In the world of Healthcare, there isn’t much room for making guesses. When you’re in charge of someone’s well being, it is critical that all of your decisions are well thought through and based on data. This is where AI can be very useful to your organization. AI, when applied properly, is very helpful in not only gathering data, but in sharing, analyzing, and storing it.
We reached out to our incredibly talented Healthcare IT Today Community for their comments and insights on the impact AI can have on data. This is what they had to say.
Ben Herzberg, Chief Scientist at Satori
From “where we are standing”, we have great visibility into what’s going on in data security. In healthcare organizations, data use is booming, with most organizations having projects of enabling data access to more users. This has a considerable impact on operational teams such as DevOps and Data Engineering, reporting that around a quarter of their time is being wasted on manual access and security policies implementation on data.
And so, many healthcare organization are either planning or in the process of automating their data access across databases, data warehouses and data lakes. As this year is going to be a lot about making the most with the resources you have, automating data security is also freeing precious engineering time for more valuable work.
Rahul Sharma, CEO at HSBlox Inc.
Technologies like AI and ML have a big role to play in data-digitization, prediction-analytics and interoperability of digital healthcare data. This will facilitate better automation of tasks and decision making processes since data-driven insights are needed in order to automate processes and data-driven insights need digitized data. According to Harvard Business Review (HBR), over 70% of the healthcare data is un-structured and exists in the form of charts/notes, images, freeform text, audio/video, wearables and in proprietary formats.
A key to implementing digital transformation is data digitization and amalgamation of that data with structured and external data sets so that a 360 degree view of the patient can be achieved to provide actionable insights to Payers/Provides and the Patients. AI technologies coupled with ML algorithms in a robust data engineering framework that enables to/from integration between systems with this digitized data are needed in order to make this a reality.
Mike Morper, SVP of Product at Virtru
As AI is increasingly leveraged to help surface additional insight in healthcare, care providers may be inclined to increase the automated sharing of information, as it will be easier to discover new patterns and form new conclusions. As in many other industries that are stewards of sensitive information, any time this data is moving between systems and, in particular, leaving the secure boundary of the provider, that data should be transported in a means that ensures only the intended recipient (system or individual) is capable of viewing it. This is a key aim of healthcare organizations: Maximizing the utility of data to deliver exceptional patient care, while ensuring compliance and protecting patient privacy.
Marilene Schofield, Director of Automation and Optimization at Ensemble Health Partners
Interoperability – effectively and efficiently moving data between provider and payor systems – is a critical first step to maximizing the potential of AI on the business side of healthcare and is where automation is making a big difference. Using robotic process automation (RPA) as an advanced interfacing tool to retrieve, normalize and share data across currently disparate systems is helping fill technology gaps, eliminate manual tasks from staff, and create the necessary data set from which we can analyze trends, build exception-based workflows, and develop predictive analytics.
Healthcare is a human system. The goal of automation or AI should not be to eliminate people from the process. The goal should be to help people operate at the top of their license – help them focus on critical-thinking tasks by getting the right data in front of the right person at the right time.
Automation is a great at handling rote processes, but not variability. Due to the inherent complexity in healthcare, specifically in the revenue cycle, there are certain processes that require human decision making and aren’t great candidates for automation. If the process isn’t standard with a consistent set of answers, automation might not be the right solution.
Healthcare is in the infancy of applying AI to make the overall system more efficient. There’s so much healthcare data available, but not a lot of it is usable. We see several healthcare providers trying to automate various exceptions in their current workflows, which leads to an army of bots managing small units of work. More sophisticated systems are using machine learning and predictive analytics to make decisions and suggestions for action, but people are still involved to validate those processes and take the recommended actions. We see significant potential and value in solving systemic problems at scale and in collaboration with payors to drive efficiency and reduce waste across the system.
The problems payors and providers are trying to solve are not extremely different. Everyone is investing time and resources to solve payment integrity issues. We’re working on ways to bridge gaps between payors and providers to make data easier to understand, access and use to ultimately simplify the payment process for providers, payors and patients.
Jason King, Senior Director of Data Science at XSOLIS
In today’s challenging environment for healthcare operations, there is increasing emphasis on turning data into actionable insights. Specialized AI techniques, such as machine learning, deep learning, and Natural Language Processing help make that happen. While clinical applications of these technologies are often in the headlines, they can also be used to tackle operational challenges – unlocking powerful data insights to improve staff efficiency, shore up revenue integrity in the face of rising costs, and reduce friction associated with transition of care coordination.
There’s a huge amount of data locked up in clinical documentation that is hard to access. AI-driven predictive models give users the ability to identify, extract, and interpret this information for key insights that would otherwise be impossible to sift through. When coupled with automation, AI offers game-changing advantages that can help alleviate the strains of operating in today’s rapidly evolving healthcare industry. Both entities working in concert with one another deliver real-time analytics that drive staff efficiencies and strategic outcomes, taking care of important but tedious tasks, and freeing up time and resource-strapped clinical teams to focus on activities that require human intervention.
Gregg Church, President at 4medica
Advances in AI and machine learning allow automation technology to reduce human intervention and error in patient matching. One way healthcare stakeholders are automating is by eliminating duplicate patient records by using a multilayered process that first runs data through a Master Patient Index (MPI) to identify and merge obvious duplicates, while a second layer uses machine learning to correct errors. Referential matching and data enrichment then further reduce duplication rates. Lastly, data analytics resolves remaining duplicates and checks for overlays. Previous steps are then rerun. From there, staff can address any unresolved questions. Using intelligent automation to improve health data produces better results faster than manually performing the tasks. It’s possible to reduce patient record duplication rates to less than 1%.
Charlie Clarke, SVP of Technology at hc1
According to the CDC, 70% of today’s medical decisions depend on laboratory test results so time is of the essence when it comes to reaching out to providers when issues arise with samples or when critical results need to be communicated. CRM automation designed specifically for labs to integrate healthcare data sources and enable quick access to collaborate and provide the insights needed to streamline internal processes is a game-changer. Not to mention, it helps provide more time for healthcare practitioners to focus on their patients and caregiving, one step in the process that can’t be replaced and should never be automated.
Venu Mallarapu, Vice President of Global Strategy & Operations at eClinical Solutions
As healthcare and life sciences increasingly adopt digital data flows, automation is a top focus within clinical development. Automation is already being leveraged to aggregate and standardize diverse data sets, setting the foundation for advanced capabilities. Next use cases are study setup in data collection and clinical operational systems, as well as content creation for submission pipelines and analytics for decision-making across the R&D value chain.
Dr. Tim O’Connell, CEO at emtelligent
Medical NLP, a specialized branch of AI, has a history of underperformance relative to the requirements and expectations of a technology for clinical use. As AI models continue to improve and deep learning progresses, and by ensuring that medical expertise guides the development of deep learning models, it is becoming possible to tap the vast amounts of clinically useful data hidden in medical text. Relevant information will be found faster and more cost-efficiently than manual data review and analysis. Further, high-quality data will fuel research breakthroughs in recognizing and treating diseases, detecting population health trends, and enabling value-based care.
Eli Ben-Joseph, Co-Founder and CEO at Regard
Artificial intelligence (AI) has made significant inroads within the healthcare industry and will become increasingly more indispensable. These tools and applications will be key to effectively revamping healthcare, allowing clinicians to focus on skilled work at the top of their license. Use cases for automation include managing population health, accurately assessing risks, and identifying potential gaps in care. Healthcare providers must not only embrace automation technology that makes data easily accessible and usable for their teams, but ensures effective and efficient use to optimize performance and improve care outcomes.
The use cases for automation and AI in healthcare are unequivocal. AI-based tools for health data extraction like Regard, for example, have proven to unlock the power of clinical data to drive revenue by significantly cutting down electronic health record (EHR) screen time, improving clinician satisfaction– thus reducing burnout, and elevating overall quality of care.
There is so much rich information that lies within data, and using EHRs in ways that promote automation will streamline processes, speed up decision-making, reduce wasted time and improve patient outcomes. There needs to be a continued prioritization of initiatives that leverage healthcare data to identify new revenue streams, optimize clinical capacity, and develop new medical advances.
Nate Fox, Co-Founder and CTO at Ribbon Health
The current provider data problem exists because there is so much inaccurate healthcare data out there that it becomes extremely messy, making it hard for providers and patients to understand what data they can trust, and where to find it. By using AI and automated models, the data can be scalably processed to determine which datasets are true and which aren’t, giving patients more accurate and personalized information on provider’s specialties, demographics, cost, contact information, and more. These automated algorithms can cluster data together, creating unique identifiers for each piece of data to help uncover the most accurate and useful information, enabling patients to make high quality care decisions.
Ophir Tanz, Founder and CEO at Pearl
The most obvious area of benefit is diagnostics––in particular, radiologic diagnostics. In dentistry, which is my company’s field, diagnostic inconsistency is a significant and well-documented problem which has serious implications for patient health and treatment outcomes. On average, dentists who use our AI when reading radiographs, detect pathology almost 40% more accurately, strengthening diagnostic foundations for care. Nearly every medical field is seeing similar benefits from AI-driven diagnostic automation.
Thank you everyone that submitted a comment, we love hearing from you! And to all those reading, please leave a comment down below or on social media. We’d love to hear what you think about this topic!
