One for all: the growth of interoperable data systems for clinical trials
As trial data grows rapidly in volume, complexity and diversity, researchers are finding it increasingly unmanageable. Elly Earls finds out why interoperable software systems could provide a real solution to the ever-increasing complexity of deciphering clinical trial data within the pharmaceutical sector.
One of the most significant challenges facing the pharmaceutical community is the explosion of genomics and other 'big data'. While researchers are more than capable of generating meaningful data from techniques such as DNA sequencing, things start to get tricky when they are faced with the proposition of managing, transmitting and analysing the increasingly complex sets of information being produced.
Moreover, according to George Komatsoulis, director (interim) of the Center for Biomedical Informatics and Information Technology (CBIIT) and CIO at the National Cancer Institute (NCI), improving patient care by advancing the approval of new diagnostic and therapeutic approaches is completely dependent on researchers' ability to better manage and make sense of these ever-growing mountains of data.
"Clinical trials-related data (in particularly genomic data that are being used to identify the specific molecular nature of a patient's disease) are growing rapidly in volume, complexity and diversity," he said. "This is driving the need for interoperable systems for collaboration and data sharing."
And it's not only researchers that need to work together; those involved in patient care are also increasingly being brought into the picture. "The integration of clinical scientific discovery and direct patient care has traditionally been referred to as 'bench-to-bedside' and is quickly become the standard of care for clinical trials," Komatsoulis explained. "Integration with electronic health records (EHRs) both broadens the categories of data available and simplifies record keeping and data collection."
Share and share alike: the advantages of interoperable clinical data systems
Traditionally, when pharmaceutical companies and research organisations have put their heads together during challenging projects, the stakeholders involved have been obliged to share data between different software systems, presenting three key problems: lack of semantic interoperability of technologies; variance of data terminologies (for example, one system uses M for male and the other uses 1 for male); and differences in structured vocabulary.
On top of this, data has historically not been released to researchers until after a clinical trial concludes. "This lag time has always limited the ability for scientific research to affect clinical trial outcomes," Komatsoulis noted. "Interoperable biomedical informatics technology provides the capabilities to ensure rapid and widespread access to trial and scientific data, allowing authorised scientists to have access to de-identified tissue, image and clinical data from trial participants, as well as scientific pre-clinical and genomics data within days to months of their availability."
Therefore, a paradigm shift from separate platforms to one interoperable system could lead to more productive clinical trials and, consequently, better patient outcomes. "Furthermore, Data and Safety Monitoring Boards will have quicker access to review and evaluate the study data for participant safety, progress and efficacy," Komatsoulis added.
Leveraging technology to streamline the clinical trial process is beneficial to big pharma and the patients it serves. Chris Lo looks into the new digital horizon for drug development.
For Charles Jaffe, CEO of standards organisation Health Level Seven International (HL7), the positive impact of using interoperable systems could potentially be felt across the entire process, depending on the level of manual input and manipulation required. "Some systems capture data at the point of clinical research and provide structured data nearly ready for regulatory submission, but this varies significantly. When there is successful end-to-end data delivery, benefits accrue to almost every stakeholder in the process," he explained.
"For the Principal Investigator (PI), data entry and adjudication (query resolution) require far less time and manual intensity. For the site management personnel and CRO management, data irregularities are reduced and disparities minimised. And, for the pharmaceutical companies, data is often cleaner and requires less manual review. Finally, from a financial perspective, the time from opening the trial to data lock can be dramatically improved."
Interoperable clinical trial data systems: what is currently available?
But is the pharmaceutical industry actually making full use of the advantages interoperable data systems can offer yet? It's certainly starting to, according to Jaffe.
"A collaborative program among HL7, the Clinical Data Interchange Standards Consortium (CDISC), the US Food and Drug Administration (FDA) and NCI visionaries is leveraging a single information model (Biomedical Research Information Domain Group, or BRIDG) to reuse data from clinical care settings in clinical research," he remarked. "For some pharmaceutical companies, this has been successfully deployed in studies that have been submitted for regulatory approval."
In this issue we explore the best strategies and technology that can help to make clinical trials quicker, safer and more cost-efficient.
Moreover, data exchange enabled by HL7's Clinical Document Architecture (CDA), a system that supports the exchange of data between providers involved in the care of patients, is now widely deployed for distributing clinical summaries, laboratory data and prescription information. "Studies are now underway to design systems that use CDA directly for clinical research," Jaffe added.
The benefits of the developments outlined above are wide ranging for both clinicians and patients. "Utilising multiple systems for data input can be cumbersome and also interferes with the natural work flow of patient care," Jaffe explained. "The systems we are building now will require physicians to learn a single interface and work in a nearly identical environment; workflow will not be disrupted."
"For patients, this means that better drugs will come to market that have more clearly defined populations, pose fewer risks for adverse events, cost less to develop and span from concept to available drug much more quickly."
Individual research associations are also making progress with implementing interoperable data systems. NCI, for example, launched the Cancer Biomedical Informatics Grid (caBIG) program in 2004, which aimed to support broad interoperability among biomedical information systems and redefine how research was conducted, care was provided and patients interacted with the biomedical research enterprise.
And, although caBIG is now being subsumed into the NCI's National Cancer Informatics Program (NCIP), the goal of the broader initiative remains the same. "We want to ensure that biomedical researchers have access to the data they require to carry out the NCI's mission to understand the causes and improve the diagnosis and treatment of cancer," Komatsoulis emphasised. "We believe the program will lower the barriers for systems to interoperate in the service of biomedical research and, ultimately, precision medicine."
There's no denying that interoperability is still in its infancy, but, equally, progress is certainly being made, with several studies currently underway to advance the use of interoperable systems. From the FDA to the NCI, the CDISC to the Partnership to Advance Clinical Electronic Research (PACeR), the issue is being attacked from many different angles and by many different bodies. A new era of interoperability is, without doubt, close at hand.