Critical Review of International Classification of Diseases, Ninth Revision to International Classification of Diseases, Tenth Revision System-Based Postoperative Surgical Complication Codes

INTRODUCTION: The 2015 transition from International Classification of Diseases, Ninth Revision to International Classification of Diseases, Tenth Revision (ICD-9 to ICD-10) in the U.S. aimed to improve diagnostic and procedural precision. However, no validated methodology exists for translating complication codes, posing a barrier to longitudinal surgical outcomes research. This study evaluates a structured approach to map ICD-9 to ICD-10 complication codes for high-risk operations, including esophagectomy, pancreaticoduodenectomy, and hepatectomy. METHODS: We developed an 8-step process using the Center for Medicare and Medicaid Services General Equivalence Mappings, incorporating manual review and clinical adjudication. ICD-9 codes representing complications, including cardiovascular, gastrointestinal, pulmonary, renal, infectious, systemic, intraoperative, wound-related, blood transfusion, and postoperative interventions, were mapped to ICD-10 and classified as exact, approximate, or mismatch. This framework was then applied to the National Inpatient Sample to evaluate postoperative complications before and after implementation of ICD-10. RESULTS: A total of 148 ICD-9 codes mapped to 774 ICD-10 codes. Clinical review identified 63.3% as exact, 33.9% as approximate, and 2.8% as mismatched. The recorded postoperative complication rates appeared lower for esophagectomy (50.6%-44.5%, P < 0.0001), higher for pancreatectomy (44.3%-46.4%, P < 0.01), and similar for hepatectomy (33.4%-32.8%, P = 0.53) with ICD-10 compared to ICD-9. The recorded length of stay and in-hospital mortality were also lower in the ICD-10 era (all P < 0.05). CONCLUSIONS: ICD-10 implementation may improve the capture of specific complications and reflect evolving perioperative care. However, significant discrepancies in code translation underscore the need for validated methods. The mapping framework developed in this study enables reliable complication tracking across coding eras and supports higher-quality surgical outcomes research.