Introduction: Hundreds of meta-analyses (M-As) are published each year and thousands more await to be performed. In addition, updating of previously published M-As and subgroup analyses are seldom done. Methods need to be developed that will make the process more efficient and powerful.

Objective: To develop RTMAS ("Real-Time Meta-Analysis System"), a system that facilitates the construction of M-As and automates their updating.

Methods: The key components of RTMAS are databases of extracted randomized control trial (RCT) data, computerized M-A protocols and the RTMAS computer interface. In the USER-CONTROLLED mode, the computer interface displays and manipulates relevant information under user guidance using a multi-level matrical format. The user explores the database in various depth of details for the availability of suitable studies for M-As or for subgroup data by specifying various inclusion criteria. Many M-As may be generated from a single database. M-As may be performed in "real-time" on pre-extracted data from selected studies. The inclusion criteria which form the M-A protocols may be stored for future use. Automatic updating of previously defined M-A protocols can be achieved.

Results: RCT databases in several ambulatory pharmaceutical areas are being developed: usage of antibiotics, treatments of hypertension and congestive heart failure, as well as in the treatment of HIV/AIDS.

Discussion: Such a system can efficiently produce new and updated M-As, as well as subgroup analyses. Standardization of the methods will produce more consistent results and ensure the quality of the process.