The analysis method will depend on the question(s) being asked and the type of data collected; however, all systematic reviews should include at least a narrative synthesis describing the results and risk of bias in the included studies.
Intervention Reviews
For a typical intervention review including quantitative data, standard effect measures will need to be chosen to compare studies. The next step usually involves determining whether statistical synthesis is possible and appropriate. This step entails determining whether the studies are sufficiently homogenous regarding clinical aspects, methodological characteristics, and statistical characteristics. Clinical and methodological heterogeneity are explored using clinical and methodological insight. Statistical heterogeneity is examined by visualizing the range of point estimates and 95% confidence intervals presented in forest plots and by calculating the I2 statistic and/or Cochran Q, which determines if the results from each study are more different from each other than one would expect due to chance alone.
Qualitative Reviews
Qualitative approaches of analysis differ from quantitative methods. Some knowledge syntheses will include both qualitative and quantitative data, and can be conducted using a variety of methods. Examples and additional information can be found in the full publication of this book.
Network Meta-Analysis
This method can be used to compare interventions that have not been evaluated in head-to- head studies and to rank the effectiveness of each treatment. Specifically, if one trial that includes all evidence (direct and indirect) compares intervention A versus intervention B and another trial compares intervention B versus intervention C, the network of evidence can be used to yield an indirect comparison of interventions A versus C.
