Report says project for wider NHS to get at GP data is overdue, over budget and under-performing
A catalogue of errors has led to the General Practice Extraction Service (GPES) joining the list of NHS IT failures, the National Audit Office (NAO) has declared.
Its newly published report on the IT system – designed to allow NHS organisations to extract data from English GPs’ computer systems – says it has come in at a cost way over the original budget and, eight years after the project began, cannot yet provide the service that was intended. So far NHS England is the only organisation to have extracted data from the system.
Although the original business case said the service would begin in 2009-10, it was not until April 2014 that the first data was provided to a customer. In the meantime, potential customers had begun to use alternative sources of data.
The NAO investigation was prompted by its financial audit of the Health and Social Care Information Centre (HSCIC), which showed the centre had agreed to pay additional charges to one of the main suppliers, Atos IT Services, when the GPES was not working as expected. This came after responsibility for the project was transferred to the HSCIC from its predecessor, the NHS Information Centre (NHSIC).
Mistakes were made in the original procurement and contract management, but the situation became more tangled in 2001, when a technical review led to changes in the design. The requirement for each GP system supplier to adopt a common query language for extraction systems was dropped in favour of logical ‘business rules’, and followed by a failure to agree detailed requirements with Atos.
Further problems arose around using a fixed priced contractual model with ‘agile’ parts.
All of this led to the cost of the programme rising from £14 million to £40 million during the planning and procurement stage, since when there have been at least £5.5 million in write-offs and delay costs.
When the system was transferred to HSCIC in 2013 it found fundamental design flaws which meant it could not deal with complex, ‘real life’ data, and was not sufficiently comprehensive to identify the problems.
Meanwhile, the time needed to design a new type of extract and restrictions in the contract severely limit HSCIC’s ability to provide data in response to requests. This has prompted the NAO to say it is unlikely that the GPES can provide the service as planned in its current form.
This leaves a situation in which the collection and analysis of primary care data is still fragmented, making it difficult to use it in monitoring quality, planning and paying for health services and supporting medical research.
Despite the damning verdict on the programme, the NAO says the GPES still has potential advantages over the alternatives, especially in extracting data types from across England.
It acknowledges that HSCIC has put in place a plan to correct the failures and says that further money is needed to improve or replace the GPES.
In response, the HSCIC issued a statement saying that since taking over the GPES it had stabilised the service to ensure it could facilitate Quality and Outcomes Framework payments to GP practices, and built upon this with faster, live reports for the addition of automated payments for conditions such as dementia and rotavirus.
“After stabilising the service we are working with organisations that have previously made extract requests to agree on their current requirements and be realistic on what can be achieved,” it said. “We expect to complete 24 different types of data extract in 2015-16.
“We are working hard with system suppliers to ensure that together we ensure GPES runs at full capacity and that new contract negotiations move forward in a positive way. We are exploiting all technological aspects of the current service and are expanding on this with our technical expertise to make the extraction process faster.”