Share, , Google Plus, Pinterest,

Print

Posted in:

Solving The Patient, Payer, Provider Conundrum With Artificial Intelligence

The healthcare industry is under tremendous pressure as it struggles to meet the specific and, often opposing, objectives of each of its key audiences. The patient, payer and provider conundrum has consistently proved deleterious to healthcare because it represents the core vectors—and their apparent opposition—within the industry.

Patients want the best healthcare services possible while minimizing costs. Payers seek profits through patient expenditures while circumscribing their own spending; providers are tasked with profiting from the administration of care. The historic inability to align these competing forces has largely resulted in the system’s escalating costs. Total expenditures for U.S. healthcare are predicted to rise to 25 percent of the economy by 2040, while the worker share of average annual healthcare premiums from 2006 to 2016 has nearly doubled.

The lack of a common incentive to align what appears to be competing interests has produced multiple effects, the most notable of which is a focus on reactive, instead of preventive, care. An analysis of that misalignment, however, reveals that what initially appears as a conflict of interests is actually a systemic series of handicaps and divergent motivating factors which, once mitigated with a shared incentive, can unite those interests, reduce costs, and reshape the focus of the entire system.

A common methodology that can benefit all three constituents is required. Technological advances in Artificial Intelligence (AI) can empower physicians, decrease costs, and aid patients in achieving their objectives; thus, creating a truly salutary system that provides the aforementioned shared incentives. The widespread implementation of AI technology from a clinically relevant perspective can transcend the handicaps limiting each partisan, provide a common point of motivation, and reshape the system from a reactionary to a preventive one that truly embraces healthy living. When used correctly, AI can meet each challenge:

Handicaps: The reverberating effects of the systematic handicaps plaguing the current healthcare system originate with the physician. The primary encumbrance limiting physicians’ effectiveness today is the marked shortage of these professionals. The 2016 update to a report from The Association of American Medical Colleges projects a shortfall of approximately 60,000–100,000 physicians by 2025, including shortages of approximately 15,000–35,000 primary care physicians and 37,000–60,000 non-primary care specialists. Physicians are scarce commodities who are frequently misallocated or misused; the effects of this limitation are felt throughout the industry. The shortage of physicians means a lack of physician time which restricts patients’ interaction with these trained professionals. The paucity of physician time also means doctors are less innovative in dealing with patients when seeking potential treatment options. This problem simply reflects the reality that there are not enough physicians today.

Although patients are most directly affected by the scarcity of physicians, the chief cause of their handicap is the reactive system which is a consequence of this shortage. The fundamental effect of that handicap is patient non-adherence; whereby, patients are not sufficiently responsible for their own care and fail to adhere to their practitioners’ recommendations. According to Forbes, the cost of patient non-adherence in the U.S. is roughly $300 billion, equal to almost 13 percent of the total spending on healthcare, which represents approximately 2.3 percent of the Gross Domestic Product. The patient adherence issue is a repercussion of the reactive healthcare system that fails to equip patients with sufficient tools to keep them informed and involved in their own care. Patients often learn of healthcare issues after ailments arise as opposed to taking active measures to ensure that such maladies don’t occur.

Additionally, there’s a dearth of feedback mechanisms available to patients for their own baseline measurements of vitals or for monitoring or preventing healthcare conditions. Thus, they have limited insight and information into their own health condition. Feedback is one of the most effective means of promoting adherence because it is motivational for patients to measure their progress based on concrete information, and to see that their behaviors affect their health. Motor vehicle tune-ups are a good analogy for this situation. Automobiles are rigorously maintained with checkups as frequently as every three months, so that owners can prevent problems or detect them before they become major issues. Ideally, patients should get checkups with a regularity and consistency. Significantly, this handicap isn’t predicated on financial concerns, but on a lack of effective personalized medical technologies for patients.

Conversely, the payer is predominantly handicapped by the reactive healthcare system in a manner distinct from that of the patient. Healthcare insurance companies are tasked with turning a profit largely by their means of diversifying risk. They can only afford expensive medical treatments for conditions such as cancer if their finances are buoyed by customers without such expenditures. Unfortunately, today’s reactive system presents the payer with an inordinate number of patients susceptible to costly healthcare conditions. Subsequently, insurance premiums increase to offset patients requiring costly medical procedures, and insurance companies are forced to cease operations because they can’t sustain their costs. Patients frequently view healthcare payers as antagonists arbitrarily raising their rates, but premium increases are merely another ramification of systemic handicaps spawned from reactive healthcare and insufficient resources: namely, physician time.

Interests: The initial interests of the payer and the patient are well aligned, perhaps even more so than the interests of the patient and physician. Neither party wants a sick patient: the patient’s objective is to remain healthy and healthy patients are vital to the operation of insurance companies that benefit from collecting premiums without additional payouts. It’s advantageous for both of these parties to detect any noxious issues well ahead of time to either mitigate or prevent them. The misalignment between these parties occurs once patients require care. In this situation, the payer’s objective hasn’t changed, it still wants to reduce costs and collect a premium. This means the payers’ priority is to remove patients from costly long-term care facilities and avoid expensive medications and procedures. Conversely, patients would rather remain in care facilities until their health improves. They desire the best medical care possible to expedite their recoveries, especially when it’s at the payer’s expense.

The physician’s interests mainly pertain to aligning their profession’s economies of scale. Specifically, the physician seeks treatment and diagnosis options that provide (a) clinical value, (b) temporal efficiency and (c) sustainable remuneration. These variables are equally important. Solutions must improve clinical value, consume little time, and provide propitious compensation for physicians.

Preventive Healthcare and AI: Medically relevant AI applications will appear in three codifications to reposition the healthcare system into a preventive one. It will aid medical diagnostics, pharmaceutical discovery, and genomic sequencing. It will furnish fast, relevant diagnostics by combining remote patient monitoring systems with AI’s learning algorithms (chiefly its advanced machine learning manifestation, deep learning, which will likely involve other neural networks too) for remote checkups. These systems will require clinically accurate, medical-grade wearable devices that continually stream patient data to fuel the quantities of data upon which AI algorithms work best. In fact, the copious amounts of data required to train AI’s learning capabilities are well suited for making those algorithms’ specialists similar to technicians in specific aspects of patient care such as cardiology or radiology. After analyzing large datasets for a particular healthcare specialization, these algorithms can be deployed in remote patient monitoring systems to screen patient data for analysis apropos to their conditions.

Thus, patients will finally access the type of routine checkups motor vehicles undergo, but can actuate them even more frequently than automobiles do. In these applications, AI will screen for analysis instead of for interpretation, which will be conducted by qualified medical personnel. Patients will leverage these monitoring systems for diagnostic purposes and for feedback mechanisms to motivate them to increase adherence. Diabetic patients, for example, can be sent a toolkit to analyze their blood sugar levels every three months to determine, before seeing a practitioner, how to alter their diet to improve their results. Intelligent patient monitoring can be used for predictive purposes across broader datasets, encompassing the entire ecosystem of patient data. It can also be used to yield personalized results that allow patients more time to prepare for detected conditions.

A Common Incentive: The shared incentive for each of these partisans is a preventive healthcare system with low costs, high value, and cost-effective physician time. Such a system is typified by expedient, reliable diagnostics which embody the physician’s interests of clinical efficacy, temporal efficiency and sustainable compensation. Quick, trustworthy diagnostics align the interests of all three healthcare constituents. Physicians gain a valued resource which reduces their time so the industry can make the most of this prized commodity. Patients readily discern their healthcare needs, leading to an increase in both engagement and adherence and a reduction in anxiety. Payers no longer squander funds on incorrect diagnoses and are freed from a reactive system that raises their expenses.

Medical-grade AI applications can contribute to timely diagnoses to become a tool for these three parties. In doing so they’ll enable physicians to streamline processes, optimize their time and concentrate on patients most deserving of it. Payers will benefit from the earlier detection of issues and ensuing conservation of funds, while patients will have faster diagnoses and a reliable feedback mechanism to increase involvement in their care management.

The analysis capabilities of AI are also beneficial for pharmaceutical discovery. They accelerate time to market and increase the effectiveness of drugs for targeted population segments. Remote patient monitoring data can be synthesized across populations suffering from certain afflictions, cross-referenced with their individual symptoms and the pharmaceuticals they’ve previously used, and deployed to locate compounds that are advantageous for those conditions. AI’s potential for this application is formidable; its deep learning algorithms are designed to detect features in huge, disparate datasets that are not discernible to entire teams of data scientists. Those algorithms can detect a handful of variables across hundreds of thousands of data points that are relevant to specific medical applications or ailments, and improve them. The propensity for neural networks to quickly traverse enormous data quantities of seemingly unrelated data sources will also enable AI to influence applications of genomic sequencing. When deployed at scale its analytics capabilities are unparalleled, which renders it immensely useful in determining variants that contribute to rare diseases, chronic conditions, and hereditary issues.

Reducing Costs and Improving Care

The widespread deployment of AI in clinical settings can align the interests of the three traditional healthcare participants and engender an industry based on preventive care. It’s perhaps the single most effective means of aiding the dwindling number of physicians and properly allocating this invaluable human resource. Too often, the lack of physician time is responsible for disorganized situations in which physicians are simply treating patients by the numbers as opposed to a hierarchic prioritization of healthcare concerns. AI will impact physicians by helping to expedite, if not automate, the daily facets of healthcare, including checkups and diagnostics, which will enable valued professionals to address more complex problems such as viral outbreaks, pregnancy issues, HIV cases, and the like. In turn, the system will benefit from the increased cost effectiveness of one of its most expensive and necessary resources.

The increased cost effectiveness wrought in the wake of AI’s emergence in clinical settings should benefit payers as well. Insurance companies will have much less need to contend with doctors’ bills for relatively mundane ailments such as cold or flu. The transition to a preventive healthcare system should also reduce the amount of patients who require costly treatment. AI should spur the medical field’s advent into personalization, helping to reduce costs in the process. Finally, patients may have the most to gain from AI’s induction into routine medical processes. In addition to getting quicker diagnostics, patients will gain a reliable coaching mechanism to give them the necessary feedback to assert more control over their own care management. AI will deliver basic feedback as a digital coach before eventually issuing feedback as part of a comprehensive, personalized medical monitoring device. Personalized medical monitoring devices will serve as a catalyst to integrate AI to deliver a continuous healthcare toolkit that drives engagement and prevention. This platform should help counteract the burgeoning patient engagement problem and ultimately result in healthier people pursuing happier lives, which is the final goal of all credible healthcare systems.