A leading NHS provider of mental health services, and major centre for clinical research.
Our client is a leading NHS provider of mental health services based in South London, and is a major centre for clinical research through the associated Biomedical Research Centre (BRC). In order to accelerate research and improve treatment and care, researchers needed a more effective way of accessing the valuable clinical data embedded across a large and increasing volume of patient records – much of it captured as unstructured text as well as structured variables.
The provision of appropriate access to electronic clinical records is non-trivial. Researchers are not allowed to see any information that would enable them to identify patients. Previously, searching and filtering of patient records used only about 20% of information which had been codified. Up to 80% of information comprises unstructured notes on patient events, assessments and care plans. These notes had to be painstakingly reviewed to remove information which might identify the patient – a highly time-consuming process which distracted researchers from their core task.
Timely and secure access to this information would enable researchers to investigate and test clinical hypotheses and identify relevant patient groups for clinical trials – providing real benefits to clinicians, patients and their carers.
Our work resulted in the design, development and implementation of a new and innovative search tool considered by the BRC leadership to be unique in the world of health research – known as CRIS (care records interactive search).
CRIS provides a secure search capability across structured and unstructured data which has been pseudonymised (and so individual patients can be identified later by appropriately authorised staff) – even when that data is embedded in text. It uses an interface which reflects the source data so minimising training for researchers. It supports ongoing updates from changing patient data so that researchers can be alerted when new patients fall within a search criteria.
This capability replaces work-intensive manual searching of patient notes enabling faster analysis of records, accelerated research and provides better and more reliable access to patients.
BearingPoint worked with the client to develop the detailed requirement and design. BearingPoint then developed and implemented the search tool. The underlying search engine uses Microsoft FAST. The front-end user interface and back-end data integration were developed using industry standard technologies. Data is merged from different sources to consolidate demographic data, patient events, results, medications and clinical notes. The ‘system’ is operationalised to take daily feeds of new data.
BearingPoint continues to work with the client on supporting and enhancing CRIS with new features around data integration, search and alerts.