(IP253) A Longitudinal Accounting of Acute Kidney Injury after EVAR Surveillance Imaging in a National Health System

Objectives: To characterize the risk of acute kidney injury (AKI) in relation to imaging surveillance obtained after endovascular aortic repair (EVAR) in VA hospitals where loss to follow-up is rare.

Methods: Records of Veterans receiving EVAR between 1/1/2000-12/31/2023 in VA hospitals were analyzed, with imaging surveillance studies (abdominal CT, US, or MRI) as the exposure of interest. The outcome of surveillance-related AKI was defined as an International Classification of Diseases (ICD)-9 or ICD-10 code for AKI within 30 days after an imaging study. Imaging studies within 30 days of a reintervention were excluded. The proportion of imaging studies with surveillance-related AKI was calculated using a broad definition of AKI, which excluded imaging studies within 6 months of a code for end stage renal disease (ESRD) or prior renal transplantation, and a narrow definition of AKI, which additionally excluded imaging studies within 6 months of a code for chronic kidney disease (CKD) and was limited to AKI codes within 14 days after an imaging study. The codes for ESRD, CKD, and for the broad/narrow definitions of AKI were chosen based on literature review and validated with retrospective chart review.

Results: During the study period, 27,292 veterans (mean age 71.7 years, 82.8% White, 99.4% male) receiving EVAR at VA hospitals had 171,799 qualifying imaging studies following their procedures. By the broad definition, the proportion of imaging studies with surveillance-related AKI was 4.1% (n=6847 imaging studies, n=4973 veterans) (Figure 1). By modality, the proportion with surveillance-related AKI was 4.0% after CTs (n=5002 imaging studies, n=3889 patients); relative to contrast, surveillance-related AKI was observed in 2.1% (n=1775 imaging studies, n=1549 patients) of contrasted studies and 6.2% (n=3546 imaging studies, n=2748 patients) of non-contrasted studies. By the narrow definition, the proportion with surveillance-related AKI was 2.1% (n=2946 imaging studies, n=2543 patients). By modality, the proportion with surveillance-related AKI was 2.2% (n=2328 imaging studies, n=2047 patients) after CTs; relative to contrast, surveillance-related AKI was observed in 1.1% (n=846 imaging studies, n=786 patients) of contrasted studies and 3.2% (n=1330 imaging studies, n=1193 patients) of non-contrasted studies.

Conclusions: Veterans who undergo non-contrasted surveillance imaging after EVAR have triple the risk of AKI compared to veterans who undergo contrasted imaging. This finding suggests that surgeons are more likely to surveil veterans they perceive as at-risk for AKI with contrast-sparing modalities, and further work may elucidate this hypothesis.