Can you explain what cloud computing is and how this can be used in clinical trials?
We use the NIST definition of cloud computing with its emphasis on a user managed configuration of services, rapid scalability and elasticity, and a metered cost model. Cloud computing, as exemplified by Amazon’s EC2 or Microsoft’s Azure services, is an extension of the traditional model of outsourced, contracted information technology (IT) provision and one which, ostensibly, holds a promise of greater user control, more immediate management and lower costs.
But, because clinical trial data includes sensitive data about individuals, and its management must be compliant with both good clinical practice and national and international requirements, the use of cloud computing facilities for clinical trial data is not at all straightforward.
Why was there a need for a workshop focusing on cloud computing?
In recent years non-commercial trials units have increasingly used external IT facilities to host some or all of their clinical data management systems (CDMS), rather than on premise facilities within the unit or its parent institution, usually a university.
These external facilities are normally provided by the CDMS vendor, and physically located at the vendor’s premises or within some external but traditionally managed IT provision. This may give only indirect control over IT services but it still provides a relatively stable and well defined IT regime.
Increasingly however, CDMS vendors are using, or planning to use, cloud computing platforms to host this data. In addition, some universities are exploring the idea of ‘hybrid clouds’ to cope with excess demand and/or backup and archiving, and in a few instances trials units themselves are looking at cloud computing as an efficient means of establishing and maintaining direct control over their IT services.
Some trials units may therefore find their data ‘in the cloud’ when this was not their intention, and may not always be clear exactly where their data, or copies of it, is located. Other units may be exploring using cloud computing facilities when the potential regulatory implications are still poorly defined.
Considering the risks identified, what do you think the potential benefits of cloud computing are for clinical trials?
Despite the potential economic benefits of using cloud based computing facilities there are several major risks that need to be understood and managed. These include:
- The possible lack of control over data location, data lifetime and data access
- Continued uncertainties over the legal jurisdictions that govern cloud based data
- Industry volatility – which may result in some cloud providers disappearing
- The lack of both clear quality standards about cloud computing
- Any clear statements on this topic from regulatory authorities.
For me cloud computing may provide high potential economic benefits together with more flexibility and scalability, provided national and international requirements for clinical trials can be fulfilled.
The KKS Düsseldorf
One way of reducing some of these risks is to encrypt the data at source, but the complexities and costs of managing encryption, especially in the long term, should not be under-estimated. At the moment therefore, any use of cloud computing for clinical trial data should be approached with great caution.
Christian Ohmann, also a co-author of the work said: “For me cloud computing may provide high potential economic benefits together with more flexibility and scalability, provided national and international requirements for clinical trials can be fulfilled.
It is the aim of the European Clinical Research Infrastructure Network (ECRIN) to bring this issue forward, involving all stakeholders (such as regulatory authorities, cloud provider, trial units and patient organisations).”
How do you think the use of cloud computing will shape research in the future?
If systems can be developed that promote confidence in cloud based services for trial data, we could see a much more distributed model of data processing emerge, with further potential benefits in costs and flexibility.
IT provision is essentially a service for clinical research and changes in the nature of that provision should not have any great impact on the scale or direction of the research itself.
If (and only if) the risks of cloud computing listed above can be eliminated or reduced, and if the position of regulators on this type of IT provision can be clarified, one would expect to see a rise in the use of cloud based facilities. And stemming from that, a more efficient use of IT resources in clinical research generally.
In the longer term, if systems can be developed that promote confidence in cloud based services for trial data, we could see a much more distributed model of data processing emerge, with further potential benefits in costs and flexibility.
What are the next steps from the workshop?
The workshop ECRIN held in late 2014 was always seen as an initial step – more workshops are therefore being planned and we welcome participation from all interested parties – especially regulatory authorities and cloud service suppliers themselves.
ECRIN is also currently revising its own standards for IT and data management in clinical trials to deal with the potential use of cloud facilities. The goal is to clarify the risks, benefits, and requirements related to cloud computing in clinical trials, so that trials units can make informed decisions about cloud computing with much greater confidence.