March 12, 2012
Fixing Electronic Data Capture
By David Yeager
For The Record
Vol. 24 No. 5 P. 20
By identifying the shortcomings of electronic data capture, the American Medical Informatics Association hopes to boost patient care and clinician workflow.
Most healthcare industry experts agree that EHRs can be vehicles for gathering immense amounts of data. For healthcare as a whole, that’s good. For clinical care providers themselves, maybe not so much.
An electronic documentation tide is threatening to swamp clinical care providers. Because of their proximity to patients, doctors, nurses, and other clinical care staff are being asked to do something they probably didn’t envision when they decided to join the medical field: collect data for billing, regulatory compliance, and other ancillary purposes.
Data collection is a source of growing frustration among clinicians because, although it’s important to the medical enterprise, it diverts time and energy from their primary focus of treating patients.
The situation has become such a concern that AMIA (American Medical Informatics Association) decided to make it the focus of its sixth annual Invitational Health Policy Meeting. With participants from small and large institutions and the public and private sectors present, AMIA convened the December 2011 meeting to discuss a set of proposed guiding principles for data capture and documentation that will be released later this year following additional stakeholder vetting.
The preliminary draft outlines characteristics of clinical documentation and data capture that AMIA considers central to healthcare: clinically driven and patient centric; efficient; accurate, reliable, valid, and complete; supportive of multiple uses; enabling team collaboration; and reflecting input from multiple sources and devices, when appropriate. The final meeting report will contain more detail about the meeting discussions and offer public policy recommendations. AMIA’s goal is to reframe the discussion about how and why data are captured and what is necessary for high-quality, effective, and safe patient care.
“We have some notes that will be many, many, many pages because they’ll put things like laboratory results and radiology reports—entire radiology reports—in the daily progress note. And all you really have to say in the daily progress note, when the patient is stable, may take several sentences, one paragraph. But it’s buried in those many pages and, basically, other doctors who go back to read it can’t find it,” says George Hripcsak, MD, MS, FACMI, chair of the department of biomedical informatics at Columbia University and vice chair of the AMIA Policy Meeting steering committee. “So the first question is can we, in effect, take back the chart for clinical purposes and make that its primary focus?”
AMIA sees several problems that need to be addressed. Perhaps the most noteworthy is that documentation requirements for billing and reimbursement often have little to do with patient care. It’s possible that a greater emphasis on value-based purchasing could improve documentation because payment would be influenced by how the patient fares. But Hripcsak is concerned that it may not be enough, noting that different payment models may simply skew documentation in a different way.
Taking Back the Health Record
One possible solution would be to stop using patient notes for ancillary purposes. Hripcsak says resident sign-out notes—the notes they exchange with each other that aren’t part of the official patient chart—are often more clinically useful than the chart itself because they’re focused solely on patient care. However, to make this work would most likely require advances in voice recognition technology and natural language processing that aren’t yet fully available.
A change that could pay immediate dividends would be to step up education efforts about what constitutes legal documentation. When it comes to regulations, Hripcsak says there’s a knowledge gap among some administrators about what is actually required. Depending on an administrator’s interpretation of the law, clinicians may be asked to provide far more documentation than what is necessary, making it even less useful for clinical needs while increasing physician workload.
But the problem of excessive documentation goes beyond the fear of liability. At its root, it’s a clinical workflow challenge. Over time, electronic systems have been developed to capture specific types of data, with design priority given to the system’s technical capability rather than the end user’s needs. As a result, instead of being a cohesive workflow tool, it’s created a series of automated systems and record repositories that aren’t always well connected and are sometimes difficult to navigate.
“What we’ve come to understand is that in the design of our electronic health record systems, historically we’ve started out from a billing, financial perspective and then we’ve kind of moved on over into the actual work product component of healthcare delivery,” says Charlotte Weaver, RN, PhD, FAAN, a senior vice president and the chief clinical officer of Gentiva Health Services. “And we haven’t followed some of the other industries that brought in the human factor, the usability, the methodologies and design principles, and the methodology of testing.”
In an effort to capture as much data as possible, documentation moved from a paper-based system to a structured electronic system with discrete data fields. This minimizes variables in text to a certain degree, but it is much more cumbersome for clinicians. Linda Kloss, RHIA, CAE, FAHIMA, president of Kloss Strategic Advisors, Ltd, and former CEO of AHIMA, says it will take time to make the transition to newer data capture techniques.
She believes healthcare organizations need to move beyond the mindset of capturing all data in a structured way with drop-down lists. A deliberate evaluation of data capture practices is needed to determine which approaches work best for specific clinicians and data types. The development of flexible data capture options will allow workflow to be tailored for particular clinical environments, Kloss notes.
With data integrity and ease of use as guiding principles, future models should allow a combination of structured and unstructured data capture. Advances in voice recognition technology and natural language processing, although not yet capable of fully realizing that vision, are moving in that direction. A wider variety of options has the potential to not only alleviate much of clinicians’ current frustration but also to better address patient needs.
“As with the adoption of any technology, whether it’s productivity tools for an office environment or electronic health records, [what’s important] is integrating technology into workflow, how people work together, and the culture of the organization,” says Kloss. “I think, for past generations, documentation was subject to the style of the clinician, subject to certain constraints of accreditation standards and other healthcare policy. But technology has brought in both new constraints and new opportunities.”
While technology does offer new opportunities, it also presents new challenges. One is that clinicians are still learning how to document electronically. Hripcsak says the ability to cut and paste on a computer has reduced the number of opportunities for clinicians to catch omissions or errors in the chart. In the paper world, a doctor could write, “Otherwise unchanged from yesterday.” Computers allow them to carry forward data fields from the previous note. Because compliance officers take a dim view of the practice, the doctor usually has to attest that he or she did everything that was prescribed in the previous day’s note with a statement at the end that is copied and pasted from the previous note.
“By letting them shortcut the process, by cutting and pasting, we were changing the thought process and maybe cutting out essential steps in diagnosis and in patient safety. That’s the first part,” says Hripcsak. “The second part is maybe there are better ways to integrate it into the computer that we haven’t really engineered yet but would address both goals. It would be efficient for doctors but also promote patient safety.”
Hripcsak says focusing on the essential aspects of the note and instituting mechanisms that force the clinician to consider whether he or she has forgotten anything—much like a surgeon taking a time out before surgery to conduct a final verification that it’s the correct patient, procedure, and site—would improve clinical documentation. With better voice recognition and natural language processing, he says it may be possible to streamline the documentation process while making it more clinically relevant.
For example, it may become possible to design a documentation system that opens like a book and encompasses free and structured text. On one side of the “book,” the clinician could type in free text. Structured fields with text boxes would be on the other side. Free text that’s relevant to one of the structured fields would cause a check mark to appear next to its box. Clicking on a box would generate free text on the other side. Boxes without check marks would be highlighted in red. This would ensure that all necessary items in a clinical note are addressed. It would also allow the clinician to choose the most efficient method of documentation—either keyboard or mouse—as people already do in some nonmedical applications or on websites.
Hripcsak says engineering new computer-human interfaces such as this should be a research priority going forward. Existing technologies such as speech recognition, keyboards, mouses, check boxes, and radio buttons can be combined in new ways to better match workflow.
Another research priority should be the study of clinical care team interactions, says Hripcsak. Because information is often documented by more than one person on a care team, developing a better understanding of each person’s role in the care process may improve coordination and lessen the need for redundant documentation. Hripcsak says it may be helpful to develop a shared interface, such as a Google document or a social networking type of tool, that allows multiple users to collaborate transparently so that every member of the team can see what’s been done and what needs to be done.
The Best Policy
Although several vendors are working on solutions to some of these technological problems, there is a need for public policy to help shape these changes. In fact, public policy was a significant area of discussion at the AMIA meeting. A burgeoning trend toward data collection from wearable and implantable devices, as well as patient-entered data, has brought a call for reexamining current regulations and policies. Because public policy was implemented in the era of paper documentation, there has been much difficulty keeping up with these changes like with workflow design.
“There’s a challenge in public policy in keeping up with the rapid pace of technological advancement. And the other potential challenge will be the need to harmonize public policy and regulations across different agencies,” says Meryl Bloomrosen, MBA, vice president of public policy and government relations for AMIA. “The discussions [at the meeting] very much included the recognition that current federal activities to enhance the role of the patient and the consumer in our own healthcare will need to also address clinical data capture, patient-entered data, and data that might be captured by wearable or implantable devices.”
Bloomrosen believes it will take time for public policy to adapt to and reflect the changing technology that’s being employed in healthcare, but she is confident it will. She says AMIA members are likely to suggest that regulations be as nimble as possible going forward. Considering the technological and organizational transitions that are occurring, patience and flexibility will be necessary. Additionally, she says evidence for which types of data capture and documentation approaches work and which do not is sorely needed to better inform future policy-making efforts.
It will be equally necessary to account for the wants and needs of consumers, Weaver notes. Because federal policy has begun encouraging consumers to play a more active role in their healthcare, additional federal guidance is needed. In addition, bringing patients into the equation may be even more difficult than streamlining clinician workflow, but it will have a significant influence on medical clinical documentation for years to come.
“We have got to bring in the consumer, the individual patient, and we’re struggling with that one. That’s truly a challenge, but it’s imperative. This is where our health policy is,” says Weaver. “We have made a commitment to allowing the patient to add information to the record, to view their information, and to have information sent to their own personal health records. You can’t do usability in any of that if you don’t include those individuals.”
— David Yeager is a freelance writer and editor based in Royersford, Pennsylvania.
AMIA’s Draft Proposed Guiding Principles
Clinical data capture and documentation should include the following:
• be clinically driven and patient centric, reflecting an individual’s longitudinal and lifetime health status;
• be efficient, enhancing overall provider efficiency, effectiveness, and productivity;
• be accurate, reliable, valid, and complete, enabling high-quality care;
• support multiple uses, including quality and performance measurement and improvement, population health, policy making, research, education, and payment;
• enable team collaboration and clinical decision making, including the patient as a member of the team; and
• reflect input from multiple sources as appropriate, including nuanced medical discourse, structured items, and data captured in other systems and devices.