Speech Recognition

Speech Recognition

In today's fast-paced health environment, Physicians are facing enough pressures with EMR/EHR adoption and other mandates. Med-scribe understands the need to focus as much time as possible on patient care, not paperwork or data entry.

Med-Scribe's back-end speech understanding improves your clinical documentation time, reduces per line costs, and improves the quality and consistency of your medical reports by allowing the patient story to be included as dictated narrative. Our tools understand the way people really speak, regardless of dialect, accents, speaking styles, dictation habits, grammar, mispronunciation or speaking rate. Med-Scribe's editing team reviews each processed report against physician dictation, and makes required edits on the physicians' behalf so reports are ready for quick review and sign-off.

Back-End Deployment with Med-Scribe Editors

Med-Scribe's back end speech understanding model is deployed to be totally transparent to the physician and supports seamless changes to the physician workflow. This technology provides accurate translations into detailed clinical narrative and saves the physician as much as 56% of their time when compared self-editing reports created by the front-end voice recognition process while decreasing costs by 21% compared to the same process.


Advantages of Med-Scribe's Back-End Approach
Front-End Voice Recognition


Failures: Software accuracy is unreliable. Documentation is more time-consuming when physicians are required to edit and "teach" the program.

Front-end speech recognition technology (SRT), where words are displayed as they are spoken, requires the clinician to be responsible for editing and signing the document

  • Even with the latest technological advancements in speech recognition software, shortcomings can induce errors in the transcribed reports and thus seriously affect the outcome of treatment.
  • In many cases, SRT mean tethering physicians to a workstation, sometimes a specific workstation that recognizes their voice.
  • For front-end SRT to be as accurate as possible, a user must immediately correct the errors made by the software so the program will "learn" the intricacies of the user's speech patterns.
  • Front-end SRT takes considerably longer to use effectively than it does to simply dictate.


Front-end SRT increases productivity by 13%
Back-end SRT increase productivity by 69%


Successes: Physician productivity is maximized, physician workflow and revenue capacity isretained.

A hybrid approach accomplishes documentation goals without affecting physician performance.

  • Dictation is the most efficient means of documenting patient encounters.
  • Dictation "normals" further the efficiency of dictation and can easily be recalled during dictation to save time.
  • Back-end SRT provides and additional quality control back-up in form of the trained eyes of the medical transcriptionist who efficiently filters out the transcribing errors made by the software.


Front-end SRT decreases costs by 28%
Back-end SRT decreases costs by 49%