Healthcare information technologies like AI-driven decision support and patient outreach tools can enable hospitals and health systems to achieve improved outcomes through proactive, predictive patient oversight, some health IT experts contend.
Why is proactive, predictive patient oversight important in today’s healthcare industry? What role does AI-driven decision support play in this kind of patient oversight? How does all of this lead to improved patient outcomes? And what kinds of outcomes can healthcare provider organizations expect?
To answer these questions, Healthcare IT News sat down with Zack Tisch, senior vice president of innovation and life sciences at Pivot Point Consulting, a Vaco Company. Pivot Point is a health IT consulting firm and No. 1 Best in KLAS managed IT services company.
Tisch has 15 years of experience leading complex, integrated healthcare information technology projects at some of the nation’s leading healthcare institutions, including Cedars-Sinai, MD Anderson Cancer Center, UCLA Health and Stanford Heath Care. He is a former Epic employee who is certified in 15 Epic applications.
Q. You argue proactive, predictive patient oversight is important in today’s healthcare industry. Please explain why.
A. Health systems across the country are finding it is increasingly difficult to compete and grow with declining revenues, new competitive pressures and increasing patient demands. It no longer is enough to simply manage the patients who organically flow into the health system for basic procedures, via emergency department visits and through referrals from primary care into specialty care.
Health systems must proactively manage the populations they serve, especially if they aspire to meet patient care expectations. Leveraging the power of electronic health records allows health systems to not only perform workup and risk assessments on the patients who are in front of their providers, but also enables the ability to risk-assess and prioritize all patients in the catchment area.
With this information, health systems can initiate a data-driven conversation with patients to encourage screening visits, staying compliant with therapy and driving early intervention to reduce cost – all while maximizing revenue opportunities for the health system. Patients feel better cared for and more aligned to their health system when they proactively receive accurate and helpful care-coordinated outreach, thereby improving patient retention.
External pressures, including new competitors entering the market, continued desire for growth, expansion of the nation’s largest healthcare providers like Kaiser Permanente, and greater patient demands for digitally driven 24/7/365 care, also contribute to this growing need for forward-thinking health systems.
Those that are able to provide proactive, predictive patient oversight will not only survive these external pressures, they will help set the bar and establish critical best practices for the future of patient/provider relationships.
Q. What role does AI-driven decision support play in proactive, predictive patient oversight?
A. Based on my understanding of the most immediate challenges and opportunities at many of the nation’s leading health systems, AI-driven decision support can play a key role in serving as a provider-productivity tool that helps organize, align and risk-stratify data so providers can make quick and efficient care decisions.
A great example of this in use is having an AI-driven algorithm do an initial risk screening on a patient for hypertension, after which – based on the calculated result range – a physician or allied practice professional can quickly take action on the most appropriate next step.
Over time, there may be scenarios where the provider role can be minimized or eliminated completely, such as low-risk dosing changes based on a calculated value. We already see similar examples occurring today in complex medical devices such as pacemakers and implanted cardiac defibrillators, where programming in the device helps to drive changes to care delivery that the device provides.
I see this same concept moving beyond physical devices and being embedded in various steps throughout the care journey. When done well, an army of AI providers can be constantly monitoring a health system’s charts and alert human providers to the most-likely-to-be-impactful events, allowing health systems to be more intentional with their resources and drive improved outcomes.
Q. What role do patient-outreach tools play in proactive, predictive patient oversight?
A. Using tools in the EHR to identify patients who are potentially at risk and have a recommended next action, such as coming in for a screening exam, is only half the journey. For health systems looking for the most value out of this effort, there needs to be strong tools, processes and procedures to reach out to the patient at the right time, with the right information, to easily drive them to the next action.
Native EHR tools are a great start, such as providing education and outreach to the patient portal or via text messages with links to helpful educational materials, research studies or patient-friendly videos. This outreach often can be automated/templated and driven based on the identified EHR worklists, allowing sites to do high-volume patient outreach with limited effort.
Health systems looking to take their patient outreach to the next level may embed external tools such as AI-driven chatbots or voicebots, much like consumers would experience when working with a large international airline, that allow patients to receive more conversational, interactive – and live – communication.
An initial text message from a health system’s virtual assistant may help a patient understand why they are at an increased risk for a particular care condition and quickly facilitate scheduling the appropriate next action, all without any staff involvement from the health system.
This not only makes the health system more efficient, it also satisfies those patients who would prefer to interact with technology rather than having to call in and speak with scheduling staff.
Q. How does all of this lead to improved patient outcomes? And what kinds of outcomes can healthcare provider organizations expect?
A. The best way to illustrate the power of a proactive, technology-driven approach is to compare the typical patient journey with and without these workflows in place. In yesterday’s healthcare workflow, a low-severity health failure patient may have been monitored by their primary care provider and progressed to intermediate-severity after a year or two with a referral to a general cardiologist.
The cardiologist would have managed and monitored the patient, performed periodic cardiac echoes, and seen the patient once or twice per year. Many patients with heart failure continue to feel worse, and their next step may be to show up in the urgent care or emergency room, often when the disease has progressed more significantly than known or understood.
At this point, based on the disease severity, the patient may have had more limited options for potential interventions and their likelihood of having the best long-term outcomes would have decreased.
Today’s technology-enabled workflow puts the patient on a heart failure registry when heart failure is first diagnosed. Algorithms run any time there is a clinical event – such as a new patient visit, a new medication or a new lab value – with personalized patient risk scores constantly being calculated.
As soon as the patient nears or crosses a risk score, the appropriate team of providers is notified to determine the appropriate next best action – which may involve bringing the patient in for a face-to-face visit, holding a telemedicine consultation, scheduling a noninvasive diagnostic procedure such as a cardiac echo, or adjusting medication dosages.
As soon as a patient crosses the risk threshold, providers are notified and can take action – even if the patient doesn’t feel any different. This allows providers to catch patients earlier in their disease progression, expanding their portfolio of intervention options, and ensuring the best possible long-term care outcomes.
By identifying patients earlier in their disease progression, tech-enabled health systems also reduce cost through avoided ER visits or reduced hospital readmissions.
This is the AI use case for healthcare, where, when done right, both the health system and the patient benefit through earlier identification, earlier intervention, reduced costs and better long-term outcomes.