Charles Kaplan
Intelligent data capture automates paperwork and promotes greater business efficiency.
At a Glance
Intelligent data capture is an application that can provide electronic access to data contained in any type of document, including clinical records and financial documents, thereby reducing or eliminating the need for manual data entry. Before implementing this technology, however, healthcare leaders should:
- Evaluate technology in the context of a business case to ensure a measurable ROI
- Communicate with employees throughout the process so they are prepared for and embrace the inevitable changes that will come with automation
- Implement the solution in phases, focusing on those document types that can deliver for a safer, less disruptive approach
- Achieve maximum benefit to the organization by reengineering business processes to fully leverage technology rather than simply automating existing manual processes
Mayo Clinic, one of the nation’s largest not-for-profit healthcare providers, was struggling with its invoice processing in early 2008 with a consistent backlog of 5,000 invoices, representing 10 to 14 days of manual labor. The cumbersome task of manually inputting information from stacks of paper, routing them to various departments for approval, and reconciling discrepancies impeded the organization’s ability to deal with this backlog.
With the addition of a new facility that promised to increase invoice volumes by 12 percent over the next year, and internal initiatives seeking to reduce costs by 6 percent, Mayo also saw little hope of achieving needed process improvements through manual keying.
Mayo found its solution, however, in intelligent data capture—an optical character recognition (OCR)-based application that sorts documents on the basis of their type and content (e.g., invoices, explanations of benefits [EOBs], orders, contracts, patient records, and correspondence). The application automatically extracts more than 90 percent of the data fields and line item detail, and validates the data against reference sources, such as purchase orders, vendor master lists, payer lists, and other data repositories. By automating data extraction from paper documents and transferring it to an organization’s information systems electronically, the application reduces or eliminates the need for manual data entry. This streamlining enables knowledge workers to forgo routine data entry and instead focus on handling exceptions. Elimination of backlogs creates immediate visibility into payment information on invoices, for example, enabling better cash forecasting and working capital optimization.
Using this technology, Mayo achieved remarkable results: Despite a 28 percent increase in invoice volume since 2008, the organization has been able to reduce its cycle time to less than 48 hours. The cost of processing each invoice has dropped below recognized best-in-class rates, with more than 97 percent of invoices being processed electronically, up from 48 percent in 2008.
And Mayo has achieved these results without requiring its vendors to alter their invoice formats, or conform to a specified electronic standard. In fact, more than a third of invoices are now received via email, eliminating paper entirely.
The improvements also enabled Mayo to reallocate half of its data entry team to other internal departments. Employees in accounts payable (A/P) now focus solely on outliers—cases in which specific information is unintelligible or inconsistent, requiring employees to verify that data are correctly processed into the enterprise resource planning (ERP) system. With 90 to 95 percent of invoice data processed hands-free, employees can focus on complex, long-term priorities. As a result, employees have expressed increased job satisfaction.
The Paper Jam
The mission of healthcare providers is to provide excellent patient care, but the industry spends billions every year on manually processing financial documents, such as invoices and EOBs, patient records, clinical documents, and lab orders. Processing costs pose a burden to providers, shareholders, medical personnel, and patients, consuming resources that could be devoted to bedside care, state-of-the-art equipment and facilities, research, and development.
The key to unlocking these resources is automation. Intelligent data capture offers healthcare providers a single advanced data extraction platform to handle any document type.
As providers update medical data and delivery processes to qualify for federal “meaningful use” initiatives, back offices remain mired in outdated, manual routines that drain valuable resources from operations. Intelligent data capture technology offers healthcare organizations an opportunity to supplement electronic health record automation trends and create efficiency across the organization, freeing up dollars and personnel to better serve their core mission.
How Intelligent Data Capture Works
Intelligent data capture enables fast and accurate extraction of data from any type of document. The software automatically sorts documents based on content, creating an automatic index into a provider’s document management system or enterprise content repositories. It then extracts specific fields and line item data from transactional documents, validates the data, and forwards the data to the ERP system.
Previous incarnations of OCR technologies were limited in the ability to understand and retrieve data from imaged documents. Each vendor and remitter used different forms with different layouts, and early OCR-based solutions required samples of each to build templates of each document layout. Such preparations required a considerable investment of time and labor before even marginal automation could be observed. In effect, the “solution” placed a greater burden on overtaxed IT departments that hoped to find process efficiency eventually. For Mayo Clinic, which processes invoices from more than 430,000 vendors, such a proposition was prohibitive.
Intelligent data capture solutions remove this legwork from the equation. Using “fuzzy logic” to interpret and process documents, the solutions do not require users to configure anchors, zones, templates, or keywords for each document type or layout. With a small sample set, the system uses advanced pattern matching technology to distinguish one document type from another and understand what to do with each document—extract this information, validate it against these data, and export it here. This generic learning capability drastically reduces reliance on IT personnel to maintain or update the solution, which allows users to observe ROI immediately. More important, vendors don’t need to adopt new practices, as automation takes place internally.
These solutions integrate with legacy processes and workflow systems and lift more than 90 to 95 percent of data automatically, saving the key strokes associated with manual data entry. Cycle times are reduced, enabling A/P departments to capture all early payment discounts. Errors are virtually eliminated as data are automatically checked and reconciled. Invoice line items are paired with purchase order (PO) line items automatically, enabling many invoices to be processed automatically, thereby freeing staff to focus on exceptions or other priorities. The software also scales to an operation’s volumes and can be used as a platform for automating other documents, including EOBs and patient records.
Case Study: Baylor Health Care System
Baylor Health Care System, headquartered in Dallas, is a not-for-profit, faith-based network of hospitals, primary care centers and practices, rehabilitation clinics, senior health centers, and affiliated ambulatory surgery centers. Including the Baylor Research Institute, the system has more than 20,000 employees and 3,400 beds. Baylor serves 1.4 million patients annually.
Baylor’s finance, supply chain management, and IT departments implemented an intelligent data capture solution for A/P automation after observing A/P staff overloaded with stacks of paper. The team was spending considerable time producing individual invoices and enduring the costs and logistical difficulties with the physical storage of paper documents.
The departments’ goal was to become as paperless as possible by minimizing the manual keying of invoice data. They also wanted to increase staff productivity by reducing phone calls from vendors and customers inquiring about the status of payments.
The data capture solution debuted in July 2010. The team charged with the implementation, which included IT personnel and representatives from among the solution’s end-users, ensured the software interfaced properly with the organization’s enterprise content management (ECM) and ERP platforms. In addition to using the software’s built-in, out-of-the-box capabilities—which included the enabling of automatic two- and three-way matching of line-item data within the ERP system, as well as the ability to establish parameters for exception handling—the team also customized the application to handle employee expense functions, at Baylor’s request. A reporting interface would provide valuable metrics and analytics—including field extraction rates, straight-through processing rates, current document status by event or project, and exception reports—for better cash flow management and compliance.
Baylor proceeded to reconfigure its internal processes to best align with the new technology, and users were thoroughly familiarized with the new application, with instruction based on their roles within the A/P process. As a supplement to the initial implementation, Baylor’s contract with the vendor would also include long-term support and maintenance as needed.
“In six months since integrating intelligent data capture into our A/P operations, we’ve seen considerable improvement in turnaround time for invoices,” says Dorothy Craig, financial services support center manager at Baylor. “We’re on track to reduce labor costs by $500,000 in the first year alone, with an 85 percent increase in staff productivity due to less keying and simplified resolution of discrepancies. We expect the overall cost of processing an invoice to settle at around $2.10, consistent with best-in-class rates.”
Lessons Learned
Baylor learned several lessons in transforming a highly manual process into an automated task with intelligent data capture.
Evaluating technology and its value proposition requires due diligence. As with any significant capital expenditure project, strong executive leadership is necessary. Management should agree that document volumes are significant enough to justify the expense and a reasonable ROI can be demonstrated. Providers should move forward on the project only if the business case is sound on paper and the technology has been studied by financial and IT specialists.
Communication with employees is critical. Many employees accustomed to a rigid daily routine will fear change and may perceive “automation” as a job-killer. The truth is quite different. Employees working with intelligent data capture technology report that their routines are less monotonous, more interesting, and less stressful in general, because they can focus more on exceptions and must deal with significantly fewer inquiries, as fewer processing errors result in fewer disruptions downstream. By earning the confidence of those who would use the application daily, management will find less resistance to automation, better positioning the project for long-term success.
Implementation in phases, rather than all at once, can be safer and less disruptive. By automating the simplest and most common documents first, users can ease into this new technology, establishing a basic protocol before diving into more complex documents. Such an approach is also more likely to find users realizing the perks of automation, rather than feeling overwhelmed. Incremental implementation also allows management to detect and address any concerns early in the process. This approach enables a stronger understanding of and confidence in the software moving forward.
As with any automation solution, intelligent data capture technology is not a “silver bullet” that magically fixes the underlying process involved. Automation augments an efficient process, but does not fix a broken one. It is important to work with a vendor that understands best practices for data processing in A/P, accounts receivable, records management, or any other paper-intensive routine. This technology is a tool for achieving efficiency and transparency, but an organization must continue to pursue effective time and labor management protocols to maximize its usefulness within the context of the healthcare mission.
Benefits to Providers
Intelligent data capture technology allows a healthcare organization to focus its resources—its people and funds—more on its care delivery mission by making the organization’s paper-based processes faster, more accurate, and less costly. The automated process supports improved vendor and customer relations, reduces risk of supply chain disruption, and promotes greater employee productivity. Intelligent data capture also supplements EHR incentives tied to the American Recovery and Reinvestment Act and positions healthcare providers to better meet the demands anticipated as a result of the Affordable Care Act.
The value of intelligent data capture is aptly expressed by Jason Hergenroeder, senior director of finance, for Cleveland Clinic, a recent adopter of the technology: “By integrating quickly with our legacy applications to streamline processing while enabling greater compliance and accountability through process transparency, we believe intelligent data capture technology will help Cleveland Clinic to remain on the cutting edge for years to come.”
Charles Kaplan is vice president, marketing, Brainware, Inc., Ashburn, Va. (Charles.Kaplan@Brainware.com).
Sidebar
The Benefits of Automation
Manual data entry has been widely shown to contribute to process inefficiencies and increased costs in all industries, and health care, in particular. In almost every case, automation provides a strongly viable solution. Consider the following findings.
A 2010 study by The Hackett Group, an international business advisory firm, finds the median cost of processing an invoice among healthcare operations is $5.24, while those achieving world-class performance spend $2.56. These findings suggest that an operation with average performance that pays 1 million invoices annually could save almost $2.7 million annually by achieving world-class performance—dollars better spent on clinical care.
High costs are often due to inefficiencies resulting from paper processes. According to Aberdeen Group, A/P specialists spend half of their time searching for documents. Costs to recover misfiled
documents average $12, and costs to replace lost documents average $220.
Cost is only part of the problem, however. The Association for Work Process Improvement finds that one in 40 keystrokes results in an uncorrected error, creating problems or discrepancies in communication, payment, billing, or more.
Moreover, such errors can translate into payment errors. The International Office Management Association (IOMA) estimates that 1.6 percent of all payments are erroneous, resulting in underpayments (potentially straining vendor relations or disrupting supply chain) or overpayments (i.e., loss of revenue). And International Accounts Payable Professionals and The Institute of Management & Administration estimate that for every $1 billion a business spends annually, $1 million in duplicate payments are made.
Although these findings do not specifically attribute such problems to manual processing, evidence strongly suggests that the pairing of automation and best practices can go a long way toward eliminating these problems. In addition to streamlining the process, freeing personnel from repetitive data entry routines allows them an opportunity to specifically address anomalies—many of which can prove quite costly, as the above findings suggest—as opposed to overlooking (or even inadvertently causing) them.