Is a new system needed for GP referral?

This content relates to the following topics:

Referrals are important to everyone. To GPs, deciding whether or not to make a referral is a core responsibility. To patients, the consequences are profound. And to taxpayers, the costs of GP referrals to the NHS came to £15 billion last year, and have been rising year-on-year for some time. No wonder so much research has been conducted on the area.

Getting under the skin of GP referrals is notoriously tricky. Researchers have invested much energy in assessing the quality of referrals, understanding variations between GP practices, and testing ways of improving referral processes. The evidence base is large and not always consistent, but we can say some things with confidence.

First, decision-making around referral is a complicated process, balancing several competing concerns and sources of information. There are many different reasons why a GP might make a referral, including the patient's need for reassurance or simply to 'do something'.

Second, there is wide variation in referral rates – up to ten-fold between practices. At least some of the variation is accounted for by non-clinical factors such as GPs' willingness to tolerate risk and uncertainty, sensitivity towards patient pressure, or fear of accusations of malpractice.

Third, the quality of some referrals could be improved. Referrals are not always directed to the most appropriate place, and there is evidence of late referral in specialties such as cancer or nephrology. Referral letters sometimes lack the details specialists need.

Fourth, we know that there are systemic barriers making it harder for GPs to make high-quality referrals. Importantly, many GPs feel it is increasingly difficult for them to contact specialists informally to seek advice before deciding whether or not to make a referral. 

It is clear that while many GPs make excellent referrals, there is considerable scope for making further improvements. But what should be done? Our research on the quality of GP referral and referral management sheds light on what approaches are most effective.

Firstly – what not to do. In the current financial climate there is a danger of becoming preoccupied with reducing the volume of referrals. An unsophisticated approach to this, for example setting target referral rates, could do more harm than good.

Several areas have established referral management centres which check all GP referral letters before directing them to the most appropriate destination, or rejecting them as unnecessary. These centres may have had some benefits but do not appear to be cost-effective and in some cases undermine referral quality.

Instead, an approach based on peer review and audit of referral patterns among groups of GPs, coupled with a system for harnessing feedback from hospital consultants, holds the greatest promise for improving the quality of referrals while also controlling costs.

The proposals in the government's recent NHS White Paper may help encourage this, by bringing GPs together into commissioning consortia and giving them the incentive to take responsibility for the implications of their referral decisions. As this new system develops, it will be important to ensure that GPs look at the quality of referrals rather than taking a more limited focus on controlling referral volume.


Jon Levick

Oakfields Health Centre
Comment date
13 August 2010
Thankyou Chris for this excellent report.
It mirrors the work we did within our local GP group (then Derwentside PCT) some years ago, which showed clearly the benefits of small group peer review of referrals in all areas, both from en effectiveness and efficency point of view.
I have always been dead against central referral management schemes that sieve out referrals... This devalues the complex decison making process at the the Dr -patient consultation.
Far more importantly, with small group peer review, the clinicians involved felt it was beneficial from a learning point of view. It not only changed individual behaviours, but also provided the drive for system changes to aid both the referral process and development of new or alternative pathways.

We had a rule that 'no referral was inappropriate', in order to get rid of the threat that a manager led top down referral review was there to bash GP's (or Nurse Practitioners).
More to discover why if 10 individual GP's were presented with a problem, there could be massive variation in how that problem was managed.. none of whish were incorrect, but brought into play all the complex issues that are taken into consideration at the consultation.
The consclusions showed that simple pathways and guidance only play one part of the decison making process, and that knowledge of alternatives, confidence of management of risk, skills and experience, patient demand and demands on GP time in the consultation all played a significant part.
The conclusion is that this complex decison making process cannot be controlled just by sending out new pathway diagrams or algorthims, but can be helped enormously by small group peer review and is typical of complexity theory learning.
However to do it requires strong clinical leadership, set aside time for Practices to committ to it (across Practices), strong feedback mechanisms and follow up of system improvements that arise out of it. In this age of financial cuts, have the organisations got the vivion to do it effectively?

Secondary Care input is invaluable at adding credibility to the process but the key to it is about GPs reviewing their own decisions making process in that core area of their own speciality... the consultation.

I would be happy to share the results of our work at the time.
Incidentally, for the sceptics and the commissioners, when done properly we showed that there was a consistent 30% altenative way of managing a problem that if you asked the Consultant directly (the endpoint of a referral) was accepted as 100% appropriate. It confirmed that the GP referral decision making is a complex process that requires structured peer review in a learning environment to improve.

Dr Jon Levick
GP, Oakfields Health Centre

Elizabeth Evans

former Medical Advisor and former GP,
Aneurin Bevan Health Board
Comment date
19 August 2010
Yes it is an excellent report.
We involved consultants from the start (in 6 weekly cluster meetings with GPs from different practices) and we found their input invaluable. However, as Jon says there are caveats. Only consultants who had an interest in whole systems approaches could appreciate that their expertise was not always required. We had at least one consultant who wanted many things referred which GPs were happy to manage themselves - fortunately in that speciality there was a reduction in demand despite this! The disagreement provoked a careful reconsideration of the guidelines recommended to GPs,by the Trust clinicians, and although a consensus was not reached on this occasion it meant that the problem was flagged up for further evaluation.
Sometimes consultants were not willing to take part in joint meetings, and this reflected a dismissive attitude to GP led decision making. This at times seemed very defensive, as though their comfort zone was being challenged.
Overall however the 6 weekly cluster meetings were much more constructive and enjoyable when consultants were there (as they were on all but one occasion over 2 years), and both sides learned a great deal.
We also invited other health care providers - people running alternative community based services and the tri-partite dialogue worked very well.
Don't underestimate the difficulties though. There are some GPs who do not believe that the effectiveness of the system as a whole concerns them at all - only the patient in front of them. They needed a lot of persuasion to come on board and really look at their referrals constructively, and see if there are better ways of dealing with the problem.

Rob Findlay

Comment date
25 August 2010
An excellent and timely report.

There was just one point at which I felt that different thinking might be helpful. The report says: "any referral management strategy needs to include a robust means of managing the inherent risks at the point when clinical responsibility for a patient is handed over from one clinician to another (so-called clinical hand-offs)"

I would argue that this accepts the concept of the "clinical hand-off" too readily. Referrals should not automatically be fire-and-forget, rather the referring GP should remain available as the patient's advisor after the referral has been made (and in rare cases remain actively involved, such as when the patient is "stuck" in a hospital bed or outpatient follow-up loop).

After all, patients must give their informed consent to every step of their treatment, and both the consultant and the GP have a role to play in informing them.

Richard jones

Consultant cardiologist, chief of medicine,
Portsmouth hospitals nhs trust
Comment date
29 August 2010
very useful comments. In secondary care we are seeing many patients who could be managed by GP's. Specialists, however, need to be far more available for informal discussion by telephone/email etc. We need to share the risk and uncertainty. Specialist time released by seeing less referrals should be directed to recreating the close ties that once existed between the two care sectors. The Facebook generation are better connected than ever but we have somehow allowed far less sophisticated software to drive a wedge between easy GP/specialist communication.

Dr Roger Davies

GP Partner rural Shrophire,
Meadows Practice
Comment date
21 November 2012
Interesting. As with others we are just about to attempt peer review of our referals in South Shropshire, 12 practices. Have the usual variation between us, and seem to have a qenuine willingness to look at our work,time will tell....will feed back in 2 months.

Add your comment