D'Arcy Gue


ICD-10-CM and the Challenge of Data Analytics

March 25, 2014


ICD-10 2 Minute Read

The ICD-10-CM transition is coming shortly and many EHR vendors are now providing the technical infrastructure to ensure all Health IT systems will be able to accept ICD-10-CM diagnosis codes and be able to report the appropriate codes to the government and insurers for reimbursement.

The vast majority of the training for ICD-10-CM has focused primarily on the professional HIM coders and secondarily on clinicians.  The purpose of the education is to familiarize the essential users on the new coding system as well as introduce the new terms and rules necessary to select the proper ICD-10-CM code.  The government aided this transition by providing the General Equivalent Mappings (GEMs).  Some suppliers have developed additional mappings and client knowledge to facilitate the transition as well. This has resulted in vendors, payers, and providers being at various stages of readiness for the changeover. Understanding how to effectively use the codes, testing strategies, reporting, work flow changes, and finding qualified knowledgeable employees are all potential challenges for the industry.

One challenge that will be faced by all hospitals, clinics, and insurers is the hundreds of thousands of reports that use ICD-9-CM to run modern healthcare systems.  Mapping that has been developed using current ICD-9-CM coding may lead to inaccuracies when using ICD-10-CM.  Trending and comparisons with subsequent revenue estimates may become problematic.  Further, these may result in resources focused on reworking data and reports, processing delays, rejections and subsequent appeals.  Healthcare organizations will then be facing lower productivity in the billing / reporting areas and a potential increase in A/R days.  Worse yet, decisions may be made based on inaccurate or unsubstantiated data.

As an example, if an organization utilizes a simple analytical approach using GEMs, diagnosis code 282.64 (Sickle-cell diseases with crisis) maps forward and backward to D57.219 (Sickle-cell/HB-C disease with crisis, unspecified) See the figure below:

icd-10 cm mapping

However, D57.212 and D57.211 also map back to 282.64.  As we know, additional information is encoded in the ICD-10-CM codes.  Depending on the purpose of the report, if the organization desires to know the total number of sickle cell patients, one would map back to 282.64.  But if a manager requires a report for the number of Splenic sequestration patients to evaluate the need for surgical care, one would map back to 289.52.  This clearly demonstrates the mapping of 282.64 is convolutedWe have found approximately 36% of ICD-9-CM codes are convoluted, presenting a profound challenge for data analysts using analytics as a basis for executive decision-making.  A recent publication described the mathematics behind convolution related to this transition and analyzed all ICD-9-CM codes.  A free tool is provided for everyone to help visualize the convolution of the codes.

With October quickly approaching, resulting in limited time and available resources in preparation for ICD-10-CM, the concept of convolution provides another challenge in the preparation for the changeover.



Related Posts