QA Quandaries: Quality Assurance — Substance and Style
By Criss Spoto
For The Record
Vol. 30 No. 2 P. 8
Top-quality health care documentation means absolute accuracy in the content and context of each and every document representing patient encounter information. It is incumbent upon all document gatekeepers to have quality assurance (QA) measures, methods, and programs established to keep quality in the forefront and to detect and address any cracks in that foundation.
QA programs can differ from organization to organization. Also, because there are always gray areas that elicit opinions and judgment calls, individual QA specialists may have different definitions of what constitutes an error. How do these factors spill over into style preferences?
The Book of Style for Medical Transcription, Third Edition, published by the Association for Healthcare Documentation Integrity (AHDI), serves as the industry standard for best practices in health care documentation and a guide for how to best express medical details in a way the reader clearly understands.
Nevertheless, facilities, practices, and other organizations sometimes branch off into their own set of style rules. It's important to institute succinct and sensible style expectations for consistency; however, it would be a mistake to presume style rules apply only to transcriptionists/editors. Guideline effectiveness and reliability wane considerably unless everyone involved in the process of document creation—from clinicians and scribes to editors and QA specialists—is familiar with and abides by these expectations.
The more health care documentation technology evolves and the closer the industry gets to verbatim transcription, the more effective that technology becomes, thereby decreasing the effort it takes to edit documents and narrowing the margin for error. Speech recognition technology boasts an ability to complete documents in half the time it takes to traditionally transcribe a document in full (encompassing everything from production-based billing to payroll) since it is most commonly based on document line counts and driven by turnaround time.
To take full advantage of speech recognition, style habits and style demands must be reconsidered. It no longer makes sense to recast a grammatically awkward sentence to one that harmonizes with formal writing rules and styles. The technology was not designed to reasonably accommodate such changes.
Did speech recognition make the mistake of assuming we all speak in grammatically perfect sentences? No, the technology predicates that the primary focus is on medical content accuracy and prompt turnaround, not perfect grammar.
This raises the following questions: How does this change industry style requirements? Which style infractions are to be considered weighted errors?
One Style for All
How are style rules determined? Besides the industry standards and guidelines provided in The Book of Style, AHDI's online Healthcare Documentation Best Practices toolkit can help diminish the gap between dictation and the resulting transcript. In all cases, style should not impede medical congruence, with style requirements conveyed to all involved in the documentation process.
Unfortunately, style requirements are often imposed on only those charged with finalizing the transcript. This presents a disconnect between the dictating clinician and the health care documentation specialist (HDS). As a result, the dictation may be subject to translation and alteration in order to fit the style requirements. For example, a document is more likely to read "The patient was seen in the emergency room" rather than "The patient was seen in the ER" when the dictating clinician is aware of the requirement for "ER" to be expanded to "emergency room."
If a facility requires the medications section of reports to be formatted in a numbered list down the left margin but the physician dictates in discussion/paragraph form, the HDS must edit the verbiage in a manner that accommodates the directive.
The situation can become further complicated when one HDS interprets this requirement differently from another. On top of that, a proofreader or a QA associate may come along with yet another idea of how the dictation should have been translated and assign weighted errors against the quality score of the document. QA staff should be coached to accept various ways of translation (so long as it doesn't interfere with medical meaning or intention), but this conflict could be avoided if all involved with patient care documentation are aware of and held to the same style requirements.
Making the Grade
Quality scores are a reflection of an individual's skill level. Their primary purpose is to identify an individual's suitability for a type of work and determine whether there's a need for further training and/or coaching.
Is it fair to down-score a document when the transcript follows exactly what was dictated by the clinician and contains no medical errors? If the transcript shows "P.R.N." as dictated, but the facility style rule is to expand P.R.N. to "as needed" is this a weighted error on the HDS's part? Did this blemish or jeopardize patient care, patient safety, or outcomes? Is this the best, most responsible and reliable yardstick to measure the HDS's skill level or prowess to work on this document type? Would point deduction for unintentionally overlooking a style preference be considered petty and punitive?
If QA is looking for and taking time to address and penalize style missteps, there is a greater risk of missing the true medical errors that must be remedied in a document.
Beyond the QA Specialist's Bounds
Speech recognition is sometimes blamed for errors and inefficiency, but the technology's processing relies heavily on how well humans understand it. Many HIM professionals have heard about the "speech wrecks" and "bloopers" generated by speech processing. Case in point: The SR Errors—Funny or Fatal? Facebook page is filled with such gaffes.
QA specialists must assume and expect that any number of defects will end up in speech-recognized drafts; it's their mission to find and correct them. When "bipolar disorder" is dictated but speech recognition captures "pineapple disorder," or when "5-year-old first grader" is dictated and the text indicates a "55-year-old first grader," QA experts must catch and correct such errors.
Speech recognition matches sound to text whether or not it makes sense and whether or not it is what was intended by the dictator. It is QA's mission to anticipate and resolve these crucial issues without judging and correcting the way a dictating clinician chooses to speak. While the dictated phrase "patient claimed to be feeling good" undoubtedly can be improved upon, changing the clinician's wording to such a degree is unwarranted and unnecessary. Further, it adds to the inaccurate coupling of audio to text for ongoing speech recognition processing.
Bringing It All Together
With the evolution of health care documentation technology comes the need to adjust the approach to and perception of quality vs style. Style and format requirements have their rightful place in this equation but it's vital that every individual contributing to the medical report be versed in these expectations for them to be meaningful and reliable.
Establishing stable methods by which to protect and preserve patient safety and document integrity through quality control measures is compulsory for success. If HDSs stay focused on the true objective of their role within patient medical care and well-being, it will be easier to recognize the difference between chargeable offenses (errors that potentially affect patient care and medical decision making) and inadvertent sidestepping of obscure style preferences that stray from the text dictated by clinicians.
The end result? True improvement to the quality of health care documentation.
— Criss Spoto has worked in a variety of roles in the health care documentation industry, including proofreading, team leadership, training, implementation, and most recently director of quality assurance and HIPAA privacy officer.