Presentation (CTI)

OBJECTIVES: In the absence of a disease‑specific ICD‑10‑GM code, identification of primary immunoglobulin A nephropathy (IgAN) in German claims data depends on heterogeneous coding practices. This study compared coding patterns between 2022 and 2024.
METHODS: This retrospective analysis of 2024 data used German statutory health insurance claims data from the population-representative InGef research database (~4 million individuals, ~5% of German population). Primary IgAN cases were identified using two algorithms: a restrictive lower‑bound (LB) definition based on biopsy‑associated ICD-10-GM codes and an inclusive upper‑bound (UB) definition additionally incorporating less specific codes. Coding patterns in 2024 were compared with those in 2022 by care setting, diagnosis position, and recorded codes. Physician specialty associated with coding was analyzed for 2024.
RESULTS: Among incident patients, the proportion of UB cases with ≥2 outpatient diagnoses remained stable (48% in 2022 vs. 50% in 2024). In contrast, inpatient coding shifted toward primary diagnoses (33% to 52%), while secondary diagnoses decreased (43% to 25%). No notable changes were observed among prevalent patients. Code usage patterns also evolved. Among incident cases, the UB definition suggested increased use of biopsy-associated codes in 2024, alongside a reduced reliance on less specific codes. The proportion of patients with at least one biopsy-associated code rose from 38% to 50%, while the share with exclusively less specific codes declined from 62% to 50%. No such trend was observed among prevalent patients. For prevalent UB patients, biopsy‑associated codes were predominantly assigned by nephrologists, whereas less specific codes were more frequently recorded by general practitioners.
CONCLUSIONS: Coding practices for primary IgAN in German claims data appear to be evolving, with a shift toward more specific coding and increased use of primary inpatient diagnoses. These changes may reflect growing clinical awareness of primary IgAN in recent years, and highlight the potential value of more specific coding for improving case identification.