November 7, 2011

Writing Effective Physician Queries
By Julie Knudson
For The Record
Vol. 23 No. 20 P. 14

Consistency and attention to detail are key when crafting these odes in support of better documentation.

Complete and accurate documentation within each patient’s medical record is one facet of the mission to ensure patient safety and quality of care. Querying is a vital part of that documentation process, but it’s also a skill that can be difficult to master.

“It’s about the phraseology and getting it right, and making sure that their documentation truly reflects that patient in the bed—no more, no less,” says Gail Marini, RN, MM, CCS, manager of clinical documentation at South Shore Hospital in Weymouth, Mass.

As the saying goes, it’s just that easy and it’s just that hard.

Jean Elsener, RN, CPUR, CRN-C, CRN-A, CCDS, clinical documentation and core measures performance improvement nurse at Hudson Valley Hospital Center in Cortlandt Manor, N.Y., says a good query highlights information already included in the record that pertains to the question being asked. Also, it presents the information and question in a way that gives the physician what she calls “an ‘aha’ moment.”

Concise questions offer specific clinical indicators and are constructed to add completeness to the medical record, Elsener says. Examples of bad queries include those that use “statements and inferences not documented in the record to lead the physician to the desired answer,” she says.

Vague questions are also a problem. “Without clarity, the physician has no idea what information is being requested,” Elsener notes, adding that other no-nos include any query that introduces a new diagnosis or questions the physician’s judgment.

Poor content isn’t the only roadblock to submitting effective queries. “Just like we would look for the legibility of the physician documentation in the record, we need to make sure that we’re writing clear, legible queries,” says Abby Steelhammer, MBA, MHA, RN, a clinical documentation improvement manager at Novant Health in Charlotte, N.C.

In addition to legibility, Steelhammer says comprehension is also paramount. “If the query development is fragmented, the wording doesn’t flow, we’re not clear with the question that we are posing, then sometimes that can be ineffective because the physician doesn’t understand what kind of clarification we’re needing,” she says.

How do you know a query was ineffective? “A lot of times, we’ll get a physician response, but it’s not the response we were looking for or that we felt like we were asking,” Steelhammer says. While such a response often points to an ineffective query, she believes it may still be a good sign because “we want to make sure that the physician is accessing their clinical judgment and we’re being nonleading.”

If her team receives an answer that’s unanticipated or incomplete, Steelhammer says additional follow-up is conducted and the query is examined more closely to learn what could be done differently the next time. “It helps us to look at the big picture and then make adjustments based on that,” she says.

A Recipe for Success
Besides knowing what makes a query ineffective, it’s important to be aware of the ingredients of a well-constructed documentation request, one that will provide coders and clinical documentation improvement (CDI) specialists with the information needed to create a complete record.

According to Richard D. Pinson, MD, FACP, CCS, principal of HCQ Consulting and coauthor of the CDI Pocket Guide, effective queries and documentation requests contain three universal elements. “First, it ought to mention what condition or diagnosis established already in the medical record is the subject of this question,” he says.

The second element is “a statement of additional data that might be lab reports, diagnostic test reports, consultant notes, even nursing notes or nutrition notes—things that are contained in the medical record,” Pinson says. It’s imperative that these additional data are pertinent to the question being asked, he adds.

The third component is the question itself, which Pinson explains should center on eliciting better clarification or more specificity of the information already contained in the record. It can be posed in a multiple choice format if appropriate.

Pinson adds that there should be a statement to the effect that the physician is using his or her medical judgment and that the question is not being asked with a specific answer in mind. “This is sort of a regulatory component that reminds the physician he’s not required or expected to give any particular answer,” he says.

Providing the right information—and enough of it—is critical to receiving a sound physician response, says Robert S. Gold, MD, CEO of Atlanta-based consulting firm DCBA Inc. The mix of information in a good query should be designed to help physicians provide the clarification or specificity that’s missing from the record in the limited time they have available. “This way you provide the physician with enough information that he doesn’t have to go running through a chart, doesn’t have to go running back to the computer, have to go to the medical records and look the thing up. It’s just a real inconvenience,” Gold says. “You provide him with all the information that’s needed.”

Forms and Formats
According to the “2010 Physician Query Benchmarking Report” released by the Association of Clinical Documentation Improvement Specialists, more than 80% of respondents reported using templates. Elsener notes that standardized forms can offer advantages and drawbacks depending on the situation. On the plus side, she says, “It would restrict the possible error of asking a leading query. It also saves time to review more records when standardized forms are used.”

However, Elsener says preprinted queries aren’t sufficient in every case. “Forms do restrict the questions that can be asked, especially if there is no area on the form to document clinical indicators that would help support the question,” she says, adding that CDI specialists and coders shouldn’t be limited to using only form queries.

“When documentation in a medical record needs to be specific and complete for quality, the query should be unique for each case,” says Elsener, who believes the use of clinical documentation query tags offers better flexibility and provides physicians with queries that are easy to read and interpret.

The team at Novant Health uses a mix of templates and free-form queries. “I have folks on my team that really prefer to write their queries themselves, and they feel like they get more physician attention that way because they can be clear and concise,” Steelhammer says. “Then I have some that really prefer to use the templates. They get to know their physicians and they’re having conversations with [them], so they feel like when they use that template, the physician easily recognizes what it is we need.”

Steelhammer says templates often enhance the standardization of the information her team seeks to clarify and promote consistency in physician responses. When a free-form query is needed, she says her team uses an established set of guidelines and criteria to craft the request.

Pinson says while standardized queries offer consistency in application and can be made “unequivocally compliant,” there will always be instances in which a free-form query is necessary. “The challenge is when the question doesn’t fall within the standardized areas,” he says. “Then somebody has got to write their own free-form query.”

The same elements and rules that apply to template requests should also be used for free-form queries. Pinson emphasizes the need to closely follow AHIMA guidelines. “AHIMA says that in order to use a multiple choice format, you have to include all clinically reasonable choices. That gets hard for a coder who may not have very much clinical experience to know what all the clinically reasonable choices might be, so there’s another impediment to doing this effectively,” he says

Pinson often finds it helpful to establish elements one and two and then to pose a question that seeks clarification. He offers examples such as whether there’s a more specific diagnosis related to the findings, what underlying condition may be the cause of symptoms, or if, based on the data, there’s a condition that’s clinically significant that hasn’t been documented.

Gold is generally unimpressed with many of the templates on the market because they contain generic, widely available information that physicians are unlikely to read. Instead, he suggests a good free-flowing query may be a more effective solution in many cases. Three or four sentences that are direct and to the point will often provide all the clinical information that’s important to the case, but Gold cautions, “Yes, you do need practice and, yes, you do need training in order to be able to determine what the clinical aspects of particular diseases are in order to create an effective free-text query.”

Road to Improvement
Improving query effectiveness can take several forms, and the best approach often depends on the individual and the situation. “I would certainly say it’s a combination based on identified needs at the time,” says Steelhammer, whose team employs quality monitoring to audit internal processes in combination with input from consultants.

“When we identify that there may be some misunderstanding and/or some gaps in how effective they are, then we’ll design some training,” she notes, adding that the different types of queries are always being monitored. “We can identify which ones we typically might run into some confusion on.”

“Training and education is the only way to teach the correct way to present a query, both via classes and practical demonstrations,” says Elsener, adding that coders and CDI specialists have various resources available to help advance their skills. “The big focus should be the reviewing of coding guidelines,” she notes, while also mentioning the importance of an in-depth familiarity with AHIMA guidelines and “learning the specificity needed for quality care and to meet quality care measures.”

Pinson, too, highlights the importance of studying the AHIMA guidelines, saying they provide invaluable direction on compliance issues as well as other information. “They give direction on how and when to [write a query]. They sort of help conceptualize the reasons for asking a query, and they give some good examples,” he says.

Pinson uses the term “education” frequently when discussing the tools needed to craft more effective queries. In addition to basic knowledge from the AHIMA, regular reviews of highly compliant and effective standardized queries can provide a better understanding of how good queries are structured. With these resources available, he says, “I don’t think you need to attend courses or classes or any of that stuff to learn these things.”

The amount of clinical training coders have received may not be a primary factor in the effectiveness of their queries, Gold says, who dismisses the notion that coders can’t be CDI specialists because they have no clinical training. “A coder who knows the clinical aspects of diseases and procedures can be as adequate in a concurrent or retrospective process as any nurse,” he says, a concept which he acknowledges some nurses and consultants won’t accept.

When it comes to education, Gold says internal programs can be effective, but not many hospitals breed such environments. “I’ve seen it done at individual hospitals where the chiefs of service or somebody who is interested in the hospital’s data will get together with the coders and go over certain diseases, the primary diseases that this particular service line deals with,” he says. “The coders will grill the doctor on the multiplicity of the kinds of questions that might be asked about those diseases, and the doctor will explain where this all comes from.”

Journal articles and webinars are additional tools for coders to boost their grasp of queries.

“I’m a big education person,” Marini says. “I never want the program perceived as if it didn’t come from a good point of resource.” At South Shore, queries that go unanswered or garner incomplete responses are evaluated to see how the process can be improved. Educational material, including articles and relevant guidelines, are regularly distributed, and coders may stop and chat with physicians.

“I think resources are key,” Marini says. “You should be confident enough to ask the question and smart enough to resource the answers.”

— Julie Knudson is a freelance business writer based in Seattle.