Translating ICD-9 Final Diagnosis Codes to ICD-10 for Bio-Surveillance Systems

BACKGROUND:  

In 2009, the Department of Health and Human Services mandated that all HIPAA covered entities transition from the International Classification of Diseases version 9(ICD-9-CM) to version 10(ICD-10-CM/PCS) on October 1, 2014.   Impact of this transition upon surveillance systems such as BioSense will be significant.  In order to receive, analyze, interpret, and report upon ICD-9 and ICD-10 encoded data across the transition period BioSense must modify the existing database structure, data extraction rules, messaging guides and develop Master Mapping Reference Tables (MMRT) to bridge the gap across the two encoding systems.  The resulting ICD-10CM codes can then be incorporated into new syndromic surveillance visit (case) definitions that will be incorporated into the overall BioSense system. The most challenging aspect of the ICD-9/10 transition for Public Health surveillance systems will be developing flexible and standardized solutions to accommodate analysis across two different code sets after transition and to create a meaningful baseline that can be used for benchmarking after the transition. Another challenge will be testing and evaluation of the solutions for these issues without real ICD-10CM encoded data prior to the transition.

Objectives: Describe the process of developing MMRT for ICD-9/10 transition and synthesize an ICD-10 based data set to assess the impact of transition on BioSense.

METHODS:   We compiled a list of ICD-9CM codes binned to select syndromes from the list of syndromes vetted by the BioSense community and updated each to 2014 ICD-9CM codes using an online tool. Then we mapped the resulting ICD-9CM codes to ICD-10CM codes.  Subsequently, we followed a reverse translation validation process, to ensure that the appropriate codes were correctly identified at the onset; the resulting MMRT relates syndromic classifications to each code groupings. A synthetic test dataset was developed by extracting Emergency Department visits received by BioSense from Tarrant County, Texas from June to November 2013 using MMRT.  

RESULTS:   Due to the greater complexity and higher level of specificity in ICD-10CM codes; ICD-9CM codes pertinent to syndromes within BioSense cannot be automatically translated without expert review and input. Existing translation tools are insufficient and frequently provide results that are either incorrect or incompatible with the syndromic surveillance concept under review.  

CONCLUSIONS:   The MMRT can be used as a ‘lingua franca” to re-define existing syndromic surveillance definitions and develop computed referential baseline data. Further, data synthesized using MMRT can be used to assess the impact of ICD-9/10 transition on surveillance systems, months prior to the transition deadline.