Center-Level Variation in the Development of Acute Kidney Injury Following Cardiac Operation

Abstract

Background

Acute kidney injury (AKI) is a frequent complication following cardiac surgery. However, factors associated with AKI remain poorly understood. In this national study, we evaluated centre-level variation in the incidence of AKI after elective cardiac surgery.

Methods

Adult patients undergoing elective coronary artery bypass graft or valve operations with normal baseline renal function were identified in the 2010-2020 National Inpatient Sample. Multilevel mixed-effects models were utilized to rank hospitals based on estimated rate of AKI. The intraclass coefficient was used to estimate the level of variation attributable to hospital factors. High AKI centres (HACs) were defined as those within the highest decile of estimated AKI rate. The association between HAC status, in-hospital mortality, perioperative complications, length of stay, and hospitalization costs also were analyzed.

Results

Of 1,324,083 hospitalizations across an annual average of 703 centres, 4.9% of patients received their operation at an HAC (annual average of 70 centres). Compared to non-HACs, HACs had a lower annual cardiac case volume (62 [interquartile range: 40-115] vs 145 [interquartile range: 80-265] cases; P < 0.001) and served a larger proportion of non-White patients (20.0% vs 15.1%; P < 0.001). After adjustment, HAC was associated with increased odds of respiratory complications (adjusted odds ratio [AOR] 1.72, 95% confidence interval [CI] 1.57-1.90), infectious complications (AOR 1.57, 95% CI 1.40-1.76), and cardiac complications (AOR 1.27, 95% CI 1.18-1.36). Additionally, HAC was associated with an incremental increase in hospitalization costs (β coefficient +$4151, 95% CI $2305-$5997).

Conclusions

We demonstrated significant hospital-level variation in perioperative AKI. HACs were associated with inferior clinical outcomes and higher levels of resource utilization.