The application of AI and other advanced technologies; it’s all about the data
All health systems are hungry for artificial intelligence (AI) to drive their analytics and workflows. Health systems need easy-to-use analytics on demand in order to make operational and clinical decisions that improve cost and quality of care—ultimately helping manage the health of the populations they serve.
Currently, analytics-derived data is helping drive accountability in Accountable Care Organizations (ACOs) and care management teams and medical professionals are relying on data at the point of care to make decisions regarding disease prevention and ways to reduce the costs. In addition, patient engagement tools and platforms—such as interactive portals, wearables and telehealth solutions—are emerging to engage patients where they are vs. requiring them to visit a care location.
AI solutions are becoming more mainstream
These solutions have become more affordable and are more readily available than ever before. The practice of using data for predictive or prescriptive analytics have become more widely adopted. Solutions are now available that use advanced algorithms and can alert organizations to patients within their database that may be at risk for conditions such as congestive heart failure, respiratory compromise and sepsis, to name a few.
Adoption of Real-Time Location Systems (RTLS) are enabling health systems to manage their assets, locate staff and track their patients as they move throughout their locations. While these systems were first introduced to help manage assets and minimize the cost of replacing lost or stolen items (such as IV Pumps), they are now being used to track the movements of a patient with an infectious disease.
These systems are also helping increase patient satisfaction through locating and moving staff to where they are most needed at any point in time. The Internet of Things has made the introduction of wearable technologies and remote technologies possible to treat patients remotely. Mergers and acquisitions of healthcare organizations have meant the aggregation of more data to manage.
Anticipated trends in use of AI
According to a 2019 Gartner report identifying the top-10 trends in Data and Analytics Technology purchases, machine learning and neuro-linguistic programming (NLP) are the dominant drivers of new analytic platform purchases.
Machine learning and AI can help greatly reduce manual data management tasks by taking over the correction of typos, invalid data entry and blank fields. Industry analysts are predicting that more than 75% of large organizations will use AI to reduce brand and reputation risk by enabling them to easily identify any potential bias.
Due to the complexity involved in building these solutions, the majority of new solutions will be created with commercial platforms vs. open source. Half of all queries will be generated via search, with voice or auto-generated queries available in the next two years. Use of Intelligent Assistants such as Siri, Alexa, Google and others continue to grow. Major players in Healthcare IT are already beginning to experiment and pilot solutions with Intelligent Assistants embedded. The foundational technologies that lie underneath continue to learn from our habits and daily activities and are now beginning to anticipate where we want to go or what we may want to do. An example is your favorite social media feed “magically” presenting you with advertisements for restaurants because you happened to search for a place to go out to eat last weekend.
Value of AI for health systems
Solutions that are AI-enabled will help health systems use the data from all of their contributing systems to see patterns and make predictions about patients and their conditions that are otherwise not possible. AI will help organize and structure the data, then analyzing that data to determine patterns and personalizing information to present to the clinician.
A simple and current example of this would be using a CT Scan to rule out a tumor on a patient’s kidney. The system is loaded with all the data necessary and taught what a normal kidney looks like. Variations from that are presented to the reading radiologist for further investigation and interpretation.
For healthcare IT professionals to be successful in the introduction of these technologies, data elements need to be clearly defined with these definitions consistently applied across the organization’s technology stack. Thought must be given to the design and implementation of systems to ensure that data is being captured in a useful format and is available for use by the organization’s Business Intelligence systems. Proper data governance, the process of oversight in the management of these data assets within any given organization, is a key component in data enablement for the application of AI and machine learning.