The NHS needs more data on activity if it is to address the crisis facing general practice

General practice is in crisis – that much is clear. But, despite the increasingly loud voices of GPs, the crisis has – up to now – been all but invisible to national policy-makers.

While the pressures facing acute hospitals are obvious through almost real-time data on activity and performance, by contrast there is almost no nationally collected activity data – and certainly none available in real time – for general practice (or for community and mental health services). It’s unthinkable that information about activity in A&E would only be available through retrospective research studies examining data that is more than a year out of date – yet that is exactly the case for general practice.

Currently, there is no routine public reporting of GP activity data and no standardised national dataset. The only data available is extracted for secondary analysis from the research databases developed in collaboration with the three main providers of GP clinical information systems (TPP, EMIS and In Practice Systems). National bodies have been using an extrapolation of data published in 2009 from QResearch (taken from practices using EMIS clinical information systems) to estimate growth in the number of GP consultations nationally. A recent National Institute for Health Research-funded study, published in The Lancet, used data extracted from the Clinical Practice Research Datalink (CPRD), and the Nuffield Trust has also analysed data from CPRD.

Over the past nine months we have been working on a project designed to address the data deficit in general practice and have now published our comprehensive analysis of the extent and causes of the pressures in general practice. We used a combination of quantitative and qualitative research to create a detailed picture. We’ve analysed 30 million patient contacts taking place over five years from 177 practices using data from ResearchOne, a research database drawn from TPP’s SystmOne; carried out a workload survey of 43 practices for week in October 2015; conducted detailed interviews in four diverse practices; and surveyed more than 300 GP trainees.

All this means that we have been able to assess how much activity in general practice has increased since 2010 and what the nature of that activity is. Our analysis found that the number of consultations taking place in general practice grew by more than 15 per cent between 2010/11 and 2014/15. By far the biggest growth in activity was in contacts with people aged over 85 who have increasing numbers of chronic conditions. We also explored other factors that are having an impact on general practice, such as: people presenting with increasingly complex needs; changes in acute, community nursing and mental health services; and the introduction of new services and treatments. We also looked at ‘supply-side’ issues such as changing work patterns, methods of funding and commissioning.

However, while we were able to identify some interesting trends and assess some areas of activity, the main quantitative part of our analysis was challenging. All the clinical information systems used in general practice are designed for practices to use in their everyday work and not for the purpose of data analysis. Unlike secondary care, there are no national standards for data entry about activity, which means that when researchers look at the data they need to make lots of assumptions about how practices code their data and, because of widespread differences in approach, some data just isn’t useable. For example, information about the average length of appointments is notoriously difficult to assess. When we looked at our data, a typical contact appeared to be unrealistically long. Having spoken to staff who used the system, we discovered that they often left the appointment record open until any administrative tasks had been completed, even if this was some time after the consultation had actually ended, thus giving a misleading record of consultation length. So we couldn’t draw any conclusions about changes in consultation length over the study period and have not included this area in our report.

Our report provides a comprehensive analysis of changes in activity between 2010 and 2015, but without an ongoing understanding of changes in activity and demand or supply, neither national policy-makers nor local commissioners can be sure whether general practice has the capacity to allow people to access care when they need it. NHS England now urgently needs to consider how it can get a comprehensive and ongoing picture of the workload and capacity of general practice so that the challenges facing general practice can be robustly addressed.

Keep up to date

Subscribe to our email newsletters and follow @TheKingsFund on Twitter to see our latest news and content.


#547570 Occam's Razor

Some good work here and it certainly gets closer to the nub of the problem than anyone else previously.

It truly is relentless and the decision fatigue is what gets you. My PC resembles a battle station when I am on-call. This increased intensity of potentially life altering decisions is why 5 days at the coal face is probably dangerous.

I would also like to address one of the proposed solutions that I feel is significantly contributing to the problem rather than solving it and that is the increasing use of 100% phone triage. Despite what you may read GPs completely hate doing it and it increases the decision density still further. I firmly believe it is no coincidence that the massively increased rate of GP burnout is linked with the increased use of these phone triage systems. I certainly have seen a lot of GPs who have left practices because of these systems.

I also significantly worry about the GP training schemes. They are not equipping the trainees with the skills for the modern workplace.

The number one priority should be retaining the current workforce - of that there is no doubt.

#547575 Dr J

As a GP in an area where the service appears to be in meltdown, I am relieved that this report is getting some more publicity for the crisis.
However, I am not at all sure about some of the content of this post.

Firstly, we don't need yet more data. GPs and their staff have been collecting more and more data for the last 10 years, but most of it is worthless without the resources to act upon the data gathered. And politicians love to have the excuse of gathering more data to avoid having to act on the data already well known to anyone who wants to see it: GPs cannot continue to manage the current workload, and are increasingly left with no choice but to vote with their feet.

Secondly, please do not recommend approaches to the crisis that have not already been proven with robust research. I have only limited experience with health coaches but so far have been far from convinced that they can reduce the number of GP consultations. In fact, some seem to delight in finding "new" problems, and are very poor at assessing what problems are urgent, yet a patient told by their health coach that they need to see their GP urgently is seldom pleased to be told that the problem is not in fact an emergency.
Experience with the 111 service sugegsts that high numbers of patients are sent to A+E, often by 999 ambulance, by a risk-adverse system, but that despite this serious and sometimes fatal illnesses are missed due to the inexperience of staff - so where are large numbers of experienced health coaches going to appear from?

The 111 issues also show what any clinician could tell you - telephone (or email) triage, as also being advised, is extremely difficult to do safely. So much of communication is non verbal, and thirty seconds of watching a poorly child can often tell us more than ten minutes of discussion by telephone.
Of course, GPs have already passed on at least as much of the workload that Can be safely done by triage or by other staff, and I am sure that many of my colleagues share my fear that due to time cosntraints we often overdo this. Tasks that twenty years ago would have been done by the GP have long since been passed to practice nurses, and many of those on to health care assistants. Of course, this also has the effect that the average complexity of the GP consultation increases, so not only are we doing more consultations, we are doing more complicated ones.
I am afraid that many of the "solutions" being suggested for the crisis are ones which, in reality, are liekly to make general practice even less attractive.

#547576 Dr J

Discussing this with my wife, I have just been reminded of another example of the limitations of triage systems and inexperienced ancillary staff. At the start of the swine flu outbreak, rapid testing showed that only a minority of those with significant symptoms actually had swine flu. However, from the reports we received, virtually everyone who phoned the triage service with any sort of viral symptoms received a prescription for tamiflu.
Now, it may be that this scattergun approach to dishing out the tamiflu Did play a role in preventing a worse outbreak, but I have my doubts. It certainly was extremely poor at distinguishing likely swine flu from minor upper respiratory tract infections (but did have the advantage for our workload that everyone with a minor viral illness phoned the advice line rather than seeing us, and got tamiflu for their coryza. Of course, it must have been good treatment, as it had a 100% cure rate....!)

#547593 Mike J

Yes, the Great Tamiflu Giveaway of 2009 was very Bad Medicine. As is telephone triage when used to avoid seeing people face-to-face....

#547596 steve black

It is good that the King's Fund are now highlighting the importance of having the right data in order to understand the problems of primary care. But it is critical that the right data is collected: that is is needs to be insightful enough to help decide how to change the system where the system has problems and to identify what the problems are in the first place.

Narrow thinking about problems limits the ability to recognise effective new ideas. For example, telephone triage is unpopular with many GPs and the evidence for its benefits is mixed. But this is largely because the wrong question is being asked. For example, Telephone Triage itself does nothing to improve the workload for GPs or the satisfaction of patients. But phone triage combined with rethinking how the actual needs of patients are met by practice activity seems to be more effective. The issue is being more flexible with meeting patient needs than the assumption that every appointment requires 10 minutes. Rethinking how the service operates, driven by the data from triage, can lower workload and improve satisfaction. Triage alone isn't enough which is why testing whether phone triage itself works often shows no benefit.

Nationally we need better data about what is happening. Locally GPs need better data about what the nature of their demand actually is (e.g. do all patients needs 10 mins with a GP?). Combined, these can help the centre plan for the future and help practices match their capacity to patient needs.

Add new comment