January 23, 2012

Accessing the Impact of ICD-10 with Data Analytics

By Paul Van Fossan

By now we all should be aware that the transition to ICD-10-CM/PCS will be a significant event that will impact not just technology, but processes across entire organizations for providers and payers alike. As we prepare for this event it will be important for organizations to understand the projected impact of ICD-10 on their financial and operational performance. Specific areas of change, and therefore, potential risk, will need to be assessed. These may include, but are not limited to:

 

  • Reimbursement - For inpatient claims reimbursed under a prospective payment system (PPS), codes determine reimbursement. To what extent will these claims be at risk of grouping to different DRGs under ICD-10?
  • Documentation and coding performance - How should providers focus their efforts to develop new processes and training programs targeting coding and documentation in preparation for ICD-10? Which codes are of the greatest concern?
  • Contracting - Which contracts are at risk of change and may need to be modified to accommodate new ICD-10 codes in order to retain proper context?
  • Operations - How will days in AR be influenced due to claims that may be pended by payers with coding issues?
  • Policies and procedures - To what extent will policies and procedures currently defined by ICD-9 codes be in need of modification or reconsideration?

To understand the answers to these questions, organizations will need to look at claims data, analyze risks and develop priorities for remediation. This can be a daunting task, as there is no simple way to translate from ICD-9-CM to ICD-10-CM/PCS.

Remember, this transition is not simply an expansion or refinement of ICD-9-CM, it is a wholesale replacement of one coding system with another. Differences are especially noteworthy for procedure codes, where we are going from about 3,800 ICD-9 codes to about 72,000 ICD-10 codes.  The GEMs (General Equivalency Maps) developed by CMS clearly demonstrate the complexity involved with moving between coding systems, given the types of relationships that have been defined - which include the following:

  • Individual ICD-9 codes that map to several ICD-10 code alternatives;
  • Individual ICD-9 codes that map to a set of two of more ICD-10 codes;
  • Two or more ICD-9 codes that map to individual ICD-10 codes;
  • ICD-9 codes with no representation in ICD-10;
  • ICD-9 codes with an exact match in ICD-10; and
  • Individual ICD-9 codes that map to codes with similar but not identical meanings in ICD-10

Given the aforementioned translation complexity, how should organizations go about assessing risk? The following approach provides the best way forward:

  • Translation algorithms should be developed and applied to claims data to create a test bed of claims that can be used for modeling and assessment. The test bed should include translated ICD-10 claims as well as the original ICD-9 coded claims.

 


 

  • Business rules (representing reimbursement rules, policy decisions, contracting rules, etc.) should be applied to both ICD-9- and ICD-10-coded claims to identify changes between systems. For example, if we're assessing reimbursement risk for DRG-reimbursed claims, we would want to group and price claims in both populations to identify areas of potential change.
  • The impact or change of specific rules between systems should be measured in terms of frequency (and, when applicable, financial impact) to help prioritize where to focus remediation efforts.
  • Codes associated with each measured change should be assessed to determine their applicability and status in the GEMs.
  • Remediation efforts should be prioritized, centering on areas of real risk.

Assessing anticipated impact to your organization can seem like a daunting task, considering the complexity of this change and the sheer number of processes it will touch. Developing a sound assessment strategy that includes the application of analytics can help you identify the most pressing areas of risk to ensure that you have a plan for addressing the issues that matter most to your organization.

About the Author

Paul Van Fossan is a director in the Financial Risk, Optimization and Growth section of OptumInsight, where he is responsible for managing a team that focuses on Prospective Payment System (PPS) analytics, facility reimbursement and benchmarking solutions.

Paul also recently has been involved with developing analytics solutions to determine payer and provider impacts the transition to ICD-10 will have on reimbursement as of the Oct. 1, 2013 implementation date and beyond. Paul holds a bachelor's degree in aeronautical engineering from the University of Minnesota.

Contact the Author

To comment on this article please go to