Why pay for performance may be incompatible with quality improvementBMJ 2012; 345 doi: https://doi.org/10.1136/bmj.e5015 (Published 14 August 2012) Cite this as: BMJ 2012;345:e5015
All rapid responses
The table presented by Roland is amusing and probably broadly correct.
However, technically the incentive scheme in UK primary care is pay for recording and not pay for performance.
What doctors would do in this system, unrestrained by any other obligation, is maximise administration and minimise patient centred care. (or as little as possible for as many people as possible, which is no different from capitation except for the ability to control the number of target patients)
This is where upcoding and widening target populations come into their own.
Competing interests: Designed a range of "turbo templates" for data collection to maximise data gathering while minimising the impact of the quality scheme on patient care.
It was like a breath of fresh air reading the BMJ editorial cited above. I have for some eight years now viewed some of the 'quality' recommendations in the 2004 G.P. Contract with caution. For one thing the word 'quality' appears akin to a newly discovered term not known of hereto! For another, its implications appear to be implicitly related to volume output and related income generation. Where for a designated disease category, a Practice has smaller numbers that do not match the national/local prevalence figures it is viewed with suspicion irrespective of the fact that the Practice has closely researched it. Also, the funding mechanism is geared towards 'illness' and, therefore, consequential treatment. The more 'ill' patients' you have the more the opportunity to meet the criteria for 'treatment' and generate QOF points that equate to money. By contrast, the healthy patient who does not fit into a categorized illness category is merely worth the per capita fee. The system is further bedevilled by the fact that lowering of some diagnostic threshold enlarges the numbers of potentially 'ill' patients in the category. It is not surprising, then, that it pays to have larger numbers of 'ill patients' on one's list and to be rewarded for treatment that yields the highest figures for the maximum payment in the categorized illness category.
Although it is common knowledge that most people are reluctant to label themselves as ill, the reverse would apply to those who stand to gain financially if the volume pool of "illness" was enlarged. This is what must have prompted one Maryland rehabilitation hospital to have diagnosed 287 cases of "kwashiorkor" on the basis of "protein malnutrition " alone.
I find it greatly perturbing that if the threshold diagnosis for some illnesses continues to be lowered it would only be a matter of time before every single individual could be labelled 'ill' and requiring 'treatment'. Also, to meet the highest level of payment in a categorized QOF illness category, and where the laboratory figures are not in the 'desirable range' most practitioners respond by increasing the strength and dose of prescribed medications. 'Quality' is,then, perceived to have been attained on the basis of laboratory figures alone. I wonder how many prescribing practitioners have ever considered the detrimental impacts on the long-term health of their patients which is magnified in those on polypharmacy. This is one event yet awaiting to unfold.
One does not have to look far to see why the national drug budget is becoming unsustainable. Rather than examine the relevancy or otherwise of so many receiving treatment, the trend has been to switch to generics and/or reduce the frequency of drug administration to a once daily dose in an attempt to contain costs. Clearly, this is no more than scratching the surface.
The oddity of some of the reward criteria in the 2004 Contract is also striking. It appears that the emphasis is on rewarding those who are viewed as 'bad Practices'. What logic is it that reasons that if you have no patient complaints in the Practice, then, you receive neither recognition nor payment! Or that, if no 'significant events' have occurred in the Practice, then no reward payments are forthcoming. Could it be that only those viewed as 'bad doctors' need encouragement by way of financial reward. How could the point have been missed that 'good doctors' also need to be rewrded to continue to be 'good' or even better doctors. After all, that is surely perceived as the desirable end point!
Or, is the distorted logic consistent with the prevailing view that you get more funding if you overspend and are penalised by having your funding cut if you are efficient enough to make a saving on your budget!
The Cochrane Review's finding on the treatment of mild hypertension is an eye-opener and this may not be the only area of illness category that might prompt a rethink. Straight-line thinking has encouraged the lowering of B.P. to 'desirable' set limits. A phenomenal slice of the national drugs budget is devoted to this activity and financial inducement provided to meet the banding levels. Likewise, the drive to lower blood lipid levels is relentlessly applied and no lower limit seems to be low enough. This is,again, consuming a significant slice of the national drugs budget and needs a discriminatory critical review in terms of value for money.
Some may remember the false economy in the issue of free spirometers to general practitioners so that they could do the procedure themselves on their COPD population thereby obviating the need of the employed technicians. What was not perceived was that the quality of data was dependent on the aptitude and experience of the operator; the device alone could not deliver the goods!
Time has passed and we have lived but have we learned?.
Competing interests: No competing interests
Paul Glasziou and colleagues  present a good summary of some of the issues to be considered when introducing pay for performance schemes. However, in the accompanying editorial  Steffie Woolhandler and her colleagues are oversimplistic in their criticism of pay for performance.
First, there is no good reason why doctors should take home less money for providing good care. This was the situation for general practitioners in the UK prior to the P4P scheme in 2004: if they invested in their practices (e.g. by employing an extra nurse) the profits of their practice went down and hence their take-home pay. This applies to many other physicians in private practice worldwide.
Second, there is no perfect way of paying doctors. I’m endebted to Bonnie Sibbald for pointing out the consequences of the four main ways of paying doctors (see table).
The conclusion of this table is that all payment systems can have perverse consequences and that we therefore rely on the professionalisms of doctors to minimise these adverse effects. This underlines the importance of ensuring that incentives of any type are as closely aligned to professional values as possible.
1. Glasziou P, Buchan H, Del Mar C et al. Pay for performance: How to make it worth doing. BMJ 2012; 345: e5047
2. Woolhandler S, Ariely D, Himmelstein D. Why pay for performance may be incompatible with quality improvement. BMJ 2012; 345: e5015.
Competing interests: I advised the UK government on the development of a P4P scheme for general practitioners (the Quality and Outcomes Framework) between 2001 and 2003.
An issue which is neglected, is selection of patients. Clearly the patients with the most obscure or serious ailments go to the specialists in the field, whose results may not reflect their ability or dedication to medicine. Unless someone has a large number of "ordinary" patients, his (her) results will not be outstanding, and the system which rewards the most dedicated will fail to reward the right physician!
Competing interests: No competing interests