Since moving to electronic medical records, doctors spend hours each day typing clinical notes. For many providers, the charting demands of EMRs have been a cause of burnout. Others have responded by turning to medical dictation software and other workflow solutions that reduce the need to type. When considering typing vs. dictation the obvious question is “Which is faster?”
Whether you type or dictate patient notes, it is ultimately a matter of minutes. There are 24 hours in a day, and the average primary care physician spends 1.5 of those finishing paperwork after clinical hours. Doctors interested in a healthy work-life balance need to use their EMR as efficiently and accurately as possible.
So which is faster – typing or dictation? Just as importantly, which is more accurate? Here is a summary of the numbers based on existing research and estimates.
Dictation and typing speeds compared
Short answer: Dictation is faster.
Long answer: Dictation speeds vary depending on the type of dictation. Many physicians who dictate use a third-party transcription service, which receives a recording of the dictated note. The recording is then transcribed by a computer and edited by a professional transcriptionist. This method can be called “back-end dictation.”
Another option is fully automated and immediate transcription using speech-recognition software. With this type of “front-end dictation,” the physician dictates directly into free-text fields of the EMR and edits the transcription before saving.
In both cases, dictation is faster than typing.
“The average US physician could reduce documentation time by about seven hours per week by switching from typing to dictation.”
Speech recognition software can easily transcribe over 150 words per minute (WPM), while the average doctor types around 30 WPM. Professional transcriptionists editing a dictation type around 50-80 WPM, an average that is also much faster than physicians.
Stanford researchers studying typing versus dictation found that speech recognition was nearly three times faster while also producing fewer errors.
We did the math using conservative estimates and found that the average US physician could reduce documentation time by about seven hours per week by switching from typing to dictation.
The bottom line is that you speak much faster than you type. As with typing, getting a dictation perfect the first time requires speaking skills that you develop over time. However, a small amount of effort to refine a new dictation workflow will literally eliminate days of typing over the course of a year.
What about accuracy?
Short answer: It depends on your workflow, but dictation can be as accurate as a carefully typed clinical note.
Long answer: Accuracy of the clinical note is a major concern of every physician. While doctors increasingly opt for front-end dictation using speech-to-text software, this isn’t universal. Many providers still send dictations to third-party transcription services where notes are edited and reviewed by a professional transcriptionist off-site.
Unfortunately, researchers have found high average error rates using this back-end transcription. This may be because it took physicians in the study an average of four days to review and sign a dictated clinical document that was sent away for transcription. The lag time created by third-party transcription services creates room for transcription errors, which are often clinically significant.
While voice recognition software has improved immensely in recent years, dictation errors are inevitable. Existing research underscores the importance of manually reviewing your notes to ensure accuracy.
The problem of accuracy is largely avoided by front-end medical dictation using software on a computer or mobile device. Using this method, physicians dictate the clinical note and their words appear instantly on the screen. They can quickly detect any errors and make edits that are immediately saved in the EMR.
Using the latest speech recognition software for instant dictation, physicians can ensure a clinical note that is 100 percent accurate, just as it would be with careful typing.