Value-based care weekly 2: Artificial Intelligence
Hospitals and physicians look to AI technology to help improve outcomes
This week, we focus on artificial intelligence technology and how it may be useful to improve outcomes in value-based care. We'll hear from thought leaders, hospital administrators, physicians and technologists in the sampling of articles below. This is a powerful tool, and hopefully, it will help physicians and hospitals meet the new criteria for providing value-based care and related compensation.
(Editor's note: Join us for a highly-curated gathering to discuss value-based care challenges, opportunities and innovations. Check out our SplashX Invent Health small gathering, co-hosted by HP.)
Hospitals use AI software to help improve outcomes
This post points out that without better clinical documentation processes and technology, hospitals and physicians risk receiving low clinical quality scores and lower reimbursement rates. "More than 80% of physicians find queries for information after they entered the [clinical] note or after the patient is discharged [by clinical documentation improvement specialists] disruptive and time consuming."
To improve the quality of clinical documentation, and related coding and reimbursement, AI can provide real-time feedback to physicians in 2 ways: (i) natural language software that analyzes clinical notes for combinations of words to resolve common coding problems at the front-end, and (ii) computer-assisted physician documentation that analyzes patient medication information in the clinical notes to capture all diagnoses and medical care in ICD-10 terminology.
Thought leaders view AI and machine learning as critical enablers
At the Leaders of Global Health & Technology (LIGHT) Forum hosted by Stanford Medical School, 250 leaders in healthcare and technology focused their conversations on the impact of data, AI and machine learning on the global healthcare system. According to this post, they discussed the change in focus from products to value and outcomes in the healthcare system. The program director, R. John Glasspool, said "[t]here is an amazing alignment emerging among the diverse stakeholders in the healthcare innovation and delivery system, as they embrace value-based care strategies. Technology to manage and understand data will be a critical enabler." One of the takeaways from this forum is to create and implement policies that make data accessible and enable its use.
AI Market and Use of AI in Healthcare
AI for healthcare is anticipated to be a significant part the overall AI market. From 2016 to 2022, the total AI market is projected to grow at a compounded annual growth rate of 62.9%, or up to $16.6 billion. In addition, “spending on AI and machine learning is expected to rise to $31.3 billion by 2019.” This creates opportunities in three cognitive categories as follows: (i) human decision making, (ii) engagement with patients, and (iii) automation of tasks to support providers. While AI has already been used to create big data platforms, frameworks and real-time analytics, it is anticipated that AI is going to be used in healthcare for data mining, accurate diagnoses, risk-based treatments, personalized patient care and AI-assisted documentation.
AI cognitive solutions currently being used in healthcare
In a 2016 report, Frost & Sullivan predicts that the AI market in healthcare will grow to $6 billion by 2021, an increase from $600 million in 2014. One of the explanations for this growth is that hospitals and physicians are looking for new ways to improve patient care and efficiencies as healthcare moves to a value-based reimbursement model.
AI provides cognitive solutions that can support physicians in assessment of big data on patients, decisionmaking and clinical workflow. AI is currently being used in healthcare as follows:
- Diagnosis and treatment – IBM’s Watson reviews vast amounts of current medical research and finds patterns for a physician’s patients, with the goal of helping physicians make better diagnoses and treatment.
- Disease management – Watson is using cognitive computing to assess a patient’s personality, socioeconomic data and demographics to predict their response to treatment. Other companies have created tools that help patients with chronic illnesses to manage their illness, with the goal of reducing future medical emergencies and improving outcomes.
- Clinical trials – Watson reviews thousands of clinical trial criteria to help physicians find clinical trials for patients who have exhausted available treatments.
- Wellness – apps use natural language processing to help patients with post-surgery recovery plans or other goals and prompts them to meet their program’s goals.
- Reimbursement – certain software platform uses natural language processing, database of clinical guidelines and predictive analytics to help payers determine whether any treatments were potentially unnecessary. Another advantage of this platform is that it is cloud-based and accessible to large health centers and small medical practices, and can be used internationally.
Current news to think about
Could value-based care services require more man power? One study shows the workload on care providers at community health centers can be exhausting. An author of the study and a natural senior scientist at RAND Corp. said that “things got worse across the board over a short period of time, which is concerning.” Physicians and staff at these centers who worked on achieving medical home recognition reported a decline in professional satisfaction by 10% and feeling burned out increased by 8%.