The Economic Rebalancing Opportunity for Health Systems
Providence St. Joseph Health CEO, Rod Hochman, describes today’s “healthcare industry financial statement,” in slightly tongue-in-cheek terms, like this: technology companies, pharmaceutical companies and medical device companies own the profits, insurance companies own the balance sheet, and health systems own the debt because of the capital intensity of investing in hospitals. Dr. Hochman goes on to point out that the industry will need to transform dramatically in the coming years due to economic pressure from patients, employers and the government. The economics will likely shift more toward health systems as data, artificial intelligence (AI)/machine learning (ML), and patient engagement will start to matter more. Every segment has work to do and will need to partner more closely to help transform healthcare.
Providers will need to deliver better outcomes at a lower cost and engage patients about their health between episodes of care — not just when they’re sick. Pharmaceutical and medical device manufacturers will need to develop more targeted therapies and devices at a lower cost to the system with better efficacy, and share economic risk. Insurers will need to be more effective at managing population health. Data, AI and digital patient engagement platforms are catalysts that will help aid in this transformation.
Health systems, in particular, have structural advantages in data, the ability to harness AI effectively, and because of their direct relationship with patients, make digital patient engagement more effective. Health systems that make investments in these areas will be more valuable partners to other segments of the industry. At Providence St. Joseph Health (PSJH), we are making deep investments in all three: Data, AI/ML and digital patient engagement.
The Health System Advantage in Artificial Intelligence
While the following is an oversimplification, ML, a discipline within AI that is very useful in healthcare applications, requires two things to be highly effective.
1. The more complex the problem being solved/algorithm being trained, the more clean, well-structured data is needed.
2. The data needs to be “tagged” as to the outcome by humans so that the machine can learn from training data to improve predictions.
Health systems have privileged access to both: large data sets and expert clinicians who can tag a series of data (symptoms, complaints) at the point of care with an outcome (diagnosis).
Insurance company data, while broader in care continuum coverage, however, can only make indirect inferences about a patient’s health status based on claims data collected from providers. Pharmaceutical and medical device companies have limited direct access to patient data. All three segments — health systems, pharmaceutical/medical devices, and insurers — will need this data to improve their value-add in the industry.
To become a valuable partner, health systems must learn how to capture, organize and leverage data and AI/ML in three ways:
Strengthen the Core: Health System Operations and Clinical Effectiveness.
AI/ML can aide in improving both clinical and operational effectiveness, thereby improving the “core” of what a health system should deliver to patients — high quality, low cost healthcare. AI/ML will be valuable in everything from improving operating room scheduling and throughput, to using deep learning in genomics for personalized treatments, to improving disease detection in radiology, to identifying high-risk patients and delivering targeted interventions in population health. AI/ML can also help improve the efficiency of core operations like revenue cycle and supply chain operations.
Add Digital Patient Engagement Data to the Mix.
Patient data captured in the Electronic Health Record (EHR) that health systems maintain represents only a small portion of the vast pool of healthcare data that will become available about consumers over time; an estimated 2,314 exabytes of healthcare data will be generated by 2020. The consumerization of healthcare is creating a new set of valuable data as providers become more digitally savvy — transacting, interacting and engaging with their patients online. Improved digital patient engagement helps health systems lower the cost of access to care and improve population health outcomes by being continuously engaged with patients about their health between episodes of care.
At PSJH, we are investing heavily in digital convenience and digital consumer engagement. Our strategy is to make the digital experience convenient to patients, to allow them to easily work with us and continuously engage them about their health. We’ve developed a platform for low-acuity, episodic conditions, such as flu and minor injuries, called Express Care. It allows patients to access care in conveniently located clinics, via telehealth, or they can summon a provider to their home. We’ve developed a women’s health patient engagement solution called Circle. We’ve implemented theKyruus platform, which matches patients and physicians. Patients can check validated patient ratings and reviews about their provider using technology from Binary Fountain. We’ve also spun out a company, Xealth, which allows clinicians to digitally recommend any digital content, app, product or service in their workflow, just as they would a pharmaceutical, making it easier for the provider to engage digitally with their patient between episodes of care. These platforms are now scaling at PSJH and are also being licensed by other health systems. They lower the cost of access, add convenience, and improve the efficacy of population health efforts.
AI in the near term can further improve digital patient engagement. For instance, AI can help make it easier for patients to find the right care solution in the health system, navigate patients to the appropriate site of care, and serve up information that is useful along their care journey. We at PSJH, however, are even more excited about the ability to combine insights around patient engagement data and clinical data using AI.
Develop Cross-Industry Partnerships.
Because health systems are uniquely positioned to draw correlations between usage and efficacy of drugs and medical devices, this makes them more valuable partners to pharmaceutical and medical device companies in clinical research. Health systems can also partner to develop economic risk arrangements around the effectiveness of their drugs and devices. These partnerships, while nascent, are starting to materialize, as pharmaceutical and device manufacturers understand that they need to deliver both clinical and economic value. Better data may also enable health systems to help determine pharmaceutical and medical device safety and efficacy once they are in market.
Health systems can work with insurance companies to manage the health of populations more effectively by sharing data and leveraging AI. These partnerships will take several forms, including acquisitions such as United/Optum and mergers such as CVS/Aetna, in addition to traditional arms-length partnerships between insurers and health systems. These partnerships, when data is shared effectively, can help lower unnecessary utilization and effectively target care management programs. Finally, there is an opportunity for health systems and insurers to leverage AI to simplify the payment/adjudication process between them, eliminating unnecessary administrative costs and a poor patient experience.
Where to Begin: 5 Steps to Harnessing the Value of Healthcare Data
Harnessing the power of data can be overwhelming, but there are steps providers can take to put themselves on the right path.
Think About Data Broadly.
What will be truly powerful is the combination of clinical, claims, behavioral, socioeconomic, environmental, genomic, and consumer behavior data, and the value that can be derived through understanding these interactions and correlations.
Invest in Data Infrastructure.
The collection of data is only the beginning. Build the infrastructure and capabilities to package, prepare and process your data to increase your ability to extract its value.
Invest in Digital Patient Engagement.
The amount of data that can be collected about consumer behavior online vs. offline is vastly different. As a result, the competitiveness of companies who are online with their customers is also vastly different than their offline competitors. Think of the advantages Amazon has over its offline retail competition in terms of knowing their customers, predicting their needs and improving their services.
Treat Data as Intellectual Property.
Develop data as intellectual property. Further, develop a set of tenets on the use of this data and a methodology for putting a value on the data you have, and apply these tenets and methodologies in partnering around your data. These tenets should first and foremost include making sure patient information is safe and their trust is never violated.
Build Data Partnerships.
Partnering on data is a new skill set that health system executives must develop with research organizations, technology companies and clinicians. Partnerships with big technology companies can accelerate delivery by capitalizing on the vast technology assets and capabilities they have developed both specific to healthcare and for other industries. Finally, internal partnership with clinicians is also essential in order to mitigate adoption barriers in healthcare.
Ultimately, the effective use of data, AI, cross-industry partnerships and digital patient engagement could be the core differentiator for health systems in terms of delivering on healthcare’s quadruple aim of better value, improved quality, and enhanced patient and provider experience. Regardless, it will be important. Health systems that get organized and make the right investments have an opportunity to become more valuable partners by catalyzing the change needed in the healthcare industry.