Value-based reimbursement is forcing hospitals and health systems to reduce costs and to improve outcomes. Whether they are trying to manage population health or to improve clinical pathways in hospitals, healthcare organizations need a way to understand and harness the huge amounts of data that could potentially be applied to achieving these goals. Artificial intelligence (AI) systems offer an approach that can derive actionable insights from large, complex datasets at the scale required by healthcare enterprises. AI can also uncover subtle predictive trends that traditional analytics platforms may miss.
An AI solution must aggregate and normalize the financial and clinical data from healthcare information systems, along with claims data from payers, in some kind of cloud-based infrastructure such as the Hadoop framework. The AI software must be live in an organization’s business processes and must be able to process data in near-real time to have an impact on clinical decision making and business decisions.
Here are the five components that an AI platform needs in order to deliver the results that hospitals and health systems seek.
To start, an AI platform must be capable of performing unsupervised learning. Unsupervised learning is critical because, in the large and complex datasets that define healthcare, the odds of asking the right question of your data is effectively zero. AI needs to discover all of the patterns or relationships that exist in the data – without requiring a practitioner to craft a question.
For example, a healthcare organization might employ AI to automatically discover groups of patients who share certain kinds of characteristics. These groups—e.g., low-income, opioid-addicted, obese patients who live alone and have two or more chronic diseases—might be targeted with personalized interventions and care paths. AI can identify these kinds of subgroups without being told what to look for—dramatically accelerating a doctor’s ability to begin crafting a care plan.
Using supervised learning methods, AI can use large datasets to predict what is likely to happen in the future with an exceptionally high degree of accuracy. In healthcare, predictive analytics can be used in health risk stratification, predictions of disease progression for individuals, the probable effects of various interventions on particular patients, and forecasts of the financial impacts of the disease burden within populations and subpopulations.
While clinicians remain the key clinical decision makers, AI can give them and their organizations a superior level of predictive ability that provides insights into the future needs, costs, disease burdens, and risks of patients.
Doctors and nurses are right to be skeptical about the recommendations of computer algorithms, particularly when they diverge from a provider’s experience and existing standards of care. As a result, it is imperative that the results of AI modeling include explanations of what the features in the model are doing in terms that are familiar to clinicians. For its predictions to have value, AI must be able to justify and explain its assertions and be able to diagnose failures.
Justification and transparency build trust among users. For example, AI solutions must always provide a “why” to back up decisions on classifying patients as high or low risk. Without a thorough understanding of the variables that AI used in arriving at a conclusion, there is no reason why clinicians should trust it. AI must justify its predictions, discoveries, and actions so clinicians can feel confident enough about its conclusions to act on them.
AI recommendations must be actionable to be of any use in clinical decision support. For example, if AI states that a particular care path is best for a certain patient, it must give specifics on the steps required to implement that care path and explain why it is likely to produce an optimal outcome.
AI must provide these insights within a time frame in which clinicians can act on them. That is why intelligent applications that use AI must be live within the clinical workflow. They must ingest new data and automatically executive the loop of Discover, Predict, and Justify at a frequency that provides value in the clinical decision support process.
Intelligent systems are designed to detect and react as the data evolves. An intelligent system is one that is always learning and constantly improving.
AI learns from the data and from feedback about its errors. As it discovers trends, predicts events, and recommends actions, it must constantly improve to be of value to healthcare organizations. AI solutions are most effective when embedded in clinicians’ daily workflows as they care for patients. AI must be designed to create a positive user experience, which facilitates both decision support and the collection of feedback to facilitate continuous learning.
This framework represents a starting point for any healthcare organization looking to deploy Artificial Intelligence. Whether it is evaluating a point solution for radiology or a platform on which to build specific applications, these five components should be present. With them, healthcare organizations can accelerate their journeys to value based care.
Gurjeet Singh is the Executive Chairman and Co-Founder of Ayasdi, an enterprise-grade artificial intelligence platform that lets you create intelligent business applications and make fundamental discoveries with big data.