We often get asked for advice on data migration, and the first thing we then ask is: “What are you migrating from and what do you want to migrate to?”

This question is often met by a deathly silence.

Whilst this may appear to be a simple question, many hospital Trusts, who are in the process of exiting the National Programme for IT (NPfIT) for the provision of Radiology PACS, struggle to answer the question. And that is because a large number of hospitals don’t understand what their fundamental issue really is. They talk of procurements to replace PACS and RIS solutions when the issues they face do not actually necessitate that approach.

The biggest challenge these hospitals  have to tackle is data localisation and NOT data migration. It’s about regaining control of their data and repatriating it such that it is held locally on disk. The reason this  is the primary issue is due to the fact that Trusts have allowed (or been encouraged to allow) their Radiology PACS data to be ’flushed‘ to the Local Service Provider’s (LSP) Central Data Store (CDS). Hospitals now have so much data flushed to the CDS that it is impossible to localise that data via the contractual route – DICOM Query/Retrieve. Our estimation is that this can only be achieved at the rate of 1TB/month, and this would have an impact on the performance of the production PACS environment.

The issue has been largely dealt with for hospitals within the CSC LSP contract as there was no contractual extension offered beyond 30th June, 2013 – consequently, Trusts have had to be decisive. However, Trusts within both the Accenture and BT LSP contracts are now faced with similar decisions –. these need to be made very quickly if those hospitals are to succeed in exiting the LSP contracts at the dates they wish.

For Trusts within the Accenture LSP contract, there are more options to successfully localise their data than in the BT contract. Often the flushed data is also held locally on tape and BridgeHead have the unique ability to rewrite that data to local disk storage below the DICOM level, such that there is no disruption at all for the Agfa PACS solution. This eliminates the need for the hospital to repatriate their data via DICOM Query/Retrieve and is, therefore, much quicker. This approach has been tested and proven to work at one of the biggest Trusts within the Accenture LSP contract, with a sample of the images successfully written to local disk storage, and successfully retrieved through the PACS workstation. As a result, a number of Trusts are openly engaging with BridgeHead to provide this service to them.

This then begs the second fundamental question: what manages the data locally on disk storage and how do Trusts stop pushing more and more radiology images to the CDS (which they are contractually bound to do). There is no point in localising the historic data if, at the same time, new images are being flushed to the CDS –Trusts simply do not have room on their SAN, and are forever reducing the size of the local cache for the storage of new images. We know of one Trust where the local cache is down to only three months of data.

The answer is to deploy a Vender Neutral Archive (VNA) or, at the very least, a DICOM Store – a location where both new images can be written to, and historic images can be localised to. This will enable Trusts to solve the two biggest issues they really face: localising and regaining control of their historic PACS data; AND not making the problem any worse by writing a copy (or two) of new images to a VNA to store the data efficiently and to protect it from misuse or loss.