Can Quality Improvement Reduce the Risk of Contrast-Associated Acute Kidney Injury?

Cardiovascular disease continues to be a major health problem, resulting in more than a million coronary angiograms performed in the United States each year. Contrast-associated AKI (CA-AKI) is a common complication of coronary angiography, with an average incidence rate of 12.8% (9%–27%) and higher incidences in patients with CKD.¹ In patients undergoing coronary angiography, CA-AKI is associated with prolonged hospitalizations, more rapid progression of CKD, and short-term and long-term mortality.^1,2 In addition, CA-AKI results in increased costs of care.³

Given the adverse clinical events and economic consequences of CA-AKI, the majority of publications on CA-AKI are focused on preventive strategies: volume expansion with intravenous fluids, reduction in contrast media volume, and pharmaceutical therapies. For the most part, pharmaceutical therapies have not been shown to be consistently effective as prophylactic strategies. Multiple studies support fluid administration as the primary preventive strategy for CA-AKI in patients undergoing coronary angiography.^4,5 Lower contrast volumes during coronary angiography result in a reduced risk of CA-AKI, and recent publications describe novel methods to reduce the contrast volume during coronary angiography and percutaneous coronary intervention (PCI).^6,7

Despite an abundance of studies supporting volume expansion with intravenous fluids and contrast volume reduction as effective strategies for mitigation of CA-AKI and published guidelines on the basis of these observations, there is evidence that physicians do not uniformly apply these effective preventive strategies.^8,9 The exact reasons for this variation in practice are unclear but likely include a failure to appreciate which patients are at high risk for CA-AKI, a lack of recognition of the adverse clinical and economic costs of CA-AKI, and an unbalanced focus on optimizing evaluation and management of the coronary artery disease.

To further improve patient safety and reduce costs of CA-AKI, new approaches are needed. One such approach involves the implementation of successful quality improvement (QI) initiatives, which are becoming increasingly recognized as effective strategies for improving patient safety and reducing health care costs. The application of QI in clinical practice has been facilitated by the adoption of electronic medical records and third-party reimbursements rewarding economical use of hospital resources and improved patient outcomes; however, the successful adoption and implementation of QI strategies remains uneven.

There have been a few QI initiatives targeting the prevention of CA-AKI. Brown et al.⁹ applied a high-intensity QI intervention in eight hospitals with goals of reducing practice variation and improving CA-AKI outcomes. The intervention focused on standardizing intravenous fluids orders and minimizing contrast volumes. This QI initiative reduced the incidence of CA-AKI after elective PCI by 21% and in patients with CKD by 28%. More recently, James conducted a randomized trial examining the effectiveness of clinical decision support for the prevention of CA-AKI in patients undergoing coronary angiography.¹⁰ Their intervention included a brief educational session on CA-AKI prevention and the provision of clinical decision support, which contained individualized patient AKI risk prediction, safe contrast dose volumes, and automated intravenous fluid volume targets on the basis of left ventricular end diastolic pressure measurements. In a multivariable analysis of the data adjusted for risk factors, the results by James demonstrated a significant odds reduction of 0.72 in CA-AKI.

In this issue of CJASN, Brown et al.¹¹ expand on these earlier QI implementation studies addressing CA-AKI by conducting the IMPROVE AKI trial. IMPROVE AKI is a hybrid type 1 effectiveness implementation trial that uses a 2×2 factorial prospective cluster-randomized design. A type 1 hybrid trial examines the effectiveness of an intervention while simultaneously exploring the implementation of the intervention in the real world. The trial randomized patients at 20 Veterans Affairs medical centers (VAMCs) to either a Collaborative team-based training or Technical Assistance intervention with or without information support through Surveillance dashboards for an 18-month period. Each site had a participating team comprising cardiologists, nephrologists, nurses, and technicians. The primary outcome was AKI (≥0.3 mg/dl or ≥50% increase in serum creatinine) occurring within 7 days after coronary angiography with or without PCI. IMPROVE AKI used an AKI Prevention Toolkit, which included interventions of standardized order sets, an increase in intravenous and oral fluids, and a reduction in contrast volume. VAMCs assigned to Technical Assistance participated in individual site monthly scheduled calls with an AKI improvement specialist to review and discuss the bundle of interventions and to provide consultation on implementing the AKI Prevention Toolkit. VAMCs assigned to the Collaborative arm experienced an enhanced implementation strategy of the AKI Prevention Toolkit. Each Collaborative site was assigned a QI and Collaborative improvement specialist, participated in monthly training calls, and used a structured agenda that focused on education of the clinical interventions in the toolkit along with a review of the Surveillance dashboard (if applicable). Sites assigned to Automated Surveillance Reporting were provided information support through a dashboard providing site-specific risk-adjusted AKI performance over time, comparisons to benchmarks, and individual patient-predicted AKI risk.

The 20 VAMCs provided eligible data on 4517 procedures. Approximately, 30% and 50% of the procedures were performed in patients with CKD and diabetes, respectively. Coronary angiography alone was performed in 55%, and PCIs in 45% of the procedures. The average contrast volume was 97 ml (range 52–160 ml). Patient baseline risk factors and procedure characteristics were well balanced across the randomized clusters. CA-AKI complicated 510 of the procedures with 235 CA-AKI events in the patients with CKD. The severity of AKI was stage 1 in 86% of both the total and CKD populations. Statistical analysis demonstrated no significant interaction between the implementation strategy (Collaborative versus Assistance) and the presence or absence of Surveillance support. The primary outcome of CA-AKI was reduced by 46% (absolute risk reduction of 5%) for the Collaborative and Surveillance cohort compared with the Assistance cohort alone using a multilevel logistic model.

The IMPROVE AKI trial has several impressive strengths. This was a well-designed prospective randomized trial involving multiple hospitals with an analysis of 4517 procedures. Population and procedural characteristics among the four cluster groups were well balanced. Multiple interventions to reduce CA-AKI were evaluated and aimed at medical centers and not individual cardiologists. The authors designed a sophisticated, multifaceted AKI Prevention Toolkit on the basis of years of their own previous QI work that can be adopted by other catheterization laboratories. There was near real-time Surveillance using benchmarks to provide progress feedback to individual medical centers. Finally, the results of this trial were similar after adjustment for potential imbalances in risk factors between sites.

The IMPROVE AKI trial also had some limitations and unexpected findings. Obtainment of baseline and postcontrast serum creatinine values at nonstandardized times may have resulted in inaccurate incidences of CA-AKI. However, both of these limitations should have been equally distributed by randomization. Not surprisingly, the incidence of CA-AKI in patients with CKD was higher at 17.8% compared with 8.6% in non-CKD patients. However, it was surprising that the risk reduction in CA-AKI with the Collaborative and Surveillance interventions in the patients with CKD failed to achieve statistical significance. One explanation for this unexpected finding is that cardiologists in the study were more likely to already use the components of the AKI Prevention Toolkit in a high-risk CKD population as part of previously established practice patterns. In the non-CKD population, the 8.6% incidence of AKI seems to be higher than one would expect to see for AKI solely because of contrast media in patients with normal kidney function, suggesting that a significant proportion of the AKI in this study was due to factors other than contrast nephrotoxicity. The successful reduction in AKI in this population with the QI initiatives may have been due to factors (e.g., volume expansion countering hypotension) other than a direct prevention of AKI from contrast media.

How should the results of this study be evaluated by physicians caring for patients undergoing coronary angiography? Hospitals and interventional cardiologists that currently closely adhere to the preventive strategies of volume expansion with intravenous fluids and minimization of contrast load in high-risk patients may not see a significant risk reduction with a formal QI initiative.⁹ For other hospitals or practices, it may be worthwhile to collect data on their CA-AKI incidence, and if the results are similar or above currently reported averages or benchmarks, then these institutions may benefit from a QI initiative similar to that described by Brown et al.¹¹ Interested physicians need to understand that local QI expertise, development of a structured plan on the basis of evidence-based standards, the presence of clinical champions, and physician and institutional engagement are all pre-requisites for successful implementation of QI initiatives. Institutional resources are needed to protect the time of physicians, nurses, and technicians to allow participation in the QI process. Another requirement, as noted by the authors, is the development of a timely and accurate Surveillance dashboard from an electronic database of clinical and procedural patient‐specific data. This is important for accountability and reinforcement of behavioral changes.¹⁰

In conclusion, Brown and colleagues have convincingly demonstrated that intensive QI interventions using collaborative team-based training and surveillance dashboards can be effective in reducing CA-AKI events after coronary angiography and PCIs. Given the clinical and economic consequences of CA-AKI, this trial provides a useful framework to guide the implementation of evidence-based QI interventions for hospitals and physicians that seek to improve their processes and outcomes for CA-AKI. What remains unclear at this time is how many hospitals and physician groups will have the committed interest, resources, and other requirements to successfully implement these strategies and whether the improved outcomes can be sustained when the QI initiative ends. Increased institutional funding, outside investment, and future studies examining the sustainability, scalability, and the clinical and financial impact of this work should be encouraged.

Disclosures

M.R. Rudnick reports honoraria from UpToDate. The remaining author has nothing to disclose.

Funding

None.

Author Contributions

M.R. Rudnick conceptualized the study and wrote the original draft, and M.R. Rudnick and C.M. Chaknos reviewed and edited the manuscript.