This site is intended for health professionals only

GPs asked to contribute patient records to NHS super-data service

GPs are being asked for their consent to have their practice data extracted and streamed to a central network by the largest suppliers of practice systems, under a Government scheme to create a central NHS patient data service.

The supplier EMIS wrote to its customers last week seeking consent for their data to be streamed to its central reporting service network, in a move that has concerned GPs privacy experts.

EMIS is the first supplier to sign a contract with the NHS Information Centre for this purpose and will feed into the General Practice Extraction Service (GPES), which will replace the QMAS reporting system for QOF payments from April 2013.

But as well as being used for QOF payments, the anonymised data can then be used for secondary purposes, and an independent advisory panel will oversee other requests for data, including research.

GPs will be asked to approve all data requests, for instance the 12 data extractions required for QOF each year, or block approve requests from a particular 'customer'.

EMIS chief executive Sean Riddell said the central reporting service would help ‘improve patient care'. ‘The data extraction service will be underpinned by the highest levels of security and information governance,' he said.

Dr Russell Brown, GPC member and a GP in Polegate said as long as the Information Commissioner was ‘doubly vigilant', practices had anything to worry about.

He said: ‘Provided the principles are followed, it should not be something to be concerned about.'

But Dr Adrian Midgley, a GP in Exeter, Devon, said the centralisation of patient data made theft more likely.

He said: ‘If you collect large amounts of data in an organised fashion, its easier for people to steal it in large chunks, or steal a particular patient's records.'

‘GPs have a reputation for being difficult about answering questions they should not be asked, whereas Government has a reputation for losing large chunks of data.'