The Paper Problem

Despite trillions of dollars of investment and decades of effort, healthcare still has a paper problem and most likely will for decades to come. 

One survey reports that 90% of healthcare providers still use paper and manual processes for patient collections, resulting in slower and less successful payment collection. In another study, 76% of healthcare organizations surveyed, that use a common EHR platform, still print consent forms instead of using electronic signature applications. And, according to an analysis by an IT consulting firm, the average 1,500-bed hospital prints 8 million paper pages a month, if that hospital is still using printed consent forms instead of electronic signatures.

Simply capturing the data or automating manual tasks with legacy data isn’t enough to significantly reduce the burden of paper on the back office. People are still struggling with finding the information they need.

It’s no secret that search engines and digital assistants, like Google and Alexa, have fundamentally changed the way people access information and get answers to their questions. It should then be no surprise that workers are being conditioned to search for exactly what they want and are now expecting answers tailored to their specific needs. Yet, these same semantic technologies that are powering search engines and digital assistants are still underleveraged in the workplace.

Building Context with Intelligent Document Processing (IDP)

IDP is a software service that extracts important data from digital and physical documents through data capture technology. Typically, the steps consist of ingestion, capture, classification, extraction, validation and exportation of the data into another system. This is often the first step to any digital transformation or hyperautomation project and it makes the data from documents usable, searchable and actionable. 

If we take IDP solutions a step further and use the data that is discovered, data scientists can create knowledge graphs to optimize data and create better outcomes within healthcare (and other industries). A knowledge graph represents a collection of entities, or data of interest, along with how those entities are related and information about them, known as metadata. Today, knowledge graphs are used extensively in anything from search engines and chatbots to product recommenders, cognitive automation and other AI-based services. They are the foundation for building context with your data.

While the use of linked data and knowledge graphs have been slowly growing in more sophisticated healthcare organizations, most of the focus has been around trying to help with making diagnosis and predicting outcomes. But advances in AI, semantic technologies and enterprise knowledge graphs hold the promise of radically transforming the back office of healthcare. Context holds the power to make healthcare more efficient and lower the costs of healthcare.  

As advances in AI have made search engines smarter, these breakthroughs have been fueled by ever-expanding knowledge graphs. Healthcare organizations can meet the growing productivity challenges by building and integrating their own knowledge graphs throughout the organization. Let’s look at several examples.

Use Case: Employee Onboarding

By 2030, the World Health Organization predicts a worldwide workforce shortage of about 18 million healthcare workers. Organizations are going to spend an increasing amount of effort and resources on recruiting and onboarding new hires. Welcoming new employees involves vast amounts of administration and record-keeping. Automating these processes lets you focus on finding the right employees for your healthcare organization without the hassle of lots of paperwork. Automating critical onboarding tasks will be vital.  

Collecting federal and state-specific government new hire documents from employees and validating against application data and past work history.

Before hiring any healthcare professional, it is an absolute must to conduct a sanction check. This is the only way to prevent hiring sanctioned individuals. Pre-employment background checks require validating data across multiple data silos. Knowledge graphs can be used to link data across state, federal and private data repositories to create a 360 degree view of the prospective employee backgrounds.

Use Case: Mailroom Automation

Many healthcare organizations are inundated with both physical and electronic mail, including invoices, contracts, administrative correspondence and many other types of documents. Digital mailroom automation can increase efficiency while delivering or processing mail and email quickly. 

Intelligent document processing solutions can not only identify the type of mail it is (invoice, lab results, records, etc.) but it can extract the data needed from that document type to expedite the delivery of it. 

When connections and semantic data is applied to the various types of mail, recipients can prioritize the type of mail it is to improve efficiency and boost customer experience. For example, patients can get test outcomes from outside labs faster and receive expedited treatment.

Use Case: Electronic Medical Records (EMR)

The EMR Adoption Model measures, from stage 0 to stage 7, show the degree to which a hospital system adopts EMR functions. Organizations that are not completely paperless cannot meet HIMSS Stage 7, which requires all clinical documents to be available electronically within 24 hours of creation or receipt. Due to the amount of paper still in use, most hospitals are in Stage 5 and Stage 6.

In the rush to be paperless, many hospitals have adopted a point of service (POS) approach to scan the documents into the EMR in near real-time. While this approach has helped to eliminate paper and expedite the creation of a single EMR, critical clinical information is now sitting unstructured and dark in the EMR – unavailable to the systems that need it to derive insights or create better patient outcomes.

By applying a semantic data-driven approach (using knowledge graphs) to capturing these clinical documents, we can map them to clinical taxonomies to understand the entities and relationships within the documents. This allows us to apply meaning to the entities and data elements and link to related data within the EMR and other data sources ultimately allowing us to create an enterprise knowledge graph on the patient. Having a 360-degree view of the patient can fuel a number of healthcare initiatives like Patient Engagement, Population Health Management, Precision Medicine and clinical decision-making.

The Bottom Line

The ramifications for using legacy systems, too much paper and resisting modernizing without context for your data can be steep. Not only in terms of costs but for patient care, employee well-being and the environment (beware: these statistics for 2022 may change your mind on why printing less paper will benefit the world). 

The first step towards automation should be done with the foresight of using an intelligent document processing solution that uses the power of AI and knowledge graphs to build context with your data. It’s ok to start small or even with one department to test the outcomes and value, just start.