater accuracy.
Partners HealthCare announced its selections for the fifth annual “Disruptive Dozen,” the 12 emerging artificial intelligence (AI) technologies with the greatest potential to impact healthcare in the next year. The technologies were featured as part of the World Medical Innovation Forum held in Boston to examine AI in clinical care including a range of diseases and health system opportunities.
“Understanding state-of-the-art medical technologies enables us to anticipate the future of clinical care,” said Gregg Meyer, MD, chief clinical officer, Partners HealthCare and 2019 World Forum co-chair. “The Disruptive Dozen technologies can offer physicians and patients a renewed sense of optimism about Artificial Intelligence and its impact on diagnosis and treatment.”
The 2019 Partners HealthCare Disruptive Dozen are:
1 Reimagining medical imaging – AI is transforming radiology and imaging, including mammography and ultrasound, to bring improvements in clinical care and diagnoses to patients worldwide. Researchers envision AI transforming mammography from one-size-fits-all to a more targeted tool for assessing breast cancer risk, and further increasing utility for ultrasound for disease detection and rapid acquisition of clinical-grade images.
2 Better prediction of suicide risk – Suicide is the 10th leading cause of death in the U.S. and the second leading cause of death among young people. AI is proving powerful in helping identify patients at risk of suicide (based on EHR data,) and also examining social media content with the goal of detecting early warning signs of suicide. These efforts toward an early warning system could help alert physicians, mental health professionals and family members when someone in their care needs help. These technologies are under development and not cleared for clinical use.
3 Streamlining diagnosis – The application of AI in clinical workflows such as imaging and pathology is ushering in a new era of AI-enabled disease diagnosis. From identifying abnormal and potentially life-threatening findings in medical imaging, to screening pathology cases according to the presence of urgent findings such as cancer cells, AI is poised to aid the diagnostic, prognostic, and treatment decisions that clinicians make while caring for patients.
4 Automated malaria detection — Nearly half a million people succumbed to malaria in 2017, with the majority being children under five. Deep learning technologies are helping automate malaria diagnosis, with software to detect and quantify malaria parasites with 90 percent accuracy and specificity. Such an automated approach to malaria detection and diagnosis could benefit millions of people worldwide by helping to deliver more accurate and timely diagnoses and could enable better monitoring of treatment efficacy.
5 Real-time monitoring and analysis of brain health – a window on the brain – A new world of real-time monitoring of the brain promises to dramatically improve patient care. By automating the manual and painstaking analysis of EEGs and other high-frequency wave forms, clinicians can rapidly detect electrical abnormalities that signal trouble. Deep learning algorithms based on terabytes of EEG data are helping to automatically detect seizures in the critically ill, regardless of the underlying cause of illness.
6 “A-Eye”: Artificial intelligence for eye health and disease – Not only is AI is helping advance new approaches in ophthalmology, it’s demonstrating the ability of AI-enabled technologies to enhance primary care with specialty level diagnostics. In 2018, the Food and Drug Administration approved a new AI-based system for the detection of diabetic retinopathy, marking the first fully automated, AI-based diagnostic tool approved for market in the U.S. that does not require additional expert review. The technology could also play a role in low-resource settings, where access to ophthalmologic care may be limited.
7 Lighting a “FHIR” under health information exchange — A new data standard, known as the Fast Healthcare Interoperability Resources (FHIR) has become the de facto standard for sharing medical and other health-related information. With its modern, web-based approach to health information exchange, FHIR promises to enable a new world of possibilities rooted in patient-centered care. While this new world is just emerging, it promises to give patients unfettered access to their own health information — allowing them to decide what they want to share and with whom and demanding careful consideration of data privacy and security.
8 Reducing the burden of healthcare administration — use of AI to automate routine and highly repetitious administrative functions. In the U.S., more than 25 percent of healthcare expenditures are due to administrative costs, far surpassing all other developed nations. One important area where AI could have a sizeable impact is medical coding and billing, where AI can develop automated approaches. The goal is to help reduce the complexity of the coding and billing process thereby reducing the number of mistakes and minimize the need for intense regulatory oversight.
9 A revolution in acute stroke care — Stroke is a major cause of death and disability across the world and a significant source of healthcare spending. Each year in the U.S., nearly 800,000 people suffer from a stroke, with a cost of roughly $34 billion. AI tools to help automate the diagnostic journey of ischemic stroke can help determine whether there is bleeding within the brain — a crucial early insight that helps doctors select the proper treatment. These algorithms can automatically review a patient’s head CT scan to identify a cerebral hemorrhage as well as help localize its source and determine the volume of brain tissue affected.
10 The hidden signs of intimate partner violence – Researchers are working to develop AI-enabled tools that can help alert clinicians if a patient’s injuries likely stem from intimate partner violence (IPV). Through an AI-enabled system, they hope to help break the silence that surrounds IPV by empowering clinicians with powerful, data-driven tools. While screening for intimate partner violence (IPV) can help detect and prevent future violence, less than 30 percent of IPV cases seen in the ER are appropriately flagged as abuse-related. Healthcare providers are optimistic that AI tools will further complement their role as a trusted source for divulging abuse.
Non-clinical factors can account for up to 80 percent of the health outcomes for patients. Such factors, including socioeconomic conditions, healthy behaviors, and physical environment, may vary drastically for each patient and can significantly impact health outcomes such as poor medication adherence, frequent visits to the ED, and more. Thus, it is essential to consider these factors while creating care plans to ensure that the specific needs of patients are addressed.
Additionally, healthcare’s transition to value-based care is pushing organizations to lead more efficient population health management programs that address every clinical and social need of the population in which they serve. The challenge, however, is that organizations don’t usually have the means to capture the social needs of the patients or address them beyond the four walls of a hospital to ensure that no care gaps remain unplugged.
Innovaccer offers to assist healthcare organizations in a stepwise approach, starting with surveys for patients to complete in order to evaluate their social needs, such as access to food, housing situations, or economic conditions. Additionally, Innovaccer’s solution allows care teams to send as many surveys as needed with multiple language support. Based on the answers received from the survey, the solution helps care teams find suitable community resources to assign to the patient from a pre-built national database.
The solution’s AI-assisted closed-loop referral process to community resources enables care teams to ensure patient-centric care, even after an encounter is over. This closed-loop referral process gives physicians and social workers complete visibility into the social needs of their patients, which allows them to refer their patients to the most relevant community resources. In fact, patients are also kept in the loop in such a way that they can track their referrals, give feedback, and coordinate with their providers at any time, all through a single mobile application.
Innovaccer’s primary aim with this solution is to empower physicians and care teams with visibility into the social needs of their patients, right in the moment of care. The solution also triggers automated and real-time alerts to care teams if a patient’s needs are found to be urgent, such as high social risk or missed follow up. Additionally, the insights from the survey are available to the physicians right at the point of care within their EHR workflows, ensuring that they have a holistic picture of their patients.
“For organizations under value-based contracts, establishing a culture of wellness is a priority to keep their business model financially viable. Social determinants of health are a gamechanger in this regard and organizations who leverage them put themselves in the driver’s seat,” said Abhinav Shashank, CEO at Innovaccer. “We hope that our solution is instrumental to healthcare organizations as they tie their efforts to address social determinants of health and create similar strategies to maximize care and cost outcomes.”
Only recently, Innovaccer also launched its first-ever in-house research authored by Dr. David Nace, CMO at Innovaccer, around the social vulnerabilities of the population across the US. The research paper named “From Myth to Reality- Revolutionizing Healthcare with Augmented Intelligence and Social Determinants of Health” discusses a revolutionary way of leveraging advanced algorithms to determine the social vulnerability of the zip code-level population.
To learn more about Innovaccer’s SDOH Management solution, click here.
By Abhinav Shashank, co-founder and CEO, Innovaccer.
U.S. healthcare is nowhere near what technology made us dream of a decade back. Healthcare technology was meant to act as a means of reducing costs, eliminating burnout, and making care delivery patient-centric. Cut to today, where a broken leg can cost a patient as much as $7,500, seven out of 10 physicians do not recommend their profession to anyone, and we rank poorly among other developed countries in terms of the number of preventable deaths.
Why did technology fail?
While disruptive technology solutions did flood healthcare in the last couple of decades, many of them required physicians to go the extra mile to comprehend those sophisticated systems. Today, physicians are still crunching large data files day in and day out, nurses are doubling up as technical executives, and patients are perplexed by the fact that their providers hardly have time for them.
It’s time for technology to care
If a technology solution is not assisting organizations in improving care quality, reducing costs, and optimizing utilization levels, then its very relevancy is questionable. Healthcare organizations need technologies that can help them actuate their data, realize their strategic goals, and bring patients closer to their providers.
Health IT solutions should make the lives of providers easier. Any health IT solution that puts an additional burden on providers is unjustified and unacceptable. Providers are not data analysts, and expecting them to train tirelessly to understand an IT system and spend a couple of hours each day navigating through complex interfaces can drastically reduce physician-provider time and pave the way for physician burnout.
In with ultimate integration. We need to bring together EHRs, PHMs, payer claims and HIEs and put it all in the palm of the providers’ hands. Whether it’s quality management or data management, it should be simple.
In with relevant insights right at the point of care. Providers are tired of wading through complicated EHRs and excel sheets. What we need now is to seize the nanosecond and realize truly automated care delivery that helps boost the clinical outcomes.
In with 100 percent transparency and bi-directional interoperability. Healthcare providers are often forced to access bits and pieces of electronic healthcare analytics and referrals on disparate applications. Physicians need to capture real-time care gaps, coding opportunities, patient education opportunities, and more; the only problem is that they don’t know how exactly to accomplish this. Providers should be able to capture the gaps in patient care right when they need to and enhance the patient experience of care.
In with true patient-centric care. Healthcare is not just providing episodic care to patients, it is about building relationships with them. In a world where the quality of care directly influences the financial success of an organization, providers should look forward to aligning the needs of their patients to their treatment procedures.
Healthcare of the 2020s needs reliable data activation platforms
“If you can’t explain it simply, you don’t understand it well enough.” — Albert Einstein
Buzzwords like innovation, intelligence, and analytics make sense in today’s time; however, unless the user experience is seamless, the charisma of back-end development does little good for healthcare professionals.
We’re moving into an age of intelligence, and in this age, successful organizations do one thing right- they know the worth of their data. This is the same thing that we need to do in healthcare. Organizations have to switch from a makeshift approach to engage patients and find a concrete strategy that is suited to their advantage, but this needs to be done with the support of data.
With more than 4 million nurses, the largest segment of the U.S. healthcare sector, nurses have indisputably demonstrated an ability to improve healthcare outcomes. We are just beginning to utilize Healthcare IT data and AI to improve patient outcomes. One of the key benefits of AI will be the ability to leverage the data from nursing care plans and nursing diagnoses to perform work load balancing for nursing staff. This is a key solution to future management of the problem of the shortage of nurses.
Another problem that needs attention is the possible disconnects which can result from nurse to nurse hand offs with the use of virtual nurses who remotely monitor patients. They enter data into their own EHR system – not the same one in use by the hospital where the patient is located. We will discuss here the nature of the data, technologies and frameworks, the nursing information model and the structure of the data elements needed to provide care needed to implement solutions for staffing, interoperability and workflow improvements.
The National Academy of Medicine’s committee background report on the Future of Nursing 2020-2030, Activating Nursing to Address Unmet Needs in the 21st Century, found the worsening health profile in the United States requires “more than a traditional medical response.” As professionals in the care team, nursing documentation requires a standardized framework to achieve consistent data quality in healthcare communications about the work of nurses. This standardized framework recognized for professional nursing documentation is the American Nurses Association (ANA) Nursing Process. This ANA framework is essential to nurses for managing and improving healthcare outcomes, safety and reimbursement as proposed by the Institute for Healthcare Improvement (IHI).
In most electronic health information record systems, the standard nursing data implemented (sometimes called the system terminology, data dictionary, or nomenclature) is proprietary with a pre-existing data structure/framework. The proprietary framework acts as a barrier to nursing documentation by constraining the available concepts for nursing documentation and the nursing care plan fields.
Without interoperable electronic data concepts available for documentation, nursing care notes become unstructured free-text and are not included in coded health information exchanges. Due to the highly structured design of EHR systems, nursing practice is determined by the system’s terminology and ontology framework configuration. If nurses do not select the ANA framework; nursing care data takes on the sedentary shape of the local proprietary data structures, rather than nesting in a flexible, portable and universal tool to enable nurses and other episodic care providers to improve future nursing interventions, practice and care outcomes.
The American Nurses Association (ANA) describes the common nursing framework of the documentation of professional nursing practice as the Nursing Process. The Nursing Process is the foundation for the documentation of nursing care. Yet, in the EHR, nursing documentation is reused during the patient’s stay, over and over, with the documentation being done from the nursing assessment as if the documentation was a template. The Nursing Process is the framework and essential core of practice for the registered nurse to deliver holistic, patient-focused care.”
Producing effective EHR systems for nursing requires a deep understanding of how nurses create and conduct cognitive documentation as well as task-oriented documentation. Most EHR systems dictate rather than adapt to nursing workflows and nursing information is not organized to fit the ANA model of care. The EHRs often assume a nursing care delivery model that is represented as algorithmic sequences of choices, yet nursing care is iterative with reformation of patient goals, revising interventions and actions and updating care sequences with individual patients based on encountered condition changes and constraints. In the dictated workflow of EHRs, nursing data is collected as care assessments with nursing diagnoses, interventions and actions in formats used to create single patient encounters.
As the country’s baby boomer population continues to age, the healthcare industry is gearing up for a whole new level of demand that it has never before gone through. With greater numbers of people requiring doctor visits and hospital care, the industry is looking for ways to be even more productive and efficient to ensure that the quality of healthcare that people are receiving doesn’t suffer.
One of the most exciting advances to hit the health sector is artificial intelligence or AI. This technology is looking to have a huge impact, not just on healthcare in the immediate future but moving forward. Here’s a closer a look at just how it’s changing the course of the industry.
Medical records and data are benefiting from the technology
When it comes to the areas that AI is having the largest impact, medical records and data keeping is a big area to focus on. When you think about the vast amount of information that needs to be collected, stored, and analyzed for each and every patient it can seem rather overwhelming. This is exactly why data management has become such a priority for AI.
Robot technology is now being used to actually collect the information, store it, find specific data when required, and allow for quick and seamless access across the board.
Wearable medical devices
Wearable medical devices are another area where AI is having an impact and bringing about some really exciting and promising products. It’s not just about devices that provide potentially life-saving alerts and information, it’s also devices that can help the wearer better their own personal health by tracking various details. Devices such as the Apple Watch and Fitbit are great examples of this kind of technology that can be useful to everyday people.
Now as for the devices that can actually offer life-saving capabilities and tools, look to options such as the Bay Alarm Medical which is a great medical alert system. While this device isn’t going to track any information or take readings, it can be worn 24/7 and with the push of the button, it connects you to a live operator that can get you the help you need.
The megalithic healthcare conference, HIMSS19, has come and has gone from the vast former swampland of central Florida. While I’m a relative newcomer to the show’s trajectory – I’ve been to four of the annual tradeshows since 2011 – this year’s version was, for me, the most rewarding and complete of them all. This could be for one of several reasons. Perhaps because I no longer represent a vendor so sitting in the exhibit hall in a 30×30 booth with a fake smile wondering when the day’s tedium would end and the night’s socials would begin may impact my rosy outlook.
Or, maybe I was simply content to engage in the totality of the experience, attend some quality sessions, meet with many high-class people and discuss so-called news of the day/week/year. Doing so felt, well, almost like coming home. Or, perhaps my experience at the conference this year was so good because of running into former colleagues and acquaintances that drove me to such a place of contentment while there. No matter the reason, I enjoyed every minute of my time at the event.
Something else felt right. An energy – a vibe – something good, even great, seems/ed about to happen. Something important taking place in Orlando, and I was blessed to be a part of it. Kicking off the week, CMS created news – like it does every year at about this time – with its announcement that it will no longer allow health systems and providers to block patients from their data. This was a shot across the bow of interoperability and the industry’s lack of effort despite its constant gibberish and lip service to the topic.
Another fascinating thing that finally occurred to me: no matter the current buzzword, every vendor has a solution that’s perfect for said buzzword. Be it “patient engagement,” “interoperability,” “artificial intelligence,” “blockchain”; whatever the main talking point, every organization on the exhibit floor has an answer.
But, no one seems to have any real answers.
For example, after nearly a decade, we still don’t have an industry standard for interoperability. Patient engagement was once about getting people to use patient portals for, well, whatever. Then it was apps and device-driven technologies. We’re now somewhere in between all of these things.
AI? Well, hell. It’s either about mankind engineering the damnedest algorithms to automate the hell out of everything in the care setting (an over exaggeration) or that AI/machine learning will lead to the rise of machines, which will help care for and cure people – before ultimately turning on us and killing or enslaving us all (again, I’m overly exaggerating).
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.