Meta-Analysis of Clinical Interventions

Meta-Analysis of Clinical Interventions

The main idea of the current paper is to provide a meta-analysis of the article titled “Meta-analysis of thoracic epidural anesthesia versus general anesthesia for cardiac surgery” written by Svircevic, van Dijk, Nierich, Passier, Kalkman, van der Heijden, and Bax in 2011. The primary goal of the paper is to conduct a detailed analysis of the research design used in the article.Meta-Analysis of Clinical Interventions

Study Design

The exploratory meta-analysis hypothesis is used in the research. The use of exploratory hypothesis is explained by the research purpose. The article studies the differences in the research outcomes with regard to the combination of general anesthesia and the thoracic epidural anesthesia and the analysis of the ways in which the already existing research affects cardiac surgery outcomes. The research includes the randomized studies retrieved from such databases as PubMed and MEDLINE (1699-2010), Science Citation Index Expended and Social Sciences Citation Index (1988–2010), EMBASE (1983-2010), CINAHL (1982-2010), CENTRAL, and Web of Science. After the collection of the data, only 28 publications were selected out of 1,390 initially identified. In fact, all the studies used in the current research are relevant. With regard to the chosen strategy of study design, it helps to discover detailed and interesting aspects of the previous research and to identify the strong and weak sides of the combination of general anesthesia and the thoracic epidural anesthesia while performing cardiac surgery.Meta-Analysis of Clinical Interventions


The investigator of the study under analysis provides a detailed description of the methodology used in the study. The first step in the research methodology is a detailed description of the search process with the identification of the inclusion criteria which helps to avoid double counting with regard to the statement of the final data used for the study. Since the research is based on the peer-reviewed articles and previous studies, the risk bias assessment was conducted and the detailed description of the aspects included in the bias assessment was provided. The methodology description also includes the data extraction form and principal endpoints. Moreover, the discussion of the statistical method is included. At the same time, the methodology does not include the specific analysis of the methods covered in the 28 used research studies.

The research methodology includes the detailed analysis of the risk of bias evaluation. The authors of the research used such criteria as “method of randomization; concealed treatment allocation; blinding during pre-, peri-, and post-operative care; blinded data collection and analysis; blinded adjudication of study endpoints; and completeness of (follow-up) data” (Svircevic et al., 2011, p. 273). Despite the fact that the inclusion criteria was taken into consideration, the articles used for the current study were not assessed in regard to other quality aspects. However, there is a full range of inclusion criteria, and the low-quality studies could not appear in the articles used for the research.Meta-Analysis of Clinical Interventions

The current study did not include any characteristics of the subjects included. The research stated the number of participants covered in all studies; however, there was no analysis of patients. On the contrary, the research results included in this meta-analysis presented a detailed analysis of the research outcomes. Thus, there were discussed in detail such analysis results as mortality, supraventricular tachyarrhythmias, myocardial infarction, respiratory complications, and neurologic complications. It should be stated that the metaregression did not show any associations “between the study outcome and factors varying over the years of execution of the individual studies or risk of bias items for any of the outcomes” (Svircevic et al., 2011, p. 276). Furthermore, the treatment peculiarities in each study are not described in detail.


Assessing the outcomes of the studies used for the meta-analysis, the investigators had all the necessary information about the authors of the articles, the subjects, and the limitations of the research projects under analysis. In fact, the identity of the authors of the investigation was completely clear and evident. The investigators of the meta-analysis used the outcomes of each article utilized while conducting the research. The discussion of the results of the meta-analysis involved some limitations and the outcomes of the previous study to show the consistency of the current study and to prove the validity of the assessment utilized. Moreover, the authors referred to the discussion of the endpoint in each research included in the meta-analysis to show the measures taken and to give the information about the final statements. Since the research results were gathered into specific groups to show the number of research projects included under the particular outcome, the definition of the endpoint affected the results. Having a purpose to compare and contrast the outcomes of the general anesthesia and thoracic epidural anesthesia, the major findings were presented.Meta-Analysis of Clinical Interventions


The research results of the studies used in the meta-analysis were consistent since the investigators took into consideration this point and developed very detailed inclusion criteria. In fact, research results consistency was one of the inclusion criteria for the articles to be used in the meta-analysis. Therefore, the meta-analysis used the research projects where the authors managed to reach the results that could be trusted. Having included a number of different research projects and being interested in heterogeneity of the findings, the meta-analysis included the statistical assessment of heterogeneity of outcomes using I2 measure and the DerSimonian–Laird two-step between-study variance estimate, t2. Since the heterogeneity results were immensely small, I2 and t2 equaled 0, several meta-analyses were used to justify the results. The number of participants included in the investigation allowed one to perform calculations; however, the results showed that the data sample had to be increased from 2,731 patients used in the current research to about 10,000 to guarantee the full consistency of the results and blind application in practice. Having grouped the research outcomes and having measured the consistency of each finding, the heterogeneity was found in some cases. The sensitivity analysis used in each group had either little or no effect on the results. Discussing the research results and the significance of the outcomes, the investigators state that the “meta-analysis showed statistically significant reductions in the incidence of supraventricular tachyarrhythmias and respiratory complications after TEA … [with] no significant differences in the incidences of mortality, myocardial infarction, and stroke” (Svircevic et al,. 2011, p. 276). However, despite these findings and statistical significance, the authors warn that the results have to be considered with caution due to a small sample included.Meta-Analysis of Clinical Interventions



Discussing the research results, the authors of the meta-analysis recognized that the larger sample sizes will increase the statistical power of the research findings and will add to the clinical knowledge. The authors of the meta-analysis suggest increasing the sample to 10,000 patients; however, it can be difficult to accomplish it, as in this case, the limitations connection to the time period can increase. The use of the larger sample can add to the limitations because it will be necessary to extend the dates of the research. However, the practices used in the past are not implemented nowadays. Moreover, the modern practices have contributed greatly to the patient’s condition. Therefore, the extension of a research sample appears to be problematic, since no examination was provided of the impact and characteristics of outliers.

The data used in the meta-analysis report is adequate. In fact, the researchers have grouped the outcomes in order to report adequate results. The subgroup analysis was performed on the selective basis being guided by the predetermined study questions initially discussed. Being interested in the specific outcomes of the general anesthesia and thoracic epidural anesthesia, the researchers have set the parameters for analysis. The researchers did not determine a fail-safe N. Therefore, no discussion is provided about the number of additional average size investigations with zero effect size possible to guarantee the validity and statistical significance of the research results.Meta-Analysis of Clinical Interventions


The research under analysis has proven that one should use thoracic epidural anesthesia with caution, as the sample used in the meta-analysis does not allow generalizing either harmful or beneficial effect. To be able to judge the harm or profitability of the thoracic epidural anesthesia compared to the general anesthesia, the sample of at least 10,000 patients is required. Therefore, the current meta-analysis does not allow stating either the harm or beneficial effect of thoracic epidural anesthesia. However, having shown a failure to determine an average effect size, it still provides some generalizing conclusion, which does not point that single studies may be more relevant to a particular patient or a particular setting.

The investigators recognize that the meta-analysis allows better extrapolation beyond the data than single investigations do because in case with meta-analysis, it is possible to obtain research results compared and assessed from different studies, and, therefore, the justification of the research results is more reasoned. The relation between the outcomes in different studies is tested that allows better extrapolation of the findings. Having grouped the findings from the studies under analysis, the investigators have added to the validity of the research having assessed the statistical significance and heterogeneity of each of the groups.Meta-Analysis of Clinical Interventions