Transcription QA Tools – Part Two

A few weeks ago, I wrote this post entitled, "Real QA for Medical Transcription."  To revisit the concept briefly, true quality management activities include data collection of information such that we can provide a picture of root causes of problems, thereby finding true solutions to those issues.  Nagging MTs to untangle terrible dictation is a good example of an industry not using root-cause analysis.  Here's an example of one of my old errors that, had true quality management efforts been in place, my error would have been corrected years earlier.  I was transcribing "phosphorous" where it was appropriate to transcribe "phosphorus."  One is a noun, one is an adjective, yes?  Read more about that here.  I was 5 years into medical transcription before that error was pointed out.  Callus/callous and mucus/mucous are more examples of that.  My own proofreading would never find that error because, obviously, I didn't have the knowledge about its proper use.  I didn't know what I didn't know, you know?  This is ONE problem with us proofreading our own work, a topic for another discussion, like this one, "Why It's Hard to Catch Your Own Typos."  With proper proofreading and quality management, my phosphorous error would have saved our American medical records from hundreds, maybe thousands, of the now permanent inclusion of my error in patients' medical records.  The moral of the story is, not enough of my work was being reviewed for a very long time by different companies and hospitals, me transcribing this one error over and over.  Ugh.  No real QA.  No feedback to me.  I'll show you in the process I'm describing below how this would have been caught and corrected 5 years earlier.  But first, this.

How Much Should We Review?

I think we review 100% of our work as part of the regular process.  I think we spend our time on accuracy, on collecting data about the quality of dictation on all reports such that we can identify trends of poor physician dictation; that cannot be accomplished with spot checking.  How much time and frustration do we spend on terrible dictation, poor speech drafts, 20 or more years trying to decipher lithotripter reports done in the lithotripter room by a physician that no one will ask to step away from the loud banging, 2-3 MTs spending their time to sort it out, and not getting paid to do so, by our production pay scheme.  You might imagine this has personally happened to me.  In the 10 years that I have been transcribing, only one or two companies have even attempted to do any type of routine checking of my work, the most aggressive process looking at 5 of my reports a month.  I hear so many say, "You can't change physician behavior."  Oh yes you can, but you have to be willing to counsel them and it will be helpful if you have data and some enforcement.  How much time would be saved identifying problems at their root, making recommendations to fix them, having the backing of administration, and enjoying future quality dictation?  This is all possible with proper quality management, but we cannot identify trends by looking at 5 records/month of one particular MT.  Proper quality management is done concurrently and is not a postmortem analysis of MT errors, but a procedure that assures accuracy in the first place and one that assures patient safety.  This is the mind switch I'm suggesting.

It Must Be Specific

Proper quality management identifies specific root causes.  Trended data provides a true picture of the problem.  Was phosphorus simply a typographical error or a knowledge error?  When seen once, it might be a typo, repeatedly, obviously a knowledge error, maybe a shortcut error, as well.  Is the quality of Dr. Smith's dictation poor because she speaks too fast, too low, her accent is difficult, or she is using a phone in OR #3 that doesn't work well?  Are the VR drafts coming out poorly for one physician over another?  Why are Dr. Brown's VR drafts in better order than Dr. Parker's?  When we identify the specifics of these things, we now have the information to correct them.  If we collect data about the quality of dictation, maybe we will find that MT Amy seems to be able to do Dr. Smith's reports with better ease than MT Barbara.  Why is that?  Possibly Amy has been doing Dr. Smith's reports for 10 years, while Barbara is new to them.  How frustrated are you with the current QA processes that are punitive by people who don't appreciate that it takes some time to develop an ear for some dictators, where we are tempted to guess at a thing to avoid the wrath of a scolding reviewer?

I have talked to Sherry Doggett since I created the "Real QA" blog in January.  I have been critical about not being able to get the ear of AHDI professionals, so I want to always make sure I do the opposite when I have that experience too.  Sherry kindly reached out to me when I wrote about my frustrations with Speech Wreck in November 2013.  Read about our discussion here.  Here's what I love about Sherry. She is honest, forthcoming, and not afraid to say, "Okay, I see what you're saying."  Sherry had sent me an email stating that she had worked with an AHDI group to develop a QA Toolkit found here.  I had, indeed, reviewed the toolkit before, but went to have a look again as a refresher.  While I believed that the information described the theory of QA well, when it got down to the nuts and bolts, all I continued to see were examples of scoring MTs with minor or major errors, still no trends identifying root causes, still no information about other aspects of this process.  We exchanged a few emails, then Sherry said, and I'm paraphrasing, "You know, I think you might be rightfully challenging us to take a better look at this."  We had a second conversation this past Wednesday, and I really love this lady, much like my love affair with Laura Bryan!  We are scheduled again to talk on February 24 to see how my ideas can be implemented into their toolkit effort.

Sherry and I also talked about 100% review.  As a retired director of a large university transcription system, Sherry communicates, and she is not the first, to say there is no way a C-level administrator is going to approve the addition of staff to do 100% QA on medical transcription.  I get this, but then again, I don't.  Let's compare ourselves to other document-producing industries.

-How often do you see an error in a published book?
-How about a newspaper article?
-Legal documents (court reporters hire and pay upwards of $20/hour to have their work proofread).

No one in these industries are spending their QA/editing efforts on scolding typists or transcriptionists. No, they simply employ a process that assures accuracy, it's important.  Why isn't it just as important when it comes to our medical information?

I'm asserting this.  Proper quality management of important medical record documents, and EHR entries, by the way, should be viewed as part of the process, not a luxury after the fact.  The MT is the person translating speech to text, the scribe is translating speech to text in another way, clinicians are inputting data in realtime, but the realtime data is for a different purpose than the dictated discharge summary.  Digest that for a moment.  All are one-time entries that are subject to error.  The end game is accuracy.  We have RHITs in medical records reviewing for completion, signatures, the presence of Joint-Commission-required documents.  It is part of the process. Imagine if we just looked at a sampling of those items.  Imagine if we looked at 5 of Dr. Sullivan's records because he usually signs all of his progress notes, but we can't spend the money to assure that he does.  Why is medical transcription any different, and why do we, as an industry, feel we must build in that cost savings in the methods we promote, and if we insist on continuing to do that, let's call it that.  I believe it is our job to promote the highest level of accuracy, then leave the implementation to individual healthcare facilities.  And currently, to be honest, the MT absorbs much of this cost because we are not paid to proofread (nor should we be proofreading our own work).

Do you know that ALL surgeries are reviewed by QA for appropriate indications and that trended data of that is required? Do you know that all blood transfusions are reviewed for the same reason, appropriate indications?  The use of antibiotics, unexpected returns to the operating room.  Yes, these things are reviewed.  Why not the accuracy of medical records?  I contend because quality standards are not enforced by The Joint Commission in nonclinical areas.

picture worth 1000 wordsSo now, this.  These three forms have been developed to show you exactly what I mean.  First, we collect data, then we enter the data into a database so we can sort and identify trends, then we pull out problems we identify, document them and make recommendations to fix the problems found.  These are samples and can certainly be adjusted to focus on particular problems in different healthcare facilities.  The important part is to COLLECT DATA, IDENTIFY TRENDS, MAKE RECOMMENDATIONS, THEN FOLLOW UP to make sure the identified problem has been resolved.  One call to Dr. Smith may not do the trick.  A formal reporting system with data will surely get you there with the backing of executives who are "over him/her."

Form #1:  QA Report Data Abstract
Form #2:  Transcription Trends Sample
Form #3:  Transcription Problems and Recommendations

As part of the healthcare facility's quality improvement process, form #3 would likely be a hospital-wide form used by all departments to report their findings to medical staff departments, the medical record committee, executive committee, and board of directors.  When problems are identified and recommendations made, these QA bodies can provide the enforcement necessary to make change when feedback is simply not enough to deal with the physician who "won't change."

sherry doggett
Sherry, thank you once again for valuing and validating my message and being willing to simply reach out and have meaningful conversations about this.  I am so grateful.  Sherry and I will be chatting again on February 24, and I am looking forward to the conversation.




Leave a Reply

Your email address will not be published.

You may use these HTML tags and attributes: <a href="" title=""> <abbr title=""> <acronym title=""> <b> <blockquote cite=""> <cite> <code> <del datetime=""> <em> <i> <q cite=""> <s> <strike> <strong>