Combining aggregate data and individual patient data in meta-analysis : an alternative method
It has been shown that, in cases where both the AD and IPD studies are available, combining these two levels of data could improve the overall meta-analysis estimates, compared to utilizing AD studies alone. However, the coverage probability of estimates based on combined studies are relatively low...
| Main Authors: | , |
|---|---|
| Format: | Conference or Workshop Item |
| Language: | English English |
| Published: |
2015
|
| Subjects: | |
| Online Access: | http://irep.iium.edu.my/46966/ http://irep.iium.edu.my/46966/ http://irep.iium.edu.my/46966/3/IREP_ProgBook.pdf http://irep.iium.edu.my/46966/2/Slides_NIK.ppt |
| Summary: | It has been shown that, in cases where both the AD and IPD studies are available, combining these two levels of data could improve the overall meta-analysis estimates, compared to utilizing AD studies alone. However, the coverage probability of estimates based on combined studies are relatively low compared to the AD-only meta-analysis, when the existing standard method was used to combine these studies. The aim of this paper is to introduce some modifications to the existing two-stage method for combining the aggregate data (AD) and individual patient data (IPD) studies in meta-analysis. We evaluated the effects of these modifications on the estimates of the overall treatment effect, and compared them with those from the standard method. The influence of the number of studies included in a meta-analysis, N, and the ratio of AD: IPD on these estimates were also examined. We used percentage relative bias (PRB), root mean-square-error (RMSE), and coverage probability to assess the overall efficiency of these estimates. The results revealed that the proposed method had been able to improve the coverage probability while maintaining the level of bias and RMSE at par to their existing counterpart. These findings demonstrated that the technique for combining different levels of studies influenced the efficacy of the overall estimates, which in turn is crucial for drawing reliable and valid conclusions |
|---|