The Role of Race on Acute Kidney Injury Following Cardiac Surgery

Background
Acute kidney injury (AKI) frequently complicates cardiac surgery and is more common among Black patients. We evaluated determinants of race-based differences in AKI rates.
Methods
Serum creatinine-based criteria was used to identify adult cardiac surgical patients developing post-operative AKI in the PERForm registry (7/1/2014-6/30/2019). Patient characteristics, operative details and outcomes were compared by race (Black versus White) after excluding patients with pre-operative dialysis, missing pre- or post-operative creatinine, or other races. A mixed effect model (adjusting for demographics, comorbidities, surgical factors) used hospital as a random effect to predict post-operative stage 2 or 3 AKI. Propensity score analyses were conducted to evaluate robustness of the primary analyses.
Results
The study cohort included 34,520 patients (8% Black). More Black than White patients were female (43 versus 27%, p<.001), had hypertension (93 versus 87%, p<.001) and diabetes (51 versus 41%, p<.001). AKI >Stage 2 occurred among 1,780 (5%) patients, more often among Black than White patients (8 versus 5%, p<.001). Intra-operatively, Black patients had lower nadir hematocrits (23 versus 26, p<.001), and were more likely to be transfused (22 versus 14%, p<.001). After adjustment, Black (compared to White) race independently predicted odds for post-operative AKI (adjOR 1.50, 95% CI 1.26-1.78). The multivariable findings were similar in propensity score analyses.
Conclusions
Despite accounting for differences in risk factors and intra-operative practices, Black patients had a 50% increased odds for developing moderate-severe post-operative AKI compared to White patients. Additional evaluations are warranted to identify potential targets to address racial disparities in AKI outcomes.