Supplementary material-Supplementary_material_R1.docx (305.75 kB)
Download file

Supplementary Material for: Correlation between the Incidence and Attributable Mortality Fraction of Acute Kidney Injury: A Systematic Review

Download (305.75 kB)
posted on 22.01.2020, 12:35 by Komaru Y., Inokuchi R., Iwagami M., Matsuura R., Hamasaki Y., Nangaku M., Doi K.
Introduction: The incidence of acute kidney injury (AKI) as diagnosed by international standardized criteria as well as its mortality has undergone extreme variations. Although AKI is a significant worsening mortality factor, a higher prevalence may lead to better patient management, thereby lowering mortality. We investigated the correlation between AKI incidence and its associated mortality. Methods: We conducted a systematic review of studies on AKI reporting its incidence and mortality. Literature searches were performed in ­MEDLINE, EMBASE, and Cochrane Library, within the time frame of 2004–2018. Studies with small number of participants (<500 for adult cohorts, 50 for pediatric cohorts) were excluded. The correlation among AKI incidence, mortality, and AKI-attributable fraction of mortality was evaluated using a regression model. The trend test was used to analyze the effect of publication year and country gross domestic product (GDP). Results: A total of 4,694 manuscripts were screened, from which 287 cohorts were eligible (adults: 203 cohorts comprising 7,076,459 patients; children: 84 comprising 69,677 patients). Within adult cohorts, AKI patients’ mortality increased (R2 = 0.023, β = 0.12, p = 0.03) but the attributable fraction of mortality decreased (R2 = 0.27, β = –0.43, p < 0.001) with the increasing AKI incidence. Both more recent publications and higher GDP countries had a lower crude AKI patients’ mortality, although AKI-attributable fraction did not decrease. Conclusions: Cohorts with high AKI incidence had a relatively low AKI-attributable mortality fraction, which suggests an advantage of more experienced AKI management. Further study is needed, however, to address the heterogeneity of included cohorts and to confirm the causality. (Registered in prospective register of systematic reviews database; CRD 42019129322.)