Improved Survival after Acute Kidney Injury

Over the past two decades, it has been progressively recognized that patients with acute illnesses complicated by AKI have increased health care costs and are at higher risk for mortality and kidney failure. Despite the utilization of KRT (i.e., dialysis), mortality with AKI is substantially higher than in other critically ill patients, even after accounting for other comorbid conditions. Furthermore, the incidence of AKI among acutely ill patients is as high as 50% (1), and the burden of AKI on health care systems has resulted in overwhelming costs and resource utilization (2).

In the study by Sohaney et al. (3) published in this issue of CJASN, the authors included >2.6 million hospital admissions of >1.6 million patients within the Veteran Affairs system over 9 consecutive years to assess the temporal trends of mortality among the patients who developed AKI. They reported the overall incidence of AKI at about 20% of all included hospital admissions, with crude mortality among those with an episode of AKI of six times (6% versus 1%) higher in the hospital than those without an episode of AKI and twice the mortality (28% versus 14%) at 1-year follow-up. They also showed that the risk of AKI-related hospital mortality was lower by roughly 1%–2% for each calendar year of admission during the study period. Meanwhile, the temporal trend of 1-year follow-up mortality was slightly higher or no different throughout the study. Indeed, following adjustments in patients without AKI, hospital and 1-year follow-up death rates after hospitalization increased by about 0.5%–1% per calendar year during the study period.

In the past two decades, efforts to raise awareness regarding AKI and its effect on clinical outcomes have exponentially increased. Along with greater awareness, the quality of care for patients at risk of or with AKI has improved, particularly with the development and validation of the consensus definition of AKI by the Acute Disease Quality Initiative and by Kidney Disease Improving Global Outcomes (KDIGO) (4); the publication of several practical clinical guidelines for the management of AKI (e.g., KDIGO) (4); increased focus by the National Institute for Health and Care Excellence; and increasing investment from the health care industry and governments in advancing knowledge regarding AKI pathophysiology, prediction, diagnosis, and management. In addition, other care processes in the management of primary etiologies of AKI have improved, including sepsis management and fluid therapy. Some of these changes may be responsible for the findings of Sohaney et al. (3) related to improved AKI-associated hospital mortality. Nonetheless, further work is needed to determine more specifically the underlying reason for the observed improvement in AKI-related mortality in the hospital. Determining the specific practice or care processes that are most impactful on the observed improved kidney outcomes is an essential target to consider for future clinical trials and quality improvement projects.

The authors have highlighted several other points worth emphasizing and exploring further. First, there was a decline in hospital AKI-related mortality over time, whereas the rates of comorbid conditions, mechanical ventilation, sepsis, and other exposures did not seem to change throughout the study in either AKI or non-AKI groups. In several other studies, a significant decline in mortality risk among patients with AKI has been observed in specific patient populations, including those with sepsis, patients following cardiac surgery, or those with acute myocardial infarction (5–7). However, the underlying reasons for these observations have not been thoroughly investigated. Factors that may have led to this improvement include increased awareness of AKI; implementation of care bundles for AKI (e.g., limiting nephrotoxin exposure and hemodynamic management); implementing prevention bundles for illnesses with a high risk of AKI (8) (e.g., heart failure, sepsis, and cardiac surgery); and earlier detection of AKI by using biomarkers (9), electronic surveillance tools, or prediction models (10). Second, in this study, the observed hospital mortality among patients without AKI increased slightly during the same period as the improvement was observed in patients with AKI. This further emphasizes that preventive measures for AKI-related death are likely more effective than those for non–AKI-related death and provides motivation for future interventions that target improved outcomes of acute illnesses to focus on AKI.

Although this study and other similar investigations mentioned above provide encouraging news of lower hospital mortality among patients with AKI, some precautions are still warranted. In the study by Sohaney et al. (3), the AKI incidence across three time periods (2009–2011, 2012–2014, and 2015–2017) remained steady at about 20% per hospitalization. Although the temporal trends of AKI incidence remained steady throughout the study, the number of patients with AKI increased over time (179,127 during 2009–2011, 183,195 during 2012–2014, and 186,460 during 2015–2017). Also, the ratio of AKI admissions to non-AKI admissions progressively increased by 0.4% for each of the 3-year periods. The previous reports in the literature clearly indicate that AKI incidence among hospitalized patients, particularly those admitted to intensive care units, continues to rise. This is likely due to multiple factors, including aging populations with higher comorbidity burden and a shift in clinical practice in selecting progressively sicker patients for hospital admission over time. The advent of effective oral antibiotics, laparoscopic or outpatient surgeries and other technologic advances, and the added burden of health care costs have led to the selection of sicker patients for hospitalization with a shorter length of stay in the hospital. Therefore, progressively higher comorbidity burden and higher severity of acute illnesses may partially explain the rising incidence of AKI among hospitalized patients. In addition, the aging population and more intense and aggressive procedures and therapeutic options (e.g., surgeries and chemotherapies) may also contribute to the rising AKI incidence (11). Thus, although the improvement in mortality of patients with AKI is an encouraging development, the number of patients at risk for AKI continues to rise, and thus, the resultant number of patients who will die with AKI remains a substantial issue.

Another important consideration is that mortality is a blunt and crude metric for characterizing AKI outcomes. In this era, with advances in hospital medicine and critical care, including highly effective organ support therapies, hospital death is often an administrative decision between patients, their families, and physicians to withdraw support. Thus, the effect of AKI on hospital mortality is difficult to disentangle from the considerations to withdraw support that parallel the increasing need for organ support therapies, including dialysis. In this regard, longer-term survival after hospital discharge is likely a more important factor for determining the effect of AKI on overall health status. Additional long-term outcomes with critical value after hospitalization are the development of CKD or kidney failure requiring long-term dialysis, quality-adjusted life years, patient satisfaction, and health care costs.

By definition, all patients with AKI have acute kidney disease (AKD). However, about two thirds of patients who meet the criteria for AKD may not reach the defining indicators of AKI (1). AKD is defined on the basis of a decline in GFR by ≥35% or an increase in serum creatinine level of ≥50% in 7–90 days after the insult and often has markers of kidney injury (e.g., proteinuria) (4). Many patients with AKD without AKI have similar outcomes as AKI but are not tagged as AKI. In a large cross-sectional population-based study, James et al. (1) determined AKI incidence as 2% over the year 2008, whereas the incidence of AKD without AKI was 4% in the same year. The outcomes of patients with AKD without AKI are similarly grim as the outcomes for those with AKI, including higher risk for mortality, de novo CKD, or CKD progression and development of kidney failure (1). The incidence of AKD without AKI among high-risk hospitalized patients is likely even higher (1). Even if we assume a similar frequency of AKD without AKI in the study by Sohaney et al. (3), there could potentially be >1,100,000 admissions of AKD without AKI that were categorized in the non-AKI group. This, in turn, could inflate the adverse outcomes in the non-AKI group. Because the diagnosis of AKD without AKI is often missed or delayed in clinical practice, these patients may not benefit from the same preventive measures that patients with AKI receive.

As informed by the study by Sohaney et al. (3) and other similar investigations, there needs to be a focus on interventions that can improve AKI outcomes. This will require first identifying the root causes of improved AKI mortality to target as preventive measures in a more focused manner to achieve even better outcomes. A comprehensive assessment of mortality rates among those with subclinical AKI or individuals with AKD without AKI is also needed. This could establish a more robust understanding of the deterministic factors for death after kidney injuries and identify more effective preventive measures.

Although changes in AKI awareness and related practice patterns have resulted in a modest improvement in the mortality of patients with AKI, the current landscape for development and validation of novel management or therapeutic strategies remains wide open. In addition, a comprehensive understanding of the pathophysiology of AKI at molecular and biologic levels would allow for the identification of therapeutic targets that could potentially have a larger effect on preventing the harm that AKI imposes on patients.

Disclosures

K. Kashani reports employment with the Mayo Clinic; consultancy agreements with AM PHARMA; research funding from La Jolla Inc.; honoraria from Nikkiso; patents or royalties from MediBeacon; and an advisory or leadership role with Baxter, GE, La Jolla Inc., and MediBeacon Inc. A.D. Rule reports employment with Mayo Clinic; serving as an associate editor of JASN, as a section editor of Mayo Clinic Proceedings, and on the National Institute of Diabetes and Digestive and Kidney Diseases Urological Diseases of America Contract Management Board; and other interests or relationships with UpToDate.

Funding

None.