Cogstack

A surging virus and disparate data

King’s College Hospital faced a significant challenge during the COVID-19 pandemic: managing an influx of unstructured patient data while tracking the spread of the virus and providing quality care to patients. Its traditional systems were inefficient and unable to process the disparate data types quickly, leading to potential clinical risks and delayed decision-making.

Professor James Teo, Clinical Director of AI and Data and Professor of Neurology at King’s College Hospital says, “With so many ways of expressing words in patient records, there was a risk that we couldn’t keep track of basic information such as how many people were infected with the virus, how sick they were, and what treatments they were on.”

Uncovering real-time insights

Using Elastic, combined with the CogStack platform, King’s College Hospital was able to consolidate multiple data sources into one searchable database supported by natural language processing (NLP). This allowed healthcare staff to quickly access relevant data without the need for specialized terminology or coding. Staff got real-time insights that accelerated decision-making, supporting the hospital’s ability to respond to public health crises while managing patient risks.

Among other benefits, King’s College Hospital saw an increase in:

• Operational efficiency: Elastic enabled the processing of vast amounts of unstructured data in record time, streamlining workflows.
• Patient safety: Clinicians were able to identify patients at risk of reactions from specific medications in just hours, significantly reducing potential harm and improving patient outcomes.
• Research ability: With decades of data to interrogate, it would have been too time-consuming and expensive to undertake certain research initiatives. Researchers are now able to extract data like medical diagnoses to determine what types of treatments work for different groups of patients.

“Clinicians really enjoy using CogStack powered by Elastic. They can search in the way that they write and talk, rather than having to use coding or specialist data skills.”

— Professor James Teo, Clinical Director of AI and Data, Professor of Neurology, King’s College Hospital

Using Search AI to improve outcomes

Through the precision of search and intelligence of AI, King’s College Hospital improved healthcare outcomes by reducing the time it takes to start treating patients. At a time when uncertainty was the norm, healthcare staff could analyze vast amounts of data to make informed, data-backed decisions that helped save lives.

With the scalability of Elastic’s Search AI Platform, the hospital was able to meet user demand while expanding to cover the needs of a large medical audience. This meant more hospitals could use the platform to help meet their patients’ needs and improve their outcomes.

“In the case of a major medical safety issue, CogStack and Elastic enabled us to find all the relevant medical records in just one afternoon, when previously it would have taken months or even years. As a result of our efforts, thousands of women were able to avoid taking a medication that could have harmed their unborn babies,” says Professor Teo.

1. Elastic’s Search AI Platform makes data searchable and usable for artificial intelligence to work. With so much data to parse through, King’s College Hospital needed a solution that could find the most relevant data and bring it to the forefront.
2. Healthcare providers need a coherent and comprehensive history of the patient. They need to troll through different notes and systems to get that view. With Elastic’s Search AI Platform, no matter what format the data is in or in which environment it lives, King’s College Hospital can search it and use AI to find the relevant information.

By building a platform using Elastic Search AI to visualize populations based on diagnoses, King’s College Hospital was able to ensure those individuals had physical health checks done in a timely manner. This type of supportive outreach model helps healthcare coordinators achieve a standard of care through an easy-to-use clinical platform.