SOLUTION PROVIDER Q&A Sponsored Content
The Future of AI & Analytics – A Conversation with HealthITq
Sheriker Obijaju Revenue Cycle IT Project Manager
HealthITq
Lori S. Wilson, MBA Infor Business Systems
HealthITq
Our reader survey found the following: 14.48 percent of readers say their organization has made no progress at all so far, but they plan to begin engaging in efforts around AI and machine learning in the near future; 24.83 percent are in early planning stages; 22.07 percent have gone live in at least one area with algorithms; 11.72 percent are significantly advanced in their journey; and 6.9 percent consider themselves advanced. Meanwhile, 20 percent have no plans to engage in such work. What is your reaction to these survey results? RQ: Healthcare complexity makes it extremely difficult to design and effectively engage with AI and machine learning and these results represent the current imma- ture and varied healthcare definitions of AI. SO: The percentages do not surprise me at all when con- sidered in the revenue cycle areas as I have witnessed many opportunities where healthcare providers are not taking advantage of automation via AI or machine learning and continue to rely on staff to provide informa- tion via manual effort.
Robecca Quammen DBA MBA FACHE Founder/Chief Executive Officer
HealthITq
Also, 22.68 percent are doing work in the diagnostic area; 22.48 percent in process-related clinical work; 33.3 percent in clinical operations; 31.78 percent in efforts to advance efficiency improvement; 13.95 percent in the surgical arena; 17.05 percent in real-time asset tracking and management; 25.58 percent in analytics to improve the patient and family experience; and 21.71 percent in finance/ revenue cycle management. Thoughts? RQ: Not surprising as much analysis and effort is routinely expended in these topics in healthcare as they are typically measurable and have data sets that are not particularly sensitive or disputable – unlike clinical diagnostic data that is subject to privacy concerns and inherent bias due to limited data sets based on geogra- phy, race, gender, and socio-economic indicators.
Based on your observation of developments in the industry, where do you think relatively “easy” wins come (if anything is easy)? Where will the hardest challenges present themselves? SO: The easiest and most productive place to implement AI is the beginning – patient access – where organizations capture payer information and patient demographics that identify propensity to pay, while also providing greater insight into critical social determinant of health attributes. LW: As costs continue to escalate, early adopters of AI will begin to see savings and applying those within sup- ply chain are practical and can be integrated into supply chain operations now.
Where are the leaders of patient care organizations stumbling right now? Where are they moving forward with the most success? LW: Success is coming via ERP systems that are cloud based, allowing for real-time data access. The largest ERP vendors are ahead of the curve, building complex technologies like natural language processing, intelligent automation, machine learning, and voice user experience into technology platforms. SO: With advanced AI we could accomplish more in the revenue cycle life cycle with fewer staff. At this time, without AI technology in place, leaders must rely on manual data collection and analysis that is time-consuming, error prone, and potentially subjective given staff skill and experience.
How does ambient clinical intelligence fit into the broader artificial intelligence discussion and landscape? RQ: Ambient clinical intelligence is an attractive and alluring solution to many of the problems inherent in current EHR documentation scenarios. Capturing the interaction between care providers and patients with- out the need to physically type or dictate responses into the EHR has long been considered the missing component to successful use of those systems and collection of actionable data. AI virtual assistants exist today in limited deployment to aid physician documentation. AI driven continuous monitoring and diagnostic capabilities are also available in contex- tual situations. Future ambient computing environ- ments require changes in social, legal, and political constructs. Trust, potential for bias, and the need for massive data sets are all barriers to successful imple- mentation in the near term.
How will this landscape evolve forward in the next few years? LW: There are companies and researchers advancing cre- ation of synthetic data simulations to augment traditional and existing healthcare data. Human and environmental factors must also be recognized beyond that of technology and algorithms. Significant advancement requires talent and roles not currently existing in healthcare. RQ: Without a high degree of knowledge and sophis- tication regarding emerging AI technologies, organiza- tions will struggle and engage in costly failures. There is an exciting future ahead where AI removes routine tasks and everyday obstacles, but it must become a focused topic in the C-suite and viewed as essential for survival.
www.HealthITq.com
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32
