Are the efforts of financial incentivisation in improving healthcare outcomes,

worth undermining the morality of clinical behaviour within UK primary care?

 

The National Health Service [NHS] in the UK has long yearned to accomplish alignment of its goals at organisational, community and individual levels. And over the last three decades or so, its most predominant means in achieving said alignment has been using financial incentives and punitive fines to condition clinical

practice in primary care. The carrots-and–sticks approach, originating from 1980s public sector reforms, enforces policies in measurable health metrics to ensure efficiency amidst the NHS’ struggling financial predicament. Via incentive schemes like the Quality and Outcomes Framework [QOF] for primary care, incentivisation in healthcare has stirred up a conflict between meeting system-imposed targets and

maintaining patient-centred, virtuous medical practice at the decision-making level. Schwartz and Sharpe in “Practical Wisdom” [31] state that the ‘rules undermine the skill and incentives undermine the will’, a phenomenon coined in behavioural science as the crowding-out effect.

 

This review aims to evaluate the following hypothesis:

 

While incentivisation may lead to improvements in specific clinical metrics, it does not lead to significantly improved patient outcomes holistically, all while undermining the moral integrity of medical practice by promoting a culture of compliance rather than care. The corruptive influence of financial in/disincentivisation turns patients into a means to an end, rather than the ends themselves.

 

     This review will aim to uncover policymakers’ rationale for, and methods of utilising incentivisation to improve clinical performance, as well as evaluating the available evidence supporting its efficacy. This will then be compared to incentivisation’s effects on medical professionalism, clinicians’ autonomy and holistic patient outcomes, interpreted through ethical and behavioural frameworks to posit causative links. Considering the evidence discussed, this literature review aims to make recommendations for reform – wherever and however the evidence points.

 

Rationale and Methods of Financial Incentivisation

​

     The NHS’ usage of financial targets and incentives sees its root in the New Public Management (NPM) reforms, which swept the UK’s public sector in the 1980s. More specifically, the radical reforms of the NHS under Margaret Thatcher were driven by NPM theory. The decentralisation of the NHS through the creation of the “internal market” saw health authorities [e.g. general practices] became micro-purchasers of services from a main body to provide for its patients. The NHS shifted its focus onto clinical targets and the carrots-and-sticks approach thus falls under the larger umbrella of the NHS’ reorganisation into a business post-NPM reforms. Since the 1990s, the policy agenda has shifted away from this business-like approach to

internal competition, focussing on clinical governance and quality standards. An NPM-driven system anchors on penalisation of over-spenders and over-referrers, with a strong disregard for equally enforcing quality and safety standards. The post- NPM era of NHS policymaking now takes a more hybridised approach.[33]

     QOF was introduced in 2004 to ‘enhance ... patient outcomes by incentivising continuous improvement in clinical care’ [Department of Health & Social Care, 2024]. It scores general practice performance against 76 indicators [clinical, public health and quality improvement], allowing them to tally points up to a total of 635 points. The value of a QOF point in 2024/25 was £220.62, meaning that general practitioners can, in theory, gain access to up to £140,000 annually through hitting QOF targets.[31] QOF constituted 8% of practice income in the years leading up to 2019.[26] NHS England additionally utilises financial penalties to mechanically improve patient outcomes. Its Cancer Plan, rolled out in 2005, introduced time-limited referral standards for suspected cancer patients, with percentage threshold targets enforced by financial penalties

incurred per patient exceeded.[23]

     The incentivised QOF indicators are in and of themselves, evidence based, and the framework also aims to maintain clinical breadth by its inclusion of diverse processes of care and disease register maintenance. Longitudinal analysis of achievement rates for both incentivised and non-incentivised activities[8] examined the impact of QOF’s introduction on 148 UK general practices through comparison to projections from pre-incentive trends for 42 activities [23 incentivised]. The analysis found that 22 out of the 23 incentivised indicators outperformed projected achievement trends, by margins ranging from 1.2%-37.7%, and this

mitigated significant underperformance from 4 non-incentivised indicators. Figure 1 shows how each of the 4 categories performed pre- and post-incentivisation.

 

[See Figure 1 below]

Figure 1: A graph of mean achievement rate, % [y] against time, financial years [x]

for incentivised and non-incentivised prescribing and measurement indicators of

quality of care [QOC]. Source: Doran, T. et al. [2011]

 

     The analysis is subject to limitations. The randomised and large sample size adds validity to the results, but the 42 indicators were not randomised from a possible 428 indicators of quality of care [QOC], as the study required indicators solely incentivised by QOF’s induction – a heterogeneous mix of care processes would increase reliability. The analysis doesn't account for confounding factors biasing certain processes of care, which incentivisation can magnify. Of the five incentivised indicators outperforming projection by 30+%, four were smoking-related, perhaps reflecting public health efforts in changing cultural attitudes. 

A cohort study conducted using data from the National Diabetes Audit [NDA] for England and Wales[16] assessed the association between mortality in diabetes patients and the number of recommended care processes they received. The study included over 80% of diagnosed diabetes cases in England and Wales, meaning the sample size is extremely representative of the population. The study equally does well to identify all important confounding factors [e.g. smoking, age] and utilise analysis-of-variance to assess their impact. With mortality outcomes well-defined by the Office for National Statistics’ criteria[7] and with an

adequate 7+-year follow-up, the results are very reliable. The study is only observational but suffices given the ethical considerations categorically disallowing experimental studies. The results found that diabetes patients with five or fewer recommended care processes carried out routinely, had a higher hazard risk of all-

cause mortality by almost one third, compared with those with all eight recommended care processes.

     Other observational UK studies[24] add supporting evidence for incentivising clinical care processes in the treatment of diabetes. NHS policymakers then use financial incentives as vehicular enforcement

of evidence-based practice [see Figure 2].

 

[See Figure 2 below]

Figure 2: An extract from QOF’s guidance for 2024/25, showing point and threshold

values for incentivised activities in the treatment of diabetes mellitus. Source: NHS

England [2024]

 

     NHS Scotland’s abolishment of QOF in 2016, to remove political barriers taking up clinicians’ time, is an interesting comparator. Subsequent comparative studies associated the removal of QOF with a sustained reduction in recorded quality-of-care indicators [Morales et al., 2023] (unclear citation??) e.g. a 40.2% absolute difference between England and Scotland in mental health care planning processes.

     In sum, the evidence supports the positive impact of financial incentivisation towards well-evidenced care processes in improving patient outcomes.

 

Impact on Medical Professionalism

​

     Professional priorities trickle down from institutional targets. When policymakers financially incentivise specific care processes towards patient outcomes, they skew the balance of clinical decision-making. A retrospective observational study[30] utilised the Cattaneo et al. Manipulation density test, to  observe whether the use of disincentivisation to enforce standards in the Cancer Plan [2005] would deprioritise clinical need in lieu of avoiding fines. The study saw a statistically significant threshold effect in the 63-day specialist referral standard - patients nearing this threshold were inexplicably prioritised over those with

objectively more clinical need. The study thus shows that, faced with the prospect of financial penalties, clinical decision-making skews towards avoiding punishment and away from patient-centred principles. A recent rapid review, comprising 41 mixed studies chosen and screened independently by 2 authors, identified measure fixation, tunnel vision and gaming as the most common negative consequences of

incentivising performance in healthcare.[19]

     Incentivisation may be a strong means to improve QOC, but the evidence here adds a sinister tone to the decline in non-incentivised care processes. Difference-in-differences analysis was applied to a 10-year dataset of 7079 UK general practices to test the effects of incentive schemes on dementia diagnoses thresholds from QOF and compare this to four measures of patient experience from the annual GP Patient

Survey.[20] The quantitative cohort study found that incentivisation was associated with both a positive impact on QOF quality outcomes, and a negative impact on patient experience indicators in continuity of care and doctor-patient relationships.

     Qualitative evidence on incentivisation proves useful in directing further research. Themes emerging from focus grouping 25 primary care clinicians in Australia[18] seemed to suggest that they struggled with their financial dependency on incentives and fiscal need to avoid penalties, obviously reducing their financial autonomy and clinical autonomy likewise. Whilst the evidence isn’t as easily applicable to the UK, this mirrors the sentiment expressed behind NHS Scotland’s abolition of QOF, with the Scottish health secretary announcing, ‘This ... marks a significant shift towards placing more trust in the clinical judgement and

professionalism of GPs’.[4] NHS Scotland subsequently championed their vote of confidence through their new GP contract which saw all former QOF incentive-based funds being relocated to core GP funding.

     A behavioural experiment conducted on German primary care physicians[6] found that if their prime objective was patient health benefit, they provided a significantly higher QOC compared to those who stated being additionally motivated by their own profit. Figure 3 shows the former still produce a significantly higher quality of care when incentivised, while the latter saw a subsequent reduction.

More interestingly, the analysis observed crowding-out of patient-regarding behaviour, with 14% of the optimally treated patients provided with less QOC once performance pay is introduced.

 

{See Figure 3 below]

Figure 3: A table showing regression results on the impact of pay-for-performance

[P4P] and intrinsic motivation on quality of care under capitation [CAP] and mixed

[CAP+P4P] models. Source: Brosig-Koch, J. et al. [2024]

 

      The use of incentivisation thus exhibits naivety in failing to consider negative lateral

effects on the decision-making of its practitioners.

 

Impact on Ethical Practice

 

     An important consideration is how financial incentivisation coexists with Beauchamp & Childress’[35] four pillars of medical ethics - justice, benevolence, non- maleficence, autonomy. Available literature seems to focus on whether incentivisation successfully enacts distributive justice and how it affects clinical

autonomy.

     Incentive schemes are utilitarian in nature, meaning that their ethical validity stems from the ends [improvements in patient outcomes] justifying their employed means. This begs the question – at which point do the ends not justify the means?

      Arguments from utilitarian distributive justice[10] are able to morally defend employing negative methods to achieve overall good, such as the aforementioned threshold effect; practices preserve more resources [practice income] to provide more clinical care, producing more overall good. This is equally

defensible as pragmatically, if practices were to always prioritise individual clinical need blindly, they’d risk resource depletion and financial infeasibility, meaning that any ethical approach to resource allocation ought to at least consider consequentialist or utilitarian theories of good.

      Nonetheless, there is still an ethical trade-off that comes with practices instrumentalising patients towards financial targets. Previously mentioned studies showed that in practice, this can cause a moral regression to more basic ethical models, such as punishment avoidance and reward-seeking behaviours in clinical decision-makers. Kohlberg’s theory of moral development places such ethical decision-makers in Stage 1 [Figure 4] as it signifies their loss of internalised, sound principles of justice and ethics, from which to act on and apply in practice. This ethical regression is caused by incentivisation stripping primary practitioners of their clinical autonomy. The overarching top-down control of clinical decision-making through policies enforced by financial tools lends itself the moniker of “neo-paternalism”, a new age where both patients and clinicians alike are stripped of their autonomy by institutional policies and control [Klugman, 2018; Mannion et al., 2023].

 

[See Figure 4 below]

Figure 4: A diagram depicting Kohlberg’s six stages of moral development. Source:

Pritchard, M. [2014]

 

      Clinical decision-making, for example in documenting and giving lifestyle counselling to hypertensive patients, involves weighing up the conflicting factors[14] and applying an individualised approach to justice, known as ’phronesis’.[21] Incentivisation disregards ethical deliberation by reducing it to a pursuit of resource gain. An interrupted time-series analysis of 12 QOC indicators found that once financial incentives were lifted, lifestyle counselling and documentation of hypertensive patients reduced by over 60 percentage points.[25] While this appears negative, non-incentivised deliberation is more ethically clear, suggesting that incentives cause practices to act against their ethical interests.

     Incentive schemes equally do not fare well up against Kantian aspects of ethical practice.[15] Good Medical Practice is a framework that ’sets out the principles, values, and standards of professional behaviour

expected of all doctors... registered with us’[11] and is deontological in nature, focussing on a doctor’s duties. It represents moral rigidity in  clinical conduct, which cannot be counterweighed by meeting practice targets. Policymakers paradoxically bestow the heavy deontological burden of ‘treating patients as ends in themselves’[3] alongside financial incentivisation which instrumentalises those very patients, creating huge tension for clinical decision-makers.

      Rather than addressing morality through rules or consequences, Aristotelian ethics evaluates the moral character of the clinician and the cultivation of virtues necessary for good medical practice – phronesis, eudaimonia [human flourishing] and ethos [moral character] should guide ethical decision-making, rather than external motivators or even externally-oriented ethical frameworks [deontology and utilitarianism]. By focusing on metric-based compliance, financial incentives shift clinicians away from the moral virtues that underpin their ethical practice.

 

Medical Education’s Role in Reinforcing or Challenging Incentivisation

​

     Medical education is not immune to the changing priorities of the NHS, and especially not to its financial realities.[27] The field may do well to prioritise clinical efficiency in curricula,[17] to best prepare them for the healthcare system they will inevitably enter. But even now, without formal curricular changes, ethical erosion will set in through observation of normalised attitudes to target-driven practice. It is important to note that target-driven practice may well be important at the institutional level, but the cost-benefit analysis is different at the individual carer-patient level. The insertion of a reductionist, single cause-and-effect policy into a complex system, is ignorant of any consequences emerging laterally from policymakers’ tunnel vision, and these consequences induce ethical erosion in training doctors. An article published by BMJ Quality & Safety[9] posits a few consequences which become instilled particularly in younger doctors, such as expecting unsustainable levels of work from the system [staff], which induces a ’major collapse in elements of a caring culture’.

     Financial incentives can harden the next generation of doctors to care ethics and patient-focussed service provision, drawing a parallel to the hidden curriculum theory in stagnant NHS culture.[5] Ironically, as discussed priorly in virtue ethics, the deliberate instillation of strong ethical values services clinicians more

strongly than allowing cultural norms to reshape their approach to care. 

     Bourdieu’s theory of habitus[2] provides a useful framework for understanding how financial incentivisation can function to reshape clinical decision-making at a subconscious level and cause nascent clinicians to internalise negative cultural norms. Bourdieu understood human behaviour to be fundamentally cultural or social, also known as habitus – a system of ingrained dispositions stemming from re-enactment of observation [Figure 5]. In the medical Field, habitus is shaped by social, cultural and economic characteristics of the NHS. When financial targets dictate clinical practice, clinicians internalise the predominance of economic rationality in medical practice, subtly overriding the importance of ethical deliberation. Considering the high-pressure environment as well, clinicians over time gradually subdue to the norms of the NHS’ culture, unless it is actively resisted.

​

[See Figure 5 below]

Figure 5: A schematic diagram displaying the intersectionality between habitus’ 6

main determinants and the field. Source: Akram, S. [2023]

 

     On one hand, training future clinicians to properly grasp considerations of cost-effectiveness and resource allocation into their ethical decision-making best prepares them for the realities of the practice. On the contrary, in the early career of  doctors, pressure to meet targets for procedures or metrics can be internalised, breeding doctors which prioritise compliance over compassion.

 

Future extensions and considerations

 

     The ideal end-goal is to maintain the benefits of incentivisation while mitigating/reducing its negative impacts. Incentives for better ethical practice, measured by qualitative surveying of patients for example, could be a solution but considering the ethical consequences of incentives, using incentives to mitigate

other incentives doesn't bode well for the NHS’ ethical climate. QOF’s unequal distribution of points across incentivised processes was ethically criticised, but future investigations would do well to consider its potency as a vehicle for addressing healthcare inequality, as well as how the ethical discussion changes. Incentive

schemes equally lack longevity in changing behaviour[12] but risk long-term damage to clinicians’ behaviour and approach to care. The NHS does employ alternative soft approaches [e.g. internal audit policies] to target internal behaviour, but more comparative evidence is needed between the financial incentivisation of direct QOC processes and the incentivisation of self-improving processes [audits], such as cost-effectiveness studies.

     Another potential avenue could comprise mixed-modality incentive structures, which aim to make up for the shortfalls of solely financial incentive schemes, such as QOF. This could be achieved by broadening the scope of what an incentive can be, in order to address intrinsic motivations which are sorely overlooked by existing performance for pay schemes. A review of the literature on targetable aspects of clinical behaviour [professional pride, ethical commitment, patient-centred care] would inform policymakers on the most suitable, non-financial incentives in spurring change. Hybridising incentive schemes in this way could perhaps be the best structured solution to preserving the gains of existing methods whilst reducing

unintended consequences such as crowding out of professional integrity.

 

Conclusion

​

     In summary, financial incentives improve specific metrics of QOC but risk distorting professionalism and ethical practice in primary care. While the evidence is strong for frameworks such as QOF having positive impacts, this must be weighed up against the instrumentalisation of patients and the ethical regression seen in clinical decision-makers unduly influenced by incentivisation. Policymakers would do well to carefully balance their use of financial enforcement with the preservation of its ethical and professional foundation in general practice. The current injection of financial incentives into general practice towards specific metrics is a naive, reductionist approach, and it reflects an unwillingness from policymakers to honestly engage with

the complexities of the NHS system.