The past few years have seen an explosion in the amount of data available to pharmaceutical companies. A decade ago, around 600 data points would be logged over the course of a clinical trial. Today, thousands are captured in a single day. While a potent weapon, this data surge presents a challenge.
Raw information is often collected in a range of different formats and stored in systems isolated from one other. As a result, much time is spent sorting and adjusting data for comparison, rather than on constructive possibly game-changing analysis.
According to NNIT's subject matter expert within life Sciences IT, companies need to put in place strategies to manage the data at their disposal. First and foremost, because data standardisation is becoming a requirement from the authorities, but also because maximum exploitation of data is sound business. Independent standards from the Clinical Data Interchange Standards Consortium (CDISC) can provide support to companies for the acquisition, exchange submission and archive of clinical research data and metadata.
“Throughout the 1990s and 2000s, companies developed systems management strategies to ensure that they were cost-effective aligned and compliant. But it is not often that you see a full information management strategy. Companies are starting to talk about it now. asking how to standardise data and eliminate repeatable action to derive business value”
When it comes to putting these capabilities in place, NNIT employs a comprehensive five-step method. First, the company’s full clinical drug development system is analysed, followed by an examination of its day-to-day proœsses. Then a data flow study is carried out to determine the interconnections between different systems, and how a change to one can affect the others. Finally, a number of implementation scenarios are considered before one is chosen.
“It is so important to apply a structured method to standardisation, because no variable stands in isolation,” he explains. “You can have someone map the relationship between all data elements. But if you don’t know who is using the data and in what process, you can end up ruining a whole set of processes. When we go in to carry out standardisation, we open the heart to access all data processes. We have to ensure that when we leave, the patient is still alive.”
One of the building blocks in the standardisation process is SDTM, a framework that, using pre-defined templates and protocols, gives a uniform structure to data. The business benefits are many. Data can be collected quickly and automatically integrated at the end of a study, allowing the analysis of data across studies. It also makes it possible to produce standardised reports. In his view, this is especially important.
“There are companies telling us, ‘we have health authorities coming to us each day seeking information, managers who want regular updates, and they want it on the fly*, he says. “They are crying out for standardised reports because they don’t want to make a report that they can’t trust. Physicians know they are risking their neck by making a decision based on clinical data. They need to be sure it is accurate”
“It ¡s important to apply a structured method to standardisation, because no variable stands in isolation."
Standardisation is not just about gaining competitive advantage Health authorities such as the FDA have increased their focus on patient protection. All data derived from clinical trials or related to the marketed drug is the responsibility of the pharmaceutical company, which must now ensure control over their data throughout a clinical trial and this information needs to be accessed and comprehended by a wide range of stakeholders. In the long run, through good knowledge management, this standardised body of shared information could lead to much quicker drug development.
“A lot of medical staff are looking for background information on the population involved in clinical trials,” he says.
“Instead of having a clinical trial that you have to develop from scratch, why not take information from the already existing population relevant to that group? Using that as a reference group or a placebo population, you are increasing the power of your study and moving faster towards product launch.”
Changing behavior is tough, but a successful CDISC implementation relies on acceptance of all stakeholders and an organisation that is able to fully absorb the changes. Therefore, change management and governanœ strategy are part of NNIT’s CDISC implementation consultancy. This way, standards will stay standard, allowing the company to manage big data in a compliant and competitive way.