Information management

Data-mining software cuts admin time

In a nutshell
Organisation: Tower Hamlets Primary Care Trust.
Problem: Preparing monthly reports to GPs to enable them to verify secondary-care costs was laborious and time consuming.
Solution: Installation of Monarch Data Pump, a leading report-mining program.
Supplier: Datawatch Europe.
Benefits: A process that used to take two days now takes 50 minutes. Ease of use — no programming or technical skills needed by users. No commissioning of custom reports from IT departments.

Rapid return on investment — time and resource savings are quick and easily identifiable. No more manual re-keying of data.

Like other primary care trusts (PCTs) in England, Tower Hamlets has to provide its GP practices with detailed information, in the form of regular reports, that will enable them to verify secondary-care costs allocated to them each month, as well as for the year to date.

The first report prepared by Tower Hamlets PCT took Senior IT Analyst Carol McGuinness two days of painstaking manual work. Thanks to the Trust’s new report-mining program, however, it now takes her only 50 minutes.

A small London borough of 20 densely packed square kilometres, Tower Hamlets contains financial and business centres including London Docklands and a diverse population of around 210,000 people, including a vibrant Bangladeshi community. Thirty-eight individual GP practices serve this population.

Each practice has the right to receive a “firm indicative budget from their PCT at any stage” in a year and thereafter, to enable them “to monitor their referral rates, make better informed clinical decisions and understand the (financial) implications of their clinical decisions”, as set out in last year’s White Paper, Practice-Based Commissioning: achieving universal coverage (January 2006). Tower Hamlets PCT wanted to begin delivering reports to its GPs as early as possible.

The PCT met with its GPs in early 2006 to discuss and agree the form that the information needed to take, to include details of:

  • elective activity (both inpatient and day-case);
  • non-elective admissions;
  • first outpatient appointments;
  • follow-up outpatient appointments;
  • consultant-to-consultant referrals; and
  • accident and emergency admissions.

The first reports had to be prepared manually. Cutting and pasting details from one spreadsheet into 38 others — despite the fact that these were less detailed reports than those the PCT now delivers —took two days. Clearly, this was impractical for a monthly task: something had to be found to reduce the timescale. One of Tower Hamlets PCT’s managers knew about the report- mining capabilities of IT services provider Datawatch Europe’s Monarch DataPump and suggested that the IT department investigate it.

Monarch can extract information simply from a wide range of reports and data sources, including Excel spreadsheets and Access databases, in a way that many other systems cannot. It then reconfigures that data and outputs it in a new form suitable for Excel spreadsheets and many other report formats. A Datawatch trainer devised an automated process for the PCT that takes Carol McGuinness or her IT colleagues only 20 minutes to perform. Thirty minutes later, the process delivers 38 separate Excel workbooks ready to email to each GP.

With less than two days’ training, Carol or another team member can now prepare 38 GP reports in under an hour. “It is so easy to do now,” said Carol. “The trainer built an automated data-extraction process for us, so we were able to get started straight away, and also trained us so that we can change the reports or process whenever we need to”.

Regular meetings are now held between the GPs and the PCT to ensure that the reports meet GPs’ needs. In May 2006, all the GPs received a full year’s results; every month thereafter, they have received a report of the year-to-date activities. Tower Hamlets PCT has found that its new data-mining program offers a fast and simple solution to its monthly practice-based commissioning report preparation.

Jargon buster

Data mining

The process of automatically searching large volumes of data for patterns using tools such as classification, association rule mining and clustering.

Excel spreadsheet

A spreadsheet program written and distributed by Microsoft for computers using the Microsoft Windows operating system and for Apple Macintosh computers. The dominant spreadsheet application and one of the most popular microcomputer applications to date.  

 
 

To top^