The problem
High-referring practices were pushing up costs in our clinical commissioning group, particularly after a partner at a small local practice took long-term sick leave and eventually resigned. High locum turnover meant that while the practice was not struggling, referral rates were too high.
Another local practice, Staffa Health – one of the largest in Hardwick CCG in Derbyshire with a patient population of 15,000 across four sites – had been using peer review internally for two to three years. We recognised that rolling out the system across the CCG could reduce our referral rates and their associated costs, by addressing performance at those practices that fell short of targets.
What we did
Our scheme had two guiding principles. The first was that the review would result in a benefit to the practice and the CCG. Our second principle was that this is not about blame and recognises that there are many reasons why referral levels differ. They may indicate a need for information, for extra training or for commissioning of new services. The aim is to identify the underlying drivers for the present situation and then make a plan to address them.
We did have variation between practices, but generally speaking the practices were not high referrers. However, we knew that in trauma and orthopaedics we were collectively nearly 19% higher than might be expected when looking at comparative data from the rest of the country, adjusted for deprivation and demography.
Practices decided to bring GPs, nurses, advanced healthcare practitioners and practice managers to a workshop event. Each practice brought two or three trauma and orthopaedics referral letters. Practices were distributed between tables so that there was a mix at each table. The nurses and advanced healthcare practitioners formed a group of their own. A GP acted as facilitator for the table and we used GPs from a practice who had been doing events for some time so they fully understood the objectives and sensitivities and how to make it a positive experience.
The format of the event was first to understand each practice's referral profile; second for each practice to read out their own referrals and for the table to share with each other how they would handle this particular situation. The process was to share referrals and use them as prompts for discussion, think about outcomes of referral, and identify any gaps or areas for change in terms of:
• information needs
• training needs
• commissioning needs
• need for guidelines – practice or other.
The third element is to take action. In this case there were a number of commissioning queries that were taken up with the contract managers. It led to a joint health training programme, especially regarding podiatry and podiatric surgery, use of injections (for which a consultant gave a short training session) and a re-design of the carpal tunnel syndrome pathway in which the whole group took part. Staff with experience giving joint injections offered training support to help other practices develop and offer injections. Now most practices have a joint injection service.
Individual practices were then encouraged to tackle specialties where they were outliers, such as ENT, dermatology and gynaecology. One small practice tackled its ENT referrals by having a GPSI in another district look through one year of referral letters and clinic outcomes and produce a report and recommendations that they then implemented as a practice policy.
Achieving target referral levels was incentivised. The target this year was arrived at by working out the CCG average per 1,000 of weighted capitation for all GP referrals. Practices are then rewarded 20p per registered patient for referring below that level. Every practice in the scheme is funded 10p per registered patient up front; where they achieve targets, they then receive a second 10p per patient. Improved performance gets practices a further 5p per patient, but deterioration earns no further payment. This is part of a bigger scheme that also includes non-elective admissions and prescribing, and in the coming year we plan to reward outcomes for our target area of diabetes.
Lessons learned
The greatest benefit has been that clinicians learned more about each other and a culture of trust developed. One practice doing well on dermatology said they had been keen to improve, but this was the first time anyone had helped them.
Peer reviews have been flagged as a ‘points-scoring' opportunity by some GPs elsewhere, but through clear communication and a professional approach doctors can accept that peer review is not a ‘blame game', but an opportunity to share and build expertise.
Registrars and trainee GPs, while often less confident, can bring useful new clinical information from recent hospital rotations that more experienced GPs may not yet have discovered.
We keep in mind that low referral might be under-referral, but use peer review to ensure patients aren't denied the care they need. Now that we have been doing this for a few years we are low referrers overall and there will be limited further savings from focus on this area beyond ongoing monitoring. We understand we are to have historic activity-based budgets, so we're worried our work as an early adopter will have put us at a disadvantage and our low referral rate will bring a lower budget.
Outcomes
We have the lowest cost per patient in our cluster. Six of the 10 lowest referrers in the cluster are in our CCG, with none above the average. Trauma and orthopaedics expenditure in 2010/11 was 17% less than in 2006/07 when we conducted our first workshop. One practice focusing on ENT cut referrals by 20% in the first year and has now reduced them by 40% compared with year one.
In the case of dermatology we are funding one of our GPs to run up to six sessions in three of our practices. The first practice in the scheme reduced dermatology referrals by 25% in the first two months, and referrals after six months are now 38% lower than their base year average monthly figure.
One practice that joined the CCG more recently has quite a challenge to reduce referrals generally, beginning with dermatology, which shows the highest variation. A GPSI in dermatology from another practice offered it training sessions. The whole team takes part in consultations with patients who have been recruited to the scheme and the specialist GP sits in. The practice reduced referrals by 25% in the first two months and is maintaining that level of success.
Results were similarly good for gynaecology. One GP reported making 104 referrals between June 2010 and May 2011, 63 of which (60.5%) were deemed appropriate when analysed. This helped the GP create an action plan, based on the remaining referrals the GP agreed could have been managed in primary care. This included nurse training, better use of the menorrhagia pathway, extra CPD on some conditions and requests for a consultant to drop in on the practice's menopause clinic.
The future
We've transferred our workshop methodology to case studies in A&E and non-elective referrals, which is enabling us to identify lots of commissioning issues in those areas.
Moving on, peer review success won't be incentivised as it has been this year across the CCG, with the focus shifting from referrals to outcomes and pathways – their successful design and how well practices adhere to them and encourage patients to use services for lifestyle change. However, some practices, such as Staffa Health, will continue to review cases internally, having discovered the benefit to both GPs and patients. Targeted areas for peer review in 2012 include surgical, gynaecology and dermatology referrals.
We have adopted this system as an alternative to a referral management centre because we believe the outcomes are superior. Our system provides a learning opportunity along with clinical support. It provides direction for pathway development and a better outcome for patients.
Dr Ruth Cooper is a GP at the Staffa Health practice in Tibshelf, Derbyshire, and Wendy Sunney is the chief operating officer of Hardwick Health CCG
