ShadowScribe, Emdat's Speech Recognition Application

Emdat's ShadowScribe Transcription Editor (Click to view enlarged image)

Emdat's suite of applications uses M*Modal speech understanding technology to capture dictation and convert it into a structured and encoded transcription.  This technology partnership can dramatically boost transcription productivity as a medical transcriptionist (MT) evolves into a Medical Transcription Editor (MTE) – a transcription editor, rather than a transcriptionist.

Emdat's speech understanding technology, ShadowScribe, allows MTSOs to improve production time, reduce per line labor costs, and improve the quality and consistency of medical reports.  An MTE focuses on the structure and content of the draft clinical document rather than transcribing the entire report.

ShadowScribe can seamlessly and easily be incorporated into an MTSO's daily transcription workflow.

SPEECH RECOGNITION TECHNOLOGY

ShadowScribe allows the person recording a dictation to speak in a natural manner – as they would when talking with another person – not as they would speak to a computer.  The unique combination of speech recognition and natural language processing means ShadowScribe will recognize and understand natural speech patterns and automatically capture clinical facts and orders (as opposed to just text) to create structured and encoded HL7 CDA documents.

IMPROVED MT PRODUCTIVITY

For many MTs, productivity is limited by typing speed.  Physicians can speak 200–300 words per minute, but most typists can only average about 65–90 words per minute (this includes correcting typos).  Consequently, on average, an MT will spend about 4.5 minutes producing a transcript for every one minute of dictation.  With ShadowScribe, an MTE can spend an average of 2.2 minutes editing a document for every one minute of dictation.

Typically, an MT produces about 110-130 lines of transcription per hour.  With ShadowScribe, an MTE achieves an average rate of 220-250 lines per hour – a 100% increase in productivity – because they review draft reports instead of typing them from scratch.

It is important to recognize that this productivity increase takes time.  The MT must learn a new skill – editing -- in order to become an efficient MTE.  The chart below depicts the typical improvement in performance over an initial 11-week period.

The overall operational savings to an MTSO is dependent upon the volume of transcription processed by ShadowScribe and edited rather than transcribed.

ShadowScribe strives to be effective for 80% percent of an MTSOs' transcription volume.

BENEFITS OF SHADOWSCRIBE

No training or changes to dictation habits are required.  Physicians and other medical professionals maintain their dictation habits and speaking style.  ShadowScribe captures and comprehends dictation by factoring into the process the way people really speak, regardless of dialect, accents, speaking styles, dictation habits, grammar, mispronunciation or speaking rate.  ShadowScribe boasts a physician adoption rate of over 80% and utilization across all medical specialties and work types.

Context-related interpretation.  ShadowScribe puts dictation into context in relation to the facts and circumstances surrounding that patient’s particular situation or event.  It automatically converts dictation into narrative text and structured, encoded data with the proper grammar and punctuation.

Automatic content organization.  ShadowScribe represents clinical notes as dictated, while automatically placing the content in sections and lists as defined by best practices, understanding medications, dosages, measurements and more.

Continued learning.  ShadowScribe is trained through corrections to improve the recognition and understanding of an individual speaker’s dictation.  In addition, the system periodically reviews large volumes of transcription looking for edit patterns across physicians and specialties.  New medical terminology is recognized and added to the language models.

Easy Integration.  Customers do not need to install additional hardware in order to engage ShadowScribe.  Emdat provides on-going developer support throughout the workflow integration process.  ShadowScribe is hosted in a secure data center where hardware, software, back-ups, and disaster recovery are administered.  Software is updated frequently to ensure all customers are using the best speech understanding technology available.

SPEECH UNDERSTANDING IN TRANSCRIPTION WORKFLOW

Dictating users conduct business as usual.  They do not even need to be aware that ShadowScribe is being used. There’s no need to “train” the system by reading a pre-defined set of text, and users can be added at anytime.  Medical personnel simply dictate into the phone, a dictation recorder, or a computer, as usual, and Emdat’s InTouch or InSync uploads the dictation.

Dictations by medical personnel new to ShadowScribe enter “Training Mode.” In this initial stage, the clinical document is typed from scratch.  These typed transcriptions are compared to the voice file and used to build a voice model that will allow ShadowScribe to begin recognizing dictation from that physician.

Once a voice model exists, ShadowScribe generates a draft clinical document that is subsequently sent to an MTE for editing.  ShadowScribe is embedded into Emdat’s InScribe application and provides the MTE with the ability to review and make corrections to the draft.

ShadowScribe links the cursor to the text while the audio is playing – moving the cursor from word to word as it is played back.  The cursor, therefore, is in the appropriate position should the MTE see an error that needs to be corrected.

As the MTE makes corrections to the draft, ShadowScribe tracks the changes and continuously learns from the edits.  If a correction is made several times, ShadowScribe learns that the corrected form is desired and changes future drafts to match the recurring edits.  As time goes on, ShadowScribe becomes more accurate and produces more precise draft documents.  These productivity increases are the primary driver behind increased transcription margins.

Along with learning, ShadowScribe captures the keystrokes used for editing a draft document to determine if additional training can optimize an MTE's performance.  An MTE must learn how to edit efficiently in order to realize the full advantages of speech understanding.

After the MTE validates and corrects the report, InScribe automatically formats the document per the designated document template.

SHADOWSCRIBE TRAINING PROGRAMS

Customers about to employ ShadowScribe should spend some time training their MTs to edit rather than transcribe.  Transcription training programs provide the right background for editing – medical vocabulary, following physician directions, etc. -- but do not yet include information on editing vs. transcribing.

To assist customers in enhancing an MT’s skills with MTE capabilities, Emdat utilizes a "Train the Trainer" program in which ShadowScribe trainers teach MTE trainers how to edit efficiently and, in turn, how to train others to do the same.  Educational materials for an internal training program and a shortcut guide for MTs are available.  Emdat also provides post-training support.