GPs are forming commissioning groups that do not cover a defined geographic area, sometimes excluding lower performing practices.

The move appears to run contrary to instructions and guidance issued by the Department of Health.

HSJ has identified 17 primary care trust areas where practices belonging to the same consortium are spread across geographic areas rather than formed from neighbours. The affected areas cover 5.1 million registered patients.

In at least eight of these PCT areas – affecting 20 consortia covering 2.2 million people – there is evidence of strong performers grouping together.

In a further seven cases, high performing practices are going it alone. Some have confirmed they do not want to form groups with worse performing neighbours.

The Red House surgery is a single high performing practice in Hertfordshire. Practice manager Ken Spooner said it planned to stay independent and that not wanting to join weaker neighbours was “an element” of its decision. He added: “It was a concern when we looked at what is going on around us.”

HSJ assessed consortia using member practices’ average achievement on the quality and outcomes framework in 2009-10, and their success in managing demand as indicated by the historic cost to the NHS of their patients, using Department of Health practice based commissioning guidance for 2010-11.

Forming consortia from non-neighbouring practices is allowed under the Health Bill and health secretary Andrew Lansley has said GPs should make their own decisions about which consortia to join.

However, the DH’s December command paper Liberating the NHS: next steps said consortia had to “serve a defined geographic area” for commissioning accident and emergency services and to serve the needs of non-registered populations, for example.

The DH has since indicated these requirements would make it unwise to have non-geographically defined consortia and for some practices to be excluded.

A further potential problem is groups formed by poorer performers being less engaged in maintaining and improving quality.

PCT Network director David Stout said any groups divided by performance were likely to change before 2013. He said: “I am sure there will be concern about some being the equivalent of ‘sink estates’, with all the most difficult practices, most deprived populations, coming together.”

Association of Directors of Public Health president Frank Atherton said the findings raised fears of consortia “cherry-picking healthier populations”.

He said it could end up that “the more deprived and less affluent people get left behind with the poor performing GPs.

Les Brann is chair of Essex GP Commissioning, which appears to be better performing and slightly less deprived than its neighbours. He said: “There has been inequality in general practice since the NHS has been in existence, why is everyone looking at it now?

“Equality will inhibit innovation and improvements in patient care… Different areas have different needs.”

Senior policy sources highlighted the potential financial problems of poor performing groups, saying they were more likely to either have to reduce services, face overspends, or have to be bailed out.

HSJ was told the problem could partly be overcome by adjusting the new funding formula to better reflect historic spending. However, one source said greater performance variations would exacerbate difficulties already faced in setting allocations.

Niti Pall is co-chair of HealthWorks consortium, with practices in Sandwell and inner Birmingham, spread across areas where other consortia are also located. She said GPs were discussing collaborating across Birmingham and the Black Country which could resolve some potential financial problems.

Dr Pall said: “The conversation [between consortia] might have to be, ‘I am going to lend you money, but I will want it back, by the way do you want any help [to improve]?’” She also said the analysis was limited because QOF scores are often not an accurate reflection of quality.

Senior figures involved in consortium commissioning said there were benefits to like-minded, closely linked GPs working together, and that they were addressing potential problems.

DH adviser and National Association of Primary Care president James Kingsland’s practice is a member of the small Wirral NHS Alliance emerging consortium. It is mixed geographically with two other consortia but none have outlying QOF scores or spending.

Dr Kingsland said: “One reason we are small is that everyone [all practices] understands [the work] and is involved. We’ve got a collective viewpoint of what we’re trying to achieve.”

See list of affected areas

City and Hackney

East London Integrated Care Registered patients: 202,000; QOF: 93 per cent; cost per 1,000: £612,000*

Klear consortium Registered patients: 44,000 people; QOF: 96 per cent; cost per 1,000: £531,000.*

The average deprivation of the two consortia is very similar. Klear GP lead Haran Patel said its good performance on QOF and, particularly, spending per patient was due to joint working for many years, including through practice based commissioning. He said the group was not excluding practices although they would have to demonstrate engagement – for example willingness to lead on commissioning issues. 

View the map

Milton Keynes

Premier MK Registered patients: 145,000; QOF: 95 per cent; commissioning cost per 1,000: £515,000; deprivation: 10.8.*

GP Healthcare MK Registered patients: 109,000; QOF: 90 per cent; cost per 1,000: £571,000; deprivation: 21.7.*

GP Healthcare MK is noticeably, but not exclusively, dominated by doctors with an Asian background. Consortium lead Nicola Smith confirmed this was the case and said groups formed from past relationships. Premier MK is largely made up of training practices and hers is not. 

She said: “It was just GPs getting together. We are going to be focusing on deprivation. It is going to be quite difficult.”

A PCT spokesman said the consortium had “both formed out of existing relationships between local practices which took advantage of common opportunities and tackled similar challenges”.

View the map

*QOF scores are the average of practices’ percentage of potential points in 2009-10. Deprivation data is the average index of multiple deprivation of the population of each practice. Neither is adjusted for member practice population sizes. Cost per patient is based on historic patterns for some activity and prescribing in the 2010-11 practice based commissioning toolkit. All from the DH. Membership information was collected from PCTs during January, February and March.