|Tel||+44 (0)1865 289300|
I am a 2nd year DPhil student under the supervision of Rafael Perera and Sue Mallett. My research focuses on assessing whether publication order of studies in a systematic review affects the results of a cumulative meta-analysis.
It is expected that the intervention effect in a cumulative meta-analysis experiences extreme values before converging to the mean value as the amount of studies added increases. My research seeks to assess whether this is the case. I am currently using data from Cochrane systematic reviews focussing on interventions in the child health, tobacco and addiction, pregnancy and cardiology areas.
Current analysis is looking at the pattern in the intervention effect over time and using time series analysis methods to assess if a change occurred in the intervention effect. The reliability of the
intervention effect is assessed using sequential monitoring boundaries. It is hoped to be able to fit a statistical function to the intervention effect; with allowance for factors such as heterogeneity, publication bias and whether the intervention is drug/non-drug
It is hoped that my work will lead to a greater knowledge as to the amount of evidence needed in a systematic review so that the intervention effect can be concluded as being the true effect; it will not change if more studies are added. This hopefully will stop studies being carried out and designed unnecessarily.