Sticky Query Situations

By Selena Chavis for For the Record

Clinical documentation improvement (CDI) processes have made significant inroads in recent years. Now a mainstream strategy within any forward-looking hospital, CDI teams are critical players in the greater quality management picture as it relates to successfully positioning for value-based care.

Not all CDI workflows and processes are created equal, however. In fact, CDI specialists may unknowingly raise ethical concerns by sending queries that are poorly worded or constructed. And with clinical validation moving front and center, it “has never been more crucial” to follow industry guidelines, says Karen Newhouser, BSN, RN, CCM, CCDS, CCS, CDIP, director of education with MedPartners’ University division.

Clinical validation is essentially the process of reviewing a claim to determine whether reported codes depict an accurate clinical representation. According to “Guidelines for Achieving a Compliant Query Practice,” an Association of Clinical Documentation Improvement Specialists (ACDIS)/AHIMA query practice brief, “In hospitals, coders and clinical documentation specialists can’t lead health care providers with queries. Therefore, appropriate etiquette must be followed when querying providers for additional health record information.”

To remain compliant, Newhouser says coding and CDI professionals must prioritize the use of clinical indicators as the most critical part of the query process. Otherwise, queries that are not supported by clinical elements in the health record can become misconstrued or “leading” in nature. She emphasizes that ACDIS and AHIMA underscore this significance by mentioning the term “clinical indicators” more than 20 times in the query practice brief.

Melinda Kantsiper, MD, an assistant professor at Johns Hopkins University School of Medicine and the medical director for the Collaborative Inpatient Medicine Service at Howard County General Hospital, points out that many of today’s CDI workflows are burdensome and disjointed for both CDI professionals and physicians. The end result is that queries are too often retrospective, occurring after patients are discharged.

“You then have to really delve back into the chart and try to remember the patient to answer the query accurately,” she says, suggesting that the timing of a query can create issues with medical record review and exacerbate ethical concerns. “That’s why it’s so important to complete as many concurrent reviews as possible. You have a much better handle on [clinical indicators] while a patient is still in a hospital than a month and a half later.”

Fragmented query processes present HIM directors with steep challenges, especially when it comes to getting ahead of potential problems, says Marisa MacClary, CEO of Artifact Health, adding that a lack of visibility into the query process can create compliance risks.

“As part of the query process, you often have CDI specialists and coders sending and tracking their queries in different systems,” she explains, pointing to manual paper trails and Excel spreadsheets that often characterize query tracking. “It’s really difficult for a manager to come in and see exactly what’s going on.”

MacClary says fragmentation also makes it difficult to audit query processes. Managers often lack the time needed to manually pull all the needed information together. As a result, problematic areas such as queries sent with inadequate supporting clinical indicators are not easy to spot.

Common Query Concerns

While many pitfalls exist within query processes, gaining awareness can be a good first step toward improving their outcomes. The following are four potentially problematic areas that present unique ethical challenges.

Poor template construction. Newhouser says the biggest problem with template construction is multiple-choice formats that offer every possible option—even those that are irrelevant. Examples include the following:

• a pediatric sepsis option, which has no bearing in the case of a 79-year-old patient;
• “acute blood loss anemia” in a multiple-choice anemia query list when there is no evidence of any blood loss; and
• “severe malnutrition” in a multiple-choice diagnosis list when the weight and BMI support only mild malnutrition or underweight.

“The fact that they are listed isn’t necessarily the problem,” Newhouser says, adding that these scenarios become an issue when they cannot be individualized or streamlined to the patient and the clinical situation. “To quote ‘Guidelines for Achieving a Compliant Query Practice Brief’: ‘Multiple-choice query formats should include clinically significant and reasonable options as supported by clinical indicators.’ By including options that don’t meet the above criteria, provider trust is lost and the query is noncompliant.”

MacClary agrees, noting that a query template can provide a physician with a list of 10 possible answers, but the CDI specialist may include only clinical indicators that support a couple. “The better option is to allow CDI specialists to omit answers from the standard template that fall outside of clinical relevancy,” she says.

Query processes that allow a liberal amount of template customization can also be problematic. For instance, the goal of many query strategies is to keep wording standardized and compliant. “It’s a delicate balance between allowing editing of standard templates to best fit the question and ensuring compliant queries across a hospital or health system,” MacClary says.

Multiple-choice query list options that affect payment or severity that don’t include other reasonable diagnoses. Consider a scenario in which an expectant mother is admitted to delivery. Her BMI is clinically in the obese range, but she also has a full-grown fetus, placenta, and fluid on board. In this case, it would seem illogical to exclude “overweight” or “normal weight gain of pregnancy” as alternate options for an obesity query.

On the surface, Newhouser acknowledges that this practice could raise an eyebrow. However, she points out that the industry has entered an era where hospitals must focus on risk adjustment. “There are definitely diagnoses that do not overtly affect the payment or severity but do risk adjust,” she says. “However, in the end, if a diagnosis is relevant to the patient, it is the right thing to do, regardless of whether it impacts reimbursement or severity.”

According to practice briefs, multiple-choice queries should include all clinically reasonable and relevant options. However, Newhouser notes that, in today’s environment, most relevant options affect reimbursement or severity at some level. “So, the question is: Is it ethical to follow advice to include options that do not affect payment or severity?” she says.

Looking at prior encounters before querying a physician to bring diagnoses forward. As risk adjustment models such as the Centers for Medicare & Medicaid Services’ Hierarchical Condition Categories (HCC) become mainstream, Newhouser says the current coding and query advice regarding the use of codes from prior encounters must be clearly delineated for CDI professionals performing concurrent record reviews based on the fact that HCC coding looks at all diagnoses from every appropriate encounter for an entire calendar year.

She explains that some industry professionals may understandably be wary of this advice because of the interpretation of Coding Clinic, third quarter 2013, pages 27 and 28, which states that diagnoses cannot be retrieved from previous encounters. This aligns with the coding guidelines that recommend assigning codes based on documentation from the current encounter.

While the role of concurrent CDI professionals is to review clinical indicators for validation and possible clarification purposes, and not to perform code assignment for placement on a claim, “Guidelines for Achieving a Compliant Query Practice” offers the following advice: “Clinical indicators should be derived from the specific medical record under review and the unique episode of care.”

“This appears to be in discordance with medicine best practice and continuum of care, which encourages providers to look in previous encounters for details that may assist with patient management during the current encounter, such as the ejection fraction from a previous echocardiogram or past labs for comparative purposes,” Newhouser asserts. “For these reasons, along with patient safety, medical necessity, and clinical validation, it might be prudent to endorse the CDI professional in retrieving additional relevant clinical indicators from prior encounters to support a query along with clinical indicators from the current encounter to remain compliant.”

While Newhouser does not advocate including only prior relevant clinical indicators to support a query, she explains that prior indicators can augment relevant clinical indicators in the current encounter.

Queries heavily weighted to the same complications or comorbidities/major complications or comorbidities. If 90% of queries cover only five diagnoses, such as blood loss anemia, sepsis, obesity, malnutrition, and abnormal lab values, a health care organization should probably consider taking a deeper look into its CDI practices. Newhouser, who points out that the face value of these kinds of query statistics can be misleading, recommends that providers always look at the “data behind the data.”

“We need to look at the patient case mix,” she says. “Were these queries from a cardiac monitoring unit, an oncology unit, a respiratory ICU, or perhaps a cardiac surgery ICU, where the patient populations are similar? One should understand the culture of the facility as well as the policies and procedures. Have any parameters been set by leadership to focus on certain areas? Perhaps these were identified areas on a PEPPER [Program for Evaluating Payment Patterns Electronic Report] or a denials report that needed focus.”

Newhouser adds that assumptions are often made without asking pertinent questions. “If the inclusions and exclusions are not identified and all questions have not been addressed, the data can’t truly be trusted until all information is available to make an educated decision,” she says.

Improving the Outlook

MacClary says health care organizations must bring more structure and visibility into the query process to improve disconnected workflows. In addition, Newhouser suggests that CDI and coding professionals need resources to accurately identify and validate clinical indicators for each query.

To improve CDI processes, Artifact Health, a collaborative physician query tool that standardizes processes, can be implemented to help negate problematic query risks by ensuring every CDI specialist is pulling down the same template and initiating queries in a uniform, compliant manner.

MacClary points out that the system provides standard templates that allow some customization to construct compliant queries. For example, CDI specialists and coders can omit certain multiple-choice answers if they fall outside of clinical relevancy. It also ensures appropriate clinical indicators accompany a query, a critical component of improving the query process.

The query goes directly onto physicians’ smartphones, and they have attached data from the chart so they can see exactly what the objective criteria are, making it easier to accurately choose the right answer for that patient.

The system can improve concurrent review percentages and reduce the time coders have to spend generating queries and tracking down providers.

MacClary points out that the solution provides greater visibility into the query process, creating reports that allow managers to drill down and see detail of every event—who created the query, when it was sent, when a physician viewed it, and how long it took to get a response.

With this kind of information available, physicians can focus on higher-level initiatives, such as how to approach querying for quality issues.