http://hdl.handle.net/10222/13185 2013-05-23T20:37:49Z http://hdl.handle.net/10222/21721 Profile of Thomas J. Marrie Dalhousie University 2013-04-15T00:00:00Z http://hdl.handle.net/10222/16004 Profile of Steven Morris Dalhousie University 2013-02-17T00:00:00Z http://hdl.handle.net/10222/16001 Profile of Kirill Rosen Dalhousie University 2013-02-17T00:00:00Z http://hdl.handle.net/10222/15967 Crosswalk: Mapping ICD-9 to ICD-10 Randhawa, Ravin The International Classification of Diseases (ICD) is a comprehensive list of codes for medical diagnoses and procedures that is generated by the World Health Organization. It is the standard by which global health information is mandated to be captured. It defines the universe of diseases, disorders, injuries and other related health conditions. Currently, ICD revision 10 (ICD-10) is the gold standard. Although ICD-10 was defined in 1990, implementing it was and is a non-trivial issue. It is an issue that will always exist when traversing to new revisions for the codes. As new revisions are introduced, the granularity of the information captured increases. Entities using the codes are tasked at updating processes and applications, a process that can take many years. A larger problem exits. As new revisions are introduced, the granularity of the codes captured increases. As an example, ICD-9 comprises of approximately 17,000 codes whereas ICD-10 comprises of approximately 141,000 codes. There are many translators available for mapping ICD-9 Codes to ICD-10 codes and vise-verse. This paper restricts the scope of the discussion surrounding mapping ICD-9 to ICD-10 to the General Equivalence Mappings (GEM) developed by the National Center for Health Statistics (NCHS) in conjunction with the Centers for Medicare and Medicaid Services (CMS), the American Hospital Association and 3M Health Information Systems. The author’s primary function as an analyst at the Population Health Research Unit (PHRU), Dalhousie University was extracting patient data from various databases housed and maintained at the PHRU repository. The author extracted data based on specifications provided by PHRU’s Business Analysts, as well as after discussion with clients. Paramount to any extraction of data was a basic understanding of the area of research and documenting deliverables. In meeting these requirements, the author encountered issues when working with data held at PHRU. The issues are not unique to PHRU. It is universal. Data captured in the PHRU databases prior to the year 2000 is in ICD-9 codes. Post 2000, the data captured uses ICD-10 codes. The author suggests that the mapping of ICD-9 codes to ICD-10 codes is not trivial. Crosswalk files are used in mapping one revision of the ICD codes to another revision. Internship report - fall/winter 2012 2013-02-11T00:00:00Z