Introduction

Trust, as a cornerstone of human relationships, applies to the patient-physician relationship. Relationship building is a basic skill for the medical professional [1, 2]. There is evidence that trust between patients and doctors has a positive effect and, if trust is missing, leads to potentially negative consequences. A meta-analysis confirmed that trust was positively associated with improved health outcomes [3] in, for example, diabetes [4], cancer [5], and human immunodeficiency virus infections (HIV infections) [6]. Trust also increases positive behavioral outcomes in patients [7], such as treatment adherence [8, 9]. In contrast, low trust in physicians has been shown to negatively affect various patient health outcomes [4, 6, 10,11,12,13,14]. Economically, if trust in physicians is missing, it has adverse financial effects on healthcare systems [15]. Furthermore, a physician may be more likely to incur complaints when trusting relationships with patients are lacking [16].

In medicine, trust can be understood as being social or interpersonal [17, 18]. Social trust refers to individuals’ trust in institutions or systems, such as the healthcare system or physicians in general, while interpersonal trust refers to the trust between two individuals [18, 19]. Social trust is believed to affect interpersonal trust in medical settings [17, 18]. There are various theories of trust from different disciplines [20,21,22,23]. However, the most prominent interpersonal trust theory in psychology (and applied in medical settings) is from Mayer et al., who defined trust as the willingness of an individual to be vulnerable to the actions of another based on the expectation that the other will perform a particular action important to the trustor, irrespective of the ability to monitor or control the other party [24]. Their theory of interpersonal trust suggests that benevolence, integrity, ability, propensity to trust, and perceived risk are components of a trust relationship [24]. When applied to the patient-physician relationship, the physician’s ability, integrity, and benevolence act as contributors. At the same time, a patient’s propensity to trust—their willingness to trust others—and the perceived risk they take when trusting a physician are also important factors. However, the reality is likely more complex, and there are probably more contributors to a trusting patient-physician relationship than the theory proposes. While different evidence-based studies have investigated the patient-physician trust relationship, to our knowledge, there has been no synthesis of all the evidence-based contributors to the relationship. In 2000, there was a call for an empirical conceptualization of trust. Rather than single theories used to explain interpersonal patient-physician trust or studies investigating isolated contributors of trust, the idea was to synthesize empirical evidence concerning how patient-physician trust can evolve into a model [19]. A recent review on trust in the medical field has renewed the need for such an empirical conceptualization of patient-physician trust [25]. Therefore, this study aimed to summarize the empirical evidence, identify the contributors to a trusting patient-physician relationship, and integrate them into a model. This model can then be used to identify potential approaches and leverage points to improve patient-physician trust. The two main research questions were:

  1. 1.

    Which factors contribute to a trusting patient-physician relationship?

  2. 2.

    Which of these factors can act as potential leverage points to improve the patient-physician relationship?

In addition, we critically assessed contributors based on how they are already implemented in healthcare systems and medical education.

Methods

As the research questions were too broad for a systematic or scoping review, a critical review with a systematic search approach was used to answer the first research question. Critical reviews focus on empirical research [26] to evaluate what is known about a specific topic and integrate it into a framework [26, 27]. They may use a systematic search strategy to integrate the strengths of systematic and critical reviews [27], including all relevant literature, to avoid biases.

Search strategy

We searched the databases MEDLINE (Ovid), Embase (Ovid), PsycINFO (Ovid), and ERIC (Ovid) for a combination of terms (or synonyms) referring to the patient-physician relationship and trust or psychological safety. Database searches were run simultaneously as multifile searches in Ovid. For the results, Ovid’s de-duplicator was used. No study or clinical trial registries or online resources were searched. No experts were contacted, nor was a citation search conducted. A reproducible search for all of the databases is as follows:

Embase (1974 to January 13, 2022), ERIC (1965 to May 2021), Ovid MEDLINE(R) ALL (1946 to January 13, 2022), APA PsycInfo (1806 to January Week 1, 2022).

  1. 1

    (patient* adj2 physician* adj2 (relation* or alliance or rapport)).ti, ab.

  2. 2

    (trust* or psychological safety).ti, ab.

  3. 3

    1 and 2.

  4. 4

    remove duplicates from 3.

We did not use any language, time period, study design, or other restrictions for the searches, and no search filters were used. The comprehensive literature search was run on January 13, 2022 and 630 articles were retrieved. An information specialist assisted in framing the research questions and provided information on the different types of reviews. Once a first draft of the search strategy was developed, multiple feedback rounds with the information specialist were conducted until the search strategy was finalized.

Screening process

Fifty-three records were retrieved from Ovid MEDLINER ALL, 509 from Embase, 1 from ERIC, and 67 from APA PsycInfo. In total, 630 records were found. As OVID’s de-duplication process did not identify all duplicates, any remaining duplicates were removed by SPL using EndNote’s duplicate identification strategy and a manual approach. After de-duplication, 613 articles remained, which were screened in two rounds. The first round was screened according to title and abstract. In the second round, 116 articles were evaluated for inclusion based on the full texts. SPL and RH did the screening, and AL decided when there were disagreements between SPL and RH. A study selection flowchart is shown in Fig. 1.

Fig. 1
figure 1

PRISMA study flowchart

Full size image

We included studies that reported factors contributing to a trusting relationship between patients and physicians and excluded those that only reported contributing factors between patients and health professionals other than physicians (or no contributors). We also included studies that explicitly measured trust between a patient and physician either quantitatively or qualitatively and excluded those with no measure of patient trust in physicians (e.g., only generalized patient trust or trust in other health professionals). We included quantitative, qualitative, and mixed methods papers and excluded dissertations and conference abstracts. Only articles in English and German were included.

Data synthesis and categorization

We first extracted the contributors mentioned in the studies as this review focused on integrating contributors to a trusting patient-physician relationship into an overall model. Extracted data included information on the setting, patients, physicians, how trust was operationalized, and which factors had a positive, negative, or no effect on the relationship. Contributors were then categorized into patient-related, physician-related, context-related, or patient- and physician-related factors. Study sizes and methods of measurement were highlighted. The factors were then synthesized, and the modifiable ones were extracted and displayed in a model.

Results

Forty-five heterogeneous studies reported factors contributing to a trusting patient-physician relationship. An overview of these studies, including the contributors to trust for each study, can be found in Appendix 1.

Patient-related factors

Several patient-related contributors to a trusting relationship were investigated, sometimes with contrasting results from different studies. These included demographic characteristics (gender, marital status, age, ethnicity, birthplace, and country of residence), health condition, health education and literacy, socioeconomic status, religious beliefs, social environment, psychological factors, and the patient’s health condition and status.

Studies found better mental and physical health status tended to positively affect the relationship—although this result was mixed. In several studies, a good general health condition and better self-reported health status were associated with increased trust towards the physician [28,29,30,31,32]. However, other studies found no correlation between self-reported health status and trust. For specific health conditions, low-risk adults without chronic illnesses had higher trust in their physicians than adults with risk factors such as diabetes or high lipid levels [33,34,35,36,37]. Disease progression, including relapses and lack of improvement of a medical condition, was negatively associated with trust [38, 39], whereas a shorter duration of illness increased trust in the physician [29]. However, two studies found no connection between trust, disease duration [40], and healing [36].

Patient health education and literacy levels were found to promote a trusting patient-physician relationship, with higher health education [41] and literacy [42] levels contributing to trust and low health literacy [43] hindering it.

Patient socioeconomic status, including occupation, employment, educational and income levels, and the presence or type of health insurance, were all potential contributors, with high (household) income and educational levels, having health insurance, and being employed positively related to trust; although, these findings were ambiguous. Religious belief was also associated with trust in physicians [31].

The social environment, including social support and the care experiences of family members, further contributed to a trusting patient-physician relationship. In particular, poor social support negatively influenced trust [43], as did dissatisfaction with the care of family members [44, 45].

The health locus of control was also associated with trust. This describes how a person views control of their health. An internal health locus of control suggests that the person sees oneself as controlling their health, whereas an external locus means that the person perceives external factors influencing their health.

Several patient psychological factors, including a propensity to trust, their coping mechanisms and attachment style, the health locus of control, and general trust in caregivers contributed to a trusting patient-physician relationship. Individuals who see powerful others as their health locus of control (i.e., believing other people, such as health professionals, can control their health) exhibited higher trust in physicians [46]. Poor coping styles hindered trust [43], while the willingness to reframe situations (a healthy coping style) added to a trusting patient-physician relationship [44]. For the most part, a general trust in doctors, caregivers, the healthcare system, or online health communities was associated with higher trust [47,48,49]. However, these findings were ambiguous regarding the propensity to trust. One study found that a patient’s propensity to trust predicted trust in their physician [50], although other studies did not find this connection [31, 36]. Table 1 summarizes all of the evidence concerning patient-related factors.

Table 1 Overview of Patient-Related Contributors to a Trusting Relationship
Full size table

Physician-related factors

Demographic characteristics, competence, communication, exploring, caring, provisioning health education, reputation, professionalism, and availability were investigated as potential contributors to a trusting patient-physician relationship.

Demographic characteristics of the physician, such as age and gender, did not contribute to a trusting relationship, although these findings were ambiguous.

Physician competency, including the perceived competence of the physician by the patient [41, 44, 51, 55, 68,69,70], the physician being up-to-date in their specialization [71], and having more years of experience [71] helped to build a trusting relationship with patients. Communication skills, including general communication skills [29, 38, 44, 52, 70, 72, 73], compassion, listening to the patient [41, 44, 52], as well as nonverbal behavior such as good eye contact, providing undivided attention, open body language, and smiling [41, 44, 52, 73] also enhanced the trust relationship as did patient-centered [63, 74, 75], comprehensive care [30].

Physicians exploring a patient’s disease and problems [69], illness experiences [28], and the context of the patient [44, 68] promoted a trusting relationship along with caring behavior [52, 70, 75] such as empathy [50, 76] and compassion [41, 44, 69].

Provisioning health education to the patient contributed to a trusting relationship [38, 41, 64, 68, 69]; however, one study did not find any association between these factors [71].

We did identify physician reputation [71] and the reputation of their medical specialty [28, 34, 39] as contributing to a trusting relationship. Moreover, different aspects of professionalism [71, 73], such as honesty [51, 55, 69] and availability [41], contributed to a trusting patient-physician relationship, while being disrespectful, arrogant, or cynical were negatively associated with trust [41, 73, 75]. These results are summarized in Table 2.

Table 2 Overview of Physician-Related Contributors to a Trusting Relationship
Full size table

Physician- and patient-related factors

Contributors related to the physician and patient were concordance, time spent together, the patient-physician alliance, and shared decision-making.

In relation to concordance, both gender and race were tested as promoters of trust; however, only gender concordance was identified as being a contributor [35, 41].

Time spent together included time spent in a single session, the overall time spent together, and the continuity of care. Most results indicated that more time spent together in a single session [28, 71, 73] (with the physician giving the patient enough time to explain the reason for the visit [77]) promoted trust, whereas physicians appearing rushed was a barrier to a trusting relationship [44]. If the duration of the relationship with the doctor was long-term [28, 36, 77], the patient had higher rates of follow-up visits [51] and more physician visits in general [37, 57]. Nevertheless, those findings were mixed, and not all studies found an association between the duration of a relationship with the doctor [40, 45] and the number of team visits [47, 48]. However, continuity of care [51] and continuity with one physician added to a trusting relationship [30].

Within the patient-physician alliance, alliances in shared decision-making [65] and having a good rapport [71] were found to enhance trust, while a patient’s perception of a physician’s distrust was a barrier [41]. Finding common ground [28] and shared identity [52] were tested but did not show any association with trust. In contrast, shared decision-making contributed to a trusting relationship that promoted trust in most studies [41, 42, 44]. These findings are summarized in Table 3.

Table 3 Overview of Patient-Physician Contributors to a Trusting Relationship
Full size table

Context-related factors

Context-related factors such as practice/institution, physician payments, and additional healthcare services were investigated as potential contributors to trusting relationships.

Most aspects of the practice or the institution were found to contribute to a trusting relationship, with easy accessibility [30] to the practice and a good reputation [71] promoting trust, while institutional betrayal [65] hindered it. The atmosphere of the practice also mattered. A good practice or organizational climate added to a trusting relationship [35], whereas perceived chaos hampered it [29]. Patients having enough physician choice also added to a trusting relationship [48], while managed care settings contributed to mistrust [41]. Inpatient settings enhanced trust compared to outpatient settings [59]. Regarding payments, situations where patients do not know how the physician is paid or the physician is paid by the number of office visits rather than a fixed salary [30] contributed to a trusting relationship. In contrast, public disclosure of payments was negatively associated with trust [78]. Additional health services such as addiction consultations [79], preventive services [77], and the coordination of specialty care [30] also contributed to patient-physician trust. These findings are summarized in Table 4.

Table 4 Overview of Context-Related Contributors to a Trusting Patient-Physician Relationship
Full size table

Potential leverage points to improve a trusting relationship

We integrated the modifiable contributors to a trusting patient-physician relationship from each conceptual group into a model and identified potential leverage points for improving the relationship (Fig. 2).

Fig. 2
figure 2

Model of contributors to a trusting patient-physician relationship

Full size image

Patient-centered leverage points

Within patient-centered factors, health education and literacy, the social environment, and psychological factors were modifiable. A patient who is better educated about health and can understand and use this education for themselves (health literacy) may form better trusting relationships with their physicians; thus, interventions should focus on improving health education and literacy. Patient psychological factors such as coping styles and health locus of control are other potential leverage points to increase trust within the relationship. The social environment, specifically receiving sufficient social support, was a further modifiable contributor to trust, indicating that targeted interventions should aim to improve patients’ social support systems.

Physician-centered leverage points

We identified physicians’ competence, communication skills, exploring, caring, the provisioning of health education, and professionalism as modifiable contributors to a trusting patient-physician relationship. For competence, being up-to-date in the specialization and perceived as competent are leverage points that could increase trust. Communication skills, including verbal and nonverbal behavior, exploring patient health, and professionalism, can also be learned and are, hence, modifiable. Caring, including empathy and compassion, is a skill that can be increased through interventions and also used to increase trust. In addition, physicians can be taught how to provide health education, and specific material can be provided to them for health education, which is another potential leverage point.

Patient and physician-centered leverage points

We identified shared decision-making, the patient-physician alliance, and time spent together as contributors that can be modified. Although time spent together and the continuity of care is context-dependent, awareness can be raised among physicians, and specific training can help the physician allow patients to explain the reason for their visit. Alliances and shared decision-making are skills taught during medical school: therefore, potential interventions already exist. Shared decision-making also includes healthcare professionals other than physicians. Therefore, one possible intervention strategy would be to foster interprofessional education and teamwork to support shared decision-making between patients and healthcare professionals.

Context-dependent leverage points

The healthcare system, provisioning of additional healthcare services, transparency regarding physician payment, and characteristics of the practice or institution (e.g., keeping a good institutional climate and having mechanisms to prevent institutional betrayal) are modifiable contributors; however, these strongly depend on the specific country. Furthermore, only a few studies have investigated contributors to a trusting relationship within this conceptual group. Therefore, the list of context-dependent contributors may be limited.

Discussion

We conducted a critical review with a systematic search strategy to identify evidence-based contributors to a trusting patient-physician relationship and integrated the modifiable contributors into a model. Our results confirm the existing theory of interpersonal trust [24], and, in line with this theory, we found that the physician’s caring (benevolence), competence and communication (ability), and professionalism (integrity) were contributors to a trusting patient-physician relationship. In addition, the physician’s exploring and provisioning of health education also contributed to a trusting relationship. We confirmed the importance of a patient’s propensity to trust as a psychological contributor and were able to add more psychological factors, including coping style and health locus of control. We further added the patient’s level of health education and literacy, and social environment as contributing factors and confirmed that, as the risk a patient must take concerning their health decreases, the easier it is for them to trust the physician. Our model further adds physician- and patient-related factors and the institutional context. The latter indicates the importance of including social trust in understanding interpersonal trust in medicine, as suggested by Mechanic [18]. One highly prominent factor was health education, which can be addressed by the physician, patient, and the context, which suggests that fostering health education is a promising intervention to increase trust.

Patients

Patient psychological factors such as coping styles and health locus of control are modifiable contributors to a trusting relationship. Previous studies have shown that coping styles can be improved for chronically ill patients [80], while other interventions can address a patient’s health locus of control and improve their social support systems. Furthermore, social support interventions have been shown to be effective in patients with different diseases [81,82,83]. Health education could be addressed through e-learning and by provisioning self-help groups that exchange ideas about diseases [84] with educational tools and teaching materials [85]. However, these interventions are system-related as the healthcare system must offer those interventions.

Medical education

Most physician-centered modifiable contributors to a trusting relationship fall under the scope of medical education. Competence is acquired and addressed through university education, graduate school, and continuing education. Communication skills are taught in medical education courses, and professionalism is addressed as a CanMED role [86]. Exploration is an important skill that is already part of communication curriculums [87] and is based on the common-sense model of illness [88]. Physicians can be taught to provide health education [89]; however, it is a skill that medical students find difficult to achieve [90]. Further intervention possibilities could address a physician’s ability to express compassion and empathy. A recent review summarized educational methods used to address medical student empathy [91], with simulation training shown to be an effective tool [92].

A practical example that implements the described practices can be found in the Presence 5 project, which teaches physicians to better listen to patients, explore their story and emotions, and connect with them. These teachings have had positive effects on the physicians’ attitude, compassion, communication, and exploring behavior [93, 94].

Patient- and physician-related factors

As with physician-related contributors to trust, patient- and physician-related promoters of trust could be addressed through medical education. Building an alliance with patients and learning about shared decision-making are skills taught in medical school [95]. The physician can also be made aware that spending sufficient time with a patient is relevant to building trust; however, the ability to modify this contributor is dependent on the healthcare and billing system.

Context-dependent contributors

We found that a transparent billing system and institution-related contributors such as reputation, medical practice atmosphere, accessibility, and additional healthcare services contributed to a trusting patient-physician relationship. A recent discussion on making health care more accessible can be found in Gupta et al. [96].

One healthcare system that addresses many of these factors is Canada’s patient-centered model: ‘the patient’s medical home.’ Under this model, patients can choose a physician they feel comfortable with and who will continuously manage their health care over their lifespan. Each physician is surrounded by a team that considers the patient’s situation and may provide additional healthcare services when needed. This model ensures that each patient receives comprehensive and accessible care that provides sufficient time with the physician and guarantees continuity of care [https://patientsmedicalhome.ca/, 97]. Over the long term, patient medical homes have led to better care, decreased costs, and more satisfaction for providers and patients [https://patientsmedicalhome.ca/, 97]. Other positive aspects of the patient’s medical home, aside from increased continuity of care and the availability of additional health care services, may lie within the aspect of time spent together [98] or improved disease progression [99], which is also addressed within the patient’s medical homes.

Strengths and limitations

The strength of this critical review lies in the systematic search approach, which only included papers that operationalized or specifically described trust. Despite this approach, we cannot ensure that we have included all empirical contributors to patient-physician trust that have been researched. While the systematic search did limit bias in the identified contributors within the critical assessment of what could be modifiable or not, the critical assessment could be biased through the author’s background. However, we discussed the process in depth as a team.

Our search strategy included psychological safety as a synonym for trust, as well as the terms rapport, alliance, and relationship. We checked indexed search terms to ensure the inclusion of relevant synonyms. In the past, trust was more conceptualized as rapport or alliance, whereas today, it is associated with a newer term: “psychological safety.” While we tried to include relevant search terms, we might have missed some, limiting the results.

While our search was not limited to patients trusting their physicians, most papers focused on this and excluded physicians’ trust in their patients. Dyadic analyses of patient-physician trust are scarce. However, Petrocchi et al. (2019) have begun investigating patient-physician trust as a dyad [100]. Some papers only reported correlations of trust with unmodifiable, less relevant, but easy-to-gather factors, such as sex or age. Thus, more contributors to trust may have yet to be investigated.

Implications for future research

Interestingly, many non-modifiable or insignificant contributors, such as physician or patient demographics, were investigated in almost every study we reviewed. However, the most promising contributors, such as health education, were barely explored. Future research should investigate modifiable and promising contributors to a trusting relationship that have, as yet, been barely researched, including patient psychological factors and additional healthcare services. Additionally, factors that have not been investigated should be addressed, including digitized healthcare settings and how telemedicine, chatbots, and video consultations affect patients’ trust in physicians. Further research should also focus on measuring how successful physician interventions are, as previous research and interventions have not increased patient trust [101, 102]. Future interventions should also consider multiple contributors to trust, as they are all related. For such interventions, the outcomes for each contributor should be evaluated first, with trust as a secondary outcome.

As the present review aimed to create a model of patient-physician trust, only studies that included trust between patients and physicians were included, with other healthcare professionals excluded. However, research has already acknowledged the importance of trusting relationships for all healthcare professionals [103], which should be further expanded. Thus, shared contributors to trust between healthcare professionals, their differences, and potential leverage points should also be identified.

Implications for practice

Our critical review has demonstrated that there are more contributors to a trusting patient-physician relationship than the theory of interpersonal trust proposes, and the context in which the patient-physician relationship takes place is highly relevant. One way to increase trust within the patient-physician relationship is to implement healthcare systems that are organized similarly to the Canadian ‘patient’s medical homes’ model. Changing the healthcare system is also an effective tool to simultaneously address multiple contributors to trust.

At the level of the institution, enhancing trust should focus on health education, which can be addressed through the implementation of self-help and support groups, providing high-quality health educational material, and training healthcare professionals.

At the physician level, we recommend taking as much time as possible for each patient to explore their perspective and current situation, organize (as much as possible) continuity of care, and ensure patient health education.

Conclusion

Using a systematic search, our model summarizes identified modifiable contributors to a trusting patient-physician relationship. Providing sufficient time during patient-physician encounters, ensuring continuity of care, and fostering health education are promising leverage points for improving trust between patients and physicians. Future research should evaluate the effectiveness of interventions that address multiple modifiable contributors to a trusting patient-physician relationship.