Get Into Pole Position With Modern IRT Services

August 3, 2017 Andy Gurd

In high-performance car racing, being at the front of the grid at the start of the race is a significant advantage. There are a lot of factors that impact a driver's ability to get the fastest qualifying time such as weather, track temperature, choice of tires, skill of driver, and car set-up. In clinical trials, there are also many challenges that determine how quickly you can reach study start- including site selection and patient recruitment, but also how long it can take to put eClinical systems in place. 

Particularly time-critical at the start of a trial is the IRT (interactive response technology, also known as IVR/IWR - interactive voice/web response) system for patient randomization, treatment allocation and dispensation, and clinical supplies management.

Traditionally, the process for an IRT provider to provision a system to meet the specific needs of a study has had four main stages:

  • Specification: Based on the Request for Proposal (RFP) and the protocol, the IRT provider writes a requirements specification,usually consisting of text and flow diagrams to illustrate transactions.

  • Development: The IRT provider programs the software according to the agreed specification. This can include 'stitching' together of previously built and validated components for common functions.

  • Validation: The IRT provider tests the system to check that it behaves according to the specification and fixes any issues.

  • User Acceptance Testing (UAT): The sponsor/CRO tests the system to check that it behaves according to the specification and as expected. The IRT provider fixes any issues.

While this approach may appear to be very flexible and it may appear to have the ability to customize precisely to meet the study specific needs, it has some significant drawbacks. Firstly, the programming of the system can take up to 3-5 weeks, depending on study complexity, because of the extent of custom software development involved. This also tends to lead to extended testing periods. Another weakness of this approach, is that specifications can run the risk of being ambiguous or incomplete. There is the potential that the sponsor/CRO doesn't realize there's been a misunderstanding or omission until the User Acceptance Testing (UAT) stage, which can mean that costly and time-consuming rework is needed, potentially impacting the study start date.

Typical timescales for system delivery with the traditional approach are 8 to 12 weeks.

A more modern approach is where, based on extensive experience, typical transactions can be parameterized into a system that is largely configuration-driven rather than programmed. This offers significant benefits in terms of speed and flexibility, and enables the system to be evolved with frequent review cycles to enable regular, rapid feedback from the sponsor/CRO. In this process, the system effectively replaces the specification as the means to understand what is being built, which is much more intuitive and less likely to result in requirements being 'lost in translation'. A configuration document is still used in order to be able to validate the settings in the system, but this can be automatically generated. All of this reduces timelines, makes reviews and testing simpler for the sponsor/CRO, and should lead to fewer surprises, and a higher quality system at UAT.

The configuration-driven system still allows for customization if required, but this can be delivered in a much shorter time than in the completely programmed approach.

Typical timescales for system delivery with the configuration-driven approach are 4 to 8 weeks.

And the flexibility of the approach means that late-breaking protocol design changes can be made quickly with minimal disruption to the study, like a racing car coming into the pits for a mid-race tire change.

The stages and timelines for the 'traditional' and 'modern' IRT delivery processes are compared and contrasted below:

Screen Shot 2015-07-16 at 1.33.04 PM.png

You can download the full size infographic here.

A configuration-driven approach to system delivery, coupled with an experienced 'pit crew' of randomization statisticians, supply management consultants and project manager, can help ensure your study has the best chance of being in the pole position at study start and has the smoothest and quickest ride through to the finish line.

Andy Gurd, Product Marketing Director, eClinical
Previous Item
Using Analytics to Optimize the Clinical Development Process
Using Analytics to Optimize the Clinical Development Process

A recent Bloomberg article discussed how the misuse of web advertising data may have caused companies to wa...

Next Item