36th John Perry Prize
Dr Neil Paul: General Practice Business Intelligence tool called Apex
The 2016 John Perry Prize was awarded to Dr Neil Paul, GP Sandbach for his work on the General Practice Business Intelligence tool called Apex.
GPs record and code most of what they do. Not just who they see, but what was said, what was done, their thoughts, their actions, and their plans. However, despite recording all this data there is little information on what GP are actually doing behind their closed doors, that a partnership or GP federation can use to help them.
There have been some great data mining successes from a clinical point of view. The QResearch database has generated new algorithms on risks of outcome for patients based on regression analysis of potential factors. Tools displaying QOF performance have driven up outcomes, numerous tools have looked at medicines usage, warning of safety issues and driving cost effective prescribing. While these are all great and focus on patient outcomes they don’t look at who is doing the work and how effective they are.
They also don’t acknowledge that general practice is a business, a business that is going through a lot of changes. It has had little increase in its funding over the last ten years and has had to find efficiency saving after saving despite increased demand. It is also facing perhaps one of the biggest changes to the NHS for some time, the formation of GP federations and super-practices, leading to multi specialist community providers and eventually the creation of accountable care organisations which might change general practice for ever.
GPs and managers to achieve these changes of continued efficiencies and joint working need information and tools to assist them.
I have been a GP Partner for 16 years in Sandbach, Cheshire. In that time I have been a PEC member on the PCT, IT lead for the PCT and CCG, Urgent care lead for the CCG, been involved in PBC and was involved in the setup of our CCG. I decided to jump from commissioning into helping setup GP federations. Locally I managed to get all 28 practices to sign up to form a federation and with the ex-PCT Chief Exec I setup a consultancy company called Howbeck Healthcare Limited to help practices form federations using our not for profit model. We have setup 3, and worked with a number of others and helped develop a range of innovative schemes helping win them over £3Million in new income.
In our work with federations IT is at core. We instituted a health-check program trying to get practices to use their IT better and to work out what problems people were having. We commissioned a collaborative Q&A website called GPask. I got the CCG to hire its own IT manager and also got the federation to hire one as well making IT more central to policy and procedure.
As part of my conversations with practices I recognised the need for tools and information for improving business processes when I kept coming across questions such as:
- How we know who is doing the work?
- How do we know if we are getting our monies worth from our salaried doctors? Or from our nurse practitioners?
- Who are they seeing? What affect has it on who I see?
- Do we see different patients?
- Am I seeing the old crumbly ones and working harder?
- Am I seeing the younger acute presentations and having to refer more than my colleagues who just see the same few patients over and over?
Some questions were less hostile such as:
- Is there any evidence of what I do that I can take to my appraisal?
- What if I work across multiple sites?
I captured these questions and looked at how we might answer them? Despite experience in writing searches and protocols in our clinical system I felt these tools weren’t enough. They tended to generate list of patients rather than useful information and there are some difficulties in distributing the searches across practices. I had a look at some of the existing tools and none were quite right.
I have an interest in Statistical Process Control methodology from my days as a PEC member and IMT lead, I had got my PCT interested in funnel plots and run charts. I felt that what was needed was an online tool that was designed from scratch with IG built in.
It should be available on the desktop of every clinician and manager and give them customises access to information based on their role. It should be able to work across practices and generate comparative views in intelligent ways. While allowing for customisation, my previous experience with some tools is that interested users get lost in the endless possibilities and inexperienced users don’t understand them so a set of standard reports well thought out and offering simple information would be the key.
The key would be accessing data from our clinical system as near to real time as possible and utilise the benefits of the centralisation of systems onto super servers to enable data capture. I felt one of the key principles should be that the data must remain in the practices’ control. Though they would have the ability to share it or bits of it with others particularly to reduce any reporting burdens they might have.
I’m lucky to be in contact with a large network of people in the industry - my monthly column about primary care IT on digitalhealth.net probably helps. I was chatting to David Stables about what he was up to and mentioned this idea. He introduced me to Edenbridge, Ex-Emis programmers with the right skills to deliver this product. I got my local GP federation to agree to be the test site in return for free access.
The journey has not been easy. Initially we were reliant on the new data extraction service from EMIS. I think it’s fair to say it’s had teething problems. We have had to move to a client API model for most of our initial data with the data extraction adding in info as it works. As part of the partner program particularly for new services we have had to go through a lot of testing.
However, we now have a working product. Its collecting data from 6 practices. We are rolling out to all 28 in our CCG and the initial feedback from practice managers is very positive. We are talking to 55 practices in neighbouring Stoke CCG who want to use it to look at their access and our current concept is to have an early adopter panel of 300 practices who help develop the dashboards and input into what is needed in future versions. We are also looking to make it work with the other clinical systems.
An example of where Apex is helping, is locally our practices are part of a Wave 2 Prime ministers challenge fund bid. Apex has been able to show which practices are opening extra, when they are opening - and the effect it has had on their overall access. There is a reduction in 3rd next available appointment whenever practices open up capacity.
Although its early days we appear to be able to dispel the myth that the work is endless. A lot of GPs will say “if I just offer more and more appointments they will be filled” we have evidence that happens to a certain point and then you have empty slots. You are in effect seeing everyone who wants to be seen.
Apex is also starting to show the effect of sickness, approving last minute leave, days when not quite enough people are in.
Of course Apex is showing some fascinating variances in practices. There is huge variation in the use of appointment systems. Some practices are very chaotic for what initially appears to be no good reason. This has developed the thought that there is a need for consultancy advice on appointment systems. While the NHS has run projects on GP access in the past, simple searches of the internet show that they included a lot of manual data processing to understand demand and capacity. We are speaking to several organisations as to whether Apex can help them or indeed whether we need to setup some form of consultancy ourselves.
Reading the BCS website about John Perry - there is the following quote: “He continued to develop both the OXMIS codes and the GP system, with programs to help GPs manage their practices as well as the collection of clinical data for research, which was the original objective. Few GPs at that time had access to computers, and this work done with the collaborating practices pioneered much that was later transferred to commercial systems with widespread implementation.”
I believe that Apex is an extension of this principle, of helping GPs manage their practices, in addition I’m currently speaking to the University of Liverpool, School of Management, where I am an honorary Senior Lecturer about whether we can jointly bid for some Wellcome Trust funding to explore Apex’s ability to help in other areas of general practice - our initial idea is around how to use it to affect individual clinician’s behaviours.
I believe John would approve of my efforts, be keen to be involved in this project and I would be grateful if you could consider me for the John Perry prize in particular for my role in the development of the BI tool Apex.
Dr Neil Paul MB CHB MBCS GMC 4091424
GP Partner @SandbachGPs, Middlewich Road, Sandbach, Cheshire, CW11 1EQ