Why Physicians Dislike Speech Recognition and EHRs
Making electronic health records (EHRs) an industry standard is a good idea. Having a digital patient record that is accessible to multiple healthcare practitioners should improve the quality and convenience of patient care as well as the accuracy of diagnoses. Additionally, the ability to use a speech recognition product to digitally capture patient information should expedite administrative processes, saving practitioners time and money. It sounds like a win-win for healthcare practitioners and their patients. This is why I was surprised when a physician recently told me he wasn't a fan of the digital transformation.
He finds speech recognition for EHRs cumbersome, claiming he spends more time correcting his electronic transcriptions than he did with his previous pen-and-paper system. This, he argued, cut into his time with patients.
In fairness, I can understand his point. After all, a 90 percent accuracy rating for a speech recognition application might sound quite good, but if the average sentence has only 10 words (which is quite low), each sentence could average one error. That's a lot of errors, especially when patients' lives and quality of life are at stake. Clearly, someone still must check the accuracy of the patients' data after it is entered into the system.
Despite this challenge, not everyone in the healthcare industry shares his negative sentiment toward speech technology and EHRs. One healthcare industry professional told me, when it comes to EHRs, physicians "will get out of it what they put into it."
This certainly makes sense, considering that some healthcare practitioners are already noticing very good results with their use of speech technology for EHRs. Sharp Rees-Stealy Medical Group is one of them. With 400 physicians and 1,700 staff members, the medical group is one of the largest in San Diego County. It chose Nuance Healthcare's Dragon Medical 360 Network Edition to speech-enable the group's Enterprise EHR solution from Allscripts. Doing this allowed the medical group to replace the labor-intensive paper process, eliminate $900,000 in annual transcription expenses, and improve documentation quality.
Additional examples are featured in our cover story, "Speech Readies Healthcare for ACA Requirements," by Senior Editor Michele Masterson. These and other success stories have undoubtedly helped to improve the overall consensus among physicians. According to the story, 65 percent of physicians believe intelligent virtual assistants could "deliver more precise and timely data that could boost care and notify them of missing information" in a patient's electronic medical record.
While this stat might sound impressive, it also means that a sizable number of physicians (35 percent) don't believe this will happen. Clearly, the speech technology industry has some more convincing to do.
Admittedly, speech recognition and EHRs are not perfect, and there will likely be obstacles along the way. However, when it comes to improving patient care, speech recognition combined with EHRs is a step in the right direction.