Defining the emergency general surgery patient population in the era of ICD-10: Evaluating an established crosswalk from ICD-9 to ICD-10 diagnosis codes

INTRODUCTION In 2015, the United States moved from the International Classification of Diseases, Ninth Revision (ICD-9), to the International Classification of Diseases, Tenth Revision (ICD-10), coding system. The American Association for the Surgery of Trauma Committee on Severity Assessment and Patient Outcomes previously established a list of ICD-9 diagnoses to define the field of emergency general surgery (EGS). This study evaluates the general equivalence mapping (GEM) crosswalk to generate an equivalent list of ICD-10-coded EGS diagnoses.METHODS The GEM was used to generate a list of ICD-10 codes corresponding to the American Association for the Surgery of Trauma ICD-9 EGS diagnosis codes. These individual ICD-9 and ICD-10 codes were aggregated by surgical area and diagnosis groups. The volume of patients admitted with these diagnoses in the National Inpatient Sample in the ICD-9 era (2013-2014) was compared with the ICD-10 volumes to generate observed to expected ratios. The crosswalk was manually reviewed to identify the causes of discrepancies between the ICD-9 and ICD-10 lists.RESULTS There were 485 ICD-9 codes, across 89 diagnosis categories and 11 surgical areas, which mapped to 1,206 unique ICD-10 codes. A total of 196 (40%) ICD-9 codes have an exact one-to-one match with an ICD-10 code. The median observed to expected ratio among the diagnosis groups for a primary diagnosis was 0.98 (interquartile range, 0.82-1.12). There were five key issues identified with the ability of the GEM to crosswalk ICD-9 EGS diagnoses to ICD-10: (1) changes in admission volumes, (2) loss of necessary modifiers, (3) lack of specific ICD-10 code, (4) mapping to a different condition, and (5) change in coding nomenclature.CONCLUSION The GEM provides a reasonable crosswalk for researchers and others to use when attempting to identify EGS patients in with ICD-10 diagnosis codes. However, we identify key issues and deficiencies, which must be accounted for to create an accurate patient cohort. This is essential for ensuring the validity of policy, quality improvement, and clinical research work anchored in ICD-10 coded data.