DRG Mismatches

By Cheryl Ericson, MS, RN, CCDS, CDIP for For the Record

When discrepancies occur, the ensuing reconciliation process serves as a prime opportunity to educate CDI and coding staffs.

Reconciliation rates monitor both clinical documentation improvement (CDI) and coding proficiency, a key performance indicator. However, many in leadership positions overlook the importance of trending reconciliation rates and fail to develop a collaborative process to resolve discrepancies when diagnosis-related group (DRG) mismatches occur.

As the skills necessary to be a successful CDI or coding professional grow increasingly complex, DRG mismatches (which occur when the final CDI working DRG does not match the final coding DRG) and the ensuing reconciliation process reveal consequential educational and training opportunities for both professions.

First, it’s important to recognize that not all DRG mismatches require reconciliation—the process of determining which, if either, is the correct DRG—and should be billed. What constitutes a DRG mismatch and whether it requires reconciliation will vary depending on CDI’s role within the organization and its workflow.

Focus Shapes Approach

Historically, CDI has concentrated on revenue integrity, where efforts focused on identifying clinical indicators of an undocumented or ambiguously documented diagnosis classified within Medicare severity DRG (MS-DRG) reimbursement methodology as a complication or comorbidity (CC) or a major complication or comorbidity (MCC). When these situations arose, CDI professionals would query the provider in an effort to obtain clarification so a CC or MCC could be coded and reported on the claim.

Some organizations remain focused on revenue integrity while others have shifted to an approach that emphasizes the “quality” of the documentation regardless of its impact on MS-DRG assignment. A focus on quality often requires comparing DRGs at the diagnosis level as well as clinically validating reportable conditions that may affect MS-DRG and all patient refined DRG (APR-DRG) assignment, risk adjustment, and quality measure cohorts.

The mission of CDI efforts determines the level of detail expected when CDI staff develop their working DRG. Is comparison made at the MS-DRG level, the APR-DRG level, or the diagnosis level?

A diagnosis-level approach expects CDI and coding to have the same diagnoses as CC(s) and MCC(s), affecting the severity of illness (SOI), risk of mortality (ROM), and quality measures through risk adjustment (Hierarchical Condition Categories [HCCs]) and establishment of the cohort.

More organizations are understanding the value of a diagnosis-level approach as they transition to processes that align with value-based reimbursement payment mechanisms. This strategy requires CDI staff to identify opportunities to capture diagnoses classified as HCCs because of their importance in risk adjustment and those that may exclude a case from a quality measure cohort, in addition to those that influence MS-DRG and/or APR-DRG assignment. While diagnoses are impactful within all three systems, approximately 40% of diagnoses classified as HCCs within risk adjustment/payment models are not CCs or MCCs.

Incorporating a risk-adjustment approach into CDI efforts is necessary to mitigate potential adverse outcomes such as readmissions, complications, and death that could be equated to poor performance on quality measures. If performance on quality measures is within the CDI scope, then accurately identifying opportunities to capture HCCs and validate quality measure cohorts should be monitored through the reconciliation process.

In Search of Validation

Validating quality measure cohorts is not limited to identifying exclusion criteria—it starts with accurately reporting the principal diagnosis, which is important from both a revenue and a quality perspective. Accurate identification of the principal diagnosis has always been important to support appropriate payment and avoid potential denials, but it also establishes cohorts for clinical quality measures.

The reconciliation process should focus not only on which principal diagnosis maps to the highest paying DRG but also on the accuracy of the clinical scenario when multiple diagnoses appear to meet the definition of a principal diagnosis. Often the DRG with the highest payment also puts the claim within a quality cohort measure. Typically, but not always, a difference in principal diagnosis will result in a different MS-DRG or APR-DRG. Consequently, organizations must determine if they are concerned about all principal diagnosis mismatches or only those that result in a different MS-DRG or APR-DRG.

Principal diagnosis mismatches often reveal a lack of adherence to coding guidance or a clinical validation issue.

Diagnosis-Level Mismatches

For secondary diagnoses, organizations must determine if mismatches are limited to the presence or absence of any CC and/or MCC (a MS-DRG-level mismatch) or if CDI and coding must have the same CCs and/or MCCs (a diagnosis-level mismatch).

Because it’s possible for CDI and coding to arrive at the same MS-DRG with a different CC or MCC, identifying diagnosis-level mismatches is more tedious than it is at the MS-DRG level. When the health record contains multiple CCs or MCCs, a diagnosis approach validates that CDI and coding have the same diagnoses that can be classified as CCs and/or MCCs. For example, if CDI has a CC or an MCC that was not reported by the coder, it would be a mismatch.

The same concept applies to organizations using an APR-DRG-level approach. In this scenario, however, instead of trying to match CCs and MCCs, a mismatch occurs when CDI and coding have a different SOI or ROM. Unlike MS-DRG assignment, where a variety of diagnosis combinations can result in the same MS-DRG, each diagnosis has its own intrinsic SOI and ROM value. A true “quality” approach would also include mismatch cases where the billed claim features HCCs or cohort exclusion criteria that were not included in the final CDI working DRG or vice versa.

Because a typical diagnosis-level review is more time-consuming compared with an MS-DRG or APR-DRG approach, organizations may have to increase the size of their CDI staffs. Although a diagnosis-level approach requires CDI staff to identify and clarify documentation associated with a majority of diagnoses within the health record, their role should not be confused with those of coding or be classified as concurrent coding.

CDI staff are primarily concerned with identifying ambiguous documentation that affects the ability to assign a code for a subset of diagnoses (CCs, MCCs, HCCs, exclusion criteria, etc) and ensuring documentation supports documented conditions (clinical validation). This provides coders with access to a “clean” health record.

Ideally, CDI and coding work on a shared platform that includes computer-assisted coding, which can identify the impact of any diagnosis and features an integrated grouper that retains diagnosis and procedure codes selected during the CDI review process. This allows coders to validate and build on CDI’s working DRG rather than starting from scratch.

As CDI staff become proficient in identifying and clarifying the documentation associated with impactful diagnoses, the coding burden is reduced, coder productivity increases, and bill hold times decrease.

To Reconcile or Not

Determining how DRG mismatches are defined is only the first step of the process; the next step is determining whether the mismatch requires reconciliation. If the goal of reconciliation is to identify educational and training opportunities—which may occur at the individual or departmental level—the reconciliation process should target cases that reveal deficiencies.

CDI workflow must be considered when determining whether a case requires reconciliation because it may limit the staff’s ability to match the coding DRG. For example, do CDI staff assign ICD-10-PCS codes, or are they expected to only “ballpark” them? If procedure code assignment isn’t a priority for CDI staff, then DRG mismatches based solely on different PCS codes should be excluded from the DRG reconciliation workflow.

Conceivably a more important workflow consideration is determining when an impactful diagnosis is established in the health record. CDI workflows where reviews do not extend through discharge are likely to have a higher volume of DRG mismatches, but it is unlikely these cases require reconciliation. Documentation within the discharge summary can completely change the clinical picture and the associated DRG as diagnoses are added, clarified, or ruled out. These types of amendments often affect secondary diagnosis assignments. Additionally, yet-to-be-ruled-out diagnoses (eg, “likely,” “probable”) documented at the time of discharge are reportable in the inpatient setting, where they most often affect principal diagnosis assignment.

Perhaps the best way to decide which cases require reconciliation is to determine the cause of the discrepancy and how future mismatches can be prevented. If the cause is attributable to CDI workflow, a process change is necessary, and reconciliation is unnecessary. Although many organizations currently use “additional information” as a reconciliation category, it yields limited actionable data. Organizations are encouraged to create meaningful, actionable, and operationally defined reconciliation categories. For example, every organization’s plan should include categories covering coding errors and clinical validation issues.

Traditionally, most mismatches are ascribed to something that reflects a “coding error,” implying established coding guidance—which can be either the Official Coding Guidelines or the AHA Coding Clinic—was not followed. Sometimes these are referred to as “learning opportunities” when development of the working DRG by CDI did not follow coding standards.

When CDI was a young profession, many who filled these roles lacked expertise in applying coding rules, leading to a fair share of errors. However, as the CDI profession has matured, so has its understanding and proficiency in applying coding guidance. Consequently, the volume of coding errors as the cause of DRG mismatches is on the decline and not limited to CDI professionals.

Tracking coding errors helps identify trends that can differentiate between the need for departmentwide training and performance issues that may require remediation. Reconciling these types of cases with CDI staff and coders promotes learning. For example, an obscure Coding Clinic guideline can be made clear.

Reconciling diagnosis disagreements not grounded in coding guidance has become increasingly challenging, especially when considering the impact of clinical validation. CDI professionals and coders can disagree not only on documentation requirements but also on whether the diagnosis is clinically validated.

The focus of clinical validation frequently centers on secondary diagnoses classified as CCs or MCCs. However, ensuring the clinical accuracy of the principal diagnosis is just as, if not more, important because it impacts more than the DRG assignment under the Inpatient Prospective Payment System. Unfortunately, many organizations are struggling with how best to reconcile differences between the clinical perspective—which is often asserted by a CDI professional with a clinical background—and the coding perspective. It may be helpful to employ physician advisors as “tiebreakers” for contentious cases.

Trust the Process
It’s also recommended to have a process in place that allows CDI staff to communicate clinically validated diagnoses to coders in an effort to make them feel comfortable reporting the condition. This is particularly important when a clinically valid, reportable diagnosis is documented only at the beginning of an admission and not carried through to the discharge summary.

Any process should include guidelines for ruling out diagnoses that cannot be clinically validated but which are nevertheless reportable. It’s also beneficial to track “clinical accuracy” reconciliation cases, which highlight differences in complex medical conditions such as sepsis. This creates educational opportunities, promotes collaboration with providers, and can lead to the development of policies and procedures that support standardized practices among CDI and coding professionals.

To maximize the benefit of the reconciliation process, the determination of the “correct” DRG, and the cause of the discrepancy must be a collaborative effort among the CDI and coding staffs that includes feedback from the physician advisor and the compliance and denial staffs. If staff work remotely or at different locations, collaborative reconciliation can be accomplished through biweekly or monthly video calls in which relevant screens (eg, the health record, CDI’s working DRG, coding’s DRG, queries) can be viewed during the discussion.

To decrease the chances of unproductive discussions, reconciliation sessions should be structured with a time limit of two to three hours as well as a per-case time limit. If a consensus cannot be reached, a query may be necessary before reconciliation can be finalized. However, all those involved in the case should agree with and understand the next steps.

Depending on the volume of mismatches, it may not be feasible to review all of them, so organizations may want to prioritize the most challenging cases or those associated with a common theme. A list of cases should be sent to all participants at least a day prior to the meeting to allow those who worked on the record time to review the case and prepare for discussion.

A collaborative, interactive approach to reconciliation is a great informal educational opportunity that can be followed by formal education or training. A successful reconciliation process will result in fewer mismatches over time and promote true collaboration among the CDI and coding staffs.