Colloquia 31 July 2008: new ways of incentivising clinical performance

In the latest Colloquia, Health Policy Insight editor Andy Cowper and associate director Tom Smith discuss new ways of incentivising clinicians’ perfomance.

AC: In the last couple of weeks, I thought one of the most interesting developments was the announcement by Imperial College Healthcare NHS Trust that they would introduce pay bonuses for successful operations was worth discussing.

Firstly, it’s an unsurprising media triumph for their impressive communications director Nick Samuels, getting home page leads on the Telegraph, Times, and BBC News Online. But secondly and more importantly, because on the back of the Darzi Review, I think it revives issues around pay for clinical performance in healthcare.

Basically, how can a bonus system avoid creating massive and dangerous perverse incentives?

I struggle to forget Rodney Ledward, the self-styled "fastest gynaecologist in the south" when it comes to activity incentives. If he hadn't been found out, and if he had been a slightly less dangerously poor surgeon with slightly better social skills for his patients, he'd have made a fortune under Payment By Results.

TS: Well, hopefully Ledward would have been picked up through readmission rates, complications and complaints. As John Appleby and Alan Maynard often point out, the data on these things exist, it’s just that it hasn’t received the proper attention.

On the specific Imperial initiative, I was confused by the coverage it got and the fuss it caused. My instinct when I saw this was that it must have been a quiet news day – the information was released to the BBC on a Sunday and covered more widely in Monday’s papers. I was surprised it got the coverage it did.

I guess it was because the initial coverage focused on individual surgeons being paid bonuses for achieving certain quality markers. The reaction of doctors and patients – at least those selected for comment – was pretty negative. “Patients will be horrified,” said Patients Association spokesperson, Katherine Murphy. “There is a real risk that the most complicated cases, and the patients in real need, will be forgotten because they don’t get the best outcomes. Surgeons said the scheme could deter doctors from taking on higher-risk patients, such as the frail and elderly, and from carrying out complex operations.

AC: The coverage seemed to be more struck by the novelty than to engage with the idea.It’s a neutral idea until you have some details, but potentially a good one if it can be got right. But I imagine that avoiding unintended consequences will be far from easy.

Yet if we move to the greater focus on outcomes, then logic suggests that once we know outcomes and have robust data, we need to do something about it. Which surely implies rewarding the very good (in some way) and retraining or otherwise having sanction against the very bad. Although moving up the curve in the middle actually gives the most utopian result.

TS: More than a week later, we still know very little about the detail of the plan, but it has become clear that the scheme is not aimed at individual surgeons and is team-based.

The really interesting thing about the story was how the press reacted to a local management initiative. It seems the press conflated a press announcement from Imperial with the announcement in the Darzi review that the national clinical excellence awards scheme would be revised. There was a suggestion that doctors’ pay would be partly determined by this initiative and others like it.

The press were discussing a local scheme as if it was a national policy. Imperial’s is a local initiative - and there will be others like it, as trusts become confident about managing doctors in new ways. One interesting thing about the Imperial scheme is that it has been designed for doctors by doctors.

AC: That sounds slightly like self-regulation. But what are the right things to measure? It’s not as if nobody can game clinical audit already.

TS: Details aside, schemes like this are a first step towards finding metrics of quality that might be shared by doctors, management and patients.

AC: Now the watchword post-Darzi is quality, and it's a good watchword. Outcomes is another good watchword. But when the two were put together in the QOF, it led to a lot of criticism of what Phil Hammond has called "deskilling GPs into stressed-out box-ticking compliance monkeys" (with apologies to Phil, I'm paraphrasing because I don't have his very good recent book www.amazon.co.uk/Medicine-Balls-Consultations-Worlds-Greatest/dp/1845021886 to hand).

TS: The QOF framework is revolutionary in its attempt to shape behaviour according to clinical evidence. The problem is that the original framework was not very challenging, so GPs began to be paid extra for things they had previously done routinely.  Nevertheless, I think it still has a lot of scope.

The difficulty is that QOF will be revised in an unstructured way. In his speech on obesity last week, Alan Johnson said that he would be drawing on QoF to tackle obesity. In his next-steps review, Lord Darzi said that NICE would lead a review of the framework.

AC: Are incentives intrinsically inimical to professionalism? Or are we near enough to medical candour in peer review, data and evidence-gathering on outcomes and quality that this stuff could work?

TS: I think this is a really interesting debate. I am familiar with the argument that incentives and professional status do not mix. And in some areas it is plain that they shouldn’t, as incentives can distort.

Yet I think incentives are important and can play a useful role. They have the potential to support professionalism and, as in the Imperial example, aim to focus teams on clinical quality. In my experience, doctors are spurned on by competition over clinical outcomes and standards.

AC: Do we really have the capability, capacity and systems though to measure this data effectively?

TS: The real story about Imperial is that it plans to overhaul the way it collects data, so the best-performing clinical teams can be rewarded.  Even the BMA has said ‘any attempt to improve the way NHS statistics are collected is welcome’. The fear of many doctors, I guess, will relate to the accuracy of the data.

At the British Society of Gastroenterology, we are working with Liverpool University to look at HES data so as to identify key markers of quality. And the methodology is really difficult. When we showed the HES data to the consultants it represented, we found lots of inaccuracies on quite basic things, how many patients had been seen, procedures performed and so on.

AC: If there are going to be lots of initiatives, collecting data on different parts of quality, is there a risk that they will be contested or that incentives – payment for performance in different areas – may clash with one another?

TS: We are, in a sense, heading towards nationally-sponsored schemes for payment by performance. Healthcare Republic today report that Mori IPSOS is to survey patients on how they judge their GP’s listening skills, and the result will affect practice income. http://www.healthcarerepublic.com/news/GP/LatestNews/832910/Patients-judge-GP-listening-skills-80-million-survey/?CMP=EMC-DAILYNEWS

My fear in all this is that different elements will not be joined up. There are two problems: that the data used by managers to manage, by clinicians to evaluate, by patients to choose and by PCTs / PBCs to commission will not be joined up, and that the processes in place for local and national regulation will clash.

AC: But that clash already happens. One crucial issue is the lack of an equivalent in primary care for hospital episode statistics and secondary user services data. The QOF is misnamed, really, because it’s about activity and compliance. I’m not sure it’s about quality (it strikes me that much of what’s in QOF was things GPs should have been doing anyway). And it’s nothing about outcomes.

TS:One key barrier to making quality central to how the NHS works is that it is not yet plumbed in to the culture of the NHS. While we have recently heard a lot of ideas for changing this, the risk is that they will be disparate and temporary.

The NHS Act currently going through Parliament contains two parts. The first is concerned with the changes to organisational regulation, the creation of the Care Quality Commission, for example, and the merger of three regulators. The second part of the Act is concerned with professional regulation. They are joined together in legislation, but hardly anywhere else. Progress in both domains is in isolation.

This is a real problem. It is surely necessary that management and clinicians use similar metrics. Yet there is a very real possibility that the outcome measures – or proxies – that will be used by management will not be metrics that will not be used in the process of medical revalidation. If so, this will perpetuate the gulf between the two.  

The key is the future of appraisal and doctors’ representatives should be supported in developing a range of measures to gauge clinical quality and these measures should be the same that organisations are measured against. Perhaps this is overly simplistic, but surely we would want to see a strong overlap between the metrics of ‘quality’ performance in both organisational and clinical domains.

AC: Surely some of this is about reducing the burden of regulation, which should win friends all round. If managers, ministers and clinicians can agree on the measurements, we start to move in a potentially positive direction.

TS: Another problem is the potential cleavage between local and national regulation. While we are nowhere near seeing local authorities held to account for health, we have seen the introduction of joint commissioning models and also the development of Comprehensive Area Assessments, which is a process led by the Audit Commission, examining local achievements. Increasingly, there will be tension between national and local measures of performance, unless these metrics are again aligned.

AC: The idea of cleavage suggests a need for support. There are obviously areas like West Kent and Kingston where it looks as if there is a ‘Wonderbra’ relationship with the local council. The meaningful question is whether you’re being forced into that bra, or have chosen to be there.

And as you noticed (www.healthpolicyinsight.com/?q=node/117) in the Hazel Blears white paper, councils will get new powers with regards to PCTs and it looks like they will be able to force petition votes.

I see that the House of Commons Health Select Committee is going to hold an inquiry into patient safety - http://www.parliament.uk/parliamentary_committees/health_committee/hc0708pn19.cfm. Their terms of reference are interesting: they want to know what the main problems are – clinical judgement, systemic errors – and they are going to look at the different levels of regulation, at the national level, within local authorities and they are also interested in the barriers to the spread of best practice.

From a patient-centred perspective, one of the challenges will be to develop outcome measures with legs – which don’t just measure altered functional status once after say, six weeks. That is already the (bad) measure for smoking cessation in the QOF, and will be very little use in long-term conditions, mental health etc. Equally, palliative care might be a bit of a bastard to assess. And I’m reminded of the old phrase that ‘success has many parents, but failure is an orphan’.