Help GPs make the right decisions first time

General practitioners are trained to manage risk in the community and act as a filter for those who may or may not require specialist help. The best way to manage referrals? Enhance this existing resource. Don’t try to second guess GPs by implementing unnecessary and costly processes.

‘We must accept that GPs’ referral decisions are imperfect but the degree of imperfection should not be overstated’

The cost of secondary care referrals, increasing demand and pressure to meet targets mean the referral process needs to be as efficient and accurate as possible. Referral management centres are a response designed to offset such pressures but I am not a fan of them for a number of reasons.

GPs are trained to decide when to refer appropriately. This choice is influenced by a myriad factors, including the patient’s views, social circumstances, previous history and the GP’s areas of clinical expertise, , as well as the condition the patient presents with; they have all the information to hand and are in a trusted conversation involving the patient.

Referral management centres attempt to second guess this complex process and require costly infrastructure to do so, running to many hundreds of thousands of pounds per year.

We must accept that GPs’ referral decisions are imperfect, as is any complex human process. However, the degree of imperfection should not be overstated. Typically referral management centres report overriding the GP’s decision in only 5-10 per cent of cases; they don’t, however, usually report how many of these “override” decisions later turn out to be incorrect. Anecdotally this is a significant proportion, with patients being directed to treatments that have already been tried or even ending up back with the GP asking to be referred again.

Building confidence, improving performance

Perhaps more worryingly, the process of second guessing expert decision makers is likely to discourage people from using their expertise, leading to a lower level of performance in the long term. The solution is not to override GPs’ decisions but to support them to be incrementally better in making their decisions in the first place.

‘Providing access to appropriate diagnostic investigations and guides to local services gives GPs the tools to make better choices first time’

Reviewing the referral patterns of GPs is key to building knowledge. Feedback loops such as peer review, comparative performance data and reflective practice can improve GP referral skills; these are low cost measures that will improve GPs’ performance in the long run. We should feed back directly to GPs on their referral rates and performance, and keep them up to date with locally available services.

Clinical commissioners regularly review performance with their GPs, including looking at how their referral rates compare with expected levels of activity. We build the activity plan, taking account of policy changes (for example, we expect dementia referrals to increase in light of the recent push on dementia diagnosis), local demographic data, the deprivation index and disease registers.

Enforcing perfection

When we audit referrals we find that only 2-5 per cent of them are of unacceptable quality − and this mainly stems from incomplete data (perhaps an incomplete history or insufficient investigations in primary care). Referral management centres don’t solve this problem.

For this small percentage of imperfection it is much more efficient to support GPs in getting it right the first time, rather than checking 100 per cent of the work to see whether it has been done right. Providing access to appropriate diagnostic investigations and guides to local services, at the point of care, gives GPs the tools to make better choices first time.

We don’t need a duplicate system trying to enforce perfection. We need to acknowledge the imperfection of decision making in a complex risk management situation and look at how we can support our frontline experts − our GPs − to improve their already good performance.

Dr Shane Gordon is clinical chief officer at North East Essex Clinical Commissioning Group. This article also appears in a “Map of Medicine” editorial paper