Government agencies, such as the departments of Veterans Affairs and Health and Human Services, the Defense Health Agency, and state health and social benefits divisions, administer various health and human services and conduct research. As such, data and information from electronic healthcare records, clinical studies and research, contracts, referrals and patient charts are critical for lifesaving public health decision-making and for healthcare agencies and hospitals to improve patient experiences.
Recent advances in artificial intelligence-powered automation have led to document understanding software that reads documents of any difficulty, accesses critical patient information and shares data where appropriate — with great accuracy and reliability.
For federal and state healthcare agencies and providers to succeed in their digital transformation strategies, healthcare leaders must consider deploying AI-powered automated document understanding technology. Document understanding is the next generation of intelligent document processing technology, which includes machine learning, enabling the consistent interpretation of data and allowing organizations to divert human resources to high-value tasks that improve patient care. This technology equips healthcare providers with a tool to prioritize patient needs and make lifesaving decisions and a tool for policymakers and agency leaders to expedite critical public health research.
For instance, one of the Centers for Disease Control and Prevention’s Data Modernization Initiatives goals is to gain actionable insights for decision-making at all levels of public health. This goal cannot be achieved if clinicians and researchers cannot extract information from documents quickly and efficiently and share it with other team members and decision-makers.
To that end, automating various document-intensive tasks that require manual processing allows healthcare providers to dedicate more time to patient care or focus on lifesaving research, a primary objective of doctors at hospitals and clinics affiliated with the Department of Veterans Affairs.
In 2021, doctors nationwide reported spending 15.6 hours per week on average dealing with paperwork and other cumbersome administrative tasks. With the use of document understanding, this timeframe can be dramatically reduced. This is compounded by the fact that, “by 2025, the U.S. is estimated to have a shortage of approximately 446,000 home health aides, 95,000 nursing assistants, 98,700 medical and lab technologists and technicians and more than 29,000 nurse practitioners,” according to a recent report by Mercer.
AI-powered data extraction fosters critical healthcare decision-making
A comprehensive data extraction solution should provide healthcare workers and administrators with different approaches based on the types of documents being processed. To automate document processing correctly, it is important to understand the differences between structured, semi-structured and unstructured data and the required approaches to each.
For example, documents with fixed positions of data, or structured documents, will need a rule-based approach, which takes actions based on user-specified rules. Documents with different templates and less structured data will require an ML approach, which means ML models are taught to find and extract data when no static rules or templates can be applied. A hybrid approach or multi-approach uses predetermined rules to extract structured data while also using ML models to recognize and process less structured document sections.
Federal healthcare agencies process millions of forms, applications and images annually. Some are digital while others are paper-based. At the same time, documents are constantly changing, varying in quality and complexity. AI-powered automation and document understanding extracts and processes data at scale, allowing healthcare employees to prioritize patient needs while enabling critical decision-making among healthcare providers, federal healthcare leaders and policymakers.
For several years, the VA has been exploring AI-powered automation as a tool to collect patient data during clinical visits, for patient scheduling and to streamline administrative functions that enhance clinical workflows and free up providers to spend more time directly with veterans.
Since deploying AI-powered automation and document understanding, VA providers were able to expedite the pre-work of patient onboarding, such as insurance verification, referrals, data collection, scheduling and others, which ultimately improved patient experiences and sped up access to critical care.
How does document intelligence make such a huge impact so quickly? By understanding the document, integrating the automation of data entry, processing and adjudicating decisions with a human in the loop, and facilitating business workflows based on a document across multiple enterprise systems.
Document understanding at the right time saves lives
Document understanding also allows healthcare providers to help prevent hospital-acquired conditions, improving patient safety and reducing the conditions people experience while in hospital, such as hip fractures after surgery. AI-powered automation can also be designed specifically to prevent medical oversights or potential human errors, improving patient quality by having charts reviewed and analyzed in an intelligent solution that is designed with safety in mind.
Healthcare providers are inundated with stacks of documents about patients, clinical studies and research, and benefit claims. And poring over charts and paperwork takes too much time if processes are not automated. To solve these challenges, AI-powered automation will continue to be a catalyst for change where document understanding removes administrative burdens so providers can provide high-quality care that improves patient experiences and ultimately saves lives.
Todd Schroeder is area vice president of public sector at UiPath.