By Sindhu Kutty for Forbes
Medical decision-making must remain with clinicians, but why does cumbersome data entry work continue to bog down their time? Can AI be used to allow physicians to spend less time on administrative tasks and more on value-added care?
Physicians can spend approximately one-third of their time creating notes and reviewing medical records in the electronic health record (EHR), and while some of this is related to bolstering ongoing care to help patients achieve positive health outcomes (for example, ensuring continuity of care for the patient between venues), the majority is for billing documentation (financial reimbursement) and to ensure regulatory compliance. And this comes at a significant cost. As payment models become more complex, physicians are seeking ways to improve clinical documentation. AI shows great promise.
This is particularly true for the tail end of the patient-physician encounter during the clinical validation or data reviews conducted for reimbursement, research and quality improvement. For example, Cerner, a leading EHR vendor, has developed a natural language processing (NLP) engine to automate medical chart reviews by evaluating EHR data and identifying opportunities for improvement and validation of documentation for in-patient encounters in near real-time.
That is great on the tail end, but what about the first step to excellent clinical documentation in the patient-physician encounter — data entry? The ultimate success would be clinical documentation software taking structured data from the patient-physician encounter and automatically entering it into its EHR field without any human help.
Voice recognition can help eliminate the burden of data entry from the care team. Physicians have long utilized medical dictation to dictate a structured clinical note along with human-powered medical transcription services, or software such as Dragon in conjunction with EHR, to alleviate their administrative burden. However, ambient voice can now get the same data from a natural interaction between physician and patient, which removes data entry completely. AI can automate the process (by becoming an “auto-scribe”) of producing clinical notes through speech recognition and natural language processing technology, in real-time, by listening in on patient-physician conversations or from summaries provided by physicians’ post-encounters with patients.
There are a significant number of market entrants in this space. For example, Nuance Communications provides software that uses neural networks to map patient-physician conversations into a note in the EHR (AI-based speech-to-text system) but requires wall-mounted devices with microphones to record each interaction. Nuance’s latest acquisition, Saykara, uses an AI voice assistant via a mobile application on the physician’s cell phone to transform salient conversational content between physicians and patients into clinical notes, prescription orders and referrals, and then populate both structured and narrative data directly into EHRs. This documentation of encounters in real-time negates the need for manual data entry or human transcription. In addition, physicians can also use this technology from within Zoom videoconference calls to document telehealth visits.
The U.S. transcription market size is approximately $20 billion (led by its use in health care) and is bolstered by this transition from traditional to AI-powered solutions, with voice recognition playing a significant role in this forecast. Through the use of innovative technologies like voice recognition and NLP, physicians can recoup up to three hours per day back for direct patient care and in the service of healing. This could mean an increase by up to a third of patient revenue per day, which amounts to a significant financial stream for provider organizations.
Based on my nearly two decades of experience consulting with more than 20 U.S. health systems, using AI to transform the clinical documentation portion of care delivery can create more seamless experiences for both patients and physicians on the continuum of care and should be explored as part of the overall digital health strategy. AI holds a lot of promise in leaning out a number of processes and reducing the burden on already overtaxed physicians especially during Covid-19. The return on investment is there, and I encourage CIOs and other health care executives to consider building these technologies into their strategic road maps.