I was reviewing AHDI’s Healthcare Documentation Quality Assessment and Management Best Practices (updated July 2017) recently and reading some of the comments here at MT Exchange and a thought occurred to me.
If I’m being really objective about documentation accuracy and accountability, and quality assessment, I’d add a category and quality point scoring system for dictation and audio, and a way to tie it back to the transcription QA score. And as usual, that’s a different post altogether and might be covered in future installments.
What I really want to talk about is blanks and policies surrounding blanks.
I’m not going to quote what you can read yourself; there’s an entire section in the AHDI document, which is linked above, and is comprehensive. AHDI has included a list of valid reasons for leaving blanks. There are unresolvable blanks, and there are resolvable blanks; i.e., valid and non-valid blanks. Non-valid blanks are blanks that can be resolved by an third party – and I’m not talking about the dictator. One would hope the dictator knows what s/he said, or intended to say. Calling a blank “non-valid” because the dictator filled it in is specious, in my opinion. I will, however, quote the following from the AHDI Best Practices document, because it bears emphasis:
Some hospitals, facilities, or organizations, however, may restrict the essential practice of leaving blanks in favor of what they call “complete” documentation. MTSOs often face pressures from their customer base to reduce and even, in extreme cases, completely eliminate all blanks. Since blanks play a vital role in healthcare documentation when used appropriately, to attempt to resolve unresolvable blanks is an unreasonable expectation and puts undue pressure on the MTSO and its HDSs [MTs] or the facility’s scribes. It is important to recognize that such a practice may encourage the HDS [MT] or scribe to hazard a guess. Guessing or making up content just to fill a blank is unethical and would require the HDS [MT] or scribe to practice their profession in a manner contrary to their training and integrity. (Emphasis added)
I see resolvable blanks as a teaching opportunity. They give a good indication of what an MT doesn’t know because they are usually the result of lack of experience and/or training. By the same token, a client demanding a “no blanks” policy also provides both a client teaching opportunity and a dilemma to anyone being
asked told not to leave blanks for any reason. In my opinion, even limiting the number of “acceptable” blanks in a document is unacceptable because it still encourages guessing. (More on this later.) So my first question to a client making this demand would be “why?” What experience in their past has led to this policy? Perhaps they have had transcription services (or transcriptionists) they feel left an inordinate number of blanks; and if they asked why the blanks were left, were they unconvinced it was dictator or audio problems, or were they otherwise unhappy with the answer? Every effort should be made to educate them to an understanding of why blanks are a valid tool and necessary for accurate documentation. The dilemma arises if, even after using the opportunity to educate a client as to why blanks are valid and the steps used to reduce or even eliminate resolvable blanks, the client still insists on this policy. Call it a bad business decision, but for myself, this is a demand I wouldn’t be able to agree to. If a prospective client can’t see why there would be valid reasons for blanks, and how dangerous it is to require the transcriptionist to “fill in the blanks” no matter what, then that’s not a client I want. Experience tells me this won’t be the only unreasonable demand they’ll make.
So what about MTSOs and facilities with a set number of “reasonable blanks” they allow?
This is another dangerous policy that encourages guessing, in my opinion. The game that will be played is a guessing game of two types: guessing that someone is less likely to look closely at a document with no blanks, and guessing so as not to have to leave them. Of course, if a document happens to get audited and the MT used guessing to avoid blanks, the result is the same (or possibly worse) in terms of low QA score and any resultant consequences to the MT. Resolvable blanks are a good indication of what an MT knows or is able to learn. If the employer is unable or unwilling to implement teaching and training initiatives to reduce resolvable blanks, then the resolvable blanks will continue – as will errors. And in that case, why is the MT working for that employer at all? Demanding fewer blanks doesn’t resolve the problem, it only aggravates it; and, at the same time, jeopardizes accuracy. Applying a policy like this to all transcriptionists, rather than addressing the few MTs who might truly be leaving more blanks than would be expected, is more simple, but it is also more damaging to everyone, including the client. The same is true if the employer does have initiatives in place and no improvement is seen. At some point, if the MT isn’t teachable, both parties ought to reach a point where they agree it isn’t going to work out, and terminate the relationship. Letting it continue while applying varying degrees of punishment isn’t productive, and it damages morale. How much additional teaching and training an employer is willing to provide depends, I suppose, on the availability of MTs who might be able to complete the work without it. In my experience, unless fresh out of school, most MTs are hired with the understanding that they have the knowledge and experience to do the job without a lot of errors or resolvable blanks, and most employers simply don’t have the resources to do additional training. Regardless, a policy limiting blanks isn’t the solution to inadequate education and/or experience.
There are two other things I think contribute to the problem. One is the number of MTSOs classifying MTs as independent contractors. I’m not even going to get into whether or not the classification is appropriate and in what instances. For the most part, however, independent contractors (ICs) are people who are supposed to know how to do the job. For that reason, additional training/education by the company hiring the IC would be inappropriate, especially on an ongoing basis. I would caution any MTSO classifying MTs as independent contractors about stepping over what is already a very thin (possibly nonexistent) line. Although the IC may actually participate in some kind of continuing education, it isn’t likely to be specific to the accounts, which isn’t usually terribly helpful in terms of improvement for that service. And please don’t get started about IC status in comments – it’s another discussion entirely!
The other contributing factor is the production-based pay prevalent in the U.S. medical transcription industry. Let’s say everyone is an employee. I’ve already written about wage theft, and there are plenty of MTs who can detail how many different ways and times they’ve been subjected to it. Any employer implementing inservice training and education initiatives is going to have to be willing to pay for their employees to participate. Again, cost and resources work against this, as does the MTs’ reluctance to attend even when paid, as the hourly compensation for attendance is usually less than what they can make on production. Production pay also works against MTs taking the feedback (edits, corrections, resolved blanks) they receive and actually doing something with it. The most productive learning is to go back to the report and listen where blanks have been filled in, but that’s uncompensated time and therefore not likely to happen. I’d like to hear from MTs about whether or not they actually look at the feedback they receive, much less go back to the report to listen while they read. I’m going to guess that not many do. And why should they? They aren’t being paid for that time. One thing I did in my service was real-time QA. I asked MTs to have an online messaging service, and QA staff used a desktop program that unified several of the messenger services. If someone hit a wall where they’d have to leave a blank, they would message QA with the account, job number and time stamp. QA would pull the file and listen and work with the MT to see if it could be resolved. It was real-time, the job wasn’t held up on a circuitous journey to QA, and the MT was able to get an answer while listening to the dictation. Even if QA wasn’t immediately available or if it took a few minutes to research an answer, the MT could continue with the rest of the report, or save the report to complete when they had an answer, so productive time wasn’t wasted. They could also, of course, ask other questions, but I felt the greatest value was the near-instant learning experience. It also saved us time because we didn’t have to provide retroactive feedback (which probably wouldn’t have been read), and the client benefit was more accurate documentation and faster TAT.
Ongoing education and training initiatives are essential for improvement, especially when tied to specific dictators/accounts. Any dictator/account generating more-than-usual blanks and errors should be targeted for specific focus. Any single MT leaving more-than-average blanks and errors should be targeted for re-education or termination. Simply applying a “reasonable blanks” rule doesn’t solve anything. Just to be clear, a third party (QA staff) should have final say when determining whether a blank is resolvable or unresolvable. It shouldn’t be left to the dictator, or to the MT. Also of note, the AHDI Best Practices document doesn’t assign error points to either type of blank, but it does discuss the practices of no blanks and limited blanks. My conclusion would be that the authors do not believe even resolvable blanks should be factored in the accuracy score. If that’s the case, I would recommend that any document with blanks that can’t be categorized as valid be sent to QA for a second look and possible resolution, and that a system be in place to track data to follow trends that may be actionable.
So what about unresolvable, or valid, blanks? This is where a “no blanks” policy really falls apart. For one thing, the policy completely ignores audio quality and dictator input. You can’t make a silk purse out of a sow’s ears. I’m going to go out on a limb and suggest that clients who demand no blanks policies don’t have a clear understanding of the problems that can arise. I’m a fan of data (caveat: quality data). Tracking valid blanks and providing clients with a report would go a long way toward educating them on the valid reasons for leaving blanks. Start with the most objective ones: audio file distortion; clipped or cut off dictation; background noise obscuring dictation; inaudible dictation; unknown person or place; blank left in forwarded text to be copied; author requested; template variables not dictated. From there, try and categorize the more subjective reasons: dictator unintelligible; inability to verify dictated term or terminology; suboptimal dictation practices.
Whatever the type of blank, my philosophy is that blanks are okay, better than guessing, but for heaven’s sake make sure everything else is correct!