Why we use open referral analysis
Analysing referral data openly among partners can save money by reducing the number of unnecessary tests and improving the consistency of clinical decision-making, writes Dr Richard Jenkins
Analysing referral data openly among partners can save money by reducing the number of unnecessary tests and improving the consistency of clinical decision-making, writes Dr Richard Jenkins.
GPs' gatekeeper role in referring patients to non-urgent NHS services that cannot be delivered in primary care has led to referrals being targeted for resource savings.
Unfortunately, this has led to some PCTs establishing referral centres to redirect or ‘bounce back' referrals to GPs and expecting them to change their referral pattern as a result. In my opinion, this is a shortsighted and non-engaging way of managing referrals. Much better, I believe, is the concept of a practice or group of practices analysing its own referral data and acting accordingly. That is what we have done and continue to do at Salters Meadow Centre in Burntwood, Staffordshire.
We are a five-partner, 14,500-patient practice and decided to start analysing our referral data when we began to implement Choose and Book. We saw great potential for gathering information that could be used in our directed enhanced service (DES) plan for PBC, for the ongoing education and interest of the partners, improving our patients' experience and journey, and increasing the efficiency of our practice.
We had no previous experience of analysing our referral data and started from scratch, following a plan-do-study-act (PDSA) process.
Step 1: plan
We decided the partner responsible for organising and monitoring Choose and Book, who is also the IT lead, was the sensible choice to be project manager. He was given the role of developing the framework for the referral analysis and engaging the secretaries to take on the required data input.
All the other partners agreed to have their referrals scrutinised and recorded for open, non-anonymised discussion at practice meetings. We all felt open data was vital for this project to succeed and this was helped by our longstanding ‘no-blame' and ‘just' culture, which encourages open information and debate.
The initial task was developing an Excel spreadsheet onto which the secretaries would enter all the referrals being analysed from all the partners over a set period. We use a similar system to monitor our Choose and Book data. If you have Excel, there is no cost implication in setting up these spreadsheets, although there is a time commitment from the project manager and the secretaries that needs to be recognised.
Step 2: do
So how does the scheme work in practice? The first thing to agree is which referrals you are going to analyse and we decided to start by breaking them down by specialty. Then we had to decide which specialty to analyse and agreed this at a practice meeting. This is an important step and should be based on any local or national data you can obtain.
Your PCT should be able to give you referral data for your practice or you could audit your own referral data over a previous time period to highlight specialties that generate a lot of referrals or where there appears to be room for efficiency savings.
There is little point analysing a specialty that only has a couple of referrals a year. By identifying the ‘low-hanging fruits' you should get a few quick successes to boost morale and encourage further engagement and participation. Our referral selections were based around the specialties of breast surgery, dermatology, ENT and urology.
I will use our urology analysis as an example of how this system works. Over a month, all the GPs referred in their normal way, using Choose and Book and dictating a letter on to tape. The dictation would then be sent via Choose and Book by the secretaries but any urology referrals were also entered on the Excel spreadsheet and the letter printed off and held in a file.
At the end of the month, the spreadsheet date was extracted and all the urology referrals broken down by each GP. In addition, the project manager was given all the referral letters and read them. From the letters he tried to identify why the referral was being made and what the GP was wanting from the referral. This process took the project manager about two hours.
Step 3: study
At the next practice meeting the urology referral analysis was presented openly.
A written summary was included in the meeting agenda so GPs could read it beforehand, and then a verbal presentation was given at the start of the meeting.
The majority of the referrals were of high quality but reviewing your own letters and comparing them to the letters from other partners is a real eye opener. The first thing we realised is that it is often not clear in the letter what GPs actually wanted from the referral. Did they want a diagnosis, access to further investigations, a management plan or access to surgery?
This is very important in PBC because, unless you are clear about what you want from the referral, you might end up paying for things you did not want or need and your patient might end up having things done that they didn't want or need.
Another question was whether the GP had put their differential diagnosis in the referral letter and if not, why not? Is it a lack of confidence or a lack of experience, highlighting an educational need? Agreeing these basics is important as creating an agreement between the partners as to what constitutes a good referral and what we expect of each other for future referrals.
From our month of data it was noted that a locum had sent two referrals for epididymal cysts. The partners all immediately agreed this was probably unnecessary but this then opened a debate on how we manage such patients in the surgery.
One partner never organised an ultrasound scan for a scrotal lump not attached to the testes because he thought it was unnecessary, one sometimes did and three always did, citing patient concern as the main reason for doing so. So, although we all agreed we would not have referred those two patients to a urologist as the locum did, the majority of us would have referred them for a diagnostic test.
Step 4: act
To resolve this, the partner who leads on clinical governance was asked to research the clinical guidance – using the internet and any guidance from NICE or medical royal colleges – on when to ultrasound and/or refer scrotal lumps. As a result, we have all agreed that scrotal lumps not attached to the testes do not need imaging and we will not refer such patients in future.
So, a month of simple urology referral analysis highlighted unnecessary referrals for ultrasound, a lack of uniformity in the practice for managing a presenting complaint and a lack of knowledge of clinical guidelines. As a result of our open discussions, we have now reduced our diagnostic referrals, reduced unnecessary time and tests for our patients and improved our own learning and experience.
We have found similar positive results in the analysis of breast surgery, dermatology and ENT. Our referral analysis will realise financial PBC benefits, efficiency benefits for the practice, educational benefits for the partners and patient benefits.
There have only been a few problems in developing this process. Our secretaries are very busy and were having to get used to the Choose and Book system at the same time as we were asking them to undertake more work with the spreadsheets. Their work needed to be acknowledged and appreciated via regular update meetings. Additionally, the project manager's time needs to be acknowledged by the partnership and accounted for in the sharing of the practice non-clinical workload.
To conclude, referral analysis in our practice works, is interesting and has shown clear benefits for us, so I encourage other GPs to give it a try.
60 Second summary
Initiative: Open peer analysis of GP referrals via monthly meetings, concentrating on one specialty per month
Policy link: Can identify any inefficiencies in referrals, so allowing a generation of future savings for PBC via better quality referral practice
Set-up time: One month or less, to develop Excel spreadsheet on which secretaries enter referral data, to enlist staff support, and to use existing data to decide which referrals to analyse
Set-up costs: Nil
Ongoing time: Two hours per month for referral analysis; 20 minutes per month to resources required discuss analysis in practice meeting; ongoing commitment from secretaries to populate spreadsheet and file letters for analysis
Outcomes: Improved referral quality and consistency, reduction in referrals and diagnostics, and beneficial GP education
Saving: Not yet realised due to early stage in project
Contact: Dr Richard Jenkins, email email@example.com
Analysis of urology referrals
• Monthly data revealed that a locum sent two referrals for epididymal cysts
Discussion between all partners
• All agreed this was probably unnecessary
• One partner believed an ultrasound scan for a scrotal lump not attached to the testes was unnecessary, one partner sometimes did and three always did, citing patient concern
• Practice clinical governance lead GP researched clinical guidance on when to ultrasound and/or refer scrotal lumps
• Guidance confirmed that scrotal lumps not attached to the testes do not need imaging
• All partners agree not to refer such patients in the future