Voice To Text Solutions Could Turn EHR-Burdened Medical Professionals into Doctors Again

Just imagine the following: your assistant invites in Ms. Nichols, who has a migraine, a bad cough, and feels nauseated. You sit down, start talking about the symptoms, see her throat, measure her temperature, pulse rate, inquire more about the headache. In the end, you set up a diagnosis, you write a prescription for some meds, send the patient for some further exams due to the migraine, but ensure her that everything’s going to be fine.



Any doctor would just get angry upon describing such a scenario – as it is unfortunately very far from reality. While in an ideal world, a patient-doctor visit should go down like this, according to a study published in Family Medicine, primary care physicians spend an average of at least 30 percent of a doctor-patient visit working in electronic health records (EHRs). While an average visit lasts for 35.8 minutes, time spent on EHRs included 2.9 minutes prior to entering the room, 16.5 minutes of face-to-face time, 2 minutes in the room, and 7.5 minutes of non-face time. Additionally, 6.9 minutes were spent working on the EHR outside of clinic hours.

Luckily, many health technology companies recognized that innovative solutions should and could give back the time and freedom that the administrative burden caused partly by EHRs took away from doctors. The most exciting ones are definitely those innovations that promise to listen in patient-doctor visits and the “conversations” between doctors and the EHR systems and provide a transcription without the doctor typing even one letter into his computer.

Medical Siri, Alexa, and other digital assistants with the task to listen to what the doctor has to say and transcribe that correctly into administrative systems started to thrive. They mean a real alternative to medical administration done manually by doctors – and institutions agree. Research firm Technavio published a report last year projecting that hospitals globally will spend more than 72 billion by 2020, representing 6 percent compound annual growth rate.

There are already valuable tools that can listen in to patient-doctor visits and have the potential to free up doctors’ time. San Francisco-based Augmedix aims to harness the power of Google Glass to make health care more patient-centric and decrease the amount of paperwork by generating medical notes in real-time, while doctors have a conversation with their patients. Start-ups like Nuance and M*Modal provide software-based dictation services to physicians. Another, California-based company, Notable, launched a wearable voice-powered assistant in May 2018 aimed at helping doctors capture data during interactions with patients.

Another strand of voice-to-text technologies could take the burden of administration off the shoulders of doctors. Lately, voice recognition technology started to replace conventional dictation across a variety of health care information systems, too, because it’s able to eliminate transcription costs and minimize transcription errors. Cody Frew, Director of Marketing at ChartLogic, a company offering voice technology to EMRs, said that thousands of dynamic, command-based responses programmed within an EHR system can substantially reduce the time it would otherwise take to perform conventional dictation. Plus, the need for transcription is removed from the equation entirely, easily saving the average physician $30,000 to $50,000 a year.

Another company, Nuance Communications offers software called Dragon Medical One, which it claims can help health care companies record patient medical experiences using natural language processing. The company says it already helped Allina Health, a health care provider, which integrated the software into its EPIC EHR, speed up the time for its doctors to fill out electronic health records. According to the case study, Allina Health saw a 167 percent increase in how much medical documentation they were able to produce by the time of publishing. In addition, there are already smart versions of common clinical devices such as thermometers, blood pressure cuffs, and scales that automatically record readings in the patient record, so doctors do not have to type the measurements.

However, recent voice recognition solutions do not eliminate transcription errors fully, thus the need for proofreading and human check-up will still take up the precious time of medical professionals. And that should be taken seriously, as even one letter recognized differently could mean a potentially life-threatening danger for patients. That’s why automated, ANI-based voice recognition systems with similar, ANI-based proofreading programs might embody the final solution for medical administration in the future. Just imagine, only the problems flagged by the ANI system would be checked by a doctor, otherwise the entire process of administration, taking medical notes, etc. would be automated. If we could move into this direction soon, the above patient-doctor visit could easily turn from science fiction into reality.

Source