Positive Predictive Values of Procedure Codes on the Treatment of Non-Muscle Invasive Bladder Cancer in the Danish National Patient Registry.

Abstract

Globally non-muscle invasive bladder cancer (NMIBC) is a high-incidence disease. There is a large heterogeneity within NMIBC regarding recurrence- and progression risks, and large-scale studies of treatment patterns and prognoses in an everyday setting could result in NMIBC-subgroup treatment optimization, benefiting both patients and the economy. The Danish national registries provide such an opportunity if the registered procedure codes are valid. Therefore, the aim of the study was to validate the International Classification of Diseases, 10th Edition (ICD-10) codes of NMIBC treatment used in the Danish National Patient Registry (DNPR). From the DNPR, we randomly selected 200 NMIBC treatment courses identified by the dates of the course and the codes of transurethral resection of the bladder ((TURB), n = 125), photodynamic diagnosis ((PDD), n = 25), bladder instillation with Bacillus Calmette vaccine ((BCG), n = 25), or bladder instillation with chemotherapy/Mitomycin C ((MMC), n = 25). We used medical record reviews as the reference standard and estimated positive predictive values (PPVs) of all procedure codes and negative predictive values (NPVs) of PDD- and the perioperative single-shot MMC codes. We identified the medical records in 150 (75%) of the 200 treatment courses (149 individual patients). The overall PPVs were TURB: 98.9% (95% confidence interval: 93.8; 100.0%), PDD: 95.8% (78.9; 99.9%), adjuvant BCG: 90.0% (68.3; 98.8%), perioperative single-shot MMC 1/5, and adjuvant MMC: 69.2% (38.6; 90.9%). The overall NPVs were PDD: 64.8% (54.4; 73.9%) and perioperative single-shot MMC: 97.7% (92.1; 99.4%). The ICD-10 NMIBC procedure codes recorded in the DNPR are generally valid with high PPVs. The NPV of the PPD code is acceptable. However, the code for perioperative single-shot MMC is uncertain with low PPV, but a high NPV.