Health Equity Implementation Framework
In the Health Equity Implementation Framework, we proposed determinants or domains believed to predict successful and equitable implementation, seen in Fig. 1.[11] Determinants are factors, levels, or components thought to partially explain why implementation would succeed or fail.[26] In the Health Equity Implementation Framework, we adapted three health equity determinant domains to the Integrated Promoting Action on Research in Implementation in Health Services (i-PARIHS) framework,[27] which also proposes a process—facilitation—by which change in each domain would occur.[26, 27] The focus of this manuscript is on three health equity determinants, rather than facilitation, as science is still emerging on how implementation processes should be tailored to promote equity. One does not need to use i-PARIHS to incorporate health equity determinants into another preferred implementation determinants framework.
[insert Fig. 1 here – Health Equity Implementation Framework]
Domains Typical in Implementation Determinants Frameworks. Broad domains typical in implementation determinants frameworks focus on factors at multiple levels, including the individual (e.g., personal characteristics, actors of implementation, individuals receiving an innovation), organization (e.g., clinical service, school, department, factory), community (e.g., local government, neighborhood), system (e.g., school district, hospital system) and policy (e.g., state government, broader laws).[28] These domains can be further specified, such as inner setting or outer setting within an organization.[29] Domains from i-PARIHS are the basis of the Health Equity Implementation Framework and include those typical in most implementation determinants frameworks.[28] Each domain acts to enable or constrain implementation and is defined in Table 1 with illustrative examples and sample measures.
Table 1
Definitions, Illustrative Examples, and Sample Measures of the Health Equity Implementation Framework
Domain & Subdomains | Definition | Illustrative Example(s) | Sample Measuresa |
Characteristics of the Innovation[32] • Underlying knowledge sources • Clarity • Degree of fit with existing practice / values • Usability • Relative advantage • Trialability • Observable results • Evidence for the innovation[67] o Research o Clinical experiences o Patient experiences | An innovation is a treatment, intervention, or practice with unique characteristics that determine how such innovations will be applied in a particular setting. Innovations fall into one of the “7 Ps”: programs, practices, principles, procedures, products, pills, or policies.[31] The innovation should be tailored to the setting’s needs and practices for successful implementation.[32] | A study examined uptake of the Healthy Heart Kit (innovation), a risk management and patient education resource for the prevention of cardiovascular disease, in a primary care setting. They found that relative advantage (innovation was the most comprehensive tool for cardiovascular health) and observable results (evidence-based practice supports innovation) were more influential to uptake of Healthy Heart Kit than other characteristics.[68] | Quantitative: • Decision-Maker Information Needs and Preferences Survey • Electronic Health Record Nurse Satisfaction Survey[69] Qualitative: • Barriers and facilitators assessment instrument • General practitioners’ perceptions of the route of evidence-based medicine • Knowledge, attitudes, and expectations of web-assisted tobacco interventions[69] |
Clinical Encounter (Patient-Provider Interaction) | This is the nature of the interaction between patient and provider. This domain is centered on how the patient and provider choose, adapt, and coordinate the conversation to achieve their shared and personal goals concerning health related matters.[43] The interaction could be influenced by: • Predisposition features which areindividual differences that influence communication that may be objective (e.g., age) and subjective (e.g., self-concept) • Cognitive / affective influences that show how communication is related to strategy (e.g., goals), attributions (e.g., stereotypical) and trust. • Communication influences which refers to how the patient and the provider should tailor their responses to create a coherent and effective exchange.[43] | In studying recordings of HIV patient-provider encounters, there was less psychosocial talk in patient-provider encounters with Hispanic compared to white patients.[42] In a study on predictors and consequences of negative patient-provider interactions among a sample of African American sexual minority women, authors found racial discrimination was most frequently mentioned, and gender and sexual orientation discrimination were also related to negative patient experiences.[70] | Quantitative: • Patient and provider questionnaires about relevant demographics to assess concordance/ match between patient and provider • Patient rating about encounter: Interpersonal Processes of Care Survey[42] • Experiences of Discrimination Scale[71] Qualitative: • Patient qualitative interviews about their experience of care[72, 73] • Clinical encounters coded using audiotapes, analyzed using the Roter Interaction Analysis System[42] |
Recipients[32] • Motivations • Values and beliefs • Goals • Skills • Knowledge • Time, resources, support • Local opinion leaders • Collaboration / teamwork • Existing networks • Learning environment • Power and authority • Presence of boundaries | Recipients are individuals who influence implementation processes and those who are affected by implementation outcomes, both at the individual and collective team levels. Recipients can facilitate uptake of an innovation or resist its implementation.[32] | An example of recipients are stakeholders in a study to identify gaps in care for Indigenous people in Australia. In this example, the following are a few recipient characteristics: • Motivations: commitment to improve Indigenous health outcomes and desire to influence practice and policy • Values: perceived importance of group input, such as discussing data, sharing views and ideas • Knowledge: experience using Continuous Quality Improvement approaches[74] | Quantitative: • Staff implementation ratings of innovation • Point of contact implementation reports assessing the innovation[75] Qualitative: • Point of contact interview[75] • The Agency of Healthcare Research and Quality guides measuring recipient perceptions about implementation[29] |
Recipients: Providers and staff Culturally relevant factors include:[38] • Demographics (e.g., neighborhood immigrant status) • Unconscious / implicit bias • Knowledge and attitudes • Skillsets | In a healthcare setting, providers and staff are the people who administer the innovation. A providers’ objectives and beliefs about a patient affects how they behave during the patient-provider interaction.[76] Providers, especially in busy healthcare settings, may be vulnerable to subconscious bias and stereotypes.[77] | Physicians who consider themselves “liberal” spent more time giving more information to patients than those who consider themselves “conservative”.[43] Providers may engage in more detailed conversations about the health status of educated patients, yet provide basic explanations for less educated patients.[43] During a post-angiogram encounter, physicians perceived patients of lower socioeconomic status as having more negative personality characteristics that include lack of self-control and more negative behavioral tendencies.[41] | Quantitative: • Implicit Association Test to assess implicit bias[78] • Surveys of relevant practice, knowledge, attitudes, or skills[79, 80] • Colorblind Racism Scale[81] Qualitative: • Analysis of taped conversation between provider and patient[42, 78] • Participant observation[82] • One-on-one interviews[83] |
Recipients: Patients Culturally relevant factors include:[84–86, 38, 37, 36] • Medical mistrust • Health literacy and numeracy • Demographics (e.g., neighborhood, immigrant status) • Socioeconomic status, including household income, net wealth, health insurance status, education level • Expectations about therapeutic relationships • Biology/genetics | In a healthcare setting, patients are the people (individuals, families, caregivers) who will actually receive the innovation. Culturally relevant factors are associated with health and health care disparities and can include demographic factors, beliefs, information, and biological or genetic conditions related to equitable implementation. | Asian American patients in Hawaii participated less in their medical visits than mainland Americans.[87] Patients with more formal educations are more expressive and tend to want to play a role in the decision-making process than less educated patients.[43] Many patients are unsure about their role in the encounter and the appropriateness of their participation.[88] We have limited ability to predict treatment success based on whether research was conducted with patients of similar clinical phenotype.[89] | Quantitative:[37] • Telephone survey of a random sample of residents • Medical Mistrust Index[90] • Measures of underutilization of health services • Health literacy question[91] • Health numeracy question[92] • Appropriated Racial Oppression Scale[93] Qualitative: • Interview about expectations for treatment or the patient-provider-interaction[42, 94] • Interviews about experience seeking care[95] |
Inner context (local)[27] • Formal and informal leadership support • Culture • Previous experience of innovation or change • Change mechanisms for embedding innovation • Evaluation and feedback processes | The immediate local setting of implementation. Examples include: • Ward • Unit • Clinic • Hospital department | Among 303 providers working in 49 publicly funded health programs for youths, providers’ perception of certain leadership styles were associated with stronger provider willingness to adopt evidence-based treatments.[96] Pisando Fuerte is a fall prevention program linguistically and culturally tailored for Latino individuals at risk for falls. It is adapted from “Stepping On,” an evidence-based fall prevention program. Fidelity to Pisando Fuerte was subpar; when comparing fidelity between two sites, fidelity was lower in the site that did not give additional time to implement the program (poor leadership support) and had no experience in organizing programs like Pisando Fuerte (no previous experience of innovation).[97] | Quantitative: • Perceptions of Supervisory Support Scale[98] • Organizational Commitment[99] • Readiness for Organizational Change measure[100] • Validated CFIR Inner Setting measures[101] Qualitative:[102] • Site visit • Key informant interviews about inclusivity • Stakeholder meetings or focus groups with providers about their understanding of equitable care • Public forums & listening sessions • Provider and staff interviews to determine actual practice and processes[103] |
Inner context (organizational)[27] • Organizational priorities • Senior leadership and management support • Culture • Structure and systems • History of innovation and change • Absorptive capacity • Learning networks | The organizational atmosphere in which the unit or team is embedded. | Hospitals’ adoption of the Culturally and Linguistically Appropriate Services standards focused on retaining translators and adapting culturally and linguistically appropriate materials. However, this adoption did not often include engagement in broader organizational change.[104] Researchers studied a disparity-reduction program in Israel across 26 clinics and 109 clinical teams. After three years, they found different inner context configurations of factors predicting disparity reduction. One example of a successful configuration were: clinics with a large disparity gap to minimize, high clinic density, high perceived team effectiveness, focused efforts on tailoring services to their enrollees patients.[105] | Quantitative: • Measures of organizational readiness for change[106] • Cultural Competency Assessment Tool for Hospitals[104] Qualitative:[107] • Key informant interviews assessing knowledge/action of policies about equity • Key informant interviews assessing beliefs organization holds about marginalized people • Stakeholder meetings about importance of equitable care • Public forums & listening sessions[108] • Focus groups |
Outer context (healthcare system)[27] • Policy drivers and priorities • Incentives and mandates • Regulatory frameworks or external accreditation systems • Inter-organizational networks and relationships | This is the broader context defined in terms of resources, culture, leadership and orientation to evaluation and learning. There is an increasing amount of research that shows that inequities in obtaining preventative care among minorities compared with whites are due to “organizational characteristics, including location, resources, and complexity of a clinic or practice.”[38] | Researchers examined predisposing, enabling, and need factors as predictors of changes in health care utilization, and found that patients’ experiences differed by group within the healthcare system, and impacted their beliefs and attitudes about receiving healthcare, ultimately affecting the extent to which healthcare services were utilized.[70] | Qualitative: • Archival analysis, reading and documenting policies, program manuals, or procedural protocols[109, 110] • Interviews with leadership[105] Quantitative: • 15 core measures of health care quality[111] • Population surveys • Social network analysis of relationships between relevant leadership and/or teams [105] • Existing reports hospital-wide scores on assessments of care and equity, e.g., National Quality Forum or Healthcare Equality Index[112] |
Societal Influences[35, 44–46] • Economies • Physical structures • Sociopolitical forces • Up-, mid-, or downstream social determinants of health[35] | Forces outside the healthcare system that influence all other domains and subdomains determining success of implementation. May include but be broader than social determinants of health. May focus on presence of stigma and discrimination such as racism, classism, or transphobia (as examples) and the institutionalization of such discrimination in every subdomain of implementation.b,c | In piloting the Health Equity Implementation Framework, societal influences affected receipt of antiviral Hepatitis C Virus medicine for Black patients in the U.S. Veterans Health Administration. One economic facilitator was free Hepatitis C Virus treatment for patients instituted by policymakers. One sociopolitical barrier to implementation was stigma about being diagnosed with Hepatitis C Virus.[11] | See below |
Economies[49] • Traditional • Command • Market • Mixed | The structure of the city, state, or country related to the wealth and resources of people and what is exchanged for healthcare delivery (e.g., insurance). This can be divided into human resources (i.e., labor, management) and nonhuman resources (i.e., land, capital goods, financial resources, and technology).[52] | In a study assessing longitudinal effects of health insurance and poverty, researchers reported low-income, middle-aged adults in the U.S. with no insurance, unstable coverage, or changes in insurance have higher out-of-pocket expenditures and financial burdens than public insurance holders, as well as increased financial burden.[113] In a case study, the presence of chronic kidney disease indicators in the pay-for-performance system in primary care created an incentive for improvement.[27] | Quantitative: • Insurance claims data • Gross Domestic Product[114] • Gross National Product[115] • Minimum Wage[116] • Population & total employment[117] • Annual average wage level of the primary, secondary, and tertiary industries[118] • Tax revenue as percentage of total revenue[119] • Interest rate on saving deposits and inflation rate[120] Qualitative:[121] • Key informant interviews about goods and services exchanged[122] • Analysis of comparative economic structure[121] |
Physical structures • Location • Availability of public transportation • Actual environment of the point-to-care • Language spoken and/or signage • Available structures in one’s neighborhood to use innovation • Grocery stores • Health care facilities • Local businesses • Physical infrastructure | The physical environment, structure, location of services and recipients. Also known as the built environment as it relates to equitable implementation.[52] | In a study comparing Black and white people living in one U.S. city, authors reported that location of residence may affect health inequities. For example, in the racially integrated, low-income neighborhood, inequities in hypertension, diabetes, obesity among women, and use of health services significantly decreased between Black and white Americans.[123] In a qualitative study of transgender individuals’ experiences in residential addiction treatment, researchers observed that residential facilities that split the milieu and housing based on the gender binary may be stigmatizing people who identify as transgender or gender non-conforming.[124] | Quantitative: • Indices of Segregation[125]b • Public data such as: Hospitals per capita, Public transportation trips per capita, Car ownership, Revenue dedicated to parks & recreation, transportation, other infrastructure needs, Grocery stores per capita • Center on Budget and Policy Priorities data • State Departments of Finance and Administration[52] Qualitative: • Windshield & walking surveys include assessing infrastructure; surveyors are on foot and take note of the neighborhood related to physical or built environment.[126] |
Sociopolitical Forces[44, 46, 55] • Policy climate • Political support • Laws • Local culture • Social movements or structures such as racism, classism, heterosexism, transphobiac | Policies and procedures, formal or informal, in national and local governments that systemically inhibit or promote equitable health. | In a U.S. study on adoption of behavioral health evidence based treatment by states, the following were some factors that played a role: state characteristics, state fiscal supports to promote innovation adoption, and state policy supports to promote evidence based treatment adoption.[55] | Quantitative: • The State-Level Racism Index[127]b Qualitative: • INCLENS equity lens: examines whether clinical guidelines address health needs and inequities experienced by marginalized groups[128] • Interview questions with recipients about laws, policies, or social movements relevant to the innovation • Archival analysis[109, 110] |
Notes. a. Measures or data collection methods are examples from literature; for a repository of implementation science measures, see the Society for Implementation Research Collaboration’s Instrument Review Project.[129] b. For a repository of measures specific to racism, see Appendix B of Racism: Science & Tools for the Public Health Professional.[130] c. Implementation scientists should review existing measurement tools specific to health disparities in your area of interest or study to further integrate health equity into implementation. |
Innovation. Innovation refers to the treatment, intervention, or practice or new “thing” to be implemented, adopted by providers and staff, and delivered to patients.[30] The innovation may be a program, practice, principle, procedure, product, pill, or policy.[31]
Recipients. Recipients are the individuals who influence implementation processes and those who are affected by implementation outcomes, both at the individual and collective team levels.[27] In health care, recipients are often many stakeholders typically grouped into providers and other staff, and patients and caregivers.
Context. Context includes different micro, meso, or macro levels that correspond to inner and outer contexts often specified in determinant frameworks.[27] Context can include factors such as resources, culture, leadership, and orientation to evaluation and learning. In this framework, the micro level includes local inner context (e.g., specific ward or clinic), whereas the meso includes the organization (e.g., hospital or medical center). The macro level of outer context includes the wider healthcare system and effect this has on the other domains (e.g., United Kingdom National Health Service).[28]
Facilitation or Process. There are processes by which barriers in implementation domains are solved or overcome, and strengths are harnessed to promote use of an innovation in routine practice.[29] In i-PARIHS, facilitation is the “active ingredient” or process.[32] Facilitation involves implementation strategies that result in implementation coming to fruition.[33, 34]
Domains Known to Affect Health Equity. Three health equity domains have clear, strong associations with inequities in health status, access to, quality of, or outcomes of health care,[35] or there is enough evidence to suggest these domains should be considered.[e.g., 36] The Health Equity Implementation Framework incorporates these domains known to affect health equity: 1) cultural factors, such as medical mistrust, demographics, or biases of recipients, who include patients, providers, and others;[37–40] 2) clinical encounter or patient-provider interaction;[41–43] and 3) societal influences including physical structures, economies, and social and political forces.[44–46] See Table 1 for illustrative examples and sample measures.
Cultural factors of recipients. Recipients in the implementation process are individuals who will consume a particular innovation (e.g., patients, providers).[27] Recipients have unique characteristics, as single individuals and as a group, that influence whether an innovation will be adapted and adopted successfully. Implementers should tailor implementation strategies so they are culturally aligned with beliefs, behaviors, preferences, and needs of all recipients—providers, staff, and patients. Some examples of recipient cultural factors are implicit bias, socioeconomic status, race and/or ethnicity, immigrant acculturation, language, health literacy, health beliefs, or trust in the clinical staff or patient group.[39, 40] Although cultural factors will vary by group and local context, it is crucial recipient cultural factors are considered, particularly when patients belong to a group experiencing a healthcare inequity.
Clinical encounter (patient-provider interaction). This domain recognizes the transaction that occurs between patients and provider in health care, where decisions concerning diagnoses and treatment are made, and providers administer care.[47] It is important the recipients tailor and adapt the conversation accordingly to achieve their individual and shared goals.[43, 48] Factors from both the patient and provider include individual differences in, for example, age and self-concept, pre-existing stereotypes, or lack of trust that could hinder the interaction.[43] There could be subconscious, implicit, or explicit bias from either recipient due to the others’ characteristics, such as race, weight, or perceived sexual orientation. Due to the myriad factors intersecting during the clinical encounter, it is crucial to assess and address it, especially with regards to health inequities.
Societal influence. Societal influence includes three subdomains: 1) economies, 2) physical structures, and 3) sociopolitical forces. This domain is similar to social determinants of health, yet incorporates more upstream determinants (e.g., governance) that have been investigated less relative to mid- or downstream determinants (e.g., neighborhoods).[35] Societal influences also include historical or current discrimination against marginalized groups, such as racism, classism, or transphobia, that may be formally or informally institutionalized within any implementation subdomain. These factors usually occur in broadest levels of the environment (e.g., province, nation), thus affecting downstream the healthcare system, clinics, and recipients. Many societal influences are interrelated, such as a policy affecting a physical structure. It is not as important at this point to distinguish whether a factor is exclusively an economy, physical structure, social norm, or all three; rather, it is important these societal influences are considered and addressed to ensure strategies address these key drivers of societal inequities. For example, Latinx patients in community mental health care (recipients, some immigrants, some Spanish speaking) interface with a service system (purveyor of mental health services, potentially an economic factor). There may be limited purveyors of services in certain geographic regions (physical structure), with higher levels of perceived racism among some providers (social discrimination interacting with recipient factor), and limited purveyors that offer culturally or linguistically aligned care to match needs of monolingual patients who have immigrated to the U.S. (physical structure). Societal influences may not all be assessed in one study or initiative using established measures, due to feasibility constraints, but they should be documented in formative evaluations / initial diagnostic assessments of the implementation problem.
Economies. There are four typical structures of economies including a traditional economy (i.e., mostly agricultural), market economy (i.e., firms and private interests control capital), command economy (i.e., government controls capital), and a mixed economy (combination of command and market).[49] It is helpful to consider how economic structure affects access to resources for implementation. Market forces can be used to change demand for products deemed healthy or unhealthy, therefore driving policy implementation. Examples of market forces include taxes on tobacco, unhealthy food, and soft drinks, or food subsidy programs for women with low incomes.[44] As another example, one study found poverty was concentrated in the U.S. central Appalachia region where mountain top mining was the main income source.[50] These residents had increased exposure to environmental hazards combined with limited resources to address negative health outcomes as a result of those hazards. Implementation in this region would need to account for resources available to residents, their unique health concerns, and healthcare in their local context.
Physical structures. Equity can be affected by how physical spaces, or “built environments,” are arranged and how transition between those spaces occurs for health care.[44] One example in healthcare settings is type and quality of language translations of information displayed (e.g., flyers, waiting rooms)—whether it matches the language of patients served.[51] The location of the healthcare setting in a town or city is important in relation to location of patients[51, 52] e.g., is it difficult for patients to get to the point of care? An example even broader includes an air pollution intervention implemented in the U.S. in which low emission zones improved air quality and had positive effects on mortality rates for all residents, but a greater benefit for the wealthiest, widening mortality inequities.[53] Another example is the implementation of one U.S. state’s naloxone standing order in which pharmacies could distribute naloxone without a prescription.[54] Results demonstrated 61.7% of retail pharmacies had naloxone available without a prescription. However, naloxone availability was lower in neighborhoods with higher percentages of residents with public health insurance—partially a physical structure problem. This finding was particularly problematic due to an increased cost of naloxone for people on Medicaid (public health insurance) as a result of the statewide mandate. These examples suggest health inequity can be maintained or widened through implementation efforts, such that the most affluent experience a better outcome than others.
Sociopolitical forces. Other societal influences are social norms or political forces, which can include but are not limited to political support, laws, and social structures such as racism, misogyny, classism, or heterosexism.[46, 55] Public health policies (e.g., fiscal, regulation, education, preventative treatment, and screening) demonstrate positive and negative effects on health inequalities that occur across health domains (e.g., tobacco, food and nutrition, reproductive health services).[44] As another example, a study examined U.S. state legislators’ behavioral health research-seeking practices and dissemination preferences and found significant variation between Democrats and Republicans, suggesting dissemination materials be tailored to different social norms for different groups.[56] Another example is how ableism (discrimination and stigma against people with mental or physical disabilities) acts as a barrier to implementation.[46]