Medical Transcription the EMR and Speech Understanding

The Medical Records institute e-Newsletter from September contained an article by Claudia Tessier from the Medical records institute titled: "Medical Transcription and EMRs: Opportunity Lost?" that discussed the relationship between medical transcription and the electronic medical record (EMR). As the Claudia says:

...medical transcription offers a bridge to EMR adoption

But the idea that

the EMR offers the best opportunity yet to get rid of transcription and its concomitant headaches

Misses the opportunity for medical transcription and valuable data that is lost with the push towards the structured form based hunt and click style documentation. In a recent discussion with a clinician he lamented the loss of "the beauty and descriptive nature of medical language that has been used to describe medical conditions and image findings". Instead as he put it "we have turned detail rich clinical information into dumbed down fill in the blank cookie cutter reports" which do not reflect the richness of the information he wants to provide to his colleagues.

To date medical transcription is estimated to constitute 60% of the input into current EMR systems but that input is in the form of text blobs and not clinical data. The article goes on to suggest that:

...EMR vendors should ramp up their cooperation to create uniform integration. Let every one of the 300+ EMR systems allow dictation and let the market determine whether the related turnaround time, quality, costs, etc. (see below) are acceptable. Let users dictate on cell phones and dictation devices, or through laptops and tablets—whatever their preference.

Is spot on - the systems work and with better integration and more choices for input we open the doors to capturing input from our clinicians caring for patients and struggling to document for the benefit of communication with others members of the team as well as capturing sufficient information to be paid for the services they are delivering to their patients.

But this input is still not resolving the necessity to feed EMR's with clinical data which is essential for computer based systems to understand the information and be able to act on it. There are existing standards to hold and transmit this information including the Continuity of Care Record (CCR) which is "....working in collaboration with HL7 on the expression of ASTM's Continuity of Care Record content within HL7's CDA XML syntax and the seamless transformation of clinical and administrative data between the two standards.” - Rick Peters, MD

Transcription companies are already offering xml-based solutions that support structured output and the significant value this brings to EMR's is that this structured data has been checked and reviewed by medical transcriptionist/editor with expertise and knowledge to validate that content relative to the original dictation input of the clinician. More value from the validated data output from Medical transcription will make the transcription industry more of a partner and even more important in their contribution and ongoing role in the delivery of high quality healthcare.

So where does Speech Understanding come into all of this - unlike the traditional speech recognition technology which Hollywood conditioned us to expect far more comprehension on the part of the engine as captured in this classic clip from Star Trek IV - the voyage home where Montgomery Scott (Scotty) of the original series of Star Trek fame is trying to interact with a computer circa 1980.....

But his experience is typical of traditional speech recognition systems and a typical response either visually or verbally would be "Directions unclear - please repeat request"

Speech Understanding is the next generation of the technology, crossing the chasm between the need and desire of physician to dictate using all the richness and expressivity of language but that is recognized and understood and not only creates an accurate representation of the free form text but also produces a structured and encoded document. Structure is captured and stored in native CDA format and encoding is achieved using clinically relevant encoding systems such as RadLex for radiology, RxNorm for drugs, Universal Medical Language Systems (UMLS) and SnoMed for clinical terminology etc

You can have the best of both worlds and Medical Transcription will be around for years to come - albeit in a updated MTv2.0 form where the transcriptionist is a knowledge based worker proofing, editing and validating clinical data......so in the words of Spock: "Live Long and Prosper"