In health research, cluster randomized trials (CRTs) are an important means of research: A group of participants is used to evaluate where treatment contamination is expected, or in studies where individual randomization cannot be carried out. However, even CRTs are prone to bias and loss of information; as a response, this paper will discuss how to assess statistical analysis methods and how to go about handling missing data when CRTs are in use.
PubMed, Web of Science and PsycINFO, will be used to find different 86 CRTs studies published from August 2013 to July 2014. Investigation of titles, abstracts, and full-text articles will be used to distinguish relevant study matching predefined criteria. Using a standard data extraction template, two independent reviewers will gather the information needed from each research. Descriptive statistics will be used to plot out a summarization of results.
As none of the information used in the study is linked to a single patient’s information, there should be no ethical concerns. Data will promptly be analyzed upon the completion of this review, with findings examined in great detail through conference presentation and publication.