Institutional financial incentives in healthcare: a review of normative considerations

Abstract

Background

Institutional financial incentives(FIs) are used to improve the quality of healthcare services and patient outcomes. In spite of their growing use, empirical evidence regarding their impact remains inconclusive. Moreover, the normative implications of institutional FIs remain underexplored. This study aims to map the normative considerations and tensions in using institutional FIs in healthcare.

Methods

We performed a systematic review of reasons to identify and map the normative considerations in institutional FIs. Six databases, namely PubMed, Embase, Scopus, Web of Science, ProQuest Central and EconPapers, and additionally Google Scholar were searched. We included peer-reviewed literature in English on financial incentives or pay-for-performance in healthcare at the institutional level, and excluded grey literature, and studies on individual provider incentives.

Results

Sixteen articles met the inclusion criteria, representing diverse disciplines such as health economics, medicine, law, public health and health policy. Two categories of normative considerations are identified: firstly, considerations at the macro and meso level (societal and institutional level), and secondly, considerations at the micro (departmental or individual) level. The key normative principles identified are effectiveness, utility, solidarity, distributive justice, procedural justice, professional values, conflicts of interest and proportionality. The tensions inherent in each normative principle are closely linked to contextual elements such as institutional characteristics, incentive design and performance measures: for instance, the type of performance measure used (process or outcome measure, relative or absolute improvement) impacts procedural justice.

Discussion

This study is limited in including only English language literature. Notwithstanding this limitation, this review provides an overview of the normative implications of institutional financial incentives. As robust empirical evidence on the institution and societal-level effects of FI is lacking, these findings emphasize the need for further normative deliberation on institutional FIs and more caution while implementing institutional FIs as a policy measure.

Supplementary Information

The online version contains supplementary material available at 10.1186/s12910-025-01252-y.

Background

Financial incentives form an important element of healthcare providers’ payment or reimbursement in many contexts [13, 25]. That such “supply-side” incentives can shape the provision of healthcare is supported by different economic and behavioral theories [2, 23, 25, 35]. Not only could incentives influence the efficacy of services provided, but they are also increasingly believed to be effective in improving the quality of healthcare services [14, 35, 42]. Further, they are believed to have broader social impact, such as reducing healthcare inequity [17, 32, 34, 42].

Supply-side financial incentives can be structured/organized in different ways. On the one hand, they could be offered to individual healthcare professionals, such as the Contract for Improving Individual Practice (CAPI) in France or the Quality and Outcomes Framework (QOF) in the UK [20, 36]. Both programs reward General Practitioners who can prove that they (or their practice) fulfils a set of performance indicators, such as patients screened for cancer. On the other hand, they could be offered to healthcare institutions, such as hospitals or clinics, to stimulate improvement at an institutional level. For instance, the UK’s Maternity Incentive Scheme (MIS) offers financial rewards to NHS Trusts that fulfil certain pre-determined quality and safety benchmarks (NHS [28]). According to Eckhardt et al. [14], institutional financial incentives are becoming more widespread and impactful. They can aim to improve different domains of care such as acute care, care for chronic illnesses or primary care [14]. As not many valid and reproducible studies have been undertaken on existing institutional financial incentives, also known as institutional pay-for-performance or P4P, evidence on their impact remains inconclusive [23, 25].

Besides being assessed by technical standards such as cost-efficiency and improved health outcomes, healthcare should also conform to normative standards. Normative standards explore and evaluate whether healthcare policies, systems or interventions are morally acceptable, based on ethical or moral values or principles, such as fairness, solidarity or justice [24]. While technical standards are highly relevant to the question of whether a policy should be implemented or not, they can be considered “instrumental” values that could contribute to the realization of normative values such as good health or equity [24]. However, technical standards could sometimes conflict with normative values. Even efficacious and cost-effective interventions could be sometimes morally unjustifiable, for instance, contact-tracing for infectious diseases without consent from individuals, or limiting access to costly medical procedures based only on age. Evaluating a healthcare intervention normatively is crucial in democratic contexts, as it prevents morally contentious health policy from being implemented. Further, as Cribb [10] argues, healthcare interventions or policies inherently embody a set of ethical values, often implicitly. While these values themselves do not (physically) make up healthcare systems, they could be reflected concretely in the organization and delivery of healthcare. For instance, a healthcare environment that is based on the idea of “cure” might be organized differently from one that embodies the idea of “care” [10].

Whereas there has been considerable normative debate on financial incentives to individual healthcare professionals, incentives to healthcare institutions have received less attention in normative literature. This could have important implications for public health and health policy, as it may mean that institutional incentive programs are not adequately normatively evaluated before their implementation. To further explore and characterize this research gap, this study aims to map the normative considerations and tensions in using institutional incentives in healthcare.

Methods

Literature search, search strategy and databases

We performed a systematic search of literature to identify relevant peer-reviewed normative articles on institutional financial incentives. We searched 6 databases, namely, PubMed, Embase, Scopus, Web of Science, ProQuest Central and EconPapers. Additionally, to ensure that publications were not missed, we screened the first 500 results from a Google Scholar search and performed snowball sampling on all included articles. The systematic search and screening followed the PRISMA guidelines for systematic reviews [29].

The search strategy comprised two groups of search terms. The first group pertaining to institutional financial incentives and it synonyms and the second group pertaining to ethics and its synonyms. Search terms within each group were combined using the “OR” operator and the two groups were combined using the “AND” operator. The search strategy for PubMed was created as the first step, and then adapted to the remaining databases. All database-specific search strategies can be found in Supplement 1.

Searches were carried out by RJ and independently repeated and verified by KD. They were conducted in December 2023 and updated in March 2024 and July 2024.

Inclusion and exclusion criteria

We included published peer-reviewed normative literature on financial incentives or pay-for-performance in healthcare at the institutional level. Normative literature, as defined by Mertz [26] refers to “literature that (i) aims to evaluate judgements, decisions, acts, (social) practices, technologies, institutions, organizations and general states of the world from a moral or legal point of view and/or to define/set out which decision or course of action is or should be morally or legally necessary, prohibited or permitted; or that (ii) aims to develop, interpret or criticize the evaluative or prescriptive concepts required for this” (p.3). Empirical literature was also included, provided that they entered into normative discussions in sufficient depth.

According to Mathes et al. [25], financial incentives or pay-for-performance can be defined as “a payment model that rewards health care providers for meeting pre-defined targets for quality indicators or efficacy parameters to increase the quality or efficacy of care” (p. 1). Thus, incentives can include not only additional monetary rewards but also preferential systems of reimbursement based on changes in pre-defined quality measures [39]. Rewards could go hand-in-hand with penalties for not attaining targets [25]. Articles on interventions that only used penalties without the possibility of reward were excluded, as they act solely as a disincentive or a negative incentive.

Only articles on financial incentives at the institutional or organizational level were included. We identified two articles based on one scheme which could potentially reward individual professionals as well as group practices (which count as institutions). These articles were included after discussion as they captured many relevant normative tensions at the organizational level.

Articles and reports from grey literature were not included in this study. Also excluded were articles solely on individual financial incentives and studies in languages other than English.

Screening and selection

The database search resulted in 14, 830 articles, which were imported to the EndNote reference manager. After deduplication, 8118 articles remained and were screened using their titles and abstracts, after which 102 articles remained for full-text analysis. After an in-depth reading of these articles, 16 articles fit the inclusion criteria. The Prisma flow diagram for the screening process is shown in Fig. 1.

Fig. 1.

Prisma flow diagram showing the screening process

RJ screened the articles based on titles and abstract. Doubts regarding article inclusion were resolved through discussion with KD. The full text screening (n = 102) was performed by RJ, and verified independently by KD.

Data analysis and reporting

We followed the “systematic review of reasons” methodology as described by Strech and Sofaer [41]. The aim of such a review is to describe the ethical or normative reasons or arguments given in literature that support or undermine an intervention, decision or policy. Following this strategy, we formulated our research question as follows: “what are the normative principles or arguments used in literature to support, undermine, evaluate or appraise institutional financial incentives in healthcare?”.

To identify the main lines of reasoning or argumentation, we used inductive content analysis (ICA) [44]. Firstly, the articles were thoroughly read and re-read. Secondly, we performed a first round of coding with the NVivo software package to identify and categorize the broad lines of argumentation relevant to the research question. Thirdly, the contents of each category were re-read and compared to develop more fine-grained sub-categories, which were in turn compared and refined iteratively. Lastly, the categories and sub-categories were synthesized to present the overall meaning of the body of literature.

Results

Sixteen articles fulfilled all inclusion criteria and were subsequently analyzed. The included articles represented diverse disciplines, such as health economics, medicine, law, public health/health systems research and health policy. Ten of the 16 articles originated from the United States, two each from the United Kingdom and Canada, and one each from Australia and New Zealand. While some of the studies were based on specific institutional incentives in different context such as primary care, cancer care, elderly care facilities, nursing care and the Medicare insurance program, others draw on multiple interventions or provide reasoned recommendations for good practice. The articles’ characteristics are summarized in Table 1.

Table 1.

Included articles and their characteristics

	Authors	Title	Journal	Discipline (of first author)	Country (of first author)
1	(Averill et al., 2011) [1]	Paying for outcomes, not performance: lessons from the Medicare Inpatient Prospective Payment System	Jt Comm J Qual Patient Saf	Medicine	United States
2	(Brereton, 2021) [4]	Toward a Preliminary Theory of Organizational Incentives: Addressing Incentive Misalignment in Private Equity-Owned Long-Term Care Facilities	Am J Law Med	Law	United States
3	(Buetow, 2008) [5]	Pay-for-performance in New Zealand primary health care	Journal of Health Organization and Management	Medicine	New Zealand
4	(Chen et al., 2022) [7]	Society of General Internal Medicine Position Statement on Social Risk and Equity in Medicare's Mandatory Value-Based Payment Programs	J Gen Intern Med	Medicine	United States
5	(Conway & Satin, 2022) [8]	The role of pay-for-performance in reducing healthcare disparities: A narrative literature review	Prev Med	Medicine	United States
6	(Culhane-Pera et al., 2018) [11]	Primary care clinicians'perspectives about quality measurements in safety-net clinics and non-safety-net clinics	Int J Equity Health	Medicine	United States
7	(Custers et al., 2008) [12]	Selecting effective incentive structures in health care: A decision framework to support health care purchasers in finding the right incentives to drive performance	BMC Health Serv Res	Social Sciences	Canada
8	(Ginsburg, 2008) [16]	Paying hospitals on the basis of nursing intensity: policy and political considerations	Policy Polit Nurs Pract	Health Systems	United States
9	(Johnson, 2007) [18]	Pay for performance: ACG guide for physicians	Am J Gastroenterol	Medicine	United States
10	(Khan et al., 2020) [20]	A pay for performance scheme in primary care: Meta-synthesis of qualitative studies on the provider experiences of the quality and outcomes framework in the UK	BMC Fam Pract	Medicine	United Kingdom
11	(Kramer, 2012) [21]	Payment for performance and the QOF: are we doing the right thing?	Br J Gen Pract	Medicine	United Kingdom
12	(Li & Evans, 2022) [22]	Incentivizing performance in health care: a rapid review, typology and qualitative study of unintended consequences [Review]	BMC Health Services Research	Health Sciences	Canada
13	(Milgate & Cheng, 2006) [27]	Pay-for-performance: the MedPAC perspective	Health Aff (Millwood)	Policy Analysis (Medicare)	United States
14	(Scott, 2008) [39]	Pay for performance programs in Australia: a need for guiding principles	Australian health review: a publication of the Australian Hospital Association	Medicine	Australia
15	(Shurson & Gregg, 2019) [40]	Relationship of pay-for-performance and provider pay	Journal of the American Association of Nurse Practitioners	Nursing	United States
16	(Wharam et al., 2009) [45]	High quality care and ethical pay-for-performance: A society of general internal medicine policy analysis	Journal of General Internal Medicine	Medicine	United States

To understand and contextualize the normative discussion/considerations, several studies included descriptions of key empirical elements which are linked to normative aspects. We identified three such elements: institution and population characteristics, incentive design and performance measures. These empirical elements and their normative relevance are described in Sect. 1 of the results.

The inductive content analysis identified two categories of normative considerations: firstly, considerations at the macro and meso level (societal and institutional level), and secondly, considerations at the micro (departmental or individual) level, as shown in Table 2. Section 2 of the results describes each category and the normative arguments linked to them.

Table 2.

Normative considerations identified through inductive content analysis

Considerations at macro- and meso-level
1. Effectiveness	Targeting quality rather than quantity	Averill et al., 2011; Khan et al., 2020; Kramer, 2012; Milgate & Cheng, 2006; Scott, 2008 [1, 20, 21, 27, 39]
	Provides resources for improvement	Averill et al., 2011; Brereton, 2021; Buetow, 2008; Khan et al., 2020 [1, 4, 5, 20]
	Conflating performance measures with actual improvement	Chen et al., 2022; Culhane-Pera et al., 2018; Khan et al., 2020; Kramer, 2012 [7, 11, 20, 21]
2. Utility	Cost-effectiveness as compared to fee-for-service	Milgate & Cheng, 2006 [27]
	Maximizing population health	Chen et al., 2022; Conway & Satin, 2022; Li & Evans, 2022; Milgate & Cheng, 2006; Scott, 2008 [7, 8, 22, 27, 39]
	Accounting for non-monetary opportunity costs	Li & Evans, 2022 [22]
	Outcomes outside institution’s control	Custers et al., 2008; Khan et al., 2020; Li & Evans, 2022; Milgate & Cheng, 2006 [12, 20, 22, 27]
3. Solidarity	Promoting collaboration	Scott, 2008 [39]
	Creating competition	Culhane-Pera et al., 2018; Ginsburg, 2008; Khan et al., 2020; Li & Evans, 2022; Shurson & Gregg, 2019 [11, 16, 20, 22, 40]
4. Distributive justice	Resource allocation within the health system	Milgate & Cheng, 2006; Scott, 2008 [27, 39]
	Consensus on the ideal incentive	Buetow, 2008; Johnson, 2007; Khan et al., 2020; Milgate & Cheng, 2006; Scott, 2008; Shurson & Gregg, 2019 [5, 18, 20, 27, 39, 40]
	Inverse equity effect	Chen et al., 2022; Conway & Satin, 2022; Culhane-Pera et al., 2018; Johnson, 2007; Kramer, 2012; Li & Evans, 2022; Milgate & Cheng, 2006; Scott, 2008 [7, 8, 11, 18, 21, 22, 27, 39]
	Population-level health equity	Chen et al., 2022; Conway & Satin, 2022; Culhane-Pera et al., 2018; Johnson, 2007; Kramer, 2012; Li & Evans, 2022; Milgate [7, 8, 11, 18, 21, 22, 27, 39]
5. Procedural justice	Choosing evidence-based measures to incentivize	Averill et al., 2011; Custers et al., 2008; Johnson, 2007; Khan et al., 2020; Kramer, 2012; Milgate & Cheng, 2006; Scott, 2008; Wharam et al., 2009 [1, 12, 18, 20, 21, 27, 39, 45]
	Participatory approach	Johnson, 2007; Kramer, 2012; Scott, 2008; Wharam et al., 2009 [18, 21, 39, 45]
	Validity and reliability of incentivized measures	Averill et al., 2011; Culhane-Pera et al., 2018; Custers et al., 2008; Johnson, 2007; Khan et al., 2020; Li & Evans, 2022; Milgate & Cheng, 2006; Scott, 2008; Shurson & Gregg, 2019; Wharam et al., 2009 [1, 11, 12, 18, 20, 22, 27, 39, 40, 45]
	Preference for aggregate measures over individual measures	Johnson, 2007; Scott, 2008 [18, 39]
	Measuring processes or outcomes	(Averill et al., 2011; Brereton, 2021; Johnson, 2007; Li & Evans, 2022; Scott, 2008; Shurson & Gregg, 2019 [1, 4, 18, 22, 39, 40]
	Incentivizing relative improvement over absolute attainment	Averill et al., 2011; Brereton, 2021; Conway & Satin, 2022; Li & Evans, 2022; Milgate & Cheng, 2006; Scott, 2008 [1, 4, 8, 22, 27, 39]
	Transparent oversight	Culhane-Pera et al., 2018; Johnson, 2007; Scott, 2008 [11, 18, 39]
	Burden related to measurement and reporting	Johnson, 2007; Milgate & Cheng, 2006; Scott, 2008 [18, 27, 39]
Considerations at the micro level
1. Professional values	Motivation	Li & Evans, 2022 [22]
	Collaboration within institutions	Li & Evans, 2022 [22]
	Standardization of care	Khan et al., 2020 [20]
	Professional autonomy	Khan et al., 2020; Kramer, 2012 [20, 21]
	Pressure to conform to standards	Custers et al., 2008; Khan et al., 2020; Li & Evans, 2022; Shurson & Gregg, 2019; Wharam et al., 2009 [12, 20, 22, 40, 45]
2. Conflicting interests	Institutional interests and patients’ interests	Khan et al., 2020; Kramer, 2012; Li & Evans, 2022 [20, 20]
	Performance measures vs. patients’ concerns	Culhane-Pera et al., 2018; Khan et al., 2020; Li & Evans, 2022; Wharam et al., 2009 [11, 20, 22, 45]
	Focus on patient groups rather than individuals	Kramer, 2012 [21]
3. Proportionality or fairness	Individual compensation for workload	Khan et al., 2020; Li & Evans, 2022 [20, 22]
	Incentive distribution within institution	Khan et al., 2020; Scott, 2008; Shurson & Gregg, 2019 [20, 39, 40]

Section 1: key empirical elements of institutional incentives

Institution and population characteristics

Studies distinguish “safety-net institutions”, which are typically in the highest quartile when ranked on the basis of serving low-income patients [7] from “non-safety-net institutions” which serve populations who are less socioeconomically disadvantaged [7, 8, 11, 18, 21, 22, 27, 39]. Depending on the health system context, safety-net institutions could also serve diverse patient groups who face higher social and health risks, such as ethnic minorities, low-educated groups or immigrants [7]. Unlike safety-net institutions, non-safety-net institutions are likelier to be better equipped and have better baseline performances for incentivized targets, which sets them at a relative advantage. Such institutional characteristics lead to normative discussions around distributive justice.

Incentive design

The objectives and design of institutional incentives are important as they determine whether the ultimate goal, that is, improving healthcare provision, is met. How much resources are allocated for such programs and whether institutions contribute directly to the resource pool are also important considerations [18, 22, 27, 39]. The conditions for gaining the incentive also play a role: for instance, incentivizing a fixed threshold of achievement could have different consequences from incentivizing relative improvement as compared to baseline performance [4, 8, 27, 39]. Incentive design characteristics tie in with normative discussions around efficiency and utility, as well as solidarity between institutions.

Performance measures

Incentive programs use performance measures or metrics to assess the extent of institutions’ improvement. Such metrics could be tailored to measure different aspects of the care process or of healthcare outcomes. The selection of metrics and how they are assessed are linked to normative considerations around procedural justice and professionals’ conflict of interest.

Considerations at the macro and meso level

Effectiveness

Effectiveness is the likelihood of the intervention meeting its stated goal or outcome, that is, improvement in the quality of services [27]. This set of arguments explores whether and how institutional incentives could lead to higher-quality healthcare services.

On the one hand, unlike other payment mechanisms such as capitation and fee-for-service, institutional pay-for-performance/institutional incentives can target the quality of services rather than their quantity, as they use quality-related metrics to assess performance [1, 20, 21, 27, 39]. Institutional P4P is effective in improving quality because by rewarding quality, it serves to align managerial, income-oriented goals to broader institutional-level, quality-oriented goals [4, 20]. Additionally, by monetarily rewarding pre-determined aspects of quality, P4P also adds resources to healthcare institutions to support structural and functional improvement, so that patients can receive the best possible care [1, 4, 5, 20]. Authors referred to increased hiring of staff and investments in IT infrastructure as a result of P4P [4, 20].

On the other hand, as Culhane-Pera et al. [11] argue, “healthcare quality measures should not be conflated with measuring quality healthcare”, implying that improvement in quality-related metrics does not always translate into actual improvements in clinical care. This could be due to various reasons, such as the metrics being used as a “tick-box exercise”, metrics not being valid or relevant to clinical care, and inordinate attention to metrics rather than patients [7, 11, 20, 21]. A more detailed exploration of the ethical considerations around quality metrics follows in the section on procedural justice.

Utility

Related to but distinct from effectiveness, utility refers to maximizing the “net benefit” or the likelihood of an intervention delivering optimal health benefits or improved health outcomes to a group or population, commensurate with the cost of/investment in the intervention [22]. As with effectiveness, Milgate and Cheng [27] argue that by rewarding institutions for quality of care rather than quantity, P4P discourages clinically inessential/needless interventions and therefore delivers more value for money. Institutional P4P is therefore more cost-effective than fee-for-service payment systems. Further, by measuring and rewarding health-related outcomes of a population group rather than individuals, institutional P4P can maximize health at a group or population level [7, 8, 22, 27, 39]. One way to target a group or population is to use tailored metrics that are relevant to the population in question, such as measuring smoking cessation rates in a group with a high proportion of smokers [8]. Another way to maximize outcomes for a broader target group is to use metrics that are not disease-specific, but rather capture improved outcomes for multiple conditions [4, 8, 22, 27]. For instance, rewarding effective infection-control processes in a hospital would likely benefit a wider patient group than rewarding improved glycemic control in people with diabetes [4].

However, authors also raised considerations that could undermine the utility of institutional P4P. Firstly, although cost-effective from a monetary perspective, institutional P4P may not compensate adequately for the non-monetary or opportunity costs of measuring, monitoring and reporting the target metrics [22]. For instance, having to report on numerous performance measures resulted in a large “data burden” for clinicians, which detracted from their clinical tasks [22] Such opportunity costs, which are not accounted for in cost–benefit analyses, are “expensive” to clinicians in terms of time and effort, and do not result in benefits to patients [22]. Further, while non-specific metrics could be useful indicators of group or population health, they could be influenced by myriad non-clinical factors related to the individual or to the socio-economic context [12, 20, 22, 27]. This could result in healthcare institutions being penalized for outcomes not within their control.

Solidarity

Solidarity in the context of institutional incentives can be conceptualized as collaborative efforts, building new relationships and collective problem-solving to attain a common goal [22].

Scott [39] proposes that institutional P4P can promote collaborative and solidaristic behavior between institutions, through encouraging the identification and sharing of innovative ways of attaining targets. That happens in practice as well, exemplified in Li and Evans’ qualitative analysis of interviews among hospital networks in Ontario, where institutions exchanged knowledge on how to reach goals (2022). However, in other settings, institutional P4P resulted in an atmosphere of competition and resentment between institutions [11, 16, 20, 22, 40], especially when linked publicly accessible scoring systems [20].

Distributive justice

Distributive justice refers to the principle of allocating (a share of) available resources in a fair and equitable way, not only between different institutions that are incentivized but also across different areas/populations and different health conditions [8].

Li and Evans [22] point out that when incentives are structured around one part of the healthcare system, such as cancer care, parts of the system catering to other conditions receive less priority. This could result in perceived injustice in resource allocation across the system [5, 22]. Some propose adopting a “budget-neutral” or “zero-sum” system of incentives, where a proportion of the institution’s budget is set aside as a contribution to the P4P program [18, 27, 39]. This way, the intervention can be funded without disruptively raising the overall healthcare expenditure and thereby ensuring fair resource allocation over the system as a whole [27, 39]. Further, according to Milgate and Cheng [27] even relatively small contributions from institutions (such as 1 to 2% of the yearly budget) are likely to be sufficiently motivating, as they are likely perceived as being lost from the routine income. Others, however, propose contributions ranging from 5 to 10%, reflecting the absence of a clear consensus on the ideal incentive which is adequate without being disruptive [5, 18, 20, 27, 39, 40].

The most often identified threat to distributive justice is the “inverse equity effect” or “reverse Robin Hood effect”, that is, institutions that cater to low health risk populations tend to achieve targets more easily and be rewarded more consistently [7, 8, 11, 18, 21, 22, 27, 39]. This results in institutions that serve higher-risk populations losing out on rewards, subsequently having fewer resources at their disposal and thus being further disadvantaged, especially in “budget-neutral” institutional P4Ps which pool and redistribute institutions’ contributions (refer previous paragraph). Not only is this unfair at the institutional level, but it could also harm health equity at the population level by allocating fewer resources to already disadvantaged populations [7, 8, 11, 18, 21, 22, 27, 39]. Some propose mechanisms such as exception reporting and social risk-adjustment in performance metrics to account for this disadvantage. Exception reporting allows institutions to exclude ineligible patients from performance metrics, based on their social or medical condition [21, 39]. Social risk-adjustment incorporates compensations based on patients’ socio-economic characteristics and/or medical complexity into performance measures [1, 7, 11, 18, 27]. However, whether and how these can be incorporated in a fair and feasible way remains debated [1, 7, 11, 18, 21, 27, 39].

Procedural justice

The following considerations explore procedural justice, that is, whether the processes that determine resource allocation are fair and just [20]. Articles discuss different aspects of designing P4P programs, measuring institutions’ performance and regulatory oversight that support or undermine procedural justice.

In designing institutional P4P, authors highlight the importance of choosing evidence-based practices and standards to incentivize [1, 12, 18, 20, 21, 27, 39, 45]. Ideally, the evidence base should be evaluated by independent experts to ensure their validity, reliability and usefulness, so that providing incentives to attain these standards is justifiable [18, 27, 39]. However, sound evidence alone does not make for a just intervention, unless it is combined with more holistic knowledge on clinical decision-making [20, 21]. Therefore, it is crucial to involve clinicians in designing P4P interventions, using a participatory approach to ensure that it is feasible and just [18, 21, 39, 45].

In measuring institutions’ performance, the choice and use of metrics or performance indicators have important consequences for procedural justice. Four important considerations are highlighted.

Firstly, the selected metrics should be valid and reliable indicators of performance [1, 11, 12, 18, 20, 22, 27, 39, 40, 45]. This means that they should adequately capture the aspect of care being incentivized, or else they run the risk of distorting the intervention’s goals and producing irrelevant outcomes. Additionally, in order to be valid, metrics should be sensitive to institutions’ efforts to improve and should not be influenced by factors beyond clinical control [12, 20, 22, 27].

Secondly, metrics should, whenever possible, be measured and analyzed at the aggregate level rather than at the level of individual institutions [18, 39]. This is a more fair process because most often, multiple institutions are involved in treating even a single patient, such as GP practices and one or more specialist facilities. It may thus be unfair to attribute a particular outcome to a single institution. However, how to then distribute rewards fairly needs to be decided.

A third consideration in choosing metrics is whether they should reflect quality improvement processes or the final outcomes. On the one hand, process-oriented metrics (such as whether infection-control protocols are implemented) could be considered more fair as they are within institutions’ control, whereas outcomes may not adequately reflect their efforts [4, 22, 39, 40]. On the other hand, better processes may not reliably translate to better outcomes and therefore it may be more fair to measure final outcomes [1, 18].

Finally, authors explore whether measuring relative improvement is fairer than measuring absolute thresholds [1, 4, 8, 22, 27, 39]. Measuring and rewarding improvement (relative to the baseline performance) could be more fair to institutions that have lower levels of performance at baseline, especially those that serve patients with complex needs [4, 8]. However, measuring absolute thresholds can motivate institutions to attain the best possible standards of care [1, 27]. Using combinations of relative and absolute attainment based on interventions’ goals may be necessary [39]

Concerning regulatory oversight, two principles are discussed. Firstly, authors propose that oversight and governance should be transparent, and all processes for collecting and analyzing institutions’ performance should be clearly communicated [11, 18, 39]. However, whether institutions’ performance should be made publicly known is debatable; this could foster competitions and tensions between institutions [11]. Secondly, collecting and sharing data for evaluation should not create unreasonable burdens to institutions [18, 27, 39]. Options to for appeal and redressal should be available without being too onerous [27].

Considerations at the micro level

Core professional values

Institutional incentives do not directly monetarily reward individual professionals; however changes at the institutional level could have important consequences for individual professionals or teams. For instance, recognizing the importance of certain aspects of care could result in improved motivation and involvement in one’s work [22]. It could also stimulate collaboration and solidarity within institutions, as different teams or departments work together to attain outcomes [22]. Promoting evidence-based standards to improve quality of care could also result in impartial and standardized care irrespective of clinicians’ or patients’ characteristics [20]. On the other hand, promoting uniform standards could detract from professional autonomy or clinicians’ freedom to use their experience and discernment [20, 21]. Further, pressure to conform to the promoted standards could result in a culture of surveillance and monitoring, leading to stress, demotivation and fatigue [12, 20, 22, 40, 45].

Conflicting interests or obligations

Institutional incentives could result in conflicting interests for individual professionals. Firstly, by adhering to the intervention criteria, conflicts could arise between promoting institutional interests and patients’ interests [20, 20]. Secondly, there could be conflicts between addressing incentivized metrics and addressing patients’ concerns, with patients’ concerns not being prioritized unless they conform to incentivized metrics [11, 20, 22, 45]. Thirdly, professionals could face conflicts in prioritization between measured or incentivized metrics and non-measurable activities which are important for holistic patient care [11, 12, 22]. For instance, under UK’s Quality and Outcomes Framework, home visits by General Practitioners were not measured or incentivized. This caused physicians to prioritize other incentivized activities, leading to a decrease in home visits, even though they were felt to be an important element of patient care [20, 22]. Fourthly, incentivizing institutional-level or aggregate outcomes may lead to professionals’ focus shifting from individual patients to groups of patients, contrary to traditional medical ethics [21]. For instance, incentivizing lowered glucose levels as a group-level measure/public health goal could result in the group becoming the focus of medical attention, as opposed to treatment being tailored to the individual patient [21].

Proportionality or fairness

Within institutions, healthcare professionals often bear the burden of increased workload to meet the intervention targets. They may not, however, receive monetary compensation commensurate to their workload; this could be perceived as being unfair [20, 22]. Similarly, the distribution of rewards within the institution (for instance among different departments or between managerial and clinical departments) should be done fairly and should recognize the proportionate contribution of each [20, 39, 40]

Discussion

This review aimed to identify and examine the normative arguments and tensions implicit in using institutional financial incentives. From a systematic search we identified 16 articles which entered normative discussions on institutional financial incentives. Compared to the initial 8116 results after deduplication, this is a relatively small body of literature, especially in comparison with normative literature on individual financial incentives for healthcare professionals, which appears to be more extensive. However, though small, this body of literature on institutional financial incentives represented diverse disciplines, namely health economics, medicine, public health/health systems research and health policy. The fact that these disciplines contribute different and sometimes conflicting normative considerations reflects the real-world challenge faced by public health policy makers, who must navigate different perspectives and value systems while making policy decisions [38].

From the reviewed literature, we identified overall three groups/categories of empirical contextual features that shape the normative considerations around institutional incentives, namely institutional and population characteristics, incentive design features and performance measures. These categories encompass a broad range of contextual elements, from concrete program features (such as what is incentivized) to wider socio-political issues such as socio-economic disadvantage. This is in line with literature on the ethical aspects of demand-side incentives (Krubiner & Merritt, 2017; [37]), and to a lesser extent, with normative literature on individual financial incentives for healthcare professionals [3, 15, 31, 33]. This highlights the importance of context especially in economic interventions for public health, and for public health interventions in general [9].

Overall nature of arguments and comparison with literature on individual professional incentives

This review identified normative considerations ranging from systemic (such as distribution of resources across health system) to institutional (institutional culture and dynamics) to individual (individual professionals’ motivation and behavior). While the former two sets of considerations were anticipated, the normative impact of institutional incentives on individuals’ professional ethics was an interesting finding, as most institutional incentives do not, to the best of our knowledge, involve direct monetary incentives to individual healthcare professionals. An exception to this is the UK’s Quality and Outcomes Framework (QOF), in which both individual General Practitioners and Practice Groups are rewarded for quality improvements [20, 21]. However, even reviewed articles that were not based on QOF discussed normative impacts of institutional incentives on individual professionals [11, 12, 22, 39, 40], suggesting that other non-monetary mechanisms might cause a trickle-down effect of the incentive to the individual professional. From the identified arguments, we hypothesize that institutional incentives cause changes to institutional norms and standards, which in turn affect clinical practice and the individual professional, though this hypothesis requires empirical testing.

In this review, most articles forward the idea that institutional incentives are not intrinsically morally questionable, but can be made morally justifiable by adjusting their empirical features. These articles argue that empirical features, such as the intervention’s evaluation, can support or undermine normative principles. For instance, incorporating risk-adjustment mechanisms that account for baseline population characteristics into evaluation could avoid penalizing safety-net institutions and thus promote fairness. This is in stark contrast to literature on personal financial incentives to health professionals, where, based on a prelimnary scoping search, we find considerable skepticism about its overall moral justifiability. This may be due to the more evident tensions between personal financial incentives and medical ethical values such as trust, patient-centeredness and altruism. Most contributors to this literature are also related to clinical medical professions. As this review shows, such critical perspectives on institutional financial incentives seem still lacking, although they influence not only individual professionals but also institutions and the health system.

The importance of normative arguments/ethical values surrounding institutional financial incentives

Mapping the normative aspects of an intervention or policy could be useful to understand the gaps and tensions in empirical research, as shown above. Additionally, it also serves to explicate its embodied normative values, which may other wise remain implicit and/or conceptually vague [6, 24, 30]. By explicating the normative values linked to institutional FIs such as effectiveness, utility, solidarity, distributive justice and procedural justice, we aim to contribute in two ways.

Firstly, explicating norms could promote fruitful engagement and dialogue between different key stakeholders, who might otherwise have differing or conflicting beliefs about the normative values [6, 30]. This may be relevant to institutional financial incentives, as a wide and diverse group of stakeholders is involved.

Secondly, explicating the norms of an intervention could ensure that they are aligned with broader social and political contexts and norms [37]. Deliberating over the intervention’s normative underpinnings could help avoid the issue of morally incompatible policy, which might result when norms remain undeliberated. For instance, while the broader policy goal may be to promote an ethos of collaboration between institutions [19, 43] including institutional FIs as part of policy may introduce an element of competition for recognition and resources. This could be of special relevance to zero-sum incentives, where pooled resources contributed by institutions are redistributed based on performance (NHS [28]. Therefore, it is crucial that such interventions are subject to normative deliberation, where implicit ethical values are made explicit.

Where institutional incentives are concerned, it is unclear whether the practice of explicating and deliberating on normative values occurs openly and consistently. Writing about the inception of the Quality and Outcomes Framework (QOF) in the UK, Kramer [21] discusses how the initial negotiation and planning “happened behind closed doors”, leading to a lack of transparency and trust when it was implemented. They further find it “strange that a major new health policy development can be introduced to an entire population without ethical scrutiny” [21] p.217). However, subsequent editions of the QOF involved more open deliberation, which led to more trust in the program, illustrating the importance of normative deliberation [21]. Therefore, we invite policy-makers, public health experts and health professionals working with institutional incentives to reflect on their practice and further the discussion.

Strengths and limitations

To the best of our knowledge, this article is the first to systematically search and map the normative considerations in institutional incentives, and connect them with empirical contextual aspects. This could prove useful to health policy makers, especially as there is optimism surrounding these interventions. This review also briefly discusses the differences between institutional and personal financial incentives. However, this study has its limitations. All searches and included articles were in the English language, meaning that relevant articles in other languages may have been excluded. Secondly, multiple synonyms have been used to describe what we define as “institutional incentives”, such as institutional pay-for-performance, provider incentives, performance-based financing and many others. We have attempted to capture as many of these synonyms in the search string; however, some may have been missed. Thirdly, a majority of the retrieved articles (eleven out of the sixteen) are based on pay-for-performance programs, which could reflect the predominance of this type of incentive. This could also mean that the normative considerations identified pertain mostly to pay for performance programs, and indicates the need for more discussion on other incentive programs. Fourthly,as Vears and Gillam ([44], p. 117) describe, inductive content analysis requires some level of interpretation of the content on the researchers’ part. This could mean that the process of coding and clustering categories and sub-categories rely not only on the content but also on the researchers’ interpretation of it, which might be open to subjectivity. To the extent possible, the three authors have discussed the categories, sub-categories and their relationships to avoid biases and to limit this subjectivity where possible.

Conclusions

Institutional financial incentives have received relatively less attention in normative literature as compared to individual financial incentives. However, they could have a significant normative impact. Through this review, we identify key normative principles, namely, effectiveness, utility, solidarity, distributive justice, procedural justice, professional values, conflicts of interest and proportionality, that influence the macro-, meso- and micro-levels. These normative aspects are closely interlinked with empirical aspects such as the institutional characteristics, performance indicators and incentive design. Deliberating on the normative tensions of institutional financial incentives could lead to more sound and justified policy, and we invite different stakeholders to reflect on their practice and further the discussion.

Authors’ contributions

R.J. and K.D. conceptualized the study. R.J. searched literature, performed the initial coding and analysis and drafted the manuscript. K.D. verified the searches and provided conceptual input on the analysis. A.R. provided conceptual input on the analysis. All authors have reviewed and approved the submitted manuscript.

Funding

This work was supported by the Katholieke Universiteit Leuven Internal Funds: PhD Scholarships for Researchers from the Global South [ZB/21/029].

Data availability

No datasets were generated or analysed during the current study.

Declarations

Ethics approval and consent to participate

Not applicable.

Competing interests

The authors declare no competing interests.