By Benjamin Ross

November 14, 2019 | "Traditional health information systems are generally devised to support clinical data collection at the point of care," Ashfaq Gilkar, Lead Clinical Business Analyst at Guys and St Thomas' NHS Trust (GSTT), wrote in the foundation's entry form for the 2019 Clinical Research News European Innovations Awards. "However, as the significance of the modern information economy expands in scope and permeates the healthcare domain, there is an increasing urgency for healthcare organizations to offer information systems that address the expectations of clinicians, researchers, and the business intelligence community alike."

GSTT's solution is to deploy a low cost structured and unstructured information retrieval and extraction architecture called CogStack. According to Ashfaq, the system meets the need of the "3R principle" (right data, right place, right time) to address deficiencies in organizational data flow, while retaining the strict information governance policies that apply within the UK National Health Service (NHS). The system received a European Innovations Award during the recent Summit for Clinical Ops Executives (SCOPE)-Europe in Barcelona, Spain.

"CogStack architecture enables GSTT to transform and ingest a large volume of clinical data in a fashion consistent with the requirements for data reuse in business intelligence, service improvement, and research," Ashfaq wrote. "This is utilized to recruit patients into various other research clinical trials and studies within the hospital."

Ashfaq tells Clinical Research News that CogStack allows clinical researchers to focus specifically on surfacing data hidden deep in hospital systems. CogStack achieves this by using machine learning to search through the data, though Gilkar laments the system is often limited in the amount of data provided.

"It all depends on your clinical inputs," says Ashfaq. "If your clinical inputs are limited, the amount of data you can manipulate or utilize will be limited as well."

Ashfaq says hospitals in the public sector suffer the most with this lack of data input, making it difficult to integrate technology like machine learning into useful practices.

"It's often an avenue that's not pursued because most hospitals don't have the basic infrastructure in place," says Ashfaq. "If you don't have that, you can't implement a system like CogStack effectively because it's only as good as the inputs you provide."

While CogStack may be limited by the amount of data available, it certainly isn't when it comes to the types of data. Ashfaq says the system is being implemented in a wide range of research areas, from dental to clinical imaging and electronic medical records (EMRs). It is even being used in clinical coding, a process in which written clinical statements are translated into a code format.

"In a major hospital like [GSTT], we have to code for every single patient, including research patients and patients in trials," says Ashfaq. "So in order to accurately represent that information—which has been a bit of a barrier in the past—CogStack has been excellent as a means of recording that clinical coding in an efficient manner and with data quality. That will help the process of our hospital economics in terms of pricing and the cost for patients."

The system was born out of requirements from the UK's 100,000 Genomes Project to find a solution to provide relevant clinical data among large volumes of disparate data sources at a low-cost. Ashfaq calls CogStack a game changer in the healthcare system.

"[AI's use in the space] is going to expand rapidly," Gilkar says. "The [NHS] in the UK is looking to create an AI center in the innovation hub, which is a national body that will create this center that will then be applied to many hospitals within the UK. I can see this taking off. I can see this expanding and being utilized in major hospitals in the UK and throughout Europe. This is the future of healthcare, in my opinion."