CPT coding by interventional radiologists: a multi-institutional evaluation of accuracy and its economic implications.

PURPOSE: To evaluate the accuracy of Current Procedural Terminology (CPT) coding for interventional radiology procedures and the associated professional economic impact when coding is performed by operating physicians.
METHODS: Procedure reports and physician charge sheets were obtained for 549 interventional radiology encounters performed by 62 physicians at 23 hospitals and analyzed for appropriate CPT code use. Physician-selected CPT codes were reviewed by experts, who determined correct coding by consensus. Physician coding errors and professional relative value unit (RVU) impact were analyzed. Expert discordance and associated RVU impact were similarly evaluated.
RESULTS: Physicians correctly coded only 242 of 549 IR cases (44%). The overall professional RVU impact of their errors was +4.2% (overcoding). Physician coding was correct least frequently for complex arterial interventions (15 of 53, 28%) and dialysis access interventions (16 of 54, 30%) and correct most frequently for less code-intensive drainage (19 of 31, 61%) and biopsy procedures (35 of 47, 74%). Experts were initially concordant in 497 of 549 cases (91%), with only a minimal tendency (+0.3% RVU) toward overcoding. Expert coding differences were explained by simple code oversights (28 of 52, 54%), coding guideline ambiguity (15 of 52, 29%), and physician documentation ambiguity (9 of 52, 17%).
CONCLUSION: When interventionalists code their own procedures, CPT errors are common, but the associated RVU impact is small. Given the consequences of incorrect coding, physician-assigned CPT codes warrant review by experienced coders before claims submission. Although radiology practices should strive for perfect coding, expert discordance suggests that this goal is unattainable but less elusive than for nonradiology services.