ABSTRACT
We examined the incidence and predictors of acute kidney injury (AKI) in elderly patients (≥75 years) enrolled in the prospective Italian Elderly acute coronary syndrome (ACS) study and explored the impact of AKI on clinical outcome. Acute kidney injury, defined according to the Acute Kidney Injury Network criteria, occurred in 128 (21%) of 615 patients. Patients submitted to coronary angiographic procedures did not present higher rate of AKI. The only baseline variables independently associated with AKI development were creatinine clearance (odds ratio [OR]: 0.98; 95% confidence interval [CI]: 0.97-0.99) and left ventricular ejection fraction (OR: 0.98; 95% CI: 0.96-0.99). Adverse clinical events were significantly higher in patients who developed AKI. After multivariable adjustment, AKI (hazard ratio: 2.73; 95% CI: 1.87-4.0) was an independent predictor of all-cause mortality within 1 year.
Subject(s)
Acute Coronary Syndrome/epidemiology , Acute Kidney Injury/epidemiology , Acute Coronary Syndrome/diagnostic imaging , Acute Coronary Syndrome/mortality , Acute Kidney Injury/blood , Acute Kidney Injury/diagnosis , Acute Kidney Injury/mortality , Age Factors , Aged , Aged, 80 and over , Biomarkers/blood , Coronary Angiography/adverse effects , Creatinine/blood , Female , Humans , Incidence , Italy/epidemiology , Kaplan-Meier Estimate , Logistic Models , Male , Multivariate Analysis , Odds Ratio , Prognosis , Retrospective Studies , Risk Factors , Time FactorsABSTRACT
The use of meta-analysis in medicine is widespread nowadays, particularly in the field of interventional cardiology. Meta-analysis is a statistical approach aiming to combine date from a large number of patients from randomized clinical studies and/or non-randomized registries so as to obtain a pooled estimate of the results and to answer specific research questions. It is important to take the correct methodological approach in order to reach unbiased conclusions. In this article, we provide an updated review of the methodological approaches needed to perform meta-analyses of non-randomized data, and we suggest a simplified check-list of items to be considered when attempting to deploy this kind of meta-analysis.
