Jonathon Tomlinson reflects on the difficulties of measuring the quality of general practice and the prospects of improvement.
In recent years I’ve seen the NHS from a patient perspective. I’ve had a few fractured bones, been present for one son’s home delivery and another’s emergency admission after being run over. I’ve seen my dad admitted to hospital with pneumonia and dementia, and a friend’s daughter admitted with type 1 Diabetes. Quality care in the case of a fractured finger, a home-birth, a delirious older person and a child with diabetes vary so widely that it is almost impossible to compare one with another. For example, mortality rates are too low to help, diagnosis was obvious, self-management support not always needed, waiting times not really significant and so on. Nevertheless in every case the professionals involved had to spend a lot of time collecting data, and we (the family) were asked for feedback. I even made a complaint. I have no idea what, if anything, was done with the data, the feedback or the complaint to improve quality.
There is a problem in that quality is so hard to measure, in any meaningful and universally comparable way, that regulators and politicians have embarked on a Sisyphean task of measuring almost everything in the hope that somehow quality will improve. The result, as Don Berwick says, is “massive, ravenous investment in tools of scrutiny and inspection and control, massive investment in contingency, and massive under-investment in change and learning and innovation”. And he adds: “This isn’t going to work, we cannot possibly inspect our way to excellence” – a conclusion echoed in a report from the King’s Fund earlier this year.
The enormous burden of data collection and inspection in general practice, at both organizational and individual levels, leaves very little time or energy for quality improvement. In the last year my practice has had a Care Quality Commission (CQC) inspection, submitted vast amounts of data for Quality Outcomes Frameworks (QoF), Enhanced Services, Prescribing Audits, Friends and Family tests, etc. Every member of the practice has compulsory training and certification several times a year. At an individual level, last year I was revalidated and this year I have already been through two appraisals, one to be able to continue as a GP and another to be allowed to supervise a trainee. Every GP has an annual appraisal – which takes about 3 hours face to face, and a dozen hours at least in preparation. Recently we had to close the practice for 3 hours while we all had mandatory cardio-pulmonary resuscitation (CPR) training. A question that always comes up in appraisals is, ‘how do you know that you are providing good quality care / teaching?’ This question is not about quality improvement, but about quality control, and ultimately about control of the medical profession by regulation and inspection. It raises other important questions, such as ‘what aspects of quality can be measured?’ And ‘for what purposes are we measuring different aspects of quality?’
“Inspection does not achieve continual, pervasive, never- ending improvement. It doesn’t foster creativity or learning or pride, it poisons them, because the main harvest of inspection isn’t learning, it is fear.”
In 20 years of medical practice I have never been asked about or taught the methods of quality improvement, and this is the case for the vast majority of health professionals who are now regulated, examined and assessed more than ever before. And as Don Berwick also says,“Inspection does not achieve continual, pervasive, never- ending improvement. It doesn’t foster creativity or learning or pride, it poisons them, because the main harvest of inspection isn’t learning, it is fear.”
Data for patients as consumers
One justification that is often given for all this measurement is that patients are increasingly seen as consumers of health services, and therefore require information about the services available to them. But what information would this mean supplying in general practice? In general practice patients range from those not yet born, to those who are dying, and almost every stage of every imaginable disease is represented. One approach to this problem, illustrated by a 112-page-long 2015 Health Foundation report, is to use as wide a range of measures as possible. Another 2014 Health Foundation report on improving quality in general practice was commissioned to inform the design of the My NHS website, which was supposed to encourage patients to choose a practice, and practices to compete with one another. The website is even less helpful than the literature used to inform it. The NHS Choices website has been collecting data from millions of patient ratings of GP services since it was launched in 1998, most of which has never been analysed or published.
And there is little evidence that attempts to measure quality for the purpose of supporting patients’ choices and decisions are helpful. Different patients need and want different things from their GP: some value access 7 days a week, others want continuity with a GP who knows them well, or home visits, mental health support, specialist referrals, dementia care, sexual health services, and so on. Moreover, healthcare demands can change rapidly and unpredictably with the onset of serious illness, so the criteria you use when you register may be irrelevant when you find yourself needing antenatal or palliative care. I am fully in support of transparency, but data collection takes time away from other important activities and it is unlikely that we could ever collect enough to satisfy every type of patient consumer.
Quality assurance by meeting basic standards
Given the diversity of patients’ requirements and the variety of ways in which quality can be conceived, it would be reasonable to conclude that measuring everything that counts for quality in general practice is simply too difficult/ broad/ contextual a task. Yet there are some fundamental, universal, comparable, and where appropriate measurable, standards that every practice should meet, and which would give some assurance of its quality: things like good access to the building, immediate access by phone for urgent problems, safe storage of drugs and vaccines, safe waste disposal, hygiene and cleanliness, confidential record-keeping, an up-to-date practice website, and so on. I would also include providing patients with access to their records, accurate registers of patients with chronic diseases, routine clinical data, and records of serious incidents and complaints. And to provide quality assurance, all these measures should be published for patients to see.
Beyond such standardisable things, though, there are few other things that matter to patients that can also be considered ‘taken-for-granted’ standards of quality. It is not as if we cannot comprehend or describe quality in patients’ experiences; but they are so entangled in individual circumstances that they are very difficult to compare. This is why, as Angela Coulter pointed out recently in the BMJ, the Friends and Family Test is ‘a mixed bag of poorly evaluated methods that leaves patients frustrated, and doctors little wiser’. Greenhalgh and Heath’s paper on measuring quality in the therapeutic relationship concluded,
“although little useful or meaningful is achieved by trying to reducing experiences and relationships to simple metrics, we should nonetheless remember that underpinning good-quality therapeutic relationships are continuity of care, from a doctor the patient knows and trusts, and sufficient time – and these can be measured”.
As Coulter also concludes, we already know that patients value continuity of care, access, involvement in decision making and self-management support- all features of patient-centred care.
Quality as ‘patient-centred’ care
One way around the problem is to focus on patient-centred care. There is a general consensus that patient-centred care matters to patients. This term does not mean forcing unwelcome responsibilities on patients, or conversely, doctors having to agree to unreasonable patient demands, as has sometimes been supposed. Patient-centred care is about how patients relate to and interact with people, organisations and materials.
Good relationships between patients and people depend on a wide variety of factors, from simple courtesy, as promoted in Kate Granger’s “Hellomynameis”campaign, to knowing patients as people and sharing decision-making with them. It can be enabled by education, continuity of care and shared-decision making tools.
Good organisational relationships with patients depend on factors like making access to care easy, communication clear and helpful, and integrating care between different specialties and between different organisations, for example between hospitals and community providers, and organising how services are delivered. In my practice we improved organisational relationships by arranging to do a full annual review for all of a patient’s long-term conditions at a single visit instead of arranging separate appointments for diabetes, heart disease, lung-disease etc.
Good material relationships include how the built environment, equipment and so on affect patients. We invited a patient with expertise in healthcare architecture to show us how we could make improvements to make our building more welcoming and anxiety-containing by, for example, making the toilets easier to find. Just as staff can be courteous and kind and an appointment letter can be helpful and reassuring, a building can be welcoming and contain anxiety.
Attention to all three aspects of relationship-based, patient centred care can help to alleviate some of the inevitable fear and anxiety that accompanies illness. And neglecting any one aspect can undermine a patient’s experience of care, no matter how good the other aspects are.
From quality measurement to quality improvement
In a keynote presentation and a paper for Journal of the American Medical Association, Don Berwick explains that we need to shift our efforts and resources from measurement and control to quality improvement. He describes how we have moved from a historical Era 1, in which professional knowledge was shrouded in secrecy, paternalism was unquestioned and external scrutiny frowned upon, to Era 2, in which managerialism, inspection, regulation, markets and financial incentives have taken over. The majority of NHS clinicians (and not only GPs) feel overwhelmed by regulation and bureaucracy. The government is mistaken in the belief that this reassures patients or motivates professionals. Berwick proposes Era 3, the Moral Era, which retains the professional pride, commitment and beneficence of Era 1 and the transparency and patient-engagement of Era 2, leaving behind arrogance, secrecy, complex incentives and excessive data collection and burdensome regulation. We can do quality improvement in the NHS. It is backed up by forty years of research, it can be taught and we have leaders and enthusiasts waiting on the side-lines. If we can clear away the excessive and unproductive burdens of Era 2 we can make time and space to embark on the quality improvement we really need.