Data migration hurdles detailed in leaked report

The complex task of transferring data from existing GP clinical systems into systems supplied under NHS Connecting for Health (CfH) contracts is costing twice as much and taking twice as long in the North East and Eastern clusters compared to London, a confidential report leaked to EHI shows.

However, the report also found that the approach used in the more expensive data transfers was robust and scaleable; a significant factor given the volume of work due to happen under the National Programme for IT.

The report on GP data migration highlights the current slow rate of migration from practice-based systems into spine compliant systems hosted by local service providers (LSPs), and warns that at current rates it will take far longer to transfer all 9000 GP practices in England than CfH’s target of five years.

Marked "restricted – management and commercial", the report was commissioned by the National Programme for Information Technology early in 2005. It followed concerns about the pace of data transfers and variations between the East and North-east clusters and London.

CfH issued a statement acknowledging data transfer as an important and challenging activity. It said: "We commissioned a report precisely because of the importance of the issue which we want actively to manage. We do not comment on the contents of leaked documents, particularly when, as in this case, it is not the final document and contains inaccuracies. However, we have shared the information with our suppliers and we are already addressing the thrust of the appropriate recommendations."

The report says that data migration is complex, requiring clinical and computer expertise. Each transfer is unique as each system is used in different ways, it says.

Overall, the document reports that data transfers so far have been satisfactory. However, it adds that the effort is on too small a scale to reach targets.

The biggest limiting factor was the lack of capacity within practices to carry out the checks as their data is moved between systems. The report says: "This [lack of capacity] will severely limit the pace of migration."

However, it adds that the situation was aggravated by the fact that clusters and the LSPs (Accenture in the North-east and East, and the BT-led Capital Care Alliance in London) had not signed off commercial agreements and had no clear project documentation.

In London, where 17 data transfers had been undertaken by March 2005 including 11 paper-only transfers, the cost per practice was £2,500, with the data transfer process from commissioning to sign off taking three to four months.

But in the East and North-east clusters only six transfers had been completed, each costing £5000 and taking an average six to eight months. The report says this cost differential could generate £5million extra expenditure and impact significantly on PCTs. It says: "By way of example, a PCT migrating 50 practices would suffer £125k additional charges which would certainly be sufficient to influence buying behaviour if both suppliers were considered ‘safe’."

The reasons for the variance were said to include higher complexity of the target system in the East and North-east clusters, where data is being migrated into Torex Synergy Enterprise. In London it is being transferred to InPractice Vision version 3.

Another source of the variance was Accenture’s choice of what the report called a "new player to this market". EHI understands that this "new player" was primary care information specialists Newchurch. However, the company’s website indicates that data migration is a core part of the firm’s business.

Newchurch and Accenture were not prepared to comment, although several sources indicated to EHI that there were considerable reservations about this aspect of the report’s accuracy.

Despite the slower handling of data transfer to date, the report supports the use of a "technical information engineering approach" in the North-east and East. This could potentially be scaled up and therefore able to deal with large numbers of transfers. The approach also produced significant data quality improvements.

In contrast, London’s approach of using a commercial contractor to work on case-by-case basis will be limited by the size of the contractor, says the report.

But a number of other factors will be important to reach migration targets, it adds. This will include expanding the technical resources available and publishing standards and documentation to allow more suppliers with general experience in healthcare data migration experience to contribute to the migration effort.

It also calls for CfH to ensure more clinical buy-in by funding locum GPs or other staff during a transfer and to make a persuasive business case.