ABSTRACT
106The predominant difference between an individual patient data meta-analysis (IPD) and meta-analysis based on aggregate data is that the combined study results from a central re-analysis of the raw data from each study. The IPD meta-analyses involve the collaboration of the investigators. Analyst can use IPD to examine and update trial data and to carry out time-to-event analysis. The summary statistics can be calculated for specific groups of patients and thus subgroup analysis can also be produced. The problem of ecological fallacies may be avoided in meta-analysis using IPD. Ecological fallacy refers to a situation where an average estimate for a group is not valid to the individual patients. Multivariate regression analysis using IPD may provide results that are directly relevant to individual patients. However, meta-analysis using individual patient data is much more expensive to conduct.
