The 15 per cent growth in GP referrals to hospital in quarter one 2008-09 is putting pressure on both providers and commissioners (see first graph).



Suggested causes include population aging, consumerism or patients returning from the private sector, but these are long term trends and do not explain the rapid change. More plausible is some change in GP behaviour in response to a change in guidelines or to disgruntlement. But not all primary care trusts have seen this increase.

One hypothesis is that PCTs and trusts have been clamping down on consultant to consultant referrals, PCTs to control demand and trusts because there are incentives associated with the 18-week target.

Referral thresholds

A lower referral threshold is a possibility. The Q1 increase represents less than one extra referral per GP a week on a baseline of five to six a week. This would be hard for practices to notice in the short term and unlikely to lead to noticeable change in the nature of the referrals. In many places waiting times are falling as the number of first appointments rises and reattendance rates fall (see second graph) and as trusts work towards the 18-week target. Because the supply of outpatient slots appears to be increasing this could induce changes in referral behaviour.

One problem is that there is a tendency to look at the data on a year on year basis. Over a longer period the increase in Q1 looks much less striking. Taking Q1 2005-06 as the baseline, there has only been a 6 per cent increase in referrals in the last three years. So recent growth could be from referral management that has not managed to hold on to gains, or a bounce back from turnaround measures.

Some areas have contained or reduced the growth in referrals. With negative growth, Kingston PCT had the lowest rate in London (see third chart).

GPs are key

The key to its success appears to be that the referral management system is GP led and has rapid communication and feedback from a GP. It also looks at consultant initiated referrals. The feedback focuses on helpful suggestions or alternatives to referral - partly because it is difficult to remember all the different routes to treatment when each GP refers 250 patients a year to more than 40 services. A key lesson is that this type of approach needs to be continuous to keep the initial gains. This suggests that practice based commissioning is likely to be most effective where it is combined with machinery to ensure that its decisions stick, is focused on key parts of the pathway and is clinically led.