Medical Scribe vs. Dictation: Which Is Better for Reducing Physician Burnout?

In the 2021 Medscape survey, an internist hit the core problem of physician burnout when asked what contributes most to the issue: “I’m working 6 days a week, nights, weekends, and holidays!” [1] The last blog broke down the reason for this burnout problem–that there is not enough time in the clinical day to do the work of almost two full-time jobs.


Healthcare leaders and physicians alike have looked to tools to reduce burnout symptoms. Two very popular tools are medical scribes and dictation services. Which service works better? Which will be better in the long run to obtain the return on investment (ROI)? Both medical scribes and dictation offer time-saving options, and both have pros and cons. Let’s assess each to get a more defined conclusion:


Medical Scribe

Medical scribes, also known as documentation assistants, are professionals that in real-time document the physician/APP-patient interaction and enter it into the EHR, including the patient history, test results, physical examination findings, and any other pertinent information. Medical scribes are often pre-med, pre-nursing, or pre-APP students eager to learn documentation. Scrivas medical scribes are certified through Scrivas University, which is run by a medical doctor to ensure the highest level of education.



  • Being physically present in the exam room allows the scribe to enter much of the data without additional assistance from the physician since they are able to see what is going on, especially in a physical exam.
  • Medical scribes can prep notes prior to the start of the day, check on lab and test results, and communicate actively with other care team members and the physician.
  • Increased RVUs & patients seen per hour:
    • Results of a combined family and internal medicine clinic saw a 10.5% increase in work relative value units and an 8.8% increase in patients seen per hour. In addition, physicians’ time spent on the computer decreased by 27% and patient-physician face-to-face time increased by 57%.[2]
  • Increased Physician-Patient Experience:
    • Although some healthcare leaders and physicians themselves believe it could negatively impact the physician-patient experience, studies have shown the opposite. A recent study[3] focusing strictly on the impact that a medical scribe has on the patient and provider experience showed that the experience was strengthened through the utilization of a medical scribe.


Dictation Services

Dictation services can be broken down into three main categories: Speech recognition software, Artificial Intelligence (AI) dictation, and Transcription services:


Speech Recognition Software

Speech recognition (SR) software has been around for over two decades.[4] Speech recognition software has the ability to identify words that are spoken and convert them into a format on the EHR system or other computer application. There are physicians that do SR while the patient is still in the room to ensure he/she relays the most accurate information from the visit while others do it right after a patient visit and/or later in the day to keep the flow of the patients as much on time as possible.



  • SR allows for time to be saved with improved workflows.
  • SR can offer a faster turnaround time compared to transcription services.



  • SR has an abundance of challenges: variable quality due to transcription errors and grammar, difficulty of use if there is too much noise where the physician is, heavy accents, power outages, and security concerns.[5]
    • In a 2017 study, the researcher Tobias Hodgson concluded, “For clinical documentation, speech recognition was slower and increased the risk of documentation errors, including errors with the potential to cause clinical harm, compared with keyboard and mouse.”[6]


AI Dictation

For AI dictation, also known as “digital scribes,” deep learning is used to try and generate accurate notes into the EHR. To generate a medical chart, the AI must be able to record the entire patient-physician interaction, convert the audio to text, and extract the needed information from the text and summarize the information.



  • AI dictation frees some of the physician’s time to have more face-to-face time with the patients.
  • Physicians spend less time on the computer than doing the charting his/herself.



  • There are a number of challenges for AI dictation though, such as[7]:
    • Audio recording and speech recognition: A recent study found that the word error rate of simulated medical conversations with commercial ASR engines was 35% or higher. [8]
    • ASR produces a transcript that lacks the clear boundaries and structure of a proper patient-physician interaction due to the unconstrained nature of the conversations.
    • Information extraction in a clinical conversation is basic: from spoken vs. written and expert terminology, which can have words that sound similar to each other.


Transcription Services

Just like SR, some physicians dictate into a recording device while the patient is still in the room, right after the visit or later in the day, and the recording is then set to a third-party medical transcription service. The service will then transcribe the patient encounter into the EHR by a professional transcriptionist. This method is called “back-end dictation.”



  • Physicians spend less time on the computer than doing the charting his/herself.



  • Studies have shown that there are high average error rates using a third-party transcription service.[9] This could be because it took on average four days to review, submit any changes that needed to be made, and sign the chart.



Key Takeaways


  • More technology doesn’t help physicians struggling with technology.
  • A medical scribe performs complex and detailed work beyond just transcribing conversations in the exam room, which (no matter the dictation type) physicians cannot do currently.






[4] Johnson M, Lapkin S, Long V, et al. A systematic review of speech recognition technology in health care. BMC Med Inform Decis Mak. 2014;14(1):94.






[8] Kodish-Wachs, J., Agassi, E., Kenny, P. & Overhage, J. M. A systematic comparison of contemporary automatic speech recognition engines for conversational clinical speech. AMIA Annu. Symp. Proc. 2018, 683–689 (2018).