Pay for performance, or “P4P” as it is often known, is now centre-stage in primary care in the United Kingdom. P4P promotes change in clinical behaviour by offering financial rewards in return for achieving certain predefined targets. Both sides of the P4P “equation” are currently the subject of much debate: how generous should the financial reward be (the first “P”); and which performance indicators (the second “P”) should be used as the basis for calculating eligibility for the reward?

From the foundation of the UK National Health Service (NHS) in 1948, general practitioners derived their income largely from capitation. Income was related more to the quantity than the quality of care. In 1990, and against much resistance from within the profession, the first performance targets were introduced into primary care. There were just two: rates of cervical smears and childhood vaccines. All this changed in 2004, when GPs accepted a new contract that radically promoted the role of P4P. This new contract tied about 25% of GP income to the achievement of a panoply of performance indicators. Large financial rewards were within reach for GPs able to achieve targets set for 147 performance indicators (subsequently revised to 135 indicators in 2006 and further revised to 128 in 2008).

Details of the original 147 performance indicators and the mechanism for tying them to financial rewards have been well documented.1 The overall structure is known as the “Quality and Outcomes Framework” (QOF). In essence, the indicators are a mix of clinical indicators covering the management of 10 long-term conditions (eg, diabetes, coronary heart disease, hypothyroidism) and a series of indicators covering managerial, organisational, educational, prescribing and “patient experience” aspects (eg, undertaking an approved patient survey each year) of primary care. The clinical indicators are a mix of process indicators (such as creating a disease register or conducting a specific investigation), intermediate, or proxy, outcome measures (such as reduced cholesterol levels), and true outcome measures (such as improved epilepsy control). Achievement of an indicator is converted into “quality points”, which are weighted according to the perceived workload required to attain the target set for each indicator. Each quality point attracts a fee — currently set at £126 (A$260) for the average-sized general practice (5891 patients) — and the maximum attainable score is currently 1000 points.

Since 2004, the components of the QOF have been revised twice. New indicators for an additional nine long-term conditions (eg, chronic kidney disease) were added in 2006, and in its latest 2008 incarnation, there are 80 clinical and 48 non-clinical indicators.

What have been the positive consequences of P4P in UK primary care? The introduction of the QOF has demonstrated that GPs in the UK have achieved far higher quality standards than expected — at least, as budgeted for by government pay negotiators. In 2007, 5% of practices achieved the highest possible score, and the mean achievement of all practices in England was 95.5% of the available points.2 Such a demonstration of success has the potential to both affirm professional pride and to provide some evidence to the general public to justify the additional taxation that has been required to fund P4P.

As with most systems of P4P, there is evidence that rewarded activity has increased. Substantial increases have been documented in some of the intermediate outcomes, such as blood pressure, cholesterol and glycosylated haemoglobin control, and also in the proportions of heart attack and stroke patients treated with aspirin.2 The overall rise in standards has been accompanied by a narrowing of the health inequality gap (as measured by the QOF) between least- and most-deprived neighbourhoods. Taken together, these achievements should translate into substantial national public health gains.

What have been the unintended consequences of P4P in UK primary care? The success of P4P has been challenged from several perspectives. First, in the case of several QOF indicators, the achievement gains pre-dated the introduction of the QOF and, since then, the improvement rate has continued at a similar pace. For example, evidence has accumulated that overall cholesterol control was improving long before 2004.4 Second, high achievement may be interpreted as targets that were too easy to attain. For example, maximum QOF points are awarded for achieving target blood pressure control in just 60% of patients with diabetes, and the target blood pressure was set at 145/85 mmHg, which is above the value given in most guidelines. Third, GPs may indulge in “gaming” to boost their tally of QOF points — either by under-counting the number of patients on each disease register (only including those for whom target achievement is more readily accomplished) or by using the process of “exception reporting”, whereby GPs can decide whether it is “unsuitable” for a patient to be considered for a given clinical target on grounds such as “maximum tolerated therapy”, extreme frailty, or not responding to three letters inviting them to an appointment.5

More fundamentally, P4P has divided GPs on issues of professionalism. For some GPs, the electronic QOF prompts that accompany a consultation with a patient act as useful reminders and allow the GP to give more thought to deeper issues during the consultation. For others, these prompts represent the intrusion of a reductionist, points-driven approach to patient care that undermines professional autonomy.6 Furthermore, it is readily apparent that measures of patient satisfaction, patient enablement, listening skills, continuity of care, and many of the aspects of general practice that give GPs their greatest professional satisfaction lie outside the scope of any of the performance indicators. There are no measures of the psychosocial problems so common in primary care, nor of many medical conditions (such as any of the chronic gastrointestinal diseases). Indeed, even the advocates of the QOF would concede that probably only a quarter of all registered patients have any of the conditions for which QOF points are available. It is thus clear that P4P in its current form only rewards a small proportion of the overall work of a GP.

Finally, the QOF has proven costly. Should such a large proportion of GP income be determined by these indicators, many of which lack a robust evidence base? Since its introduction, a vigorous debate has continued on balancing the “pay” component of P4P with the “performance” component. The higher than expected performance resulted in higher than expected costs, which contributed to NHS overspending in the early years of the new contract (although spending has now been reined in).

What are the lessons for Australia? In the UK, the QOF was introduced as a “big bang” solution, without prior piloting. From our perspective as UK-based researchers and practising GPs, we think that perhaps the single greatest lesson from this experience is that piloting might have avoided many of the unintended consequences of P4P in primary care.

Although P4P focuses attention on the quality of care, the definition of quality should be multifaceted and derived from evidence-based guidelines and the collective views of GPs, public health specialists and, importantly, patients.

Performance indicators undoubtedly need to change over time, either because the evidence base has changed or because the original indicator has become redundant, and a transparent and robust system needs to be devised for regular updating of indicators and targets. An evolving QOF can act as a means to translate research evidence into practice. Simply put, the QOF has the capacity to “cut the implementation corner”.

Although contentious, P4P represents a bold attempt to redefine the quality standards of primary care, such that best-practice targets quickly become the norm, expected by patients, health service planners and doctors alike. Beyond this, the next challenge is how P4P can evolve into rewarding a broader, more pluralist definition of quality and not merely the narrow focus of those things that are easy to measure.