One challenge of addressing health conditions is that symptoms often aren’t apparent until significant damage is done. If you are affected by a new type of disease, you could have no signs indicating a problem. So that, managing disease requires managing by the numbers, with the health care data. With these figures in mind, it’s no wonder that managing the disease is a major emphasis for players who are seeking to improve the quality and cost of care nationwide. But day-by-day, the health outcomes and thus health care data are said to be increasing. This had become a burden for the doctors and researchers in storing and searching of such an enormous amount of data. To overcome this problem, Artificial Intelligence (AI) must be employed to predict population health risks.
What is Population Health?
Population Health (PH) is defined as the health issues of a group of individuals, including the sharing of such matters within the group. These groups are usually geographic such as nations rather communities, but can also be other groups such as employees, disabled persons, ethnic groups, prisoners, or any other defined group. The health outcomes of such groups are of relevance to policymakers in both the private and public sectors.
Note that our population health is not just the overall health of a community but also includes the division of health. Overall health could be quite important if the majority of the population is relatively healthy even though a minority of the population is enormously less healthy. Ideally, such variations would be eliminated or at least substantially reduced.
The right-hand side of the figure shows that there are many health determinants, such as medical care systems, genetics, individual behavior, the social environment, and the physical environment. Each of these determinants has a biological impact on personal and population health outcomes. The purpose of PH is concerned with both the definition of measurement of health results and the pattern of determinants.
| Recommended for you | |
| Big data to make clinical studies obsolete | |
| Ethics for AI in medicine | |
| Artificial Intelligence for beginners |
Why is it such a hot topic?
Population Health Management (PHM) is a hot topic in health care today. This is due to the unique features that it offers towards health care platform:
- The evolution of population health management strategies
- Data needs for effective population health management
- Population health business models
- Vendor solutions
A unified analytics platform for improving population health presents insights to care, case managers, providers and the individual patient. Care providers can examine which patients need primary care interventions or health screenings, setting the stage for improved preventive care and better management of chronic diseases. Patients can now be involved at a higher level to obtain their care goals through many patient meeting platforms including both active and passive participation through portals and remote monitoring devices. According to the U.S. Department of Health and Human Services, about 75% of the country’s eligible professionals and more than 91% of hospitals are on electronic health records certified for Stage 1 meaningful use. A preference considered important in achieving the aim of Population Health is to diminish health inequities or differences among different population groups due to, among other factors, the Social Determinants of Health (SDOH). The SDOH include all the factors: social, cultural, environmental and physical the different populations are turned into, grow up and function with during their lives which potentially have a measurable impact on the health of human populations. The WHO’s Commission on Social Determinants of Health, reported in 2008, that the SDOH determinants were engaged for the bulk of diseases and injuries and these were the primary causes of health inequities in all countries. In the US, SDOH were expected to account for 70% of avoidable mortality.
Why are governments so interested in this topic? (especially USA)
The PH concept represents a change in the focus from the individual-level, characteristic of most mainstream medicine. It also seeks to complement the outstanding efforts of public health agencies by addressing a wider range of factors shown to affect the health of different populations. Federal and state public health policies and programs play a significant role in the health of the overall population of a country and its states; however, as noted in the definition of public health is not the same as PH and PH offer more advantages than the public health. We all know that the US health care system is facing significant quality and cost challenges. And a major driver of cost in health care is the chronic disease. In particular, according to the Centers for Disease Control (CDC), more than 75% of our nation’s health care spending is for individuals with chronic conditions. For example, consider diabetes, which is the seventh leading cause of death in the US by affecting 25.8 million (8.3 %) of the US population. Diabetes is the primary cause of heart disease and stroke and is the leading cause of kidney failure, non-traumatic lower-limb amputations, and new cases of blindness among adults in the United States. In adults, type 2 diabetes estimates for 90 to 95 % of all diagnosed cases of diabetes. 18.8 million People have been officially diagnosed with diabetes while 7 million have not. This may be due to the high-cost perspective. The best way to start generating better outcomes for diabetes patients—and those with other chronic diseases—is to use data well. And by handling data well, you will be ready to identify high-risk diabetics and those people with diabetes with treatment and screening gaps. From this menu, you can actively prioritize and target interventions to those particular patient groups. With proper use of data, you enable each patient to manage their diabetes by the numbers better. So, governments are turning towards it. Moreover, they are also investing in PH management because it helps
- Reduce the frequency of health crises and costly Enterprise Data visits and hospitalizations.
- Minimize the cost per service through a combined delivery of care team approach which includes social workers, clinicians, physical therapists and behavioral health care, professionals.
- Increase the overall patient experience, in part by presenting improved access to care.
- Increase patient engagement and empowers patients to better self-manage their health and participate in the decision-making process.
What’s the implication of AI and Big data in potentially predicting population health risks?
PHM is considered wider than disease management in that it also involves “intensive care management for individuals at the most significant level of risk” and “personal health management… for those at lower levels of predicted health hazard. The risks may be the cost perspective, managing the data, the relation between payer and provider becomes unpleasant, or the technology must bring out a tool that must become more than a data capture and report. These players cared a lot because they had so much money on the line. They needed data-driven systems that could help identify and manage patients who might incur significant medical expenses at some coming time. The care management market was turned out of this need. Population health efforts are intensely data-driven and require more data than is captured between the walls of hospitals or clinics. AI and Big data are on the brink of becoming dominant forces in the health care industry. The health care industry serves a particularly significant opportunity for machine learning to prove its value. The sheer volume of available medical knowledge has long since outstripped even the most knowledgeable clinician, claiming supercomputers just to keep up with the advanced best practices and big data breakthroughs in genomics, population health management, predictive analytics and clinical decision support. Almost every phase of health care could, theoretically, avail from an AI approach.
Computers don’t ignore what they have learned, making them ideal helpers for the largest of big data analytics designs like personalized medicine based on genomics plus clinical decision aid for complex situations like cancer. They don’t have internal biases, so they are more likely to produce actual examinations unclouded by preconceived socioeconomic assumptions about the patient, which can produce disparities in care. They can identify shifts and patterns in data more swiftly and comprehensively than most humans, so they may be capable of predicting conditions like sepsis before the patient even starts to feel ill.
The capability of a computer to independently solve problems that they have not been explicitly programmed to address is termed as AI. Machine learning algorithms make a computer’s “thoughts” by presenting a conceptual structure for processing input and giving decisions based on that data.
An AI computer needs to be able to accept data about the problem from its surroundings, create a list of actions that it could take, and maximize its probability of achieving its goals by using logic and possibility to choose the activities with the highest chance of success.
The learning happens when the application reabsorbs its past experiences and practices that data to inform future actions. Doing this enables the AI program to prioritize the choices that result in victory more often, heightening its probability of becoming the right answer. Humans complete these types of tasks almost without thought every moment of every day, but few algorithms are sophisticated enough to mimic our natural ability to process external input, extrapolate implicit information from a query, use logic and analysis to make a decision, and predict the likely outcomes of each action before they occur. “Researchers can apply methods such as machine learning to the plenty of biological data that has suddenly become available, and use those advanced analysis tools to treat cancer better.” AI may be a welcome addition to the patient meeting and monitoring areas, as well.
As health care organizations begin to focus on customer prospects in response to rising out-of-pocket costs and value-based compensations, providers will need to learn how to personalize the reduce unnecessary expenditures, patient experience, and maintain open lines of conversation between office visits to keep patients as healthy as possible. As Big Data continues getting greater health care organizations are beginning to adopt emerging technologies such as AI and machine learning to glean insights, a transition to value-based care and advance precision medicine. This is not to say that it would be as easy as extending the latest and largest in cutting-edge software tools. Preferably, the excitement that Big Data and the associated technologies exist is accompanied by confusion among the hype, Harvard Medical School associate professor and Cyft CEO Leonard D’Avolio said. There will be an event conducted by HIMSS which addresses emerging technologies, leadership insights, challenges to overcome and practical advice about thriving in the increasingly data-rich health care world. There will be a large discussion about data usage within health care organizations by a lot of researchers.
As UC Berkeley health policy educator Stefano Bertozzi, explained, health care information is always getting bigger, and that means hospitals have the possibility to improve care not only delivery but also enterprise performance but the analytic ability is the biggest bottleneck hospitals quarrel with today.
Are there any solutions or models in use?
Yeah. There are appropriate solutions in use to predict the population health risks.
- The current focus of precision medicine features medical treatments in health care settings. However, since health in communities is driven by social, biologic, environmental and economic processes, those factors of health and health inequalities also need to be addressed including the development of precision tools to measure them. To make an impression on population health, we must explore the intersection of precision medicine with population health signs in general and cancer control more accurately to ensure that we take advantage of opportunities to develop more accurate approaches to targeted attacks for both communities and different patients. A DCCPS-wide interest group was designed to explore more thoroughly the scientific and societal junctions between Precision Medicine & Population Health, and the unique position that many population sciences and team science can work in bridging the gaps between the two.
- Interoperability is a fundamental element of population health because all of this information is never in application, database or even one data center position. Integrated systems streamline data sharing and support community health initiatives; nonetheless, many organizations don’t have a clear concept for how to meet the demands of the ever-changing health care industry.
- Analytics, which is the accumulation and aggregation of data needed to improve outcomes and improve medicine.
- The patient engagement which is the communications and interactions with the patient when they are not within the four walls of current clinical setting.
- Family planning programs (including contraceptives, sexuality education, and promotion of safe sex) play a significant role in population health. Family planning is one of the most incredibly cost-effective invasions in medicine. Family planning saves lives and money by reducing unintended pregnancy and the transmission of sexually transmitted infections.
- Amazon’s take on AI into its Echo devices, which can help to build a smart home ecosystem by using voice recognition to stimulate its omnipresent Alexa personal assistant. In a talk at the Vanity Fair New Establishment Summit, Bezos hinted that Amazon might be eyeing the healthcare industry shortly. How Alexa will enhance the healthcare experience remains to be observed, but it’s possible that the clinics of the future will have an AI listening device in every patient compartment, replacing the nurse call systems, physician pagers and hanging PA announcements of yore with an unobtrusive, intelligent and responsive communication system.
- Remote patient monitoring could also help from an artificial intelligence program taking on the responsibility of coordinating Internet of Things devices in the home for disabled, elderly or frail patients.
Since the passage of the Affordable Care Act, the University has more demand than potential, resulting in students being turned away. Executives, clinicians, and nurses recognize the need for education and experience when addressing the health of populations.
Who are the players?
Health care professionals associated with populations to improve the health of communities by promoting health, preventing disease, and addressing health inequities. The PH management includes many parts of people as its players.
- Hospitals
- Primary care physicians
- Pharmacies
- Specialists
- Reference labs and more
Some of the companies are applying AI and Big data in PH.
- IBM Watson Health is apparently the most well-known name in cognitive computing at the time, although it isn’t the only player in the realm. Watson got an immediate start in the health care industry using its normal language processing and semantic computing skills to train in clinical decision support at some of the top companies in the country. Later ingesting millions of pages of academic literature and other health care data, the system can help providers obtain decisions by offering a series of suggestions along with confidence intervals that show how relevant the course of action may be. The higher the number, the more sure Watson is that a particular drug, therapy, or diagnosis is the way to go. IBM has been stepping Watson in health care data for several years and has shelled out billions of dollars to procure big data analytics companies that will further its goal of creating a competent partner for quality care.
- Watson has its opponents in the clinical decision support space, however, and IBM isn’t the only one taking an AI path to curing cancer. Microsoft is also sloping up its efforts to apply advanced machine learning algorithms to the mysteries of human biology.
- Amazon’s take on AI into its Echo devices, which can serve to create a smart home environment by applying voice recognition to activate its omnipresent Alexa personal assistant.
- Semantic data lakes are one entry point into what may eventually become artificial intelligence, and they are already finding a foothold in healthcare. Partners HealthCare, the American Society of Clinical Oncology’s CancerLinQ and Montefiore Medical Center are just a few examples of health care focused projects employing machine learning and semantic computing methods to build semantic computing systems that can support predictive analytics, collaborative research, clinical decision support and perhaps ultimately transition into what could be considered artificial intelligence.
- AHIMA’s principles of data governance stress the role of data integrity a fundamental concept for analytics by moving providers to develop clear, consistent, and standardized policies and procedures for creating and managing data.
As many of the AI applications had come from the Silicon Valley tech giant, Many health care institutions are already working on developing their own intelligent big data analytics systems based on machine learning principles.
What does the future hold?
While this sharp vision of the future is still firmly in the field of science fiction novels and summer blockbuster movies, contemporary advances in artificial intelligence and machine learning are blowing some to admire if Isaac Asimov’s Three Laws of Robotics are running to become applicable to everyday life shortly rather than later. The interdisciplinary team of clinicians, IT, care coordinators, and business analysts now have a single source of truth and near-real-time results to proactively employ and work with their patients to accomplish care. According to their Director of Clinical Business Analytics, “What we’ve accomplished with Population Health is something we’ve been trying to do for over 20 years with our various clinics. We used to pull together reports, all with different data manually, and we had no way to monitor our populations proactively. Now, we have near real-time data that enables our care coordinators to drive preventive care and ultimately lower our population health costs.”
Moreover, US Department of Health and Human Services had sponsored a website named Healthy people 2020, which identifies 42 topics considered SDOH and consists of approximately 1200 specific goals that are supposed to improve PH.
Improvements in genomic profiling and personalized medicine will ultimately innovate how we treat chronic diseases, and a vendor-neutral population health management system that grabs all data from all sources without prejudice will be an absolute requirement.
The White House is already studying about how to address issues of safety, fairness, regulation, and security as AI systems go from laboratories to real-world settings. It may still take a few years before clinicians can rest and relax while their robot companions take a crack at diagnosing their patients, but the advancement of AI is moving quickly enough to warrant a thoughtful conversation about how these technologies will impact society shortly.
Health care is going to be one of those industries that are elevated and made better by machine learning and artificial intelligence. Everything we admire about civilization is a product of intelligence, so expanding our human intelligence with artificial intelligence has the potential of helping culture flourish like never before – as long as we manage to keep the technology beneficial. AI is also starting to become a familiar concept to most of the modern world. So, there is no room anymore for inconsistent quality and inconsistent data. Maybe in the coming decades, using AI and Big data in PH will provide a vibrant picture of patient’s health and also gains a 3600 view of the patient population within a metropolitan area. Will the future bring us?
Image credit: www.istockphoto.com
