Inter-rater Reliability of ACS-NSQIP Colorectal Procedure Coding in Canada

Abstract

Abstract The American College of Surgeons National Surgical Quality Improvement Project (ACS-NSQIP) collects risk-adjusted surgical data utilizing Current Procedural Terminology (CPT) codes. Inconsistent code assignment may affect benchmarking calculations. This study aims to assess inter-rater reliability of coding colorectal resection procedures across Canada by ACS-NSQIP surgical clinical nurse reviewers (SCNR) and the impact on risk predictions. An electronic survey was distributed to Canadian SCNRs, asking them to assign CPT codes to simulated synoptic operative reports. Percent agreement and free-marginal kappa correlation were calculated. The ACS-NSQIP risk calculator was used to compare predicted morbidity and mortality between the two most frequently chosen codes for each case, to demonstrate impact on risk prediction. 44 of 150 (29.3%) survey recipients responded. There was significant variability in the CPT codes chosen. Agreement ranged from 6.7% 62.3%. Free-marginal kappa correlation ranged from moderate agreement (0.53) to high disagreement (-0.17). The ACS-NSQIP risk calculator predicted absolute differences in risk of serious complications and mortality ranging from 0.2–13.7% and 0.2–6.3%, respectively. This study demonstrated low inter-rater reliability in the coding of ACS-NSQIP colorectal resection procedures in Canada among trained SCNRs. The resulting coding inconsistency translated to variation in risk prediction.