Facilitating and Inhibiting Factors in the Design, Implementation, and Applicability of Value-Based Payment Models: A Systematic Literature Review

Evidence on the potential for value-based payment models to improve quality of care and ensure more efficient outcomes is limited and mixed. We aim to identify the factors that enhance or inhibit the design, implementation, and application of these models through a systematic literature review. We used the PRISMA guidelines. The facilitating and inhibiting factors were divided into subcategories according to a theoretical framework. We included 143 publications, each reporting multiple factors. Facilitators on objectives and strategies, such as realistic/achievable targets, are reported in 56 studies. Barriers regarding dedicated time and resources (e.g., an excessive amount of time for improvements to manifest) are reported in 25 studies. Consensus within the network regarding objectives and strategies, trust, and good coordination is essential. Health care staff needs to be kept motivated, well-informed, and actively involved. In addition, stakeholders should manage expectations regarding when results are expected to be achieved.


Introduction
Value-based health care (VBHC) primarily aims to improve value for patients, defined as improving health outcomes against equal or lower costs. One way to achieve this is to make the reimbursement of providers partly or fully dependent on achieved patient value (outcomes/costs) (Porter, 2009), known as value-based payment (VBP). To guide health systems to accomplish high-quality, value-based care, the Quadruple Aim serves as a guiding principle. The four elements to focus on are improving patient experience, improving health of populations, reducing cost of care, and improving provider experience (Cattel & Eijkenaar, 2020;Iglesia et al., 2020). To contribute to the evidence on the effectiveness of VBP models, this review focuses not only on factors that enhance these models in any of the four focus elements of the Quadruple Aim but also on hindering factors that should be avoided.
The Health Care Payment Learning & Action Network (HCPLAN) was developed to align payment models across the public and private sectors. For this, an Alternative Payment Model (APM) Framework (2017) was created. This framework is composed of four categories. The first relates to fee-for-service models not linked to quality or value. These are the models for which VBP was developed to substitute. Category 2 refers to fee-for-service models linked to quality and value, whereas Category 3 refers to alternative payment models that build on the fee-for-service architecture. According to Damberg et al. (2014), VBP mostly refers to pay-for-performance (P4P) and bundled-payment schemes. These models are integrated into Categories 2 and 3 of the APM framework, respectively. In P4P models, health care providers receive financial incentives for achieving targets on performance measures previously defined (Eijkenaar, 2012). The improvement and/or achievement of these targets may be value-related (both costs and outcomes; Roland & Campbell, 2014), but may also solely regard quality indicators (Pay-for-Quality; Busse et al., 2019). Depending on the arrangement, providers may have to pay penalties if targets are not met (Cromwell et al., 2007). Bundled-payment schemes are integrated into Category 3 of the APM framework if they relate to clinical episode payments. In these schemes, providers are held accountable for the quality and cost of care delivered during a certain episode. Providers may benefit from shared savings if spending is within a riskadjusted budget and may incur shared losses if the budget is exceeded (Mechanic et al., 2011). This may provide financial motivation for health care professionals to coordinate care during the entire episode (Mechanic, 2015).
In the United States, the implementation of VBP models has been growing exponentially (Fisher & Shortell, 2010). This is done, among others, through accountable care organizations (ACOs). ACOs refer to health care organizations composed of physicians, hospitals, and other health care providers who work together to deliver coordinated care to patients. This payment model ties provider reimbursement to performance on quality measures and reductions in the cost of care. Providers in the ACO agree to accept a financial risk and are eligible for a share of the cost savings achieved through improved care delivery if they achieve quality and spending targets that are negotiated between the ACO and the payer (Damberg et al., 2014). According to the HCPLAN, most ACO arrangements can be placed in Category 3A or 3B, depending on whether the risk arrangement includes only upside share savings or both upside shared savings and downside risk for providers (penalties may be applied), respectively (APM Framework, 2017). Category 4 of the APM framework includes population-based payment models. In these schemes, the financial risk is fully placed on the health care providers. In our study, we focused on models in which both health care providers and the payment agency (insurance companies) share in the financial risk. Therefore, we exclude models belonging to Categories 1 and 4 of the APM framework.
Despite the increased interest and uptake of VBP models, evidence of their potential to improve quality of care and efficiency outcomes is mixed and limited (Damberg et al., 2014;Eijkenaar et al., 2013;Jha et al., 2012;Lindenauer et al., 2007;Mehrotra et al., 2009;Rosenthal & Frank, 2006;Ryan, 2009;Ryan et al., 2012). The limited evidence on impact might be due to context, among others. As indicated by Porter and Lee (2013), two of the six pillars to move toward a high-value health care delivery system are the integration of care across separate facilities and expanding geographic reach (Porter & Lee, 2013). As VBP models with a focus on only one provider might create incentives for the optimization of only part of the total patient pathway delivered by the involved provider, this might hinder patientcenteredness and the integration of care pathways between providers. To address this, VBP models are frequently applied within networks of care (NOC). These are groups of health service delivery sites, connected across all levels of care that share responsibility for health outcomes (Carmone et al., 2020). In addition, VBP models are also regularly applied within transmural care. Transmural care is a type of care provided based on collaboration and coordination between different levels of care such as primary and secondary care (Bloemen-Vrencken et al., 2005). As NOC implies a connection between health center, and transmural care connects primary to tertiary care, these concepts partly overlap but are also complementary. We found it important to incorporate both terms as the integration of care can be between any combination of primary, secondary, and tertiary care. In both NOC and transmural care, multidisciplinary teams of care providers from different levels of care and/or health service delivery sites work together, with an emphasis on collaboration. Having the same indicators across all sites could lead not only to cost savings but also, and above all, to improved quality of care (Porter et al., 2016). The complexity that is introduced by this collaboration provides a barrier to the implementation of VBP models, further hindered by the dependence on the institutional setting and the overall context (Conrad et al., 2014). This complexity might be a reason for the lack of consistent positive results from the implementation of these payment models. Analyzing the factors that hinder or enhance the quality of VBP models can help the broader implementation of these models.
Previous reviews have mostly focused on the design and effects of P4P models as well as on the complexity of the models and subsequent need for tailoring to the specific setting of implementation (Eijkenaar, 2013). There have also been reviews on the economic evaluation of P4P and on the effects of monetary incentives for physician groups (Heider & Mang, 2020). The variables of VBP models that influence costs of NOC in the United States have also been reviewed (Chee et al., 2016). What is lacking is a review of the evidence on all the factors that influence VBP models, not only within a context of NOC and transmural care, but also not restricted to one single country. Such evidence may contribute to the development of new and more effective VBP models.

New Contribution
This systematic literature review aims at contributing to the literature by providing an overview of the factors that facilitate or inhibit the successful development, implementation, and applicability of VBP models in the context of NOC and transmural care context. We also aim to make this overview without restricting our search to any specific countries. With this viewpoint, we expect organizations to take this overview into account when planning to adopt VBP models, gain some knowledge on the factors to support and the ones to avoid, and ultimately create value to their patients. In addition, we decided to take an organizational perspective in this study, rather than a policy evaluation perspective. Our objective was to focus on elements that hospital stakeholders are familiar with and can also change, improve on, and/or avoid. We did focus on the context outside the institutions, which involves the inclusion of external stakeholders such as policy makers.
Theory Greenhalgh et al. (2004) published a multilevel framework for studying diffusion of innovations in health care based on a cross-disciplinary systematic literature review. This framework was built to better organize the determinants of diffusion, dissemination, and implementation of innovations in health service delivery, considering the complexity of the situation and its several interactions. Greenhalgh et al. (2004) proposed this model to facilitate and guide a process of critical thinking about the complex aspects of the innovation and the system in which it is introduced and how these aspects might interact as adoption occurs.
The concept of VBHC has been discussed for over a decade. However, its optimal implementation is a topic of ongoing discussion and may be considered an innovation. Porter and Lee (2013) refer to a need for a completely new strategy when introducing VBHC (Porter & Lee, 2013). Ramos et al. (2021) also consider VBHC an innovation, as multiple, complex systems are present, with several actors influencing each other and others, and blurry boundaries. We aim to look back at what research has been conducted and contribute to innovative ways of implementing VBHC. When introducing VBP models, there may be requirements such as organizational changes, process changes, new policies, or new mechanisms. Facilitating and inhibiting factors of VBP models are critical aspects that affect the effectiveness of these innovative payment models. Hence, to further analyze the facilitators and inhibitors of VBP models, we adapt the framework proposed by Greenhalgh et al. (2004) .
We only extract and keep framework items relevant to the topic of our review and add new ones. Greenhalgh's model encompasses the design and implementation as two important stages in building strong links among the determinants of diffusion, dissemination, and implementation of innovations. We take these two stages as important elements in the introduction of VBP models, and add their applicability and evaluation, related to when the model is already functioning/ being applied. Therefore, our framework (see Figure 1) divides the facilitating and inhibiting factors into three main categories: development/design, implementation, and applicability/evaluation. These three distinguishable stages may have different facilitating and inhibiting factors associated with them. Making such a division may help differentiate these factors by each stage of the introduction of VBP models. To deepen the framework and systematically organize the factors found in the included publications, we divide these three categories into subcategories.
Outer context and psychological antecedents are two elements included in Greenhalgh's model, which we find important to include in the design and development stage of our framework, the former relating to the outside elements and the latter relating to the understanding of VBP models. In addition, funding and organizational structure are two elements included in other components of Greenhalgh's model, which are important to single out and include in a framework for such complex models. Although objectives and strategies, and stakeholders are not directly mentioned by Greenhalgh and colleagues, the conceptual model mentions the importance of clear goals and priorities as well as the involvement of several actors, namely, leaders and managers, which are also vital when introducing VBP models. The implementation stage of our framework is formed by the subcategories: dedicated time and resources, leadership and management, human resource issues, assessment of implications, and concerns in early use. All these subcategories are present, either as elements or as components of elements, in Greenhalgh's model. We consider that resources such as finances, time, knowledge, and leadership are of great importance when implementing VBP models and are warranted separate categories. There is a wide range of resources, and several different factors can be differentiated and allocated separately, so that conclusions can be taken. In addition, we believe assessing the possible effects of VBP models and any concerns in their early use is also important to consider for success in implementing such models. In the final stage, the applicability and evaluation, we allocated the elements of concerns of established users and network structures from Greenhalgh's model, to consider any concerns that arise from the application of VBP models as well as any factors that may influence the NOC that involve such models (an important part of our study focus), respectively. In addition, care coordination is considered by the team to be one of the most important elements in the whole framework as it encompasses any factors related to patient care and the coordination expected from adopting VBP models. The standardization and centralization of resources, services, and training are also warranted a separate category, as well as any inequalities that may arise or be solved by applying VBP models.
There may be some overlap when allocating each factor to a category and subcategory, as factors do not belong purely to a single category. However, to make an overview of facilitating and inhibiting factors and take conclusions, there needs to be allocation. The primary researcher builds the framework, and its elements are discussed among the team, until a consensus is reached.

Methods
As systematic reviews have become progressively important in health care, it is vital to ensure clarity in reporting what is done in the study and what is found (Moher et al., 2009). To achieve this, we applied the Preferred Reporting Items for Systematic Reviews and Meta-Analyses (PRISMA) flowchart and checklists (Supplemental Materials). The flow diagram includes the number of records that were screened and eligibility criteria applied as well as the number of articles that remained to be included in the study. As for the checklists, one outlines the items to include when reporting a systematic literature review, while the other outlines the items to include in the abstract of the study. Prior to the review, the review protocol was registered in Prospero (ID CRD42021259630).
Inclusion criteria are the following: • • Type of payment model-Value-based payment models under Categories 2 and 3 of the APM framework. Category 2 refers to fee-for-service with a link to quality and value, such as P4P models. Category 3 refers to alternative payment models built on fee-forservice architecture, such as shared savings models (with and without downside risk); • • Type of care-Transmural care or care provided in a network of providers; • • Setting-Networks of care, multiple providers, or any other setting that includes more than one provider; • • Analysis of the payment model-Outlining barriers and/or success factors of VBP models.
Exclusion criteria are the following: • • Type of payment model-Non-VBP models such as FFS were excluded. As we only included shared risk models, we also excluded global/budget/populationbased payment models such as capitation schemes described in the APM framework under Category 4. Studies not covering payment models at all were also excluded; • • Type of care-Not mentioning transmural care, which means only referring to primary/secondary/tertiary care singularly, and not combined; • • Setting-Single providers (only one organization); • • Analysis of the payment model-Not mentioning barriers and/or success factors of VBP models.
For a study to be included in this systematic literature review, it had to fulfill all four inclusion criteria simultaneously. If a study met one or more exclusion criteria, it was excluded. By doing so, we ensured all studies were aligned with our study aim.

Search Strategy
The search was conducted in July 2021 in PubMed, PsycINFO, Cochrane Library, JSTOR, EconLit, CINAHL, PsycArticles, and Trip Database.
Three key components were used to build the search terms for the identification of studies on the facilitators and barriers of VBP models: (a) keywords related to VBHC (e.g., P4P or bundled); (b) keywords related to provider payment (e.g., incentive or model); and (c) keywords related to transmural and NOC (e.g., multiple providers or intramural). Synonyms of those keywords and differences in spelling were considered when deciding on the exact search phrase. MeSH terms were also included, when applicable, to ensure comprehensiveness. In addition, the search was limited to publications from January 2005 onward and written in English. We selected this starting period as, based on PubMed searches, VBHC seems to have gained a lot of traction around 2005, with the example of Michael Porter's book on the redefinition of health care being published in 2006 (Porter, 2006). The search was conducted based on keywords present in the title and/or abstract (Supplemental Materials). The search phrase used in each database can be found in Supplemental Materials.

Study Selection Process
The selection process followed three steps: screening based on title and abstract, screening based on full text, and screening of reference lists. The primary researcher was the main screener in both the title/abstract and full-text screenings. In the title and abstract screening, the primary screener reviewed all titles and abstracts, and determined a study's eligibility for the next full-text screening. The inclusion criteria required that the study examined a VBP model and NOC or transmural care. Two other researchers (additional screeners) cross-checked one third of studies that the primary researcher deems "excluded" and all studies "uncertain." If all screeners could not determine if a study met the inclusion criteria, the study was included for full-text screening.
Three screeners, including the primary researcher, participated in the full-text screening. To ensure consistency, the same percentage of studies "excluded" by the primary screener (one third) and all "uncertain" studies were crosschecked by the two additional screeners. In addition to the inclusion criteria used in the first screening, studies were deemed eligible only if they also focused on facilitators/ inhibitors of VBP models. Also, the studies should be peerreviewed. Commentaries, briefs, essays, reviews, and any non-peer-reviewed publications were excluded. Finally, articles for which the full text cannot not be found were also excluded, after attempting to contact the authors and the journal in which they were published.
After the two stages of screening, the final selection included publications of VBP models in NOC and transmural care and outlining facilitating and/or inhibiting factors of these models. We screened the reference lists of all included publications. During the screening of the reference lists, potentially relevant titles that mention VBP models were included for a check. The main researcher screened the title/ abstract and full text using the same inclusion criteria as mentioned before. The included articles were added to the final selection of publications for analysis. Commentaries, briefs, essays, reviews, and any non-peer-reviewed publications, with all other inclusion criteria fulfilled, were once again excluded.
Cohen's Kappa was used to measure the interrater reliability between the primary researcher and the two additional screeners (Supplemental Materials). In both title and abstract screening (κ = 0.984), and full-text screening (κ = 0.853), there was almost perfect agreement.

Quality Assessment
The quality of the quantitative studies was assessed using the "Quality Assessment Tool for Quantitative Studies" by the Effective Public Health Practice Project (EPHPP). The Critical Appraisal Skills Program (CASP) checklist was used to assess the qualitative studies. The quality of the included studies was evaluated as strong, moderate, or weak using these two tools. If a quantitative study had no aspects rated as weak in any sections that composed the EPHPP assessment tool, it was considered a study with high quality (strong). A qualitative study was rated as strong if at least six of nine study questions in the CASP assessment tool were evaluated positively (if the answer was "Yes"). Mixed methods studies were assessed based on both checklists mentioned above. The primary researcher carried out the quality assessment.

Data Extraction and Analysis
After the screening, categories relevant to the review were identified, and information was extracted based on those categories. This directed qualitative content analysis was based on Hsieh and Shannon (2005). The key results were presented narratively and in tables. In addition to information on facilitating and inhibiting factors of VBP models under NOC and transmural care, we also extracted data on the VBP models studied and the outcomes of these models. The themes that formed the units of analysis were (a) general description of the publications and (b) facilitating and inhibiting factors in the design/development, implementation, and applicability of the VBP models studied.
For each included article, facilitating and inhibiting factors were extracted by the primary researcher. Each factor was allocated to one of the three categories formulated and, afterward, to one of the subcategories. After the totality of included publications was analyzed, and factors were allocated, the rest of the author team confirmed the classification made. Any differences in the classification were discussed between all authors until a consensus was reached. Tables, including factors and allocation to categories and subcategories, were built and filled by the primary author (Tables 2 to  4). The rest of the team confirmed their correct classification and allocation. To find each factor in the included publications, several elements were looked at. In some studies, factors were mentioned directly, while in others, facilitators and inhibitors were clear from the context in which they were described. In several papers, the factors might arise from an improvement or deterioration of the outcome measures. Our search included but was not limited to these articles. Therefore, this article focused only on facilitating and inhibiting factors, and not on outcomes. This subject is important for subsequent analysis.

Results
The search yielded 5,988 publications, of which 5,148 were included in the first screening phase, after duplicates were eliminated. Based on the inclusion criteria, 575 publications were eligible for the second screening phase. These publications had their full text reviewed, and 106 publications were kept and included in the systematic review. We also checked the reference lists of those publications, as well as the reference lists of the articles that fit the above inclusion criteria but were not original peer-reviewed articles, and 37 publications were included. In total, 143 publications were included, each including one study. More details can be found in Figure 2.

General Description of the Selected Publications
The main characteristics of the included publications are described in Table 1 Three main approaches to data collection/analysis are applied in the quantitative studies (e.g., using secondary data/patient records) and five approaches in qualitative studies (e.g., focus group discussions). In addition, there are simulation studies and case studies. Most studies (96 publications, 67%) use a quantitative analysis based on secondary data/patient records. Only two studies (1%) use simulations, and 14 publications (10%) report case studies.

Facilitating Factors of VBP Models
Design and Development. The first category of Table 2 includes publications reporting on the facilitating factors for the design and development of the VBP models, divided into six subcategories. One of the subcategories is funding, mentioned in 11 publications (8%). More specifically, sufficient funding facilitates the development of a VBP model. In total, 39% of all studies report facilitators on objectives and strategies (56 studies), and 28% on organizational structure (40 studies). When considering the objectives and strategies of the VBP model, ensuring realistic and achievable targets, associated with adequate, risk-adjusted, incentives, is the most frequently mentioned facilitating factor (Table 3). Rotating quality measures also rewards improvement and is linked with savings and with physicians involved in their design, which is another very prominent facilitating factor in this subcategory. As for organizational structure, the most frequently mentioned facilitator is diversity in the composition of professionals involved in the design and development of the VBP models. The second most frequently described factor in this subcategory is sufficient infrastructure and advanced health information technology within the NOC, to support the introduction of such a new model. In addition, 16 studies (11%) report outer context facilitators. In this subcategory, a model consistent/compatible with location and current policies is the most frequently described element to facilitate its design. In the "psychological antecedents" subcategory, all involved stakeholders having a prior good understanding of the model helps design a successful model (mentioned in seven studies). Also, four studies report facilitators related to the active involvement of stakeholders. Three of those specifically mention the involvement of consultancy firms in the management of the program as helping with the success of developing a VBP model. Table 2 includes factors to aid in the success of the implementation of VBP models. These elements are further divided into five subcategories.

Implementation. The second category in
One of those subcategories includes elements related to dedicated time and resources, present in 11 publications (8%). In total, 10 studies state that previous experience in the development, implementation, and evaluation of VBP models aids in the success when implementing the new model (Table 3). Six studies (4%) mention strong and active leadership and management as a supportive element in aligning and motivating internal and external stakeholders. In total, 12 studies (8%) describe factors in the human resource category. More specifically, high motivation, engagement, and trust among involved stakeholders are stated as elements of the success of VBP models. Only 3% of the studies report adequate access to appropriate care as a facilitating factor. This means a model is more successful in a setting that does not involve an overnight hospital stay for the patient and includes follow-up appointments and/or postdischarge care programs. This element is integrated in the "concerns in early use" subcategory.
Applicability. The third category in Table 2 relates to the applicability of the VBP models and is subdivided into six other subcategories. One of those subcategories is care coordination, in which success elements are present in 36 studies (25%). The most prominent factor in this subcategory is transparency and communication among involved stakeholders, and active involvement of clinicians, reported in 30 and nine studies, respectively (Table 3). Regarding inequalities, perceived fairness among stakeholders in how bonuses are distributed is mentioned in three studies to facilitate the applicability of VBP models. As for the "network structures" category, learning and open networks sharing and analyzing historical claims are considered in six publications (4%) as a facilitator. The same number of studies points to adequate collaboration between clinicians and hospitals, within a network, as an element of the success of a VBP model. Facilitating factors related to standardization and centralization are mentioned in 16 publications (11%). Within this subcategory, standardized training for everyone in the team is stated by nine studies as helping to make a success of VBP models. Seven publications report facilitators about target updates (5%), in which three elements are included. Making sure the  , 5, 17, 18, 21-23, 25, 27, 29, 32, 34, 46, 47, 60, 61, 65, 66, 70, 71, 80, 82, 89, 104, 110, 111, 125, 129, 136, 139, 140, 142 Rotating quality measures, with physicians involved, also rewarding improvement, and linked with savings & new clinical protocols 24 17 1, 3,11,13,18,22,23,25,27,38,39,58,66,72,104,107,121,129,131,132,137,139,141 Outer context Model is consistent and/or compatible with location (adjust for regional differences) and current policies 14 10 7, 17,32,35,65,68,69,82,86,96,99,115,117,119 High-cost/high-performing areas and regions with fewer physician groups  2, 4, 7, 10, 29, 31, 37-40, 52, 55-57, 61, 66, 71, 76, 83, 92, 95, 103, 105, 106, 125, 130, 136, 138, 140  program has the intended effects in regular evaluations and ensuring a relatively stable patient mix and size are two elements mentioned in two studies as facilitators. Rewarding physicians for excluding patients who are included in the program but do not show up, and for including patients that started using the program, is an element of the success of applying the VBP model (three studies).

Inhibiting Factors of VBP Models
In Table 2, the inhibiting factors related to the studied VBP models are reported. The categories and subcategories are the same as in Table 2 and are the same as in the facilitating factors. Only the factors inside each subcategory differ. In Table 4, a detailed version of Table 2 can be found, and the inhibiting factors found are allocated to each subcategory, alongside the reference number of those publications. There are studies that report multiple factors belonging to different categories and/or subcategories, and other studies do not report barriers.

Design and Development.
Regarding the design and development of the VBP models, Table 2 reports on the various factors that inhibit the success of the models. In total, 11 publications (8%) mention the lack of funding as an inhibiting factor. Half of the publications in this category describe inhibiting factors related to objectives and strategies (30 studies). The most frequent factor described as inhibiting the success in the design of VBP models is inadequate outcome measures or targets and/or insufficient incentives to achieve those targets (Table 4). Ten studies report inhibiting factors related to the structure of the organization interested in introducing VBP models (7%). Six of those studies mention that smaller organizations may act as a factor that hinders the  , 9, 23, 32, 40, 63, 67, 77, 84, 89, 94, 99, 102, 103, 109, 116, 121, 132, 133, 139 , 7, 8, 10, 17, 23, 27, 35, 63, 73, 79, 84, 91, 112, 117, 124, 127, 133, 136, 141 Low initial quality of care of organizations 1 1 13 Limited availability of data and/or absence of an integrated information technology system for data registration and sharing High performing organizations with already low benchmarks don't have much "fat to trim") 13 9 8, 32, 34, 46, 68, 87, 90, 109, 124, 129, 134, 139, 141 a Percentages refer to the share of all studies reviewed. Percentages add to more than 100% because studies can report multiple facilitating factors. development of these models as they lack the ability to absorb start-up costs and track medication compliance. Organizations, in turn, get discouraged from making the necessary changes to successfully implement VBP models in the future. Inhibiting factors related to the outer context are reported in six studies (4%). Four of these six studies mention an unsupportive or lack of consideration in adjusting the design of models given the local context (any regional differences that might exist). As stated in our aim, we applied an organizational perspective, and hence will not go in-depth into this element. However, it is important to point out these inhibiting factors and be aware of their effect. Finally, only three studies (2%) mention tensions between involved consultancy firms (stakeholders) as limiting the success of VBP models.
Implementation. Regarding the implementation of VBP models, Table 2 outlines inhibiting factors under this category. In total, 25 studies (17%) report inhibitors related to dedicated time. Overall, 20 of those studies mention that knowing that improvements take a long time to manifest discourages the implementation of VBP models (Table 4). Inhibitors related to human resource issues are mentioned in 14 studies (10%), and nine describe a lack of motivation and engagement among involved stakeholders or lack of uptake of required mechanisms. A lack of trust among the stakeholders involved in the implementation of the models is mentioned in five studies as impeding its success. In the early use of VBP models, there are concerns that hinder its success, reported in nine publications (6%). These concerns are related to the complexity of the structure of the program or lack of experience in implementing required mechanisms. In addition, six studies (4%) mention barriers to the assessment of implications. The most frequently mentioned factor in this subcategory (five studies) is the possibility of opportunistic behavior by providers. This behavior might be translated into selecting healthier patients.
Applicability. Regarding the applicability of the VBP models, publications mentioning these inhibiting factors are reported in Table 2. In 13 publications, inhibitors related to care coordination are reported (9%). Within this subcategory, a lack of transparency and communication among involved stakeholders, mentioned in 10 publications, is stated to hinder the potential success of VBP models (Table 4). Factors such as inequalities are reported in 28 studies (20%). In eight of those studies, a higher workload of staff after implementing the program and payment delays are stated to hinder its success. In 18 publications, there are inhibiting factors related to target updates (13%). Half of those studies mention that inadequate adjustments for differences in patient mix act as a barrier to the success of VBP models. In addition, five studies report concerns in established users (3%). Two studies specifically mention a lack of compatibility with other VBP models or with current practices in the health care system.
Another two studies mention that the inability to respond to constant payment changes may also impede the successful application of the models. Within "standardization and centralization," five studies report inhibiting factors (3%). The most prominent factor (four studies) stated that a lack of job description for involved professionals or lack of a clear and defined standardized approach hinders success. Finally, two studies described barriers related to network structures (1%). One publication specifically mentioned that, when applying a VBP model, a tightly controlled and narrow NOC may provide too few options of care for patients.

Discussion
This systematic literature review presents evidence on the facilitating and inhibiting factors in the design, implementation, and application of VBP models. In our study, there are several factors mentioning stakeholders. This concept refers to anyone with an interest, power, and/or influence in the model, inside (such as managers, nurses, and physicians) or outside (policy makers) the NOC. In addition, articles with weaker quality are not excluded, but triangulated with papers of higher quality, and no singular conclusion is taken from papers with a low quality. The findings show that 40% of the reviewed publications mention objectives and strategies, and organizational structure as facilitators to the design of VBP models. There is a considerable number of studies describing the importance of funding and human resources to the success of VBP models. A similar number of studies describe inhibiting factors in these same subcategories. In addition, an appreciable number of studies outline the presence and lack of prior experience with VBP models as an important factor in the implementation of these models. The inhibiting factor of "excessive amount of time improvements take to manifest" is one of the most prominent factors found in this study. Within the applicability of VBP models, factors regarding care coordination are most prominent, especially in facilitating factors, with 25% of all included studies outlining facilitators that belong to this category.

Targets and Incentives and Diversity of the Team
A focus on attainable targets and benchmarks, motivating incentives and measures, and evidence-based protocols are the most important elements of these payment models in the literature (Counts et al., 2019;Eijkenaar, 2013;Fisher & Shortell, 2010;Heider & Mang, 2020) and might increase efficiency in health care and quality of health outcomes. In case targets and measures are unrealistic and incentives insufficient, they might act as barriers to the design of VBP models (Counts et al., 2019). National and local legislation needs to be taken into consideration as it can affect the network of organizations and be a major differentiating factor in how to design these models (Chee et al., 2016). In addition, a diverse organizational structure composed of professionals with different backgrounds enriches the organization and contributes to an efficient design of the models, according to the literature (Conrad et al., 2014;Eijkenaar, 2013). This facilitating factor, if lacking, has not seemed to hinder the design and development of VBP models. Further study may be necessary to validate this conclusion, comparing the outcomes of using these models, with and without a diverse team.

Financial/Human Resources and Prior Experience
The results imply a possible relation between the success of the implementation of VBP models and the availability of financial and human resources, as well as prior experience with VBP models. In the context of collaboration within and between organizations, a well-functioning VBP model is characterized by engagement, trust, and motivation from everyone, together with strong leadership (Conrad et al., 2014;Eijkenaar, 2013;Kuhn & Lehn, 2015). If trust is not built among the stakeholders, or they are unmotivated, the implementation of VBP models can be inhibited (de Vries et al., 2019). In such a complex scenario of NOC and transmural care, having prior experience with these models is an advantage for the implementation of the model. This suggests that, even if VBP models are different given the different contexts of organizations, there are common elements that, if already assimilated, can reduce some of the intricacies of VBP models. Specifically, strong management involving the staff in the change process and implementation of the models may increase their sense of ownership, and hence keep motivation, engagement, and trust at high levels, as suggested by the literature (Haggstrom & Doebbeling, 2011), regardless of the conditions inside and/or outside of the organization.

Excessive Time Improvements Take to Manifest
Given the complexity of the VBP models, organizations are searching for ways to best implement them, as suggested by the results. With new processes needing to be adapted, tested, and updated, it takes time for improvements to manifest themselves (Heider & Mang, 2020), which might cause friction among the organization's stakeholders and a subsequent loss of trust in VBP models (Counts et al., 2019). A need arises for all stakeholders to manage expectations and consider that it takes time for such innovative payment models to achieve effects (Milstein & Schreyoegg, 2016). This may be done by the management explaining very clearly the details of VBP models to the organization's employees, including when and how improvements in efficiency and health outcomes are expected to take place (creating milestones).

Care Coordination
As indicated by the results, care coordination is a key element in the effective application of VBP models. Transparency and communication among the involved stakeholders are considered vital facilitators, along with involving those stakeholders in the decisions about VBP models. This suggests that, with a high level of transparency and communication, the multidisciplinary team will be more involved and more coordinated, and ultimately perform their tasks more effectively (Kuhn & Lehn, 2015), reducing unnecessary costs, and especially improving quality of care of patients. This collaboration can also go outside the threshold of the organization, namely into the NOC. With hospitals collaborating with each other, for example, by sharing data and communicating about patients, there will be a smoother flow and less waste of both time and money (Kuhn & Lehn, 2015). To achieve a high level of collaboration, results suggest a need for standardization and centralization of resources such as time and knowledge. Training needs to be available for everyone in the team so that everyone knows their tasks and the best way to perform them, keeping autonomy. It may also be that training is needed to explain the idea of VBHC and VBP models as it is not widely implemented (Lally et al., 2019). Without care coordination, communication, and/or transparency among stakeholders, the well-functioning of the VBP model appears to be hindered. This can happen because of misaligned views coming from a lack of standardized training for the team, and a lacking clear approach to take by the stakeholders. However, existing literature does not directly relate these three concepts, and it would be interesting to study how those relate to each other.

Limitations
Although our review is systematic, and we ensured its quality by following the PRISMA guidelines, there are still some limitations. The cross-check of the screening process was partial (see "Methods" section), and the reference check was only conducted by the primary researcher. This may have created some selection bias. Also, publication bias cannot be entirely excluded as the review only included studies published in academic journals. Some of the included publications did not use random sampling for their data collection, and hence their results may not be generalizable. We provided an extensive overview of facilitating and inhibiting factors, but we did not review information on their importance. Therefore, further research should be conducted, resorting to a consensus among experts, to identify which facilitators and barriers are most important.

Conclusion
This systematic literature review shows that the effects of facilitating and inhibiting factors of VBP models in the context of NOC and transmural care are recognized and that there are a variety of factors that influence them. Consensus appeared to be a common element among all factors that facilitate the development, implementation, and applicability of VBP models. There needs to be a consensus among stakeholders in the network, such as physicians, nurses, managers, and policy makers, regarding what are the primary objectives of the models, as well as the strategies to achieve those objectives. With different stakeholders' involvement in the design process, there also needs to be a culture of trust and good coordination in place. Once designed, there appear to be fewer barriers for a VBP model to be successful. Along with the entire network, health care staff needs to be kept motivated and well-informed, as well as actively involved in all phases of the model. It seems to be of vital importance for stakeholders to manage expectations regarding the effectiveness of the models as it takes time to see improvements and the overall added value of VBP models.

Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect to the research, authorship, and/or publication of this article.