Relationship of Cystatin-C Change and the Prevalence of Death or Dialysis Need after Acute Kidney Injury: a Meta-Analysis.

Yunlin Feng,Yue Zhang,Guisen Li,Li Wang
DOI: https://doi.org/10.1111/nep.12312
2014-01-01
Nephrology
Abstract:AimCystatin-C (CysC) has been demonstrated as a sensitive and reliable biomarker to predict the onset of acute kidney injury (AKI). However, there are few studies concerned about the relationship between CysC and the outcomes of AKI. The aim of the present study was to determine whether CysC elevation prior to definite diagnosis of AKI is related to higher prevalence of death and dialysis need outcome.MethodsA meta-analysis was conducted by searching PubMed, EMBASE and Cochrane Library database using the terms related to AKI combined with cystatin-C'. Bibliographies of relevant papers were reviewed manually. Eligible studies were those investigating death and dialysis need outcomes after AKI with CysC measurement, and were limited to English articles. Non-human studies were excluded. Random effect Mantel-Haenszel statistical method was used.ResultsSix studies were finally enrolled, consisting of 2332 patients. All of these studies were hospital-based prospective cohort studies. The follow-up duration varied from 5 days to 1 year. The odds ratio values for baseline CysC elevation and death as well as baseline CysC elevation and dialysis need were 2.34 (95% confidence interval [CI] 1.46-3.75) and 4.40 (95% CI 1.58-12.22), respectively (both P<0.05).ConclusionPatients with CysC elevated prior to AKI diagnosis have higher risk to develop death and need dialysis during short- and long-term follow-up after AKI, thus having worse outcomes. This population deserves more careful observation and might benefit from more frequent follow-up visits in the clinic. Future work is needed to get a consensus cut-off value defining CysC elevation.Summary at a Glance This is a meta-analysis examining the prognostic value of Cystatin-C measured prior to the development of acute kidney injury. Despite the positive finding, a cut-off value still needs to be determined.
What problem does this paper attempt to address?