There is a type of meta-analysis called a network meta-analysis that is potentially more subject to error than a routine meta-analysis. A network meta-analysis adds an additional variable to a meta-analysis. Rather than simply summing up trials that have evaluated the same treatment compared to placebo (or compared to an identical medication), different treatments are compared by statistical inference.
An example of a network analysis would be the following. An initial trial compares drug A to drug B. A different trial studying the same patient population compares drug B to drug C. Assume that drug A is found to be superior to drug B in the first trial. Assume drug B is found to be equivalent to drug C in a second trial. Network analysis then, allows one to potentially say statistically that drug A is also superior to drug C for this particular patient population. (Since drug A is better than drug B, and drug B is equivalent to drug C, then drug A is also better to drug C even though it was not directly tested against drug C.) A network meta-analysis combines multiple studies and makes statistical comparisons in a similar manner.
The problem with network analysis in regards to a meta-analysis, is that a network meta-analysis is more likely to be valid when analyzing very similar studies for very similar patient populations. Since network meta-analysis extends the number and type of studies being combined, there is even more potential for combining studies that are not adequately similar. The quality of some recent network analyses in the hypertensive literature highlights the problems of this type of analysis. (See limitations of the ALLHAT network meta-analysis-details).
In general, it may be prudent to be wary of studies using network meta-analyses unless nearly identical treatments are studied (not just one component of a treatment strategy).
copyright 2005-2020 ImprovingMedicalStatistics.com