Silico Research finds prevalent use of data warehousing technology in R&D
According to the study, virtually all those who are not deploying data warehousing technology at the moment, expect to be doing so by 2004. This implementational build-out combined with the fact that individual biopharmaceutical companies are deploying more warehouses across more functions, will account for a 36% increase in the number of warehouses in the sector this year and an estimated 150% over the next three to four years.
"Companies are adopting a strategy of deploying more warehouses rather than adding new users and functions to existing warehouses," said Emmett Power, CEO of Silco Research and lead analyst on the research. "This means that the average number of new users for each warehouse is static at around 300 users and that scalability is less of an issue than often thought. We see no evidence of the rise of the ever-expanding monolithic enterprise data warehouse."
According to Silico, the highest deployment of data warehouses is in the early discovery stages of the drug development process and in clinical trials.
"The leading role of data warehouses is in managing laboratory information management systems with 47% of data warehousing having a LIMS function," said Power. "After that, bioinformatics (47%), clinical trials (32%) and pharmacokinetics (32%) are the key points of deployment for data warehousing technology.
"We are seeing different types of warehouses evolving at either end of the discovery pipeline," he said. "This may limit the amount of warehouse-driven data, application and functional integration that is possible or indeed desirable across the drug discovery processes."
According to Power, most data warehouse deployment is currently focused at the departmental level. "Companies believe, as a article of faith, that they should link scientists and researchers across the enterprise," he continued. "How they go about doing this is another matter. We're seeing a lot of testing of federated and virtual warehouses and other middleware solutions but no clear answers so far."
Silico research claims that part of the problem is measuring the payback on the investment and cautions that simple ROI calculations are not sufficient to capture the nuances of the deployment question in the biopharmaceutical R&D environment.
"The correct methodology used to address the deployment issue is a complex question," concluded Power. "The methodology will depend upon a number of factors including where in the R&D process the warehouse is positioned and the extent to which it is part of the R&D infrastructure. Wrongly modeling the deployment calculation and the methodology will lead to a misallocation of IT resources and a lower return on R&D investments."