Risk Factors and Effect of Acute Kidney Injury on Outcomes Following Total Hip and Knee Arthroplasty

Background: Development of acute kidney injury (AKI) following primary total joint arthroplasty (TJA) is a potentially avoidable complication associated with negative outcomes including discharge to facilities and mortality. Few studies have identified modifiable risk factors or strategies that the surgeon may use to reduce this risk. Methods: We identified all patients undergoing primary TJA at a single hospital from 2005 to 2017, and collected patient demographics, comorbidities, short-term outcomes, as well as perioperative laboratory results. We defined AKI as an increase in creatinine levels by 50% or 0.3 points. We compared demographics, comorbidities, and outcomes between patients who developed AKI and those who did not. Multivariate regressions identified the independent effect of AKI on outcomes. A stochastic gradient boosting model was constructed to predict AKI. Results: In total, 814 (3.9%) of 20,800 patients developed AKI. AKI independently increased length of stay by 0.26 days (95% confidence interval [CI] 0.14-0.38, P <.001), in-hospital complication risk (odds ratio = 1.73, 95% CI 1.45-2.07, P < .001), and discharge to facility risk (odds ratio = 1.26, 95% CI 1.05-1.53, P = .012). Forty-one predictive variables were included in the predictive model, with important potentially modifiable variables including body mass index, perioperative hemoglobin levels, surgery duration, and operative fluids administered. The final predictive model demonstrated excellent performance with a cstatistic of 0.967. Conclusion: Our results confirm that AKI has adverse effects on outcome metrics including length of stay, discharge, and complications. Although many risk factors are nonmodifiable, maintaining adequate renal perfusion through optimizing preoperative hemoglobin, sufficient fluid resuscitation, and reducing blood loss, such as through the use of tranexamic acid, may aid in mitigating this risk. (C) 2020 Elsevier Inc. All rights reserved.

Automated summary

The study identified independent risk factors for developing AKI during total joint arthroplasty and established a predictive model to assist surgeons in reducing this risk. The final predictive model showed excellent performance with a c-statistic of 0.967.