Meta-analysis on COVİD-19 and acute kidney injury

Meta-analysis is a method of combining the results of multiple studies on a topic with various methods and statistically analyzing the findings of the studies. In periods when a priori information is scarce and there is an acute need for scientific information (e.g. pandemics, emerging diseases, etc.), the need for reliable scientific sources with high quality and evidence value becomes even more important. It is important to examine the factors that may limit objectivity in the evaluation of meta-analyses, which are considered the most reliable studies in the evidence pyramid. In this meta-analysis, it was aimed to investigate the acute kidney injury (AKI) that may develop in patients hospitalized due to COVID-19 conducted by different investigators worldwide between 2022 and 2023 after the COVID-19 pandemic was declared and to provide strong evidence about AKI, which is an important renal involvement of COVID-19, a new disease. In addition, we aimed to assess whether different quality classifications lead to limitations in information source and objectivity. Meta-analysis was performed using R studio statistical software (version 1.0.143) and meta, metaphor packages. I2 and Cochran's Q statistics were used to assess heterogeneity. Due to high heterogeneity, a random effect model was used. Publication bias was assessed with funnel plots and Egger's test. In our study, subgroup analyses were performed, and meta-regression was performed. A total of 51 studies that met the screening criteria were identified by literature search and heterogeneity I2 was found to be 98.7%. The overall prevalence of AKI in patients hospitalized due to COVID-19 was 34.7% (95%CI: 29.2-40.1). The quality of a total of 51 studies was scored using JBI criteria (minimum: 3 maximum: 8). Subgroup analyses were performed by grouping the studies with different quality classifications. In the subgroup comparisons made in three different quality groupings, no difference was found between the groups in terms of the frequency of AKI. Therefore, considering that studies classified as low quality could also be included in the analyses, other subgroup analyses were performed in a total of 51 studies. Due to COVID-19, 18 studies (I2 99.1%), all of which included patients hospitalized in intensive care units, and 33 studies (I2 98.5%), which included patients hospitalized in both wards and intensive care units, were evaluated as two separate groups. In the meta-analysis of studies that included only patients hospitalized in intensive care, the prevalence of AKI in patients hospitalized in intensive care due to COVID-19 was found to be 44.5% (95%CI: 34.2-55.3%), considerably higher than the other group prevalence (29.7% [95%CI: 24.5-35.6%]) (p=0.013). In gender subgroups, although the prevalence was higher in males, no difference was found. No difference was found between the prevalences according to the development index of the countries. The belief that the publications submitted to journals during the pandemic period passed through the peer review process quickly and were accepted without sufficient review has made the quality and reliability of the publications questionable. In these periods of uncertainty, evaluating the incidence of the disease, its complications, how it differs from country to country, according to the characteristics of individuals such as age and gender, together with the quality classification of publications can provide a reliable source of information and will provide valuable information at the point of taking measures at the global level. Key words: COVID-19, acute kidney injury, meta-analysis, quality classification, development index, subgroup analysis