Feature General Practice

Bigger is better for primary care

BMJ 2011; 342 doi: http://dx.doi.org/10.1136/bmj.d1754 (Published 23 March 2011) Cite this as: BMJ 2011;342:d1754
  1. Nigel Hawkes, freelance journalist
  1. 1London, UK
  1. nigel.hawkes1{at}btinternet.com

A new report into the quality of primary care finds standards are generally good but highlights the need for more cooperation both between practices and with secondary care, Nigel Hawkes reports

“At its heart, general practice in much of England remains a cottage industry” declares a new report from the King’s Fund—a judgment that, if true, questions the wisdom of putting the entire NHS in the hands of general practitioners. Are “autonomous artisans” who relish isolation and resist comparison with their peers really cut out to be the vanguard of a revolution in quality for the NHS?

The question is well worth asking, though it is less clear that the report provides the answer. General practice is often praised, less often examined. Partly, as the report makes clear, it is because of a lack of adequate data, partly a reluctance to dig too deep into the variations in the quality of care that general practitioners provide. Every GP, it is safe to say, knows of a local colleague who may not be up to scratch. Sometimes, they can even be heard muttering about it. But let’s not go there.

The report, Improving the Quality of Care in General Practice,1 was produced by an independent panel latterly chaired by Ian Kennedy. He was joined on the panel by Michael Dixon, a Devon GP and chair of the NHS Alliance, Steve Field, former chairman of the Royal College of General Practitioners, Ursula Gallagher, director of quality, clinical governance and clinical practice at Ealing Primary Care Trust, and Rebecca Rosen, a GP and senior associate of the King’s Fund.

Performance matters

Why bother? Sir Ian explains: “Most activity in healthcare takes place in general practice, but analysis concentrates on the acute sector. Quality in primary care is somewhat neglected, and very difficult to measure when the interactions are so complicated.

“The good news is that in terms of performance the standard is generally good. But we need to understand better what is meant by proper performance, how much it varies from practice to practice—and within practices—and what needs to change in future.”

The inquiry began by commissioning 10 research projects and four discussion papers, most of which are available on the King’s Fund website (www.kingsfund.org.uk/). The report pulls together their findings, looking at the quality of care currently delivered by GPs, how data available or potentially available could improve quality measurement, what practical measures might be taken to promote quality improvement, and how general practice needs to change in future.

The care provided by GPs varies in ways that are sometimes important and sometimes less important. There are believed to be big variations in the quality of diagnosis both within and between practices, especially for acute illness, but it is hard to establish just how big they actually are. One source of information is significant event audits, when practices review unusual events both good and bad to discover what might be learnt from them; another is examining claims made by doctors to the Medical Defence Union and the Medical Protection Society. The commonest failing that emerges is diagnostic delay.

Referral rates vary considerably, with major implications for NHS spending, but there seems to be little direct association between such variation and subsequent patient outcomes. Although some doctors refer some patients when it is clinically unnecessary, it may be hard to decrease unnecessary referrals without also decreasing necessary ones. Referral letters vary greatly in quality and usefulness.

Prescribing practice also varies widely, with clinical guidelines sometimes ignored. If all practices were as efficient as the top 25%, £200m (€230m; $320m) a year could be saved on statins alone, the report says, citing a 2007 study by the National Audit Office. But general practice in England has one of the highest rates of generic prescribing in the world (83% in 2008), and studies suggest that differences in prescribing are largely accounted for by case mix.

Areas for improvement

So far, this hardly comprises a serious charge sheet, but the panel is sharper in its judgments of the treatment of long term and chronic illness. It suggests that between half and two thirds of those who would be likely to have dementia are left undiagnosed, and less than a third of patients with osteoarthritis get the care in general practice that they should have had. All those needing long term care are supposed to have a care plan, but only 11% of patients report being told they actually have one, according to the general practice patient survey.2 “That’s not good enough” says Sir Ian.

Slightly more striking findings emerge from the research into non-clinical aspects of primary care. Patients are not too concerned about access or opening hours—though these vary widely, with only 42% of patients in the worst performing practices being able to book ahead for an appointment—and are prepared to wait a bit longer to see a GP they like. “It’s quite clear that patients like to see a particular doctor,” Sir Ian says. “They are prepared to trade waits for choice, and a large number are not getting to see their preferred doctor.” This aspect of care, relational continuity, has deteriorated slightly in recent years, the report concludes.

Patients are insufficiently involved with decisions made about their care, the panel believes. Such involvement “nurtures the therapeutic relationship” in Sir Ian’s words, and feedback is intrinsic in quality control—“a crucial aspect in how you are doing is what your patients say.” But it is easy to understand why some GPs take their patients for granted, given that over 80% of patients trust their GPs even in the 10% of practices that perform worst in the GP Patient Survey.

The problem of measurement

What levers are there to improve quality and narrow variation? The Quality and Outcomes Framework introduced in the 2004 GP contract is “useful but insufficient” in Sir Ian’s judgment. The framework covers less than 10% of the activities undertaken in primary care. He sees the answer in greater cooperation, more transparency, better leadership, and a greater willingness to confront poor performance. “There has not been a great deal of standing up to poor performance in the past” he says.

Will general practice commissioning provide a new way of burying poor performance? The panel warns that it may. “There is a risk that the membership arrangements and governance of GP commissioning consortia will allow practices to protect their members’ interests, rather than seeking to challenge and improve quality,” it says. “The accreditation process must be strong enough to ensure that robust governance arrangements are in place to promote internal performance management.”

But that assumes that adequate performance measures are in place. There are plenty of datasets, such as Indicators for Quality Improvement (developed by the NHS Information Centre), NHS Comparators (a set of 200 indicators made available to primary care trusts, acute trusts and general practices but not to the general public), and Practice Profiles developed by public health observatories. In autumn 2009, the report says, the Department of Health identified more than 20 indicator sets and quality improvement initiatives, not including local examples.

But what is lacking, Sir Ian says, are unifying analyses. The datasets provide “a multiplicity of partial insights into quality, but they are not effectively brought together.” So at the end of all the painstaking efforts the King’s Fund team has made, there remains a lot of uncertainty.

Pressure for change

The dominant model of general practice remains one of small, independently contracted businesses, the report says—the “cottage industries” it compares unfavourably with the industrial scale organisation it believes is needed. “Patients have much higher expectations today of what good performance is,” Sir Ian says. “The panel thinks that a radical change is required in how general practice works if it is to up its quality game.”

But if patients are happy and damaging variations in quality so hard to pinpoint, why propose dragging general practice into an industrial age that few seem to want? There is a risk here of an inappropriate referral, of failing to match diagnosis with the right treatment.

The panel believes that general practice can have it both ways, preserving its traditional virtues while gaining benefits of scale. It asserts that this approach is not about sweeping away small practices but recognising the benefits that come from being part of larger organisations or networks. “We believe there is an urgent need to bring isolated practices into more formal accountability structures, as part of larger provider organisations . . . nevertheless, there are aspects of small practices that patients value.”

Sir Ian says that many different pressures are driving primary care into a more systematic structure. Technology is one; only by the best use of information technology will practices be able to compare their results with others’. Transparency is another, because poorly performing practices have been able to hide their failings from public and patients for too long. “Greater transparency is needed to encourage challenge—that involves peer review and bench marking, and a greater commitment to use feedback from patients.”

Changing patterns of care are a third. “General practice must be a navigator and not a gatekeeper. The complexity of healthcare is such that a patient needs a guide, a trusted advocate. GPs should play that role.” The dual role of GPs as commissioners as well as providers of care, the shift of care from hospital to community, and the emergence of better informed patients all pose challenges that tomorrow’s GPs have to meet. So do new methods of providing seamless treatment through integrated care organisations for patients with chronic disease and complex needs. “The evidence presented to the inquiry strongly suggests that higher quality care in general practice will only arise when GPs and their teams begin to look beyond the care that takes place within surgery walls,” the panel concludes.

But its most resounding conclusion is a negative one. It set out to find new measures of quality in primary care and admits that it failed. “This was a difficult task and one we were unable to complete” the conclusions of the report acknowledge. Instead of measurement, it falls back on exhortation.

“If general practice is to focus on improving quality it must be able to demonstrate that it is doing things differently and better. It is no longer acceptable for professionals simply to claim there is no way of assessing what they do and that they should be trusted with the task.” But some GPs may feel that if one of the largest and most ambitious efforts ever made to define quality measures in primary care has failed to find any, trusting them may indeed remain the only option.

The state of general practice

  • Big variations in the quality of diagnosis within and between practices, especially for acute illness

  • Referral rates vary considerably. Referral letters vary considerably in quality and usefulness

  • Prescribing practice varies widely, with clinical guidelines sometimes ignored

  • Between half and two thirds of patients likely to have dementia are left undiagnosed, and less than a third of patients with osteoarthritis get the care they should have

  • Patients like to see a particular doctor and are prepared to wait to see the GP they like

  • Plenty of datasets but no unifying analyses


Cite this as: BMJ 2011;342:d1754


  • Competing interests: The author has completed the unified competing interest form at www.icmje.org/coi_disclosure.pdf (available on request from him) and declares no support from any organisation for the submitted work; no financial relationships with any organisation that might have an interest in the submitted work in the previous three years; and no other relationships or activities that could appear to have influenced the submitted work.

  • Provenance and peer review: Commissioned; not externally peer reviewed.