Data indicate that acute kidney injury (AKI) affects one-third of children admitted to a pediatric intensive care unit (PICU) and is related to the development of several complications and poor outcomes. An understanding of the most common risk factors associated with AKI can help prevent onset and related complications, explains Enrico Vidal, MD, PhD. However, few pediatric studies have focused on identifying these factors.

Assessing AKI Burden

For a study published in Blood Purification, Dr. Vidal and colleagues sought to assess the incidence rate of AKI, identify risk factors for AKI, and evaluate clinical outcomes in a large sample of children admitted to a tertiary care PICU during a 3-year study period (January 2014-December 2016). “The study was mainly a retrospective review of clinical charts of all children admitted to identify those who developed AKI during their PICU hospitalization,” explains Dr. Vidal.

The researchers reviewed the charts of PICU patients aged 20 or younger retrospectively for the presence of AKI based on both serum creatinine elevations and urinary output, comparing those with and without AKI for a series of risk factors—hypotension, multiple organ dysfunction syndrome (MODS), coagulopathy, thrombocytopenia—at admission and during their PICU stay. Risk factor presence or absence was assessed once daily for each case and recorded until AKI development, PICU discharge, or death. Use of inotropes and exposure to nephrotoxic drugs was also evaluated. Diagnoses at PICU admission were categorized as respiratory failure, shock (hypovolemic and hemorrhagic shock), cardiac disease (cardiogenic shock, congenital cardiopathy, and elective cardiac surgery), infection (septic shock, bacterial and viral infection), trauma (trauma, severe head injury, burns), postsurgical (elective and urgent non cardiac surgery), and other.

More Than One-Quarter At Risk

Dr. Vidal and colleagues identified AKI in more than one-quarter (27%) of all patients and in 18% of those aged 1 month or younger. Among those with AKI, 39% developed stage I, 24% developed stage II, and 37% developed stage III. Patients with AKI were significantly older (median age, 5 months vs 1 month) and sicker than those without AKI at PICU admission. Those with AKI were more often hospitalized for cardiac (40.5% vs. 33.5%), oncologic (10.0% vs. 6.3%), metabolic (3.2% vs. 1.0%) or renal disease (3.6% vs. 0.0%), and then admitted to the PICU because of cardiac dysfunction, shock, or infection, whereas those without AKI during their PICU stay were more often hospitalized for respiratory disease (30.1% vs. 14.1%) and admitted to the PICU with respiratory failure.

While all factors in the univariate analysis were significantly associated with AKI risk (Table), independent risk factors for AKI development upon stepwise logistic regression analysis included diuretic use (odds ratio [OR], 2.78), multi-organ dysfunction syndrome (OR, 2.68), use of two or more inotropes (OR, 2.56), age older than 2 months at PICU admission (OR, 2.43), serum creatinine greater than 44 μmol/L (OR, 2.23), coagulopathy (OR, 1.89), at least one comorbidity (OR, 1.84), and exposure to nephrotoxic drugs (OR, 1.66).

“AKI is also a strong surrogate marker of illness severity, since, when present, it was associated with an increased length of stay and risk of mortality,” says Dr. Vidal. Indeed, patients with AKI had a median PICU length of stay of 8 days, compared with 4 days for those without AKI. Crude mortality rates in the overall cohort of patients with AKI was ten times higher than that of those without AKI (12.6% vs 1.2%), with mortality rates of 18% for those with Stage I AKI, 14% for those with stage II, and 68% for those with stage III.

Careful Monitoring of Preventable Risk Factors

Dr. Vidal notes that several risk factors for kidney dysfunction in the PICU setting are preventable, adding that “careful monitoring of these risk factors in critically ill children can prevent major complications and lead to better outcomes. Nephrologists and intensivists should both be aware of this in order to tackle AKI. Thus, research in this area should focus on prediction models for AKI in children based on early biomarkers or on real-time automated alert systems that use data derived from EHRs to promptly identify at-risk patients.