Acute kidney injury in different time windows: a retrospective study of hospitalized elderly patients

Abstract Background To compare the differences between the Kidney Disease Improving Global Outcomes (KDIGO) criteria of the 48-hour window (early acute kidney injury [AKI], 3–5 day window [middle AKI], and 6–7 day window [late AKI]) in the diagnosis of AKI, as well as the relationship between the diagnosis time windows and 90-day mortality. Methods We conducted a retrospective cohort study. All elderly patients admitted to the Geriatric Department of the Chinese PLA General Hospital between 2007 and 2018 were evaluated for AKI during their hospital stay. Patients with AKI were divided into early, middle, and late AKI groups according to the time of diagnosis. Statistical analyses were performed using SPSS 21.0 statistical software. Continuous parametric variables are expressed as the means ± standard deviations (SDs), and continuous nonparametric variables are presented as the medians with interquartile ranges (25th and 75th percentiles). Categorical variables are presented as numbers (n) or percentages (%). Group comparisons were conducted using one-way analysis of variance or the Kruskal-Wallis H test for continuous variables and Pearson’s chi-square or Fisher’s exact test for categorical variables. Logistic regression analyses and a forward stepwise selection method were used to identify risk factors associated with AKI diagnosis time windows and 90-day mortality. Results During the follow-up period, 1847 patients were enrolled. Overall, 22.4% of the patients (413/1847) developed early AKI, 7.3% (134/1847) developed middle AKI, and 10.7% (197/1847) developed late AKI. Risk factors for early AKI included age, hypoalbuminemia, low prealbumin level, and the need for mechanical ventilation; middle AKI was significantly associated with age, low prealbumin, low hemoglobin, and the need for mechanical ventilation, whereas late AKI was closely associated with age, low baseline estimated glomerular filtration rate, low prealbumin, and low hemoglobin. In the multivariable-adjusted analysis, AKI time windows (early AKI, odds ratio [OR]: 6.069; P < 0.001; middle AKI, OR: 5.000; P < 0.001) and late AKI (OR: 2.847; P < 0.001) were more strongly associated with higher 90-day mortality than non-AKI. Conclusion Clinical differences and risk factors for AKI in elderly patients depend on the definition used. A better understanding of how AKI develops during different diagnostic windows may lead to improved outcomes.