NHSBSA use machine learning to improve patient outcomes, optimize internal processes and reduce fraud. They have identified £1billion in ongoing annual savings.
“Oracle Advanced Analytics’ data mining capabilities and Oracle Exalytics’ performance really impressed us. The overall solution is very fast, and our investment very quickly provided value. We can now do so much more with our data, resulting in significant savings for the NHS as a whole.”
The NHS Business Services Authority (NHSBSA) is a special health authority and an arm’s length body of the Department of Health for England. It provides a range of critical central services to NHS organizations, contractors, patients, and the public. Services include managing the NHS Pension schemes in England and Wales, managing payments to primary care dental and pharmacy contractors, and administering the European Health Insurance Card (EHIC).
The UK's Department of Health asked the NHSBSA to take a proactive role to identify opportunities to reduce costs and eliminate waste. One way to do this was to find better ways to use the vast volumes of data already collected and held within the organization to help reduce fraud and error throughout the health service.
The NHSBSA needed a new, centralized solution that would enable it to gain better value from its data which is spread across a disparate set of IT systems, data, storage, and analytical capabilities. To achieve this, it chose an end-to-end Oracle solution including Oracle Advanced Analytics, Oracle Exadata Database Machine, Oracle Exalytics In-Memory Machine, Oracle Endeca Information Discovery, and Oracle Business Intelligence Enterprise Edition.
It wanted to use new insights gathered from data to help identify cost savings and meet NHSBSA strategic goals and to identify and prevent healthcare fraud and benefit eligibility errors to save costs.
In addition, the NHSBSA needed to take advantage of existing data to transform business and productivity.
We chose Oracle because the solution could cope with very large data volumes running into billions of rows and could scale as volumes increase. In addition, the Oracle solution required no IT team support to run the queries, which enables our team of data analysts to be self-sufficient. Oracle Exalytics’ in-memory capability gave us the speed we required, and Oracle’s engineered systems accelerated deployment and reduced risk. Working with Oracle has been a very positive experience. The team has been incredibly responsive and provided a number of experts to help us get up and running as quickly as possible. With one vendor providing the whole solution, it’s very easy for us. If we need help, we know where to go.
Get the details
With Oracle's solution, the NHSBSA established its Data Analytics Learning Laboratory (DALL), investing in both technology and expertise to create insight from its data. Within the first three months of operation, the organization identified circa GBP100 million (US$156 million) in potential savings.
The NHSBSA used analytics to identify significant savings within NHS dental services and find instances of activities which do not demonstrate good value for money.
“With Oracle Analytics, it is much easier to detect anomalies in behaviors. We used anomaly detection to discover where there might be evidence of inappropriate behavior in dentists’ claims, enabling NHS commissioners to follow up and challenge their activities,” explained Nina Monckton, head of information services, NHSBSA.
The EHIC is available to all European citizens covered by a statutory social security scheme and entitles them to free healthcare while visiting other European countries.
During an analysis of EHIC data, the NHSBSA discovered commercial addresses being used fraudulently to apply for EHIC cards and uncovered the use of invalid NHS and National Insurance numbers to apply for a card.
“We used Oracle Exalytics and Oracle Business Intelligence for the EHIC application to improve the front-end validation process, prevent fraud, and blacklist addresses showing suspicious activities,” Monckton said.
The NHSBSA receives data relating to more than one billion prescription items dispensed in primary care settings each year. Previously, the NHSBSA did not have the computing power to analyze this data at a transaction level.
The NHSBSA can now analyze billions of records at one time, and by analyzing much larger sets of patient data, the NHSBSA can provide insight that is helping to improve standards of care throughout the health service.
“Previously, our information analysts did not have the ability to directly query data as it was mainly held in live operational systems. Now that we are able to transfer data to our Exadata environment, we have dramatically improved our ability to deliver value from our data,” Monckton said.
Analyzing unstructured text to measure satisfaction
Prior to implementing Oracle Endeca Information Discovery, the NHSBSA was only able to undertake limited analysis of unstructured text from employee or patient surveys. This constrained the number of open-ended questions being asked because of the time it took to analyze them. With Oracle Endeca Information Discovery, the NHSBSA can now pull out themes in the text and its analysts can then drill down immediately to check whether their initial hypothesis is correct.
“Oracle Endeca Information Discovery is a fantastic tool. It’s really helpful for open-ended questions and it’s great to be able to overlay quantitative and qualitative data. We used it to overlay the data from an employee satisfaction survey asking how engaged people felt at work with sickness data taken from the human resources department. The team uncovered a direct correlation between teams where staff reported they felt less engaged at work and levels of sickness within those teams,” Monckton said.
Improving data matching to save millions of dollars
In England, some people are entitled to free medical prescriptions or dental treatment from the NHS. The NHSBSA works with the Department of Work and Pensions (DWP) to establish that those patients declaring that they are exempt from a charge for dental treatment and/or medical prescriptions are claiming correctly. Using Oracle Exalytics to compare datasets, the NHSBSA reduced the rate of non-matching records for dentistry from 15% to just 5%.
The role of data governance
Data is now moving to the heart of all NHSBSA programs. As a result of the organization’s new analytics capability, teams have a better understanding of what they can do with the data and are more careful about what data they collect.
“We now know that if we collect the right data at the start of a program, we can measure what is working down the line. We are starting to change the culture of the organization around our data governance. There has been a massive shift. Data is now central to all our new programs, and data governance is at the heart of everything we do,” Monckton said.
Using the Data Analytics Learning Laboratory to achieve strategic goals
The NHSBSA’s data analytics investment is helping the organization to achieve its 5-year strategic goals, which include helping to save GBP1 billion (US$1.56 billion) for NHS patients, reducing unit costs by 50%, improving service and delivering great results for customers, and deriving insight from data to drive change.
“With our newly established Data Lab in place, we can add even more value to the NHS. I cannot begin to describe how significant that has been. This project is really helping us to achieve our strategic goals. In addition, we are working in a different way now and it has even helped with how people interact and function in the workplace.
“We’ve had a very positive response, and our chief executive is extremely impressed with our achievements and the results we have shown so far. As a result, management is recommending that our suppliers and partners come to see what we are doing to learn from our experiences,” Monckton said.
Over the next six months, the DALL team has a large number of analytics projects in the pipeline and is looking to help other areas of the business to better leverage their data. The organization will focus on how it can use Oracle Business Intelligence Enterprise Edition with business users. In addition, the NHSBSA is investigating how it might share data and its analytical ability with other government organizations to drive further value from its investment.
Oracle ran a proof of concept (POC) to show the speed and capability of the proposed end-to-end solution. The POC used publically available data sets for NHS prescription data. It covered 50 million prescribed items, 300 million records, and six months of data. The team concentrated on finding anomalies in the data and carrying out further analysis to understand them before presenting the findings in a clear and straightforward way.
Following the POC, Oracle worked with NHSBSA and its data center partner, Capita, to complete the implementation. During implementation, Oracle provided the NHSBSA with access to a virtual environment. This enabled the team to get some experience with the tools before completing the implementation. As such, NHSBSA was familiar and confident with using the new analytics tools from day one, saving considerable time and gaining immediate value.
NHSBSA identified which data it should use for analysis and transferred it across to its Oracle Exadata environment. To date, it has transferred more than 15 billion rows of data into Oracle Exadata. The prescription services database with 14 billion rows of data is the largest exported data source using 400 gigabytes. The export took 10 hours to complete with Oracle as the source database.
Among other wins, the agency Identified up to GBP100 million (US$156 million) that could potentially be saved across the NHS through benefit fraud and error reduction, by deploying new analytics infrastructure and implemented changes to prevent fraudulent European Health Insurance Card (EHIC) applications.
It used data matching to identify savings that can be made through the recovery of money from patients claiming exemption from charges for dental treatment or prescriptions when not eligible to do so, and also used anomaly detection to uncover fraudulent activity where some dentists split a single course of treatment into multiple parts and presented claims for multiple treatments.
The agency analyzed unstructured text to measure employee satisfaction in more detail and found a direct link between those who felt less engaged at work and those more likely to take time off sick. In addition, it analyzed billions of records at one time to measure longer-term patient journeys and to analyze drug prescribing patterns to improve patient care.
The NHSBSA also established a new Data Analytics Learning Laboratory (DALL) that uses data and analytics to drive action and significant savings for the NHS.
In all, the agency implemented Oracle Advanced Analytics, Oracle Exadata Database Machine, Oracle Exalytics In-Memory Machine, Oracle Endeca Information Discovery, and Oracle Business Intelligence Enterprise Edition to deliver fast analysis and data mining for NHS and wider government departments.