Umbrella review is a third study conducted through evidence synthesis method, based on the secondary studies including systematic reviews and meta-analyses. Although it has been widely used abroad, further understanding, recommendation and application of this type of method are still limited in China. We introduced the definition, objective, present application, the similarities and differences with systematic reviews or meta-analyses, indications, limitations, and research progress of umbrella review in the research area of traditional Chinese medicine (TCM), aiming to benefit future clinical research and treatment in practice.
Umbrella review is a research method that comprehensively analyzes the systematic reviews and meta-analysis of a research question. In recent years, the research methods of umbrella review have been widely used, but the quality of umbrella review is uneven. Therefore, this paper focuses on the production methods and existing challenges of umbrella review, in order to provide references for domestic researchers to make umbrella review.
Objective To overview the systematic reviews/meta-analysis (SR/MA) for the effectiveness of yoga on patients with type 2 diabetes mellitus. Methods The CNKI, WanFang Data, VIP, CBM, PubMed, Cochrane Library, Embase, Web of Science, JBI and CINAHL databases were electronically searched to collect SR/MA on the intervention of yoga in diabetes mellitus from inception to November 6th, 2023. Two researchers independently screened the literature and extracted data. AMSTAR was used to evaluate the quality of methodology, and GRADE was used to evaluate the certainty of evidence, and the outcome indicators were statistically analyzed. Results A total of 14 SR/MA were included. The evaluation results of AMSTAR showed that 7 articles were of high quality and 7 articles were of moderate quality. The result of GRADE showed that there were 2 items of high-level evidence, 26 items of intermediate evidence, and the remaining 31 items were low-level or very low-level evidence. The results showed that yoga could significantly reduce fasting blood glucose (FBG) (moderate confidence), glycosylated hemoglobin (HA1C) (moderate confidence) and postprandial blood glucose (PPBG) levels (moderate confidence), and was also superior to other interventions in high-density cholesterol (HDL) (moderate confidence), low-density cholesterol (LDL) (moderate confidence), triglyceride (TG) (moderate confidence), total cholesterol (TC) (moderate confidence), systolic blood pressure (SBP) and diastolic blood pressure (DBP) (moderate confidence), muscle strength (high confidence), cardiorespiratory fitness (moderate confidence) and weight (moderate confidence). Conclusion The existing evidence shows that yoga has a good effect on blood glucose control (moderate confidence) and also has a certain effect on lipid parameters (moderate confidence) and anthropometric indicators (moderate confidence), but the quality and confidence of the current research evidence are low. Future researchers should standardize the research design to provide more high-quality evidence for the prognosis and treatment of type 2 diabetes mellitus.
The overlap of literature in umbrella reviews can affect the reliability and accuracy of research conclusions, leading to results with a higher risk of bias. Therefore, it becomes crucial to assess the degree of overlapping and how to handle it. In order to avoid redundant calculations and reduce the risk of bias, researchers need to quantify the degree of literature overlap and adopt corresponding processing strategies. This paper provides a detailed introduction to the calculation methods of overlapping and different strategies for handling overlapping, aiming to provide a reference and guidance for domestic scholars' understanding and application of this method.