I went to medical school because I wanted to type all day.
– No Doctor Ever
MEDICAL TRANSCRIPTION
As medical providers, you have several choices for producing your documentation. These include medical transcriptionists, scribes, speech recognition, and artificial intelligence. An experienced medical transcriptionist is your best way to obtain accurate documentation with little need for oversight from you. In our increasingly litigious society, doesn’t it make sense to trust your documentation to experienced, accurate medical transcriptionists and not worry about the types of errors made by student scribes, speech recognition, and artificial intelligence? Why risk a large settlement or even your license?
ABOUT US
My service has been in business for over 25 years, and during that time my team and I have had experience with all medical specialties, many formats, and several different EMRs. Our goal is to ease your documentation burden by providing you with chart-ready reports that will not require hours of your time to correct. We offer the following benefits for your practice:
- Fast, accurate, reliable transcription
- Experience with ESL dictation
- Verbatim transcription or transcription with grammar cleanup, as you prefer
- Reports delivered on a secure .ftp site or placed directly into your EMR
If you already have an MT, we offer vacation or sick leave coverage for that person.
If your scribe is a medical or nursing student, you will be spending lots of time correcting errors, including homonyms (e.g., perineal and peroneal, ileum and ilium, etc.), dosages (15 mg often sounds like 50 mg), and laboratory values (again, many numbers sound the same). It isn’t the student’s fault. He or she just hasn’t studied that part yet. But you are the one who is responsible for the accuracy of your records, so you will have to fix any mistakes made by an inexperienced scribe.
Speech recognition is well known to make mistakes. There are actually several books that have been published that contain what are perceived to be funny errors made by speech recognition. They’re not funny when they appear on your patient’s chart. I actually had to correct an SR-produced report in which the doctor had dictated “cholesteatoma,” but the SR had put “Colorado State University.” I don’t think that’s funny (although it certainly would explain why the patient had ear pain with all those buildings jammed in there!). And errors in medication dosage are never amusing. If your patient suffers untoward side effects from a wrong dosage, you are the one who is held responsible, not the software program. Again, this demands more of your time to fix these reports.
Artificial intelligence is the shiny, new method of documenting your patient encounters. It shares the inability of speech recognition to distinguish between similar-sounding words. But it also comes with a new problem, what the IT experts refer to as “hallucination.” Sometimes, the AI program will add things into your report that you never said. For example, it may extrapolate from your diagnosis what your treatment course would be and add that in even when you have not dictated it. Perhaps you would agree with the treatment, but maybe you had a different course in mind. In any event, anything you did not dictate does not belong in your report. As a matter of fact, if your report says that you did something that you did not actually do, it constitutes fraud. Another problem with AI is that instead of making a report for one patient, it may blend together two or more patients’ reports into one. And here again, you must spend your time reviewing and correcting any mistakes in your report, even whole paragraphs you did not dictate.