Acute kidney injury (AKI) occurs frequently in patients with acute coronary syndromes (ACS) and is associated with adverse short- and long-term outcomes. To date, however, no standardized definition of AKI has been used for patients with ACS. As a result, information on its true incidence and the clinical and prognostic relevance according to the severity of renal function deterioration are still lacking. We retrospectively studied 3,210 patients with ACS. AKI was identified on the basis of the changes in serum creatinine during hospitalization according to the AKI Network criteria. Overall, 409 patients (13%) developed AKI: 262 (64%) had stage 1, 25 (6%) stage 2, and 122 (30%) stage 3 AKI. In-hospital mortality was greater in patients with AKI than in those without AKI (21% vs 1%; p <0.001). The adjusted risk of death increased with increasing AKI severity. Compared to no AKI, the adjusted odds ratio for death was 3.5 (95% confidence interval 1.79 to 6.83) with stage 1 AKI and 31.2 (95% confidence interval 16.96 to 57.45) with stage 2 to 3 AKI. A significant parallel increase in major adverse cardiac events was also observed comparing patients without AKI and those with stage 2 to 3 AKI. In conclusion, in patients with ACS, AKI is a frequent complication, and the graded increase of its severity, as assessed using the AKI Network classification, is associated with a progressive increased risk of in-hospital morbidity and mortality.
The Acute Kidney Injury Network (AKIN) has developed a classification system according to the serum creatinine (sCr) changes and includes 3 different stages of acute kidney injury (AKI) to provide a standardized definition of AKI. The reliability of this classification for assessing mortality risk has been validated in critically ill patients, in those with sepsis, and, more recently, in patients undergoing cardiac surgery and coronary angiography. However, no study has evaluated the relation between AKI, as defined by the AKIN criteria, and in-hospital outcomes in patients hospitalized with acute coronary syndromes (ACS). Therefore, it is unknown whether the AKIN classification can be applied and has similar clinical and prognostic implications in these patients, in whom the true baseline renal condition cannot be assessed. The present study evaluated the incidence of AKI, according to the AKIN definition, in a large, single-center cohort of patients with ACS and investigated the possible association between AKI severity and in-hospital morbidity and mortality rates.
Methods
The present retrospective observational study was conducted at the Centro Cardiologico Monzino, University of Milan (Milan, Italy). All consecutive patients with ACS admitted to the coronary care unit for >48 hours from January 1, 2002 to March 31, 2011 were identified through a search of the clinical database and included in the present study. Patients requiring chronic peritoneal or hemodialysis treatment were excluded. The ethics committee of our institute approved the present study as a retrospective cohort study.
The sCr at hospital admission and daily during the coronary care unit stay was available for all analyzed patients. The estimated glomerular filtration rate was estimated using the abbreviated Modification of Diet in Renal Disease equation. Baseline renal insufficiency was categorized as an estimated glomerular filtration rate of ≤60 ml/min/1.73 m 2 . AKI was defined, applying the AKIN classification, according to the maximum increase in sCr from baseline (hospital admission): stage 1, ≥0.3 mg/dl sCr increase; stage 2, more than two- to threefold sCr increase; stage 3, more than threefold sCr increase from baseline or sCr ≥4.0 mg/dl with an acute increase >0.5 mg/dl; or the need for renal replacement therapy, irrespective of the stage at renal replacement therapy. When occurring during the first 48 hours after admission, AKI was defined as early; when occurring afterward, it was defined as late. renal replacement therapy was performed in the case of AKI with >24-hour oligoanuria, concomitant fluid overload, hyperkalemia (>6.5 mEq/L), or metabolic acidosis (pH <7.1). In patients with AKI, renal recovery was defined as a reduction in sCr to the “no-AKI” range without the need for renal replacement therapy. The left ventricular ejection fraction (echocardiogram) was measured in all patients at hospital admission. The in-hospital mortality and major adverse clinical events were evaluated. They included major bleeding (requiring blood transfusion), acute pulmonary edema (with or without the need for mechanical ventilation), cardiogenic shock, and clinically significant tachyarrhythmias (ventricular fibrillation, sustained ventricular tachycardia, and atrial fibrillation) and bradyarrhythmias requiring pacemaker implantation.
Continuous variables are presented as the mean ± SD and were compared using the t test for independent samples. Variables not normally distributed are presented as the median and interquartile range and were compared using the Wilcoxon rank sum test. Categorical data were compared using the chi-square test or the Fisher exact test, as appropriate. The identification of the independent predictors of the considered end points (AKI and in-hospital mortality) was assessed using logistic regression analysis with stepwise selection of variables. The initial set of the potential predictors undergoing selection included age, gender, left ventricular ejection fraction, sCr, estimated glomerular filtration rate, ACS type, treatment type (medical vs percutaneous coronary intervention), diabetes, hypertension, smoking, hypercholesterolemia, previous myocardial infarction, previous coronary artery bypass grafting, and creatine kinase-MB peak value. The results of the analysis were summarized as odds ratio (ORs) and 95% confidence intervals (CIs).
To avoid spurious selection of the predictors, because the model was built and tested on the same sample, a cross-validation procedure was used. The sample was randomly split in half 200 times, and the model, including the independent predictors, was selected in the first arm (training set) and subsequently tested in the second half (testing set). For each variable, we computed the number of times it was selected in the first step and the number of times it was confirmed (deemed as significant) in the second step. We considered a predictor validated when it was selected and confirmed ≥70% of time.
A risk score for AKI was then developed. A logistic regression model was used, including all the validated independent predictors of AKI. The AKI risk score (predicted probability of event) was computed for each patient using the following formula :
predicted probability of AKI = e ( β 0 + ∑ β i X i ) 1 + e ( β 0 + ∑ β i X i )
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
