Claims-Based Studies Underestimate Adoption of Robotic Surgery

Wall-Wieler E, Lee S, Liu Y, et al. Sensitivity of Insurance Claims Codes in Identifying Robotic Assisted Surgery. Ann Surg. January 2026.

Wall-Wieler and colleagues assessed how well insurance claims procedure codes identify robotic-assisted surgery across nearly 3 million procedures (n = 2,978,390) in the PINC AI Healthcare Database, using operative or registry data as the reference standard. 

They found that overall claims-only identification had a sensitivity of approximately 0.578, meaning that only about 58% of true robotic cases were captured by robotic-specific billing codes. Sensitivity exceeded 0.8 for inpatient procedures, but was very low for outpatient cases, dropping below 0.5 by 2021 for procedures commonly performed in outpatient settings, leading to significant under-reporting of robotic utilization in administrative data. 

Importantly, although the positive predictive value was generally high (coded robotic cases were usually truly robotic), the low sensitivity means a large proportion of robotic surgeries, especially outpatient, are miscoded as laparoscopic or unspecified. This misclassification varied by procedure type and over time, resulting in biased estimates of adoption rates, costs, and outcomes when using claims alone.

For surgeons and researchers, these results imply that claims-based studies likely underestimate robotic adoption and may distort comparative effectiveness and cost analyses unless supplemented by chart validation or registry linkage.