The EHR Equation: Good Narrative = Better Results
By Isam Habboush
As part of the HITECH Act, the initiative to achieve widespread meaningful use of EHRs is aimed at promoting healthcare reform through improved quality, safety, and efficiency. In today’s predominantly digital world, increasing the adoption and utilization of EHRs remains a hurdle to overcome in enabling broader reform. In this context, the adoption of EHRs refers to the pervasive deployment and utilization of electronic systems in which clinical data can be effectively created and managed.
Despite the fact that the value of digitization and automation is ingrained in our collective personal and professional psyches, the transition to a digital healthcare system continues to progress slowly. One possible explanation for this lag is that most HIT systems were not designed with the practicing physician in mind. Rather, HIT systems’ design has largely followed in the footsteps of their counterparts in the business world, where the target audience is office workers who spend most of their working hours bound to a desk and interacting with a computer. Most office workers have excelled in the art of touch typing and have mastered navigating endless computer screens and pull-down menus with a mouse or other pointing devices.
Conversely, physicians are not office or desk bound. Physicians’ time is best spent seeing patients and delivering care, either in the clinic, the emergency department, or the hospital wards. As such, interface designs of IT solutions that may have proven successful among office workers may not necessarily translate to a physician’s day-to-day environment.
The challenge of EHR adoption goes even deeper than user interfaces. In an effort to address physicians’ concerns for usability, the HIT industry has responded with template-driven menus with preconfigured pick lists and check boxes to streamline patient documentation. Unfortunately, while addressing usability, designers have encumbered the physicians’ ability to use narrative (detailed notes or dictation) in their patients’ documentation.
Narrative is used to describe the uniqueness of a patient’s clinical story, including impressions and observations by the physician, and to describe the clinical decision process that led to a specific diagnosis and course of treatment. Most of this information doesn’t naturally lend itself to a template-based approach, where commonality is typically emphasized over uniqueness and brevity over details. Consequently, narrative remains an essential component of any clinical record and in communication among providers as part of an ongoing care episode, referrals, or consultation. Any move toward the digital record must preserve the narrative without interfering with normal workflow.
In a recent survey conducted by Nuance, 1,000 physicians were shown a narrative dictation of a patient encounter side by side with the template-driven equivalent documentation using a leading EHR system. Ninety-seven percent of physicians selected the narrative as the more valuable documentation in treating patients, and 96% expressed concern that they may lose the patient’s unique story with transition to point-and-click EHRs.
Clinical language understanding (CLU) is a new technology currently being developed to bridge the gap between narrative dictation and structured data entry into an EHR. As physicians dictate their clinical notes into an EHR, CLU analyzes the narrative and automatically extracts pertinent clinical data such as problems, symptoms, severity, allergies, medications, and smoking status. These data are automatically normalized and coded into standards such as ICD-9, SNOMED, and RxNorm, as appropriate. Physicians can review, edit, and sign off on the extracted data, which then are encoded into a Health Level Seven International clinical documentation architecture format and made available to the EHR or other downstream clinical IT systems for consumption. The final result is akin to the physician taking the time to manually enter the data into the EHR using menus and pick lists but with the added benefits of the availability of the narrative and the efficiency of doing it only once.
The initial goal of CLU is to support users as they work to comply with the requirements outlined as part of the meaningful use criteria. However, several other downstream applications are being investigated. For example, CLU technology can be used to assist in quality and regulatory reporting, coding for billing, and clinical decision-support tools.
Documenting patient visits is complex and does not naturally fit into the confines of template-based user interfaces. As CLU applications find their way into routine clinical use, the bridge will finally be established to align the ways in which physicians like to document information and how EHRs and other clinical IT systems consume and utilize clinical data, leading to greater acceptance and adoption of digital records.
— Isam Habboush is director of product management at Nuance Communications, Inc.