Conceptualizing and Measuring Trust, Mistrust, and Distrust: Implications for Health Equity and Building Trustworthiness

Abstract

Trust is vital to public confidence in health and science, yet there is no consensus on the best way to conceptualize, define, measure, or intervene on trust and its related constructs (e.g., mistrust, distrust, and trustworthiness). In this review, we synthesize literature from this wide-ranging field that has conceptual roots in racism, marginalization, and other forms of oppression. We summarize key definitions and conceptual frameworks and offer guidance to scholars aiming to measure these constructs. We also review how trust-related constructs are associated with health outcomes, describe interventions in this field, and provide recommendations for building trust and institutional trustworthiness and advancing health equity. Ultimately, we call for future efforts to focus on improving the trustworthiness of public health professionals, scientists, healthcare providers and systems instead of aiming to increase trust in these entities as they currently exist and behave.

Introduction

The relationship between trust and health outcomes has received increased attention, particularly during the COVID-19 pandemic. The pandemic revealed longstanding fissures in the U.S. public health system, and it coincided with pervasive spread of misinformation and one of the largest social movements bringing attention to racism and inequity (21, 34). The convergence of these events facilitated research aimed to understand trust in public health systems. Some of this research found lower levels of trust in public health systems reduced vaccine uptake (3). Since the pandemic began, trust in public health professionals, the government, and scientists has declined (7, 68, 97, 116). Prior research has also reported low trust in medical providers and systems among historically marginalized racial and ethnic populations, arising from ongoing and historical racism (15, 16, 23, 48, 57).

Trust and its related constructs (e.g., mistrust, distrust, and trustworthiness) are inextricably linked to vulnerability (60). Although there is no consensus on definitions, trust is often conceptualized as “willingness to be vulnerable under conditions of risk and uncertainty (12).” In medical and public health contexts, patients are in a particularly vulnerable position because placing trust in untrustworthy entities with power (e.g., providers who misuse their power) can have devastating health consequences. Patients with historically marginalized identities (e.g., due to their race/ethnicity) face increased vulnerability as they frequently experience discrimination when seeking care (18, 67, 89). Therefore, building trust in public health and healthcare systems has been cited as a component of health equity (135).

Building trust is indeed important for health equity, but there is no consensus about how trust is defined, measured, or improved. The purpose of this article is to highlight and discuss common challenges encountered in public health research on trust. We synthesize literature from this wide-ranging field that has conceptual roots in paternalism and racism and has largely focused on U.S. populations (25, 32, 66). We summarize definitions, conceptual frameworks, and considerations for measuring the distinct but related constructs of trust, mistrust, distrust, and trustworthiness. We also summarize studies describing how trust-related constructs are associated with health outcomes and results from interventions in this field. Finally, we discuss key gaps in the literature and provide recommendations for building trustworthiness and health equity.

Conceptualizing and Defining Trust, Mistrust, Distrust, and Trustworthiness

The literature on trust in health systems is replete with contradicting definitions (92). The definition of trust is sometimes tautological (e.g., trust is believing someone is trustworthy and someone is trustworthy when you can trust them). Researchers frequently do not offer the definition of trust used in their studies, nor do they explain how measures of trust relate to their operational definitions. Table 1 provides examples of definitions for trust and its related constructs.

Table 1.

Example Conceptualizations, Definitions, and Measures of Trust, Mistrust, Distrust, and Trustworthiness^*

Construct	Example Definitions / Conceptualization in Existing Literature	Example Unidimensional Measures	Example Multidimensional Measures
Trust	“Willingness to be vulnerable under conditions of risk and uncertainty (12).”	Trust in Physician Scale (13, 120)	Trust in the Health Care Team Scale (105)
Mistrust	Suspicion, sense of unease, or belief that a trustee (e.g., individual provider or medical institution) will not meet an agreed upon expectation, provide care that does not optimize well-being, or worse, actively harm a patient. Mistrust often refers to general attitudes (e.g., towards all providers or the overall healthcare system) (24, 56).	Medical Mistrust Index (71)	Group-Based Medical Mistrust Scale (124)
Distrust	Similar to mistrust, distrust has been conceptualized as unease or suspicion of healthcare entities but typically refers to attitudes about a specific object (e.g., a patient’s primary care provider) (24, 56).	Health Care System Distrust Scale (108)	Revised Health Care System Distrust Scale (110)
Trustworthiness	The extent to which a transaction between a trustor and a trustee results in an outcome predefined as acceptable by the trustor. Trustworthiness is often conceptualized as a property or attribute of an entity (12, 13).	Health Organization and System Trustworthiness Scale (11)	Perceptions of Research Trustworthiness Scale (115)

^*Note: The definitions and measures presented in this table are examples and do not represent an exhaustive list. For additional definitions and measures, scholars can consult reviews of measures and examine recent literature in this rapidly-growing field (1, 84, 85, 92, 118, 132).

The definition of trust can also vary based on scholarly discipline. Philosophers have defined trust based on the fulfillment of expectations (19), psychologists have defined trust as characterological or about attitudes/beliefs, sociologists have defined it as relational to social structures, and economists have defined it behaviorally (e.g., revealed preferences) (53). Public health researchers have benefited from these diverse perspectives and applied definitions based on the context of their studies (e.g., by applying psychology perspectives to conceptualize trust in medical providers as characterological) (118). Trust is considered inseparable from vulnerability because there is no need for trust in the absence of vulnerability (60). In healthcare, patients are in a vulnerable position due to information asymmetry where they typically have less medical knowledge than providers.

Across disciplines, objects (e.g., people or things) or categories of objects are the basis upon which trust is defined (30). The object of trust may be considered a member of a category. For example, providers are members of an occupation with public reputations shaped by the collective experiences of the communities they serve. Personal experiences can, and often do, replace general attitudes toward providers as a category of professionals to shape the level of trust one has in specific providers (58). A patient may have a negative attitude about medical providers generally but have a positive attitude toward the specific provider from which they seek care.

Trust is also defined and studied among objects at multiple levels of the socioecological system. Trust can be studied at the intrapersonal level, where a person’s personality and biology (e.g., neurochemicals) influence their behaviors, and at the interpersonal level, where trust influences relationships between individuals and groups of other people (27). Trust is also studied at the organizational, institutional, system, and societal levels. For example, the Pew Research Center administers polls to assess the percentage of U.S. adults that trust the American people to do various activities (e.g., obey laws) (101). These surveys rely on each respondent’s definition of trust. A close examination of health systems trust literature reveals several related but distinct, arguably, constructs – mistrust, distrust, and trustworthiness.

Mistrust has been proposed as a construct that is separate from a continuum ranging from high to low trust and not the same as no trust (24). Mistrust is the suspicion, sense of unease, or belief that a trustee (e.g., an individual provider or institution) will not meet an agreed upon expectation, provide care that does not optimize well-being, or worse, actively harm an individual. Although there is no consensus, mistrust and distrust are frequently described as comprising similar concepts (e.g., unease and suspicion) but differing in their objects (132). For example, mistrust is used when measuring general attitudes towards categories or clusters of objects (e.g., all providers), and distrust is used when measuring specific attitudes towards specific actors (e.g., your primary care provider) (56).

Furthermore, mistrust and distrust are often conceptualized in the context of marginalized populations where heightened suspicion and vigilance are directed towards systems and individuals that represent dominant cultures and populations with power (24). Mistrust and distrust are rational, active responses to marginalization, historical trauma, racism, and other forms of oppression. As such, scholars have argued that mistrust and distrust are ‘group-based’ because they involve suspicion of those who do not belong to one’s group or of systems where one’s group is poorly represented (24, 124). Group-based mistrust and distrust may focus on any marginalized population, but studies often focus on racial and ethnic groups.

Trustworthiness is recognized as a separate construct but is inextricably linked to other trust-related constructs. Trustworthiness is commonly referred to as the characteristics of a trustee, such as being empathetic, accessible, approachable, respectful, and humble (13, 131). Trustworthiness is also defined by observing trustee behavior (e.g., the extent to which a transaction, between a trustor and a trustee, results in an outcome predefined as acceptable by the trustor). The trustworthiness of an actor is also situational. For example, someone might rate the trustworthiness of their local public health department as high for inspecting the safety of food service establishments (e.g., restaurants) but low for inspecting transportation infrastructure (e.g., bridges).

Problematic Conceptualizations and Definitions: Paternalism and Racism

Many trust-related constructs are rooted in paternalism with assumptions that entities with power (e.g., healthcare providers/systems) will make decisions that are best for those they care for. Medical paternalism is the idea that interfering with patient choice is justified to promote their best interest (32). The implied goal of research using paternalistic conceptualizations is often to get people to trust entities with power as they currently exist when they are vulnerable with medical need. Yet evidence of racism, classism, sexism, and other forms of oppression in health systems challenges whether these systems deserve trust (48, 112, 137).

Like healthcare systems, conceptualizations of trust-related constructs are embedded in structural racism. Substantial trust literature focuses on white populations in western countries and on attributes of trusted entities, like their competence (92). Conversely, mistrust and distrust are often framed as negative characteristics of historically marginalized populations. Prior research has also categorized mistrust and distrust into dimensions of competence and values (e.g., respect, honesty, and caring) and suggests that racial differences in these constructs are largely driven by group differences in values (15). Yet, research in this field rarely attempts to measure values identified by marginalized communities as important in favor of measuring attributes researchers deem meaningful. Furthermore, mistrust and distrust are often described as cultural beliefs and irrational attitudes. For example, conspiracy beliefs about government creation of HIV/AIDS as a form of genocide against Black populations have been cited as a source of mistrust and a cultural barrier among marginalized populations (66). However, Jaiswal and Halkitis aptly argue that this framing is likely racist because it places the duty to overcome medical mistrust on populations experiencing marginalization instead of on the perpetrators of such oppression (66). Scholars studying mistrust and distrust often contest this framing and situate their work in the context of racism and marginalization (24). Yet, the larger ethos of the trust field pathologizes populations and perpetuates racist narratives that mistrust and distrust are negative characteristics of marginalized populations. Other oppressive narratives—such as ones rooted in homophobia, transphobia, and classism—are also used to conceptualize mistrust and distrust. For example, mistrust has been framed as a cultural barrier among gay populations and used to place HIV risk reduction responsibility on individuals rather than systems perpetuating inequities (66). As future work strives for cohesive definitions of trust-related constructs, it is vital that these efforts are rooted in anti-racism and center diverse voices internationally.

Frameworks and Models of Trust, Mistrust, Distrust, and Trustworthiness

Although frameworks of trust, mistrust, distrust, and trustworthiness exist, they often comprise similar concepts making it challenging to distinguish unique characteristics. And, despite growing understanding of how medical paternalism, racism, and marginalization shape trust, few frameworks and models depict structural factors influencing trust-related constructs. Furthermore, few frameworks in this field have been empirically tested or validated. Herein, we describe key frameworks in this field and note conceptual gaps and their implications for health equity.

Trust

Numerous trust frameworks have been introduced in the field. These frameworks are related but conceptually distinct in their focus. For example, the Health Systems Trust: Content Area Framework delves into health systems trust from the standpoint of analyzing measures used in the field (92). Honesty, communication, confidence, and competence were the most common content area dimensions identified in measures while fidelity, system trust, confidentiality, and fairness were included less frequently. Lemmers and van der Voort‘s Trust Framework seeks to demonstrate trust between healthcare providers and patients with their families (76). This framework identifies competence, intention, integrity, global trust, risk, and outcome as the core dimensions of trust. Similarly, Greene and Ramos identified three dimensions which integrate communication, caring, and competence to foster patients’ trust in providers (55). Another framework, The Trust Triangle, posits that trust in organizational leadership is driven by authenticity, logic, and empathy (49).

In another framework focusing on inpatient and outpatient providers, interpersonal trust and trust-by-proxy are the primary domains (81). This model postulates that system-level factors functioning as trust-by-proxy may be more effective in building interpersonal trust among providers in different settings (81). Lee and colleagues recently proposed a framework to increase trust between patients and health organizations that offers nine approaches to build trust (73). Example approaches include leadership fostering a trusting environment and measuring trust in patient care evaluations (73).

Mistrust

A few frameworks depict mistrust. The Framework for the Study of Medical Mistrust and Health posits that injustice and discrimination shape health disparities (24). The framework presents proximal pathways leading to mistrust, including factors such as knowledge, societal resources, stress, and cultural transmission that create behavioral, psychological, and physiological responses. Hammond further considers background factors, identity/socialization factors, recent healthcare experiences, and recent socio-environmental experiences that influence mistrust among African American men (61). This Conceptual Model of African American Men’s Medical Mistrust is one of the first to consider discrimination and masculine role identity as they relate to mistrust among African American men (61).

Distrust

Distrust frameworks are especially scarce. Armstrong and colleagues published the Conceptual Model of Health Care Distrust, which describes distrust in the context of patient, provider, and healthcare system characteristics (17). The model posits that healthcare distrust can affect health status by influencing factors such as adherence to medical recommendations and healthcare utilization. Angelo and colleagues adapted this framework for genomic research (14). In this context, distrust affects use of genomic sequencing results and medical recommendations (14).

Trustworthiness

A recent framework, Trustworthiness and Trust in Health Care Organizations and Systems, posits that cognitive processes and behaviors occur under conditions that require trust, and patient health goals influence trusting behaviors (12). Wolfson further identified dimensions of trust, also known as the 4Cs, to enhance health system trustworthiness (133). These dimensions include competency, caring, communication, and comfort.

While originally developed to build trust in the workforce, the Reina Trust and Betrayal Model, has been applied to healthcare (88, 103, 104). The model’s three dimensions are trust of communication, trust of character, and trust of capability with associated behaviors. Lastly, the Association of American Medical Colleges created the Principles of Trustworthiness: A Community-Academic Partnership Model using a health equity lens, which yields 10 principles to increase healthcare system trustworthiness. These principles include statements such as “You are not the only experts,” and “An office of community engagement is insufficient (36).”

Framework Gaps and Future Research

Frameworks are critical to ascertain mechanisms that determine trust, mistrust, distrust, and trustworthiness in public health and healthcare systems and to create harmonized research approaches. However, few frameworks explicitly consider the causes of health inequities despite their role in determining health outcomes for historically marginalized populations. Notably, a few mistrust frameworks center racism and discrimination and their impact on health outcomes (24, 61), but more work is needed to expand these framings to other trust-related constructs. There is also a need for new frameworks that clarify connections and distinctions between trust-related constructs and how they collectively affect health equity.

Measurement of Trust and Related Constructs

Similar to the literature on definitions and frameworks of trust, mistrust, distrust, and trustworthiness, there is no consensus about the best way to measure these constructs. In this section, we describe factors that scholars should examine when choosing a measure of a trust-related construct and applying it in health equity research.

Choosing the Construct, Object, and Perspective of Interest

Before choosing a measure, it is important to reflect on the specific construct of interest (e.g., trust, mistrust, distrust, or trustworthiness) and its definition/conceptualization. For example, research focused on understanding a trustee’s characteristics may align well with trustworthiness measures, whereas research focused on general attitudes of suspicion towards clusters of objects may fit well with mistrust measures.

It is important to carefully consider the trusted object of interest in measurement (e.g., medical providers in general or a specific provider). In some situations, the object of interest can influence the construct of interest. For example, mistrust and distrust comprise similar concepts of suspicion but differ in objects. Therefore, research examining suspicion of medical providers in general should consider a mistrust measure whereas research on suspicion of a specific provider should consider using a distrust measure.

An additional, but rarely considered, decision relates to whose perspective is considered during measurement. Most existing research focuses on the patient perspective, for example by measuring patient trust in providers (130). However, provider trust in patients may also influence health outcomes (e.g., a provider may not mention an effective treatment option if they do not trust a patient to follow it) (138). Therefore, researchers examining trust from the provider’s perspective might consult review articles on this topic (118, 130) and consider using a previously validated measure focused on provider trust (122).

Dimensionality Considerations

An additional measurement consideration is whether to use a single item, unidimensional scale, or multidimensional scale. Many trust-related studies use single-item measures (e.g., agreement with the item “I trust the healthcare system”) (132). The key benefit to this approach is brevity because a single item usually does not increase respondent burden on surveys. Single items might also be useful to gather preliminary data. However, a single item typically does not capture the complex nature of trust. As highlighted in the frameworks and models section, experts often conceptualize trust-related constructs as multidimensional (i.e., comprising multiple dimensions like competence, communication, and fairness).

An alternative approach is to use a unidimensional scale, which is a set of items each designed to capture a single construct (31). Most trust-related measures are unidimensional (85, 132). These unidimensional measures capture relevant complexity better than single-item measures and may undergo rigorous validation processes. However, unidimensional scales have been criticized for not sufficiently capturing the multidimensional nature of trust-related constructs (105). This issue may be especially problematic when trust is a main outcome. For example, intervention effects may be masked if they operate by improving a dimension that is not sufficiently measured.

Accordingly, when a trust-related construct is a primary outcome, researchers should consider using multidimensional measures. A health system using a multidimensional measure to assess patient trust in their organization may discover lower scores on some dimensions (e.g., fairness) and higher scores on others (e.g., provider communication). Such information can guide interventions, as strategies to improve organizational fairness may be more impactful than strategies focusing on enhancing patient-provider communication. In this same example, using a unidimensional scale may provide the organization with an overall trust score but little guidance on dimensions driving these scores. These benefits must be weighed with drawbacks in selecting measures. Multidimensional scales typically contain several more items than unidimensional scales, adding to respondent burden. Furthermore, caution should be taken before selectively using a few items from a validated scale as this practice may bias results.

Health Equity Considerations for Measurement

Scholars measuring a trust-related construct in health equity research must be careful to select an appropriate measure that includes relevant dimensions. Health equity research often aims to understand ways in which everyone can have a fair opportunity to be as healthy as possible, and some measures support this goal better than others (29). Indeed, prior research suggests historically marginalized populations may place more emphasis on dimensions commonly missed in trust measures, such as fairness, safety, suspicion, and secrecy (15, 92, 115, 123). Furthermore, researchers examining health equity may prefer to evaluate differences in average trust scores across populations (e.g., patients of different races, ethnicities, or genders). However, such comparisons are most valid when prior research demonstrates that a scale measures the construct in the same way across each population (i.e., measurement invariance) (100). Evidence of measurement invariance can increase our confidence that differences in scale scores across populations stem from differences in the underlying construct (e.g., levels of trust) and not measurement error. Evaluating group differences in trust scores is another area where measuring relevant dimensions is important. If a chosen measure does not include dimensions most relevant to historically marginalized populations (e.g., fairness), differences in trust scores across groups may be inaccurate and misleading.

Resources to Guide Measure Selection

After identifying potential measures, a review of evidence for psychometric properties (e.g., reliability and validity) is necessary. Ideally, the chosen measure(s) will have sound psychometric properties with prior validation for use in the population of focus. Although several dozen trust-related measures have been developed, they vary widely in their reliability and validity. Systematic reviews of trust-related measures are potential places to find reliability and validity information for existing measures (1, 84, 85, 92, 132). Table 1 provides an overview of example measures and Table 2 presents an overview of questions scholars should consider when choosing a trust-related measure. Tables 1–2 serve as a starting place to find measures with evidence of reliability and validity and not an exhaustive list of considerations or measures.

Table 2.

Questions to Consider When Choosing a Measure of Trust, Mistrust, Distrust, or Trustworthiness

Question
• What trust-related construct(s) are most relevant to the research question (e.g., trust, mistrust, distrust, or trustworthiness)?
• What is the trusted object of interest (e.g., a specific medical provider, providers in general, a specific medical organization, the general U.S. healthcare system, health insurers, etc.)?
• Whose perspective is relevant to the research question (e.g., patients or providers)?
• Is using a single item, unidimensional scale, or multidimensional scale most appropriate and feasible for the research study?
• What existing measures fit study needs (e.g., in terms of the construct, object, perspective, and dimensionality of interest)?
• Which relevant existing measures have the strongest psychometric properties?
• Which relevant existing measures have been validated in the population of interest in the current study?
• What health equity considerations are most relevant (e.g., is it critical for the chosen measure to assess constructs like fairness and demonstrate measurement invariance across populations of interest)?
• If the population includes individuals from historically marginalized groups, are the needs, priorities, and values of marginalized groups reflected in the measure?
• If no existing measure is ideal, are there opportunities to modify aspects of the study (e.g., the trusted object of interest or desired measure dimensionality) while still answering the key research questions? In this situation, what are the benefits and limitations of proceeding with a measure that is not ideal?
• If no existing measure is ideal, are there enough resources, time, and team expertise/interest to develop and validate a new measure or adapt an existing one using gold-standard approaches (43, 78)?
• What are the strengths and weaknesses of the final measure selected, and what are the relevant implications of these strengths and weaknesses for the overall study?

Options When No Ideal Measure Exists

If researchers are unable to find an ideal measure, one option is to modify aspects of the study (e.g., by using a unidimensional measure if an appropriate multidimensional scale does not exist). If modifications are not possible or appropriate, researchers can consider developing a new measure or adapting an existing one. Before taking this approach, it is important to consider whether there are enough resources, time, and team expertise to develop a high-quality measure (43, 78). Importantly, many existing measures are developed and validated in English, making it challenging to measure trust in global settings where other languages are primarily spoken and among individuals in the U.S. for whom English is not their preferred language. In these situations, translating measures into other languages may be needed, but extreme care should be taken to understand if constructs, response options, and meanings translate as intended in another context (35). Ultimately, researchers should document strengths and weakness of any chosen measure before collecting or analyzing data.

Measurement Gaps and Future Research Needs

Overall, there is little evidence available about the ‘best’ measures in this field, and few measures include dimensions (e.g., fairness across diverse racial, ethnic, gender, and socioeconomic groups) that are particularly relevant in health equity research. Prior research has also found that many existing studies using trust-related measures have not reported psychometrics or demonstrated evidence of validity (85, 92, 118). There is also a need for measures that are brief but capture the multidimensional nature of trust-related constructs. New trustworthiness measures are especially needed, as there are few existing options.

Trust, Mistrust, Distrust, Trustworthiness, and Health Outcomes

Trust-related measures have been linked to several health outcomes including attitudes and beliefs about healthcare, health-related behaviors, self-reported health, and objective clinical measures of health status (4, 22, 24, 25, 45, 52). Evidence linking trust to health-related outcomes has been most studied in populations with chronic diseases, such as diabetes and cancer. Most studies have been conducted in North America, with a few in Europe, Asia, and Australia (25). Despite previously documented differences in trust across racial and ethnic groups, studies often do not report findings disaggregated or stratified by race or ethnicity. However, a few studies focus specifically on marginalized racial and ethnic groups. More recent studies have examined trust and COVID-19 vaccine uptake and use of telehealth (3, 91, 99, 109, 123). Below, we summarize literature on the association between trust and health outcomes. For brevity, we focus mostly on the trust literature as prior literature reviews have summarized the relationships between trust, mistrust, distrust, and health outcomes (2, 24, 25, 63, 118). Much of this literature suggests that mistrust and distrust are negatively associated with health outcomes, such as cancer screening and treatment adherence (2, 24). Few studies focus on trustworthiness and health outcomes. Furthermore, few studies examine quantitative relationships between the different trust-related constructs or how trust, mistrust, distrust, and trustworthiness are differentially associated with health outcomes (95).

Attitudes, Beliefs, and Perceptions

Although not a direct measure of health, individuals’ attitudes, beliefs, and perceptions regarding healthcare experiences may result in or be related to health behaviors and ultimately health outcomes. Higher patient satisfaction scores have been associated with trust in providers among patients with diabetes (13), hypertension (126), and rheumatic conditions (51), as well as in a general population of patients (25) and pregnant women (65). Trust has also been linked to acceptance of treatment including willingness to accept antiretroviral therapy among people with HIV (10), willingness to take immunosuppressive therapy among people with Lupus (128), and willingness to undergo carotid endarterectomy (28).

Health Behaviors

Adherence to a medical regimen is one of the most studied health behaviors related to trust. Higher trust scores are associated with higher or improved adherence to medication and treatment in people diagnosed with various conditions such as diabetes, hypertension, HIV, and cancer (10, 26, 46, 63, 75, 102). A large body of evidence suggests that trust-related constructs are associated with self-reported use of health services such as cancer screenings (2, 61, 63, 86, 98). Similarly, prior research suggests mistrust of healthcare organizations predicts self-reported underutilization of health services including missed appointments, postponing care, and not filling prescriptions (71).

Self-reported Health Outcomes

Growing research has focused on trust and self-reported health outcomes. For example, health-related quality of life is positively associated with trust among people with diabetes (9, 25). Higher trust has also been linked with mental health outcomes, such as lower depressive and anxiety symptoms, in individuals living with diabetes and cancer (8, 114). One study of people who underwent total knee arthroplasty found higher trust correlated with less pain (33).

Objective Clinical Health Outcomes

Studies examining relationships between trust and objective clinical health measures yield conflicting results. Trust has been linked with earlier diagnosis of cancer (79) and laboratory values indicative of chronic disease management (74, 80). Higher trust in providers has been associated with some clinical healthcare utilization outcomes, such as HIV care retention as confirmed by medical records (54). Yet, a meta-analysis examining trust in health professionals and health outcomes did not find significant correlations between trust and clinical outcomes (25). Study authors speculate that the widespread use of cross-sectional studies may limit detection of a trust effect on clinical outcomes, which may be time-delayed and operate through mediating variables like medication adherence (25).

Marginalized Racial and Ethnic Groups and Health Outcomes

Despite previously documented differences in trust-related constructs among marginalized racial and ethnic groups, many studies do not disaggregate or stratify results by race or ethnicity. Some studies do not report race or ethnicity while others categorize diverse populations as ‘nonwhite’ or ‘other.’ For example, a study of patient satisfaction and experience found 24% of participants were nonwhite and 4% were of Latin origin; however, trust scores were not disaggregated (20).

Several studies have focused on examining trust-related constructs in specific marginalized racial and ethnic groups. For example, in a longitudinal study of Chinese older adults, every one-point increase on the Trust in Physicians scale was associated with a 2% lower risk of depressive symptoms (44). Higher race-based medical mistrust, often assessed with the Group Based Medical Mistrust Scale, has been associated with a number of health-related outcomes such as long-term lapse in screening mammography participation among African American and Latina women (83, 95, 124).

Emerging Topics and Future Research Needs Related to Trust and Health Outcomes

Many studies document the positive association between trust and vaccine hesitancy, which received increased attention during the COVID-19 pandemic (3, 62, 93, 136). For example, a study of U.S. adults found lower trust in the medical profession was associated with COVID vaccine hesitancy (111). Furthermore, a rapid systematic review of determinants of COVID-19 vaccine hesitancy found trust in experts, scientists, health agencies, pharmaceutical companies or mass media was positively associated with COVID-19 vaccinations (129).

Increased utilization of telehealth and virtual healthcare during the COVID-19 pandemic sparked growing interest in understanding barriers and facilitation of telehealth adoption. Trust emerged as a crucial factor to promote uptake and satisfaction with telehealth use (91, 99, 109). Telehealth visits were also associated with reduced provider–patient connection, dissatisfaction with care, and mistrust among Black, Hispanic, and Native American individuals (70).

Emerging research is investigating how forms of oppression rarely linked to healthcare, such as police violence and social vulnerability, affect trust-related constructs and health outcomes. Alang and colleagues found that experiences of police brutality is associated with increased mistrust in medical institutions (6) and that mistrust mediated the relationship between police brutality and unmet need for medical care (5). In another study of Asian American ethnic groups, living in higher social vulnerability index areas was associated with lower trust in public health agencies and COVID-19 vaccine accessibility (134). Emerging research has also focused on how misinformation can erode trust and result in preventable illnesses (34). Future research is needed to increase the quality of science on trust and health outcomes and expand our understanding of emerging priority topics, such as misinformation.

Interventions on Trust, Mistrust, Distrust, and Trustworthiness

Due to evidence showcasing how trust is associated with health outcomes, scholars have become interested in intervening on trust. Two systematic reviews examined effectiveness of interventions for improving patients’ trust in providers (82, 107). The latest review found ten randomized controlled trials in North America involving three provider trainings (120, 121, 127), four patient education programs (37, 38, 87, 125), and three trials providing patients with additional information about doctors such as financial incentives (59, 64, 94). Unfortunately, there was insufficient evidence to conclude that any intervention increases or decreases patients’ trust in healthcare providers.

Since these systematic reviews were conducted, additional studies have examined trust interventions going beyond North America and yielded mixed results (50, 77, 106, 113, 119). Some of these interventions found measurable improvements. For example, one study showed that trust in providers in intensive care units improved with family meetings and use of shared decision-making (77), while another found that adding a health coach to the patient care team increased patients’ trust in providers (119). Few intervention studies focus on marginalized populations. Research specifically focused on African American communities found that trust in medical researchers was improved using culturally appropriate educational interventions to increase cancer clinical trial participation (41) and through community listening sessions to disseminate research findings (40, 42). Conversely, several other studies have found no effect of trust interventions (39, 47, 69, 72, 90, 96, 117). For example, trust was not affected in studies evaluating a shared decision-making training program for providers in Japan (117) or in an online video-vignette experiment on decision-making involvement in the Netherlands (96).

Key Intervention Challenges and Future Research Needs

The measurement of trust, time, and funding are key factors that make evaluating trust interventions challenging. Concerns have been raised about the sensitivity of trust instruments, and ceiling effects are often observed given the generally high levels of patients’ trust in providers (107). Moreover, the duration of time needed to intervene on trust can be a challenge because building trust takes time. Furthermore, funding for studies targeting trust as a primary outcome is rare, which often results in trust being relegated to a secondary outcome where insufficient statistical power can diminish the study’s ability to detect changes in trust. Intervention studies are particularly needed that focus on structural factors affecting trust, such as racism. Additional studies are needed to evaluate whether interventions that improve trust are different from those that improve trustworthiness and reduce mistrust or distrust. Importantly, interventionists aiming to improve trust or reduce mistrust or distrust should carefully examine whether the entities of focus deserve to be trusted and if an intervention centered on trustworthiness is more appropriate.

Recommendations to Advance the Science of Trust, Mistrust, Distrust, and Trustworthiness

Table 3 presents key recommendations for advancing the science of trust, mistrust, distrust, and trustworthiness and their implications for building health equity. In these recommendations, we emphasize the need to develop anti-racist definitions of trust-related constructs along with unified conceptual frameworks. We spotlight the need for high-quality measure development and for studies to center health equity when assessing associations between trust and health outcomes. Our recommendations call for funding to support trust-related interventions that prioritize partnership with community members. Finally, we encourage scholars to focus on improving provider and system trustworthiness, as opposed to only focusing on increasing trust in these entities.

Table 3.

Recommendations to Advance the Science of Trust, Mistrust, Distrust, and Trustworthiness: Implications for Health Equity

Topic	Recommendation	Health Equity Implications
Conceptualizations and Definitions	1. Develop anti-racist cohesive definitions of trust, mistrust, distrust, and trustworthiness that are also measurable. 2. Partner with experts from various fields, disciplines, and communities to codevelop cohesive definitions.	Existing conceptualizations and definitions of trust-related constructs are often rooted in paternalism and racism. Such conceptualizations pathologize people, particularly marginalized populations, and often assume that entities with power deserve trust. Efforts to create cohesive definitions should be rooted in anti-racism and center diverse voices internationally, especially those who have historically been marginalized. Building consensus about definitions in partnership with community members and across disciplines will help unify the field and foster ways to measure constructs consistently.
Frameworks and Models	3. Develop unified theoretical and conceptual frameworks that can explain determinants of these constructs and mechanisms by which these constructs affect equity in health outcomes. 4. Incorporate theories and perspectives from diverse fields (e.g., psychology, political science, economics, sociology, and public health) in framework development efforts.	Theoretically informed frameworks will help scientists identify specific mechanisms that may be most appropriate to intervene on to improve equity in health outcomes. Results from intervention studies may in turn inform revisions to frameworks as the evidence grows about successful and unsuccessful interventions. These frameworks should center existing research highlighting historical trauma, racism, and other structural inequities as factors affecting trust.
Measurement	5. Create high-quality measures that address critical gaps in the measurement of trust (e.g., the need for concise multidimensional scales, measures that consider trust from different perspectives such as providers, and measures that are appropriate for use in diverse populations across languages). 6. Consistently include measures of trust-related constructs in large-scale national and international surveys (e.g., the Behavioral Risk Factor Surveillance System survey).	Ensuring measures are appropriate for use in a wide range of diverse populations will facilitate more accurate trust measurement overall and comparisons of trust across different groups (e.g., by reducing measurement error). Improved measurement can also facilitate subgroup analysis to help researchers understand populations that might greatly benefit from community outreach and other efforts aimed to build trust and address health inequities. Furthermore, when high-quality measures are consistently included in national and international surveys, researchers can track trends over time, especially for populations that are often excluded from or not reached in smaller surveys.
Associations with Health Outcomes	7. Conduct rigorous studies to better understand the association between trust-related constructs and health outcomes (e.g., by developing longitudinal studies, research that goes beyond self-reported outcomes, and studies that measure multiple trust-related constructs to provide insights about which constructs are most predictive of health outcomes). 8. Strive to recruit large, diverse samples in research studies so that findings can be disaggregated across populations (e.g., by race, ethnicity, gender, education, etc.).	Although prior studies have documented differences in trust across racial and ethnic groups and other populations, studies rarely disaggregate findings regarding the association between trust and health outcomes. Recruiting large, diverse samples will facilitate analyses that disaggregate results across populations and may help identify inequities in health outcomes that were previously unmeasured. Additionally, conducting higher quality studies can add to a more rigorous evidence base about how trust-related constructs differentially relate to health outcomes and health equity.
Interventions	9. Fund high-quality, theoretically informed interventions aimed to increase trust in and trustworthiness of healthcare entities and measure changes over time. 10. Prioritize partnership and co-leadership with community members in intervention development and evaluation efforts to build trust in the intervention.	Scholars have long posited that improving trust in and trustworthiness of healthcare providers and institutions may reduce health inequities, but there is a dearth of evidence about successful trust interventions. Funding high-quality interventions that prioritize community-academic partnerships may lead to novel interventions that build trust.
Overall Implications for Trustworthiness	11. Focus future efforts on improving the trustworthiness of healthcare providers and systems instead of solely aiming to increase trust in these entities.	Historically, researchers have considered ways to improve trust in healthcare providers and systems as they currently exist, but there is a longstanding history of medical racism and abuse that diminishes the trustworthiness of these entities. Shifting focus towards improving the trustworthiness of healthcare entities may lead to meaningful institutional changes that promote health equity (e.g., efforts to address institutional racism).

Conclusion

The literature on trust, mistrust, distrust, and trustworthiness is vast but there are still key priority areas that are undeveloped yet critical in advancing the science in this area. Currently, there is no consensus on how to conceptualize, define, measure, or intervene on these constructs. Fortunately, there is emerging evidence that can guide researchers in making high-quality decisions (e.g., on measurement) when studying these constructs. As this field grows, it is critical that scholars counter traditional racist framings of trust-related constructs, center health equity, and build a more inclusive and international evidence base. Moreover, the pervasive spread of misinformation in recent years and the erosion of trust in science and medicine underscore the need to better measure and monitor trust in healthcare. These efforts should focus on improving healthcare provider and system trustworthiness instead of aiming to increase trust in these entities as they currently exist.