Left‐Digit Bias in Serum Creatinine Levels and Contrast Administration: A Nationwide Cohort Study Using a Japanese Percutaneous Coronary Intervention Registry

Le résumé fourni par la source

BACKGROUND: Contemporary quality initiatives emphasize minimizing contrast volume as a key modifiable factor to prevent acute kidney injury (AKI) following percutaneous coronary intervention (PCI). Left-digit bias is a cognitive bias where the leftmost digit of a number disproportionately influences decisions. This bias may unintentionally affect contrast administration when interpreting laboratory values like serum creatinine. Therefore, we aim to assess the impact of left-digit bias in serum creatinine level on contrast volume in PCI. METHODS: We analyzed patients undergoing PCI between January 2018 and December 2022 using the Japanese nationwide prospective multicenter registry. The primary outcome was PCI contrast volume. Left-digit bias was assessed by comparing contrast volume differences at specific creatinine thresholds (eg, 1.0 mg/dL and 2.0 mg/dL). AKI risks were calculated by a validated risk scoring system. RESULTS: Among 735 696 PCI procedures, the median contrast volume was 117 mL (interquartile range 85-157 mL). Analysis of average contrast volume revealed a sharp decrease at baseline creatinine levels above versus less than 1 mg/dL (mean difference, 2.2 mL [95% CI, 1.6-2.8]) and 2 mg/dL (4.7 [95% CI, 1.6-7.8]) but not at 3 mg/dL (-2.0 [95% CI, -10.0 to 6.0]). Despite differences at integer thresholds of creatinine, minimal variation across AKI risk levels suggests operators prioritize creatinine values over predicted AKI risk. CONCLUSIONS: Our study provides valuable insights into contemporary trends in contrast volume for PCI and identifies left-digit bias in creatinine interpretation affecting contrast administration. Formal risk stratification is essential to optimize contrast use for AKI prevention.

Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.