New England Surgical Society
NESS Home NESS Home Past & Future Meetings Past & Future Meetings

Back to 2024 Abstracts


Development of a Novel Hierarchical Surgical Complexity Classification Score to Account for Procedural Heterogeneity Across Surgical Specialties
*Gabriela Esnaola BA2, *Thomas Gill MD3, Ronnie A. Rosenthal MD, MS, FACS1, Kevin M. Schuster 1, Robert D. Becher MD, MS, FACS1
1Department of Surgery, Yale School of Medicine, Houston, TX, CT; 2Yale School of Medicine, New Haven, CT; 3Department of Internal Medicine, Yale School of Medicine, New Haven, CT

Objective: Adjustment for procedural risk as a function of surgical complexity is needed in epidemiologic studies to accurately assess and compare outcomes following major surgery. Current approaches to classifying surgical complexity focus on narrow sets of operations within a single specialty, depend on patient-level risk factors, or are limited to datasets that use the Current Procedural Terminology (CPT) code system. These approaches, respectively, cannot adjust for the heterogeneity of operations across multiple surgical specialties, do not account for operation-specific risk independent of the patient, and cannot be applied to analyses of datasets employing the ICD-9/10-Procedure Coding System (ICD-PCS) to identify operations. Given the increasing use of ICD-PCS-based Medicare data in surgical research, our objective was to develop a novel hierarchical surgical complexity score based on factors inherent to the operation itself to classify ICD-PCS codes into objective tiers of operative risk.

Design: Development of a surgical complexity classification score.

Setting: All acute care hospitals that perform major surgery.

Patients/Other participants: The classification score applies to all patients (?18years) who have undergone major surgery. We interviewed 20 surgical attendings representing the spectrum of surgical specialties, selected via convenience sampling.

Intervention: We created a composite complexity score based on 5 factors inherent to an operation: Invasive Procedure Complexity Matrix (IPCM) category; work Relative Value Units (wRVUs); procedure subtype/objective (e.g., resection); surgical anatomical site; and Healthcare Cost and Utilization Project (HCUP) procedure class. IPCM placement and wRVUs assigned to CPT codes were mapped to 2015-2023 ICD-PCS. Complexity associated with procedural and anatomical characteristics represented in ICD-PCS code textual descriptions was defined and scored via expert interviews with surgical specialists. See Figure.

Main outcome measure: Composite ICD-PCS surgical complexity score placing each code into 1 of 3 hierarchical tiers of operative complexity: high, intermediate, or low.

Results: Our approach classified 13% of ICD-PCS codes (n=9,164) as high complexity operations, while most operations were classified as either intermediate (69%; n=48,968) or low (18%; n=12,944) complexity (see Figure). Internal validation showed that each tier of complexity, and their overall proportion, were consistent with comparable surgical complexity CPT-based classification systems.

Conclusions: We developed a novel hierarchical composite surgical complexity classification score to stratify ICD-PCS codes into 3 tiers of surgical complexity as a function of risk endogenous to the procedure, independent of patient-level risk. Following external validation, this classification score will permit investigators to better adjust for operational risk when analyzing outcomes using ICD-PCS codes spanning surgical specialties.


Back to 2024 Abstracts