Sage Journals: Discover world-class research

Abstract

Previous research has demonstrated associations between social network characteristics and health care utilization. However, understanding the role of network- and individual-level factors together remains limited. We use the Person to Person Health Interview Study (2018–2020) to show network-level and individual-level predictors of two health care utilization measures: having a regular doctor (preventive care) and seeking medical services after an acute health care problem (reactive care). We explore preventive care among the full sample (N = 2,524) and reactive care among the subsample that reported a physical or mental health problem in the past year (n = 885). We measure medical trust as an individual-level characteristic (personal trust in physicians) and a network-level characteristic (average trust in physicians among social ties). Results show that higher medical trust in doctors within one’s network positively predicts both having a regular doctor and utilizing medical services after a health care problem, even after adjusting for known predictors of utilization. Our findings provide support that a network’s pro-medical culture matters beyond individual medical trust for health care utilization.

Keywords

medical trust health care utilization social networks

Health advocates have invested great effort in encouraging people to seek medical care because these behaviors improve long-term health, reduce health care costs, and help alleviate the burden on the health care system (House 2015; McWilliams et al. 2007; Starfield, Shi, and Macinko 2005). Specifically, preventive physician visits play a vital role in maintaining overall well-being and addressing emerging health concerns (Blewett et al. 2008; Levinson, Gorawara-Bhat, and Lamb 2000). Furthermore, utilizing medical services after a physical or mental health diagnosis is critical for managing immediate health crises and preventing further complications (Bindman et al. 1995; Perry and Pescosolido 2015). Despite the multiple types of health care utilization, existing scholarship often relies on measures of help-seeking that combine preventive care and medical services that are sought in response to a new health problem (i.e., reactive care).

One’s social network, defined here as the set of ongoing personal relationships through which individuals exchange information, emotional support, and practical assistance, can influence health behaviors, treatment-seeking, and medical compliance (Christakis and Fowler 2007, 2008; Latkin and Knowlton 2015; Rosenquist et al. 2010; Rosenquist, Fowler, and Christakis 2011; Valente 2010).¹

Individuals with health problems who confide in network members supportive of medical care report better health outcomes than those who do not have networks supportive of medical care (Perry and Pescosolido 2015). Although trust in doctors is a known predictor of medical utilization, it operates both individually and on the network level (Birkhäuer et al. 2017; Boulware et al. 2003; Hall et al. 2001).² For example, whether personally trusting of physicians or not, a patient might be more likely to attend a preventive visit if family members reinforce the importance of routine care or even schedule the visit on their behalf. To address both types of medical trust, we use two trust-related variables for the analysis of health care utilization: network trust and individual trust in physicians.

In this article, we ask: How does the relationship between medical trust and health care utilization vary by level of trust and type of utilization? Our research contributes to the current body of knowledge in three ways. First, we use both social network theory, specifically, the network episode model (NEM), and cultural sociology to underscore the importance of personal networks for health care utilization. Second, we use both individual- and network-level measures of medical trust in our multivariate models, which allows us to distinguish differences between these two levels of trust in health care utilization. Lastly, by examining two different health care utilization measures, we gain insights into how individual trust and network trust may be differentially associated with preventive care versus reactive care. To accomplish this, we use respondent demographics and egocentric network data from the Person to Person Health Interview Study (P2P; 2018–2020), including a full sample of respondents that completed the positive-relationship name generators (N = 2,542) and a subsample of respondents who report a physical or mental health problem in the last year (n = 885).

Background and Theory

In the following paragraphs, we develop three arguments central to this article. First, we use the NEM to explain why we expect network medical trust to influence health care utilization even after accounting for individual medical trust. Second, we also discuss network-level medical trust and propose that the cultural theory of hybrid habitus can help explain how norms and network influences shape medical utilization. Third, we theorize potential differences between preventive care, measured by having a regular doctor, and reactive care, measured by seeking medical services in response to a health care problem in the past year.

Network Episode Model

The NEM emphasizes that individuals are socially embedded. Thus, their network interactions influence decisions regarding health care in multiple ways, including the construction of norms and treatment options (Pescosolido and Boyer 1999). Even before illness onset, individuals are engaged with their social networks, seeking information, and sharing experiences related to health. (Pescosolido 1992). The trajectory of an illness begins with a “precipitating episode,” which could range from the emergence of symptoms to a health-related problem. This pivotal moment triggers illness-specific interactions within an individual’s network, prompting exchange regarding the perceived health problem. For example, an individual experiencing persistent chest pain may perceive the symptoms as more serious and be encouraged to quickly see a doctor through discussions with family members who have experienced similar symptoms and sought prompt care. Alternatively, in a network where members have downplayed similar symptoms, the same individual may be more likely to defer seeking formal medical attention. In both cases, the network is active in constructing a shared understanding of whether the health issue requires formal health care or can be informally managed. Importantly, those who are skeptical of the medical system may still view themselves as acting appropriately if their networks reinforce norms that favor medical mistrust or avoidance.

Cultural Theory and Hybrid Habitus

Individual health, perceptions of illness, and health care utilization are culturally situated (Bourdieu 1977; Marmot 2005; McLean 2016; Pescosolido et al. 2007). Drawing on Bourdieu’s (1977) concept of habitus and Sewell’s (1992) ideas about multiple intersecting structures, Lo and Stacey (2008) coined the term hybrid habitus to capture how orientations toward health and health care are structured both by collective norms within social networks and by individual-level variation in these orientations. Hybrid habitus emphasizes that health behaviors are shaped by both one’s immediate social context and by broader, historically rooted cultural structures. This approach emphasizes that individuals are not merely passive recipients of cultural norms but actively navigate and negotiate these structures, drawing on available cultural resources to make sense of their health care experiences.

Hybrid habitus also captures the ways that social structures shape patient orientations toward health. For instance, the availability and strength of social ties, the range of supportive functions available within one’s network, and the density of connections can all influence the formation of medical trust (Berkman et al. 2000; Cornwell and Waite 2012; Lin 2002; Pescosolido 2006). Networks that provide diverse forms of support, for example, may expose individuals to a broader range of health-related norms and expectations, potentially reinforcing collective orientations toward trust in medical institutions or, alternatively, deepening skepticism if negative experiences are shared (Decoteau 2013; Sewell 1992). The capacity to act on health-related advice may be constrained by social location because individuals embedded in smaller or more isolated networks may have fewer opportunities to seek second opinions or access alternative sources of information, reinforcing their reliance on local norms and practices (Shim 2010). These structural factors shape both the opportunities individuals have to access health-related information and the social norms that guide how they interpret and act on this information.

Experimental work on cultural trust demonstrates further that collective orientations toward trust within one’s social environment can shape individual behaviors even when personal trust is held constant (Doyle 2021, 2023). Network-level trust is thus a collective phenomenon that can independently shape health behaviors, reinforcing the idea that hybrid habitus is a layered, dynamic structure that shapes both collective norms and individual experiences.

Medical Trust as Hybrid Habitus

Medical trust is one expression of hybrid habitus, reflecting both collective values within one’s social and cultural networks and individual experiences with medical institutions. Trust is a dynamic, context-dependent construct rather than a fixed disposition (Möllering 2013). Medical trust, then, should be understood as a practice that emerges in specific contexts, shaped by the quality of interactions within medical systems (Eyal, Au, and Capotescu 2024). Mistrust can also accumulate over time, often driven by experiences of exclusion, discrimination, or perceived incompetence within medical settings (Decoteau and Sweet 2024; Benjamin 2014; Washington 2006). However, mistrust is not solely a function of individual experiences. It can also reflect broader social structures and collective orientations, as evidenced in examples of those with high socioeconomic status rejecting medical advice in contexts where their networks reinforce skepticism toward medical interventions (Reich 2020).

Those embedded in networks with higher medical trust may be more likely to seek regular, preventive care, viewing it as a normative and beneficial practice. Conversely, individuals in networks with low medical trust may approach health care seeking reactively even if they personally trust doctors. This distinction between individual and network orientations aligns with the hybrid habitus framework because it captures both collective norms and individual experiences in shaping health care decisions. Medical trust is therefore not only a reflection of personal beliefs but also a complex orientation shaped by one’s position within multiple, overlapping social structures, including the availability and quality of social ties, the density of one’s support network, and the cultural expectations that circulate within these networks (Lo and Stacey 2008).

The NEM’s illustration of the illness trajectory in conjunction with medical trust as hybrid habitus highlights how networks influence health care utilization not only by influencing individual-level attitudes but also through the construction of appropriate options and norms. Based on the NEM and cultural theory, we hypothesize:

Hypotheses 1a and 1b: Even after accounting for an individual’s personal medical trust, we predict those with network members that have higher levels of medical trust are more likely to utilize (Hypothesis 1a) preventive health care and (Hypothesis 1b) medical services in response to a new health problem.

Health Care Utilization

Preventive health care and health care in response to a medical problem (reactive care) represent distinct strategies for addressing health care needs. Having a regular doctor—especially key for preventive care—allows for ongoing monitoring and early detection of potential health issues, reducing the likelihood of severe complications and the need for emergency care. Regular, preventive visits can establish a baseline for individual health, making it easier to detect changes over time and intervene early, which can lower long-term health care costs and reduce the strain on emergency services (Gray et al. 2018; House 2015; Sanmartin and Ross 2006). Moreover, those with regular doctors tend to have higher incomes, greater social capital, and overall higher health care utilization rates (Bataineh, Devlin, and Barham 2019; Devlin and Rudolph-Zbarsky 2014; Thanh and Rapoport 2017).

In contrast, seeking health care in response to a new health problem captures a different pattern of utilization. Individuals who lack a regular doctor may only seek medical care when symptoms become concerning, often addressing acute issues rather than maintaining consistent preventive care. This reactive approach can contribute to less continuity in care and may result in greater reliance on hospital and emergency services because these individuals may not engage with health care providers until prompted by immediate health concerns (Blewett et al. 2008; Koplan et al. 2009; Lambrew et al. 1996; Xu 2002). Although this approach addresses emerging health needs, it often lacks the proactive benefits of preventive care and can lead to higher overall health care utilization and a greater risk of adverse health outcomes over time.

Given that these types of health care utilization are distinct and that they may capture different populations, we expect that individual medical trust may act differently. Preventive care is dependent on established trust and ongoing relationships because individuals are more likely to engage in continuous, regular medical care if they trust their health care provider. Preventive care also involves engaging with the health care system proactively, often without an immediate or tangible need. This proactive engagement requires a level of trust that the benefits of care—such as screenings, checkups, and vaccinations—are worth the time and financial investment despite the lack of immediate symptoms.

Health care sought in response to a new health problem is often driven by acute symptoms or immediate health concerns, which can override more reflective considerations, such as individual trust (Pescosolido and Boyer 1999). In these situations, the urgency of the health problem is the motivator for seeking care regardless of one’s broader trust in medical providers (Mechanic and Meyer 2000). As a result, individual medical trust may not predict episodic care because the decision to seek treatment is often based on immediate physical need rather than established trust. Based on these proactive versus reactive differences in motivation, we hypothesize:

Hypotheses 2a and 2b: After controlling for network-level medical trust, individuals who have higher levels of personal medical trust will (Hypothesis 2a) be more likely to report having a regular doctor but (Hypothesis 2b) be neither more nor less likely to report seeking medical care in response to a health problem.

That is, we expect a positive relationship between personal trust and preventive care but a null relationship between personal trust and reactive care.

Data and Methods

Person to Person Health Interview Study

To test our hypotheses, we used the first wave of the P2P study, which employs an egocentric network questionnaire that asks an individual (ego) to identify and describe members (alters) of their social network (Perry, Pescosolido, and Borgatti 2018). The P2P study was created to explore health-seeking pathways (Green and Pescosolido 2024) and was composed of a population-representative stratified probability sample of households across Indiana (Perry et al. 2023). It uses a face-to-face survey design to study and collect information regarding participant social and physical environments, cultures, behaviors, networks, and genetics.

Sampling, recruitment, and survey methodology for the baseline interview were designed to parallel the national General Social Survey. To achieve this, the study collected data in two parts. First, NORC, the sampling design partners, developed a randomized state-level household sampling plan intended to reflect the population distribution of the state but clustered within 50 percent of counties to facilitate data collection (Green and Pescosolido 2024). Then, at the household level, an adult was asked to describe all adult members of the household. From there, a survey respondent was chosen at random to participate in the study. Participants were noninstitutionalized, cognitively capable adults ages 18 years and older. Baseline data were collected face-to-face between October 2018 and July 2021. A total of 2,685 respondents completed the survey (response rate = 29 percent).

Dependent Variables: Health Care Utilization

Regular doctor (preventive care)

Respondents answered the question: “Do you have a doctor or nurse who you usually see if you need a checkup, want advice about a health problem, or get sick or hurt?” Response categories included “yes” or “no.” Respondents that indicated yes were considered to have a regular doctor and assumed to have access to regular care (e.g., preventive care, referrals; Lambrew et al. 1996). Having a regular doctor is used here as a proxy for preventive care. It is included in analytic models as a dummy variable.

Sought medical services (reactive care)

Respondents answered two questions used to construct this variable: “During the past year, have you thought or has someone told you that they thought that you might have a (physical/mental) health problem?” Respondents that indicated yes (n = 885) were then asked: “Please tell me if you used any of the following methods to resolve your (physical/mental) health problem,” with a dropdown list of options. If respondents selected “seek medical attention at a clinic, doctor’s office, or hospital,” we coded them as 1, and if not, we coded them as 0. Seeking medical services in response to a health problem is used here as a proxy for reactive care. It is included in analytic models as a dummy variable.

Key Independent Variables: Medical Trust

Network medical trust (network level)

Respondents were asked to name people with whom they discuss problems, discuss their health, receive health advice, and spend most of their free time.³ Respondents were then asked a series of questions about each alter they name. Network-level trust was calculated as the mean of all nonmissing responses to the question: “How much does [alter] trust doctors to take care of people’s problems?” Response options included “none,” “a little,” “some,” and “a lot” and were coded so that higher values indicated more trust in physicians. This continuous variable, ranging from 0 to 3, is the cumulative average of what an “ego” reported their “alters” trust in physicians to be. Although this is a measure of an ego’s perception of network trust, previous research suggests that an individual’s perceptions can closely predict reality (Sarason et al. 1991).

Personal medical trust (individual level)

Individual-level medical trust is calculated using five questions from the reliably used and tested Trust in Physicians Scale (Anderson and Dedrick 1990; Thom et al. 1999). In this scale, respondents were asked whether they strongly agreed, agreed, neither agreed nor disagreed, disagreed, or strongly disagreed with the following statements: (1) “I trust my doctor to tell me if a mistake was made about my treatment,” (2) “I trust my doctor to put my medical needs above all other considerations when treating my medical problems,” (3) “I trust my doctor’s judgements about my medical care,” (4) “My doctor is a real expert in taking care of medical problems like mine,” and (5) “I doubt that my doctor really cares about me as a person.” Statements 1 through 4 were reverse-coded so that higher numbers indicated more medical trust and lower numbers indicated less medical trust. We then averaged all nonmissing values of these responses into a single score. This continuous variable, ranging from 1 to 5, is the cumulative average of what an ego reported their own personal trust in physicians to be.

Network-Level Variables

Strength of ties

This continuous variable was calculated as the mean of all reported responses to the question: “On a scale of 1–10, where 10 is strongest, how strong is your relationship with [alter]?” We aggregated tie strength across the total network to capture feelings of trust, intimacy, and closeness (Marsden and Campbell 1984). This continuous variable, ranging from 1 to 10, measures the average tie strength of an ego’s network members.

Network size

Network size was calculated as the total number of people named across the name generators excluding duplicates and negative social ties. This continuous variable, ranging from 1 to 25, accounts for the size of an ego’s personal social network.

Network support multiplexity

Respondents were asked: “Has [alter] done any of the following for you or helped you out?” Response options included “listen,” “care,” “advise,” “help,” and “help materially [with money or other material support].” We averaged all nonmissing counts of these responses into one network-level score. That is, we calculated the average number of support functions across an ego’s network (Perry et al. 2018). This continuous variable, ranging from 0 to 5, measures the average number of support functions that an ego’s network provides.

Individual-Level Variables: Personal Health

Medical problem

Respondents answered: “During the past year, have you thought or has someone told you that they thought that you might have a (physical/mental) health problem?” Respondents that indicated yes were coded as 1, and respondents that indicated no were coded as 0. It is included in analytic Model 1 as a dummy variable, and only those answering yes are included in analytic Model 2.

Health insurance

Respondents answered: “Which of the following health insurance options do you currently have?” Response options included: a plan through your employer; a plan through your spouse’s employer; a plan you purchased yourself; Medicare, Medicaid, or Healthy Indiana Plan (HIP 1.0 or 2.0); some other government program; or insurance from somewhere else. Respondents were coded 0 if they answered no insurance and coded 1 if they selected any other option. It is included as a dummy variable in analytic models.

Self-rated health

Respondents answered: “Would you say that, overall, your physical health is: excellent, very good, good, fair, or poor.” These five responses were coded so that a higher number indicated better health. Self-rated health is included in analytic models as a continuous variable ranging from poor (1) to excellent (5).

Individual-Level Variables: Demographic Predictors

Gender

We use the variable “sex” to approximate gender, which was constructed from the question: “What was your gender assigned as birth?” Female is coded as 1, and male is coded as 0. We use this binary variable to preserve sample size and include those who identify with a different gender than their sex assigned at birth in our analysis. It is included in analytic models as a dummy variable.

Education

Respondents were asked: “What is the highest level of education you have completed?” Response options consisted of five levels of education: less than high school, high school graduate or GED, some college (no degree), some college/technical certificates/associate degree, college or higher. Sensitivity analyses comparing models with five- and three-category versions of this variable showed no meaningful differences between adjacent categories. Therefore, we present results using a three-category classification: high school diploma or less (reference), some college/technical certificate/associate degree, and college or higher. Educational attainment is included in analytic models as a categorical variable with three groups.

Race/ethnicity

This variable is a time-invariant, four-category variable that includes White (reference), Black, Hispanic, and other category. It is created using the race and ethnicity questions. Respondents were asked: “What race or races do you consider yourself?” Response options included American Indian, Asian, Black/African American, Native Hawaiian or Pacific Islander, White, and other. Respondents were also asked: “Do you consider yourself to be Hispanic, Latino, or of Spanish origin?” Response options included yes or no. The four-category race/ethnicity variable was calculated so that White, non-Hispanic respondents were coded as 1; Black, non-Hispanic respondents were coded as 2; Hispanic respondents (regardless of race) were coded as 3; and all other racial identities were coded as 4. This four-category variable is included in all analytic models.

Analytic Strategy

To address our research question, we construct multivariate logistic regression analyses using STATA 18. Approximately 1 percent of respondents do not name any positive social ties (i.e., they only name “hasslers”); we exclude these 31 cases from analyses. Of the remaining 2,648 respondents, we drop approximately 4 percent of missing data listwise (n = 106), leaving final analytic samples of 2,542 respondents for Model 1 (having a regular doctor) and 885 for Model 2 (medical utilization after a health problem). Much of our missing data came from our key independent variables, personal medical trust and network medical trust.

To control for network characteristics, we incorporate measures of the mean size of a network, the mean number of support functions, and the mean strength of ties. Additionally, we include individual-level predictors, including demographic characteristics of gender, education, and race, and health-related measures of past-year medical problem, self-rated general health, and insurance status. First, we provide sample descriptive statistics in Table 1 to offer an overview of the full sample (N = 2,542) and the subsample that reported a past-year medical problem (n = 885).⁴ We then use binary logistic regressions to predict our two binary outcomes, having a regular doctor (i.e., preventive care) and having gone to the doctor after a health problem (i.e., reactive care).

Table 1.

Person to Person Survey: Descriptive Statistics of Model 1 (N = 2,542).

Variable	M	SD	Minimum	Maximum
Health care utilization
Has a regular doctor (1 = yes)	0.83		0	1
Sought medical services (1 = yes)	0.22		0	1
Medical trust
Average network trust in doctors	2.35	0.63	0	3
Personal medical trust scale	4.09	0.73	1	5
Network-level characteristics
Average tie strength	8.64	1.28	2	10
Average number of support functions	3.51	1.03	0	5
Network size	5.36	2.90	1	25
Personal health characteristics
Self-rated health	3.21	1.04	1	5
Has health insurance (1 = yes)	0.93		0	1
Past-year medical problem (1 = yes)	0.35		0	1
Individual-level characteristics
Sex (1 = male)	0.38		0	1
Education
High school or less	34%
Some college/associate degree/technical certificate	36%
College or higher	29%
Race
White	84%
Black, African American	8%
Hispanic	2%
Other (AIAN, NHPI, Another race)	6%

Source: Person to Person Health Interview Study (2018–2020).

Note: AIAN = American Indian and Alaska Native; NHPI = Native Hawaiian or Pacific Islander.

Results

Descriptive Statistics

Table 1 provides a comprehensive overview of the demographic, social network, and health-related characteristics of the full sample (N = 2,542). The sample includes more women than men, with 62 percent women and 38 percent men. Regarding educational attainment, 34 percent of the sample completed high school or less, 36 percent have some college or an associate degree, and 29 percent hold a college degree or higher. The racial composition of the sample roughly reflects the broader population demographics of Indiana, with 84 percent identifying as White, 8 percent identifying as Black, 2 percent identifying as Hispanic, and 6 percent identifying as other.

In terms of network characteristics, the sample exhibits strong ties, with an average tie strength of 8.64 out of 10. Networks are generally supportive, with the average network providing between three and four of the five possible support functions. Network sizes are relatively small, with an average of five to six people; 90 percent of networks comprise fewer than eight individuals.

Health-related variables indicate that 93 percent of respondents have health insurance and that 7 percent do not. Respondents provided self-rated health scores ranging from 1 to 5, with an average score of 3.22. The mean score suggests that on average, respondents rate their health as “good.”

Finally, regarding our key dependent and independent variables, 83 percent of respondents report having a regular doctor, and 22 percent report seeking medical services in response to a recent health problem. The average network trust in doctors is 2.35 (SD = 0.63) on a 0 to 3 scale, indicating moderate trust within social networks. Individual medical trust is notably higher, with an average score of 4.09 (SD = 0.73) on a 1 to 5 scale, reflecting relatively strong high medical trust.

Table 2 provides an overview of the subsample of respondents (n = 885) who reported experiencing a health care problem in the past year. The health care utilization measure indicates that 64 percent of respondents sought medical services following their health problem and that 86 percent report having a regular doctor.

Table 2.

Person to Person Survey: Descriptive Statistics of Model 2 (n = 885).

Variable	M	SD	Minimum	Maximum
Health care utilization
Has a regular doctor (1 = yes)	0.86		0	1
Sought medical services (1 = yes)	0.64		0	1
Medical trust
Average network trust in doctors	2.29	0.63	0	3
Personal medical trust scale	4.04	0.81	1	5
Network-level characteristics
Average tie strength	8.30	1.36	2.67	10
Average number of support functions	3.50	0.97	0	5
Network size	5.65	2.92	1	20
Personal health characteristics
Self-rated health	2.85	1.03	1	5
Has health insurance (1 = yes)	0.92		0	1
Individual-level characteristics
Sex (1 = male)	0.35		0	1
Education
High school or less	33%
Some college/associate degree/technical certificate	38%
College or higher	29%
Race
White	85%
Black, African American	8%
Hispanic	2%
Other (AIAN, NHPI, another race)	6%

Source: Person to Person Health Interview Study (2018–2020).

Note: AIAN = American Indian and Alaska Native; NHPI = Native Hawaiian or Pacific Islander.

Compared to the full sample, the subsample reports slightly lower average levels of self-rated health (M = 2.85 vs. 3.22 in the full sample), consistent with having recently experienced a health issue. Network trust in doctors is also slightly lower (M = 2.29, SD = 0.63), as is individual medical trust (M = 4.04, SD = 0.81). The subsample’s average network size is slightly larger (M = 5.65 vs. 5.36), and tie strength is somewhat lower (M = 8.30 vs. 8.64), possibly reflecting more less intimate networks during periods of health strain. The demographic composition of the subsample is similar to the full sample: 65 percent are female, and 92 percent have health insurance. Educational attainment is consistent, with 33 percent having a high school education or less, 38 percent having some college, and 29 percent having a college degree or higher.

Table 3 presents the results of two logistic regression models predicting health care utilization in odds ratios (ORs). Model 1 examines the likelihood of having a regular doctor (N = 2,542), and Model 2 examines the likelihood of seeking medical services after a health care problem (n = 885).

Table 3.

Main Effects, Logit (Binary) Regression Models Predicting Health Care Utilization in Odds Ratios.

	Has a Regular Doctor (N = 2,542)	Seeking Medical Services After a Health care Episode (n = 885)
	Model 1	Model 2
Medical trust
Mean network trust	1.23(0.11)*	1.33(0.16)*
Individual medical trust	1.94(0.15)***	1.14(0.11)
Network-level characteristics
Mean strength of tie	1.21(0.06)***	1.03(0.06)
Mean number of support functions	0.82(0.05)**	0.89(0.08)
Network size	1.04(0.02)	1.07(0.03)*
Personal health characteristics
Self-rated health	0.76(0.05)***	0.80(0.06)**
Has insurance	6.37(1.10)***	4.30(1.23)***
Past-year medical problem	1.41(0.19)*	-
Individual-level characteristics
Male (reference = female)	0.52(0.06)***	0.81(0.13)
Education (reference = high school diploma or less)
Some college/associate degree/technical certificate	1.44(0.20)*	1.52(0.27)*
College or higher	1.30(0.20)	1.27(0.25)
Race (reference = White)
Black	0.73(0.15)	0.84(0.24)
Hispanic	1.15(0.51)	0.36(0.17)*
Other race	0.42(0.09)***	0.49(0.15)*
Constant	0.02	0.17
Pseudo R²	0.15	0.07
χ²	354.08	79.1
ll	−972.03	−540.09

Source: Person to Person Health Interview Study (2018-2020).

Note: Standard errors are in parentheses. ll = log-likelihood.

p < .05. **p < .01. ***p ≤ .001.

Medical Trust

In both models, mean network trust is a significant predictor of health care utilization. In Model 1 and in support of Hypothesis 1a, a 1-unit increase in mean network trust is associated with a 23 percent increase in the odds of having a regular doctor (OR = 1.23, p = .020). In Model 2 and in support of Hypothesis 1b, mean network trust is also significantly associated with seeking medical services after a health care episode, with a 33 percent increase in the odds (OR = 1.33, p = .018). Individual medical trust significantly predicts having a regular doctor in Model 1 (OR = 1.94, p < .001) but does not significantly predict seeking medical services after a health care episode in Model 2 (p = .159). These results support both Hypotheses 2a and 2b.

Network-Level Characteristics

Among the network-level characteristics, mean strength of tie is a significant predictor in Model 1. A 1-unit increase in the strength of tie is associated with a 20 percent increase in the odds of having a regular doctor (OR = 1.21, p < .001). However, this characteristic does not significantly predict seeking medical services after a health care episode in Model 2 (p = .635). The mean number of support functions is negatively associated with having a regular doctor in Model 1 (OR = 0.82, p = .001) but does not significantly influence seeking medical services in Model 2 (p = .180). Network size is marginally significant in Model 1 (OR = 1.04, p = .078) and becomes a significant predictor in Model 2, where a 1-unit increase is associated with a 7 percent increase in the odds of seeking medical services (OR = 1.07, p = .021).

Personal Health Characteristics

Self-rated health is a negative predictor of health care utilization in both models. In Model 1, a 1-unit increase in self-rated health is associated with a 24 percent decrease in the odds of having a regular doctor (OR = 0.76, p ≤ .001), and in Model 2, it is associated with a 20 percent decrease in the odds of seeking medical services after a health care episode (OR = 0.80, p = .004). Health insurance strongly predicts health care utilization. Those with insurance are more than 6 times more likely to have a regular doctor (OR = 6.37, p ≤ .001) and more than 4 times more likely to seek medical services following a health problem (OR = 4.30, p < .001). The variable indicating whether a respondent had a recent medical problem was included in Model 1 and increased the odds of having a regular doctor (OR = 1.41, p = .010); it was omitted from Model 2 because Model 2 includes only those who reported a health problem.

Individual-Level Characteristics

Gender significantly affects health care utilization in Model 1: Women are nearly twice as likely as men to have a regular doctor (OR = 1.94, p < .001). However, gender does not significantly predict whether someone seeks medical services after a health care episode in Model 2 (p = .165). Education is a significant predictor in both models, although the effects differ. In Model 1, respondents with some college or an associate degree are 44 percent more likely to have a regular doctor compared to those with a high school education or less (OR = 1.44, p = .010). In Model 2, the same group is 52 percent more likely to seek medical services after a health care episode (OR = 1.52, p = .018). Respondents with a college degree or higher are more likely to have a regular doctor in Model 1 (OR = 1.30), although this finding is marginally significant (p = .095). This education level does not significantly influence seeking medical services in Model 2 (p = .226).

Race shows some variation in patterns of health care utilization. Hispanic respondents do not differ significantly from White respondents in having a regular doctor in Model 1 (OR = 1.15, p = .762), but in Model 2, they are significantly less likely to seek medical services (OR = 0.36, p = .033). Respondents identifying as other races are less likely to have a regular doctor in Model 1 (OR = 0.42, p < .001) and are less likely to seek care after a health care episode in Model 2 (OR = 0.49, p = .023). No other racial group differences reached statistical significance in either model.

Discussion

In this article, we asked: How does the relationship between trust and utilization vary by level of trust and type of utilization? Consistent with Hypotheses 1a and 1b, we found that respondents with networks that had higher amounts of medical trust (compared to those who had networks with lower medical trust) were more likely to have a regular doctor and seek medical services in response to a new health care problem. We, as expected, found evidence for Hypothesis 2a, that individual medical trust predicted having a regular doctor, and for Hypothesis 2b, that individual medical trust did not predict utilizing health care services in response to a new health care problem.

Both individual- and network-level factors matter for individuals’ health care utilization. Individuals with higher network trust were most likely to have access to health care services through having a regular doctor and accessing medical care after a health problem even after controlling for known predictors, including individual trust. Individual trust in doctors mattered for predicting having a regular doctor but not for accessing medical care after a health problem.

We also found that different network characteristics predicted different types of health care behavior. People with closer, more emotionally connected relationships (i.e., stronger ties) were more likely to have a regular doctor. Close ties could play a key role in encouraging routine, preventive care—whether by sharing health advice, modeling regular checkups, or simply reinforcing the importance of staying on top of one’s health. Surprisingly, those whose networks provided more types of support were less likely to have a regular doctor. One possible explanation is that people who can rely heavily on informal support from their networks might feel less need to maintain a relationship with a regular health care provider. People with larger networks were more likely to seek care after a health problem, possibly due to increased access to social and financial resources (Cornwell and Waite 2012).

Those who had insurance were more likely to have a regular doctor and have had gone to the doctor within the last year, which is consistent with prior literature reporting that not having insurance is a frequent barrier to accessing medical care (Baicker et al. 2013; Shami, Tabrizi, and Nosratnejad 2019). Also consistent with prior literature, women were more likely to be higher medical utilizers (Bertakis et al. 2000; Bird and Rieker 2008; Courtenay 2000; Umberson 1992). Self-rated health was negatively associated with both outcomes. People who felt less healthy were more likely to have a regular doctor and to seek care after a health problem, reflecting greater medical need among those in poorer health.

Our study aligns with and underscores the theoretical importance of the NEM, where health care utilization is conceptualized as a series of network problems wherein individuals navigate their social networks to access health care resources (Pescosolido 1992). It does so by demonstrating that network trust plays an independent role in individuals’ decision to seek out medical services to address a health care problem, thereby potentially facilitating ongoing access to health care services. Although individual trust plays a role initially facilitating access to health care resources, its influence on specific health care behaviors after a health problem may be better explained by network trust and medical culture through the mechanism of meaning making in their networks.

By examining the relational ties and network structures within which individuals are embedded, social network analysis allows for a nuanced understanding of how network trust operates (Scott 2000). Our use of social network analysis in tandem with individual- and health-related predictors revealed that individuals with higher levels of network trust were more likely to have access to health care services through the establishment of a regular doctor-patient relationship. Our findings also highlight the need for further research to explore how different types of social network cultures may exert varying influences on health care utilization and other health behaviors.

Theoretically, we contribute to a cultural understanding of medical trust. We do so by demonstrating that trust is not merely an individual trait but a layered orientation shaped by both personal experiences and norms within social networks (Decoteau 2013; Lo and Stacey 2008; Möllering 2013). Our findings are strengthened by the concept of hybrid habitus, which refers to the ways in which individuals navigate complex, intersecting social structures to make sense of their health behaviors. The fact that network trust predicts health care utilization even after controlling for individual medical trust underscores the idea that medical trust is a collective phenomenon.

Our results also hold implications for a stronger understanding of how structural inequalities shape trust. Variations in health care utilization by race, gender, and socioeconomic status in our models suggest that hybrid habitus is not simply a shared cultural orientation but one that reflects differential access to social capital and medical resources (Lo and Stacey 2008; Sewell 1992; Shim 2010). Individualistic models of medical trust do not account for the broader social contexts that shape how individuals engage with health care systems.

Public health interventions can be more effective if they engage with the social networks individuals are embedded. Community health workers (CHWs) and peer educators that operate within the same cultural and social contexts as their patients can act as brokers, reinforcing positive health norms, addressing mistrust, and connecting patients to medical resources (Balcazar et al. 2011; Rosenthal et al. 2010; Witmer et al. 1995). Workers can facilitate trust by building on the existing social and health capital within communities and acting as trusted intermediaries between patients and health care providers.

Effective interventions should address the barriers faced by different racial, ethnic, and socioeconomic groups. For example, public health programs that train CHWs to address medical mistrust related to historical abuses or language barriers can be particularly impactful in marginalized communities (Betancourt et al. 2003; Brach and Fraser 2000; Washington 2006). This could include providing translation services, culturally sensitive health care materials, and community outreach programs that align with the specific needs and concerns of targeted populations. Additionally, interventions should recognize that mistrust is not confined to marginalized groups (Lupton 1997; Reich 2016). Tailoring interventions to address varied forms of mistrust can enhance their effectiveness and ensure that public health messages resonate across social contexts. Finally, increasing network trust requires ongoing efforts to acknowledge historical medical harms, foster transparency, and build lasting patient-provider relationships (Benjamin 2014).

Limitations

Although all hypotheses were confirmed and our results support and extend existing literature, there are a few study limitations to consider. One concern is the reciprocal relationship between individual medical trust and having a regular doctor. Given that our data are cross-sectional, we do not make any claims regarding causality. However, we also note here that we cannot establish the exact direction of this relationship (Elwert and Winship 2014). Research suggests that individuals who trust medical professionals are more likely to establish a relationship with a regular doctor, which, in turn, can reinforce and strengthen their trust over time (Boulware et al. 2003; Hall et al. 2001). Despite this, it is important to recognize that this relationship is well documented in the health care utilization literature as bidirectional. For example, consistent with the behavioral health model (Andersen 1995), individual characteristics, such as trust, are known predictors of health care access, including having a regular doctor. At the same time, ongoing interactions with health care providers shape trust (Gray et al. 2018). Prior research indicates that these constructs remain analytically distinct.

Another limitation of the study is the temporal ordering of survey measures. Specifically, we use a present-time measure of individual medical trust to predict retrospective medical utilization for having a regular doctor and having gone to the doctor within the last year. However, because this study is descriptive rather than causal, the observed differences in network medical trust remain valid. Networks are relatively stable, meaning that the influence of network trust on health care utilization is very likely to be stable within the year that trust was measured (McPherson, Smith-Lovin, and Cook 2001). Moreover, individual trust in one’s doctor functions as a control variable in our models rather than the primary explanatory factor, allowing us to isolate the enduring influence of network medical trust. This analytical approach mitigates concerns about reverse causality while preserving the theoretical and empirical significance of networks and health care utilization.

Finally, we acknowledge the distinction between personal trust in a specific doctor and broader perceptions of what influences medical trust. Although these may capture related but distinct dimensions of medical trust, our primary focus is on network-level trust and its implications for health care engagement. By controlling for individual trust and leveraging established theoretical frameworks, our analysis remains robust in capturing broader structural dynamics. Future research could address these limitations and explore other measures of health care utilization.

Conclusion

Compared to those with networks that had lower medical trust, individuals who had networks with higher medical trust were more likely to both have a regular doctor and seek medical services in response to a health care problem. This difference persisted even after controlling for known predictors of medical utilization on both the individual level and network level, including a measure of individual medical trust.

These findings contribute to a stronger understanding of how trust-related factors, network characteristics, and individual-level variables predict two measures of health care utilization. This differentiation between the effect of a network’s trust and individual trust allows us to better tailor interventions and policies to effectively address trust-related barriers to health care utilization. By examining network trust, we can uncover how the collective trust within social networks shapes health care utilization patterns, resource allocation, and access to care. By illustrating how different trust affects different types of utilization, we can better tailer interventions at the network level for preventive care to be utilized.

Lastly, our study underscores key tenets of cultural sociology. Cultural sociology posits that individuals’ health care behaviors are not solely determined by personal characteristics but are also shaped by broader cultural norms, values, and social networks (Swidler 1986). In our analysis, we observed that network trust in doctors significantly predicted health care utilization in both models, highlighting the importance of beliefs and attitudes in shaping preventive and episodic health care seeking. We additionally underscored the role of individual trust in facilitating access to preventive care. These results bolster claims of a hybrid habitus that reflects the value of both network-level and individual medical trust.

Future research with large populations of interest and longitudinal surveys can consider how different types of networks may play a larger influence than others or how different mediating individual-level qualities may make individuals suspectable to network influence. Qualitative and mixed-methods work in this domain would also be elucidating. How might individuals understand their trust relative to their networks? Qualitative work could further illustrate details in the meaning-making process and contribute to a strong constructionist understanding of medical trust.

Footnotes

Appendix

Table A1.

Person to Person Survey: Descriptive Statistics by Medical Problem Status (N = 2,542).

	No Medical Problem (n = 1,657)	Medical Problem (n = 885)	p Value
Health care utilization
Has a regular doctor (1 = yes)	0.82 (0.39)	0.86 (0.35)	**
Sought medical services (1 = yes)	0.00 (0.00)	0.64 (0.48)	***
Medical trust
Mean network trust (0–3)	2.40 (0.63)	2.294 (0.63)	**
Individual medical trust (1–5)	4.12 (0.68)	4.04 (0.81)	***
Network-level characteristics
Mean strength of tie (1–10)	8.82 (1.20)	8.30 (1.34)	***
Mean number of support functions (0–5)	3.51 (1.06)	3.50 (0.97)
Network size (0–25)	5.20(2.90)	5.70(2.91)	***
Personal health characteristics
Self-rated health (1–5)	3.41(0.10)	2.90(1.02)	***
Has insurance (1 = yes)	0.93(0.26)	0.92(0.27)
Individual-level characteristics
Sex: female	0.60(0.49)	0.65(0.48)	**
Education
High school diploma or less	0.349(0.48)	0.333(0.47)
Some college/associate degree/technical certificate	0.36(0.48)	0.38(0.49)
College or higher	0.295(0.46)	0.29(0.45)
Race
White	0.84(0.37)	0.85(0.36)
Black	0.09(0.28)	0.08(0.26)
Hispanic	0.02(0.12)	0.02(0.15)
Other race	0.061(0.24)	0.06(0.23)

Source: Person to Person Health Interview Study (2018–2020).

Note: Standard errors are in parentheses. Significance based on p values from t tests (continuous variables) and Pearson’s chi-square tests (categorical variables).

p < .01. ***p ≤ .001.

Funding

The author(s) disclosed receipt of the following financial support for the research, authorship, and/or publication of this article: This research was supported by a grant from the Robert Wood Johnson Foundation – Health Policy Scholars Program. The funder had no role in the study design, data collection, analysis, or interpretation of the results.

Ethics Approval

Ethical approval for this study was obtained through the Person to Person Health Interview Study’s Institutional Review Board (IRB) at Indiana University (IRB Protocol No. 1803431862).

ORCID iDs

Tania Ravaei

Megan Bolton

1

For this analysis, we focus on personal networks composed of social ties (alters) with whom an individual (ego) discusses important matters, including health, and/or spends the most time with. These egocentric networks capture the core social relationships that individuals rely on for health-related advice and support rather than broader, more diffuse forms of social connectedness, such as online communities or organizational affiliations ().

2

We define medical trust as the confidence individuals place in medical institutions, health care providers, and the health care system. It can include beliefs about the competence, integrity, and benevolence of health care professionals and perceptions of the effectiveness of medical treatments and services ().

3

The P2P study contains five name generators, including four that capture positive social ties and one that captures negative social ties. We exclude egocentric networks that only include negative social ties (i.e., people that hassle, cause problems, or make life difficult).

4

Appendix contains results of two-sample tests of proportions comparing the sample characteristics of respondents without a past year medical problem (n = 1,657) and those who do report a past-year medical problem (n = 885).

Author Biographies

Tania Ravaei is a PhD candidate in the Department of Sociology at Indiana University. She works with the Kinsey Institute for Sex Research and the Irsay Institute for Sociomedical Sciences Research. Her research interests lie broadly in medical sociology, population health, and political sociology. More specifically, she focuses on sexual behavior and population health, sociopolitical attitudes, and mixed methods. Ravaei is also a health policy research fellow with the Robert Wood Johnson Foundation, and her work has been generously supported by the Andrew W. Mellon Foundation. Tania’s research has addressed topics such as public opinion on health and education policy and adolescent women’s sexual behavior in relation to mental health status. Her dissertation explores agency and decision-making in intimate relationships, particularly how gendered expectations shape life course trajectories and reproductive choices. She is an alumna of UCLA, where she earned degrees in anthropology and sociology.

Megan Bolton, PhD, is a research scientist at the Irsay Family Institute at Indiana University Bloomington, where she earned her doctorate in sociology in 2021. Her research focuses on medical sociology, stigma, and social networks, with particular attention to how interpersonal relationships shape mental health and access to care among emerging adults. Her dissertation examined how young people with mental health challenges navigate stigma and maintain social ties across the transition to adulthood, using a mixed-methods approach that included social network analysis, longitudinal survey data, and in-depth interviews. Through this work, she explored how perceptions of mental illness influence self-concept, help-seeking behavior, and the structure and meaning of social support networks. More broadly, her research agenda investigates how stigma operates not only at the individual level but also within relational contexts and how these dynamics influence inequality in health, social integration, and overall well-being across the life course.

References

Andersen

Ronald M.

1995. “Revisiting the Behavioral Model and Access to Medical Care: Does It Matter?” Journal of Health and Social Behavior 36(1):1–10.

Anderson

Lynda A.

Dedrick

Robert F.

1990. “Development of the Trust in Physician Scale: A Measure to Assess Interpersonal Trust in Patient-Physician Relationships.” Psychological Reports 67(3, Pt. 2): 1091–100.

Baicker

Katherine

Taubman

Sarah L.

Allen

Heidi L.

Bernstein

Mira

Gruber

Jonathan H.

Newhouse

Joseph P.

Schneider

Eric C.

, et al. 2013. “The Oregon Experiment—Effects of Medicaid on Clinical Outcomes.” The New England Journal of Medicine 368(182013):1713–22.

Balcazar

Hector

Rosenthal

E. Lee

Brownstein

J. Nell

Rush

Carl H.

Matos

Sergio

Hernandez

Lorenza

. 2011. “Community Health Workers Can Be a Public Health Force for Change in the United States: Three Actions for a New Paradigm.” American Journal of Public Health 101(12):2199–203.

Bataineh

Hana

Devlin

Rose Anne

Barham

Vicky

. 2019. “Social Capital and Having a Regular Family Doctor: Evidence from Longitudinal Data.” Social Science & Medicine 220:421–29.

Berkman

Lisa F.

Glass

Thomas

Brissette

Ian

Seeman

Teresa E.

2000. “From Social Integration to Health: Durkheim in the New Millennium.” Social Science & Medicine 51(6):843–57.

Benjamin

Ruha

. 2014. “Race for Cures: Rethinking the Racial Logics of ‘Trust’ in Biomedicine.” Sociology Compass 8(6):755–69.

Bertakis

Klea D.

Azari

Rahman

Helms

L. Jay

Callahan

Edward J.

Robbins

John A.

2000. “Gender Differences in the Utilization of Health Care Services.” The Journal of Family Practice 49(2):147–52.

Betancourt

Joseph R.

Green

Alexander R.

Carrillo

J. Emilio

Ananeh-Firempong

Owusu

II . 2003. “Defining Cultural Competence: A Practical Framework for Addressing Racial/Ethnic Disparities in Health and Health Care.” Public Health Reports 118(4):293–302.

10.

Bindman

Andrew B.

Grumbach

Kevin

Osmond

Dennis

Komaromy

Miriam

Vranizan

Karen

Lurie

Nicole

Billings

John

, et al. 1995. “Preventable Hospitalizations and Access to Health Care.” JAMA 274(4):305–11.

11.

Bird

Chloe E.

Rieker

Patricia P.

2008. Gender and Health: The Effects of Constrained Choices and Social Policies. New York, NY: Cambridge University Press.

12.

Birkhäuer

Johanna

Gaab

Jens

Kossowsky

Joe

Hasler

Sebastian

Krummenacher

Peter

Werner

Christoph

Gerger

Heike

. 2017. “Trust in the Health Care Professional and Health Outcome: A Meta-analysis.” PLoS One 12(2):e0170988. doi:10.1371/journal.pone.0170988.

13.

Blewett

Lynn A.

Jo Johnson

Pamela

Lee

Brian

Scal

Peter B.

2008. “When a Usual Source of Care and Usual Provider Matter: Adult Prevention and Screening Services.” Journal of General Internal Medicine 23(9):1354–60.

14.

Boulware

L. Ebony

Cooper

Lisa A.

Ratner

Lloyd E.

LaVeist

Thomas A.

Powe

Neil R.

2003. “Race and Trust in the Health Care System.” Public Health Reports 118(4):358–65.

15.

Bourdieu

Pierre

. 1977. Outline of a Theory of Practice. Translated by Nice

Richard

. Cambridge, UK: Cambridge University Press.

16.

Brach

Cindy

Fraser

Irene

. 2000. “Can Cultural Competency Reduce Racial and Ethnic Health Disparities? A Review and Conceptual Model.” Medical Care Research and Review 57(Suppl. 1):181–217.

17.

Christakis

Nicholas A.

Fowler

James H.

2007. “The Spread of Obesity in a Large Social Network over 32 Years.” New England Journal of Medicine 357(4):370–79.

18.

Christakis

Nicholas A.

Fowler

James H.

2008. “The Collective Dynamics of Smoking in a Large Social Network.” New England Journal of Medicine 358(21):2249–58.

19.

Cornwell

Erin York

Waite

Linda J.

2012. “Social Network Resources and Management of Hypertension.” Journal of Health and Social Behavior 53(2):215–31.

20.

Courtenay

Will H

. 2000. “Constructions of Masculinity and Their Influence on Men’s Well-Being: A Theory of Gender and Health.” Social Science & Medicine 50(10):1385–401.

21.

Decoteau

Claire Laurier

. 2013. “Hybrid Habitus: Toward a Post-colonial Theory of Practice.” Postcolonial Sociology 24:263–93.

22.

Decoteau

Claire Laurier

Sweet

Paige L

. 2024. “Vaccine Hesitancy and the Accumulation of Distrust.” Social Problems 71(4):1231–46.

23.

Devlin

Rose Anne

Rudolph-Zbarsky

Jamie

. 2014. “Social Networks and the Probability of Having a Regular Family Doctor.” Social Science & Medicine 115:21–28.

24.

Doyle

Joshua

. 2021. “The Effect of Cultural Trust on Cooperation in Two Behavioral Experiments.” Social Psychology Quarterly 84(3):246–66.

25.

Doyle

Joshua

. 2023. “The Effect of Cultural Trust in the Presence of Second Order Trust on Cooperation in a Behavioral Experiment.” Advances in Group Processes 40:49–69.

26.

Elwert

Felix

Winship

Christopher

. 2014. “Endogenous Selection Bias: The Problem of Conditioning on a Collider Variable.” Annual Review of Sociology 40:31–53.

27.

Eyal

Gil

Larry

Capotescu

Cristian

. 2024. “Trust Is a Verb!: A Critical Reconstruction of the Sociological Theory of Trust.” Sociologica 18(2):169–91.

28.

Gray

Denis J. Pereira

Sidaway-Lee

Kate

White

Eleanor

Thorne

Angus

Evans

Philip H.

2018. “Continuity of Care with Doctors—A Matter of Life and Death? A Systematic Review of Continuity of Care and Mortality.” BMJ Open 8(6):e021161. doi:10.1136/bmjopen-2017-021161.

29.

Green

Harold D.

Pescosolido

Bernice A.

2024. “Social Pathways to Care: How Community-Based Network Ties Shape the Health Care Response of Individuals with Mental Health Problems.” Social Psychiatry and Psychiatric Epidemiology 59(3):431–42.

30.

Hall

Mark A.

Dugan

Elizabeth

Zheng

Beiyao

Mishra

Aneil K.

2001. “Trust in Physicians and Medical Institutions: What Is It, Can It Be Measured, and Does It Matter?” The Milbank Quarterly 79(4):613–39.

31.

House

James S.

2015. Beyond Obamacare: Life, Death, and Social Policy. New York, NY: Russell Sage Foundation.

32.

Koplan

Jeffrey P.

Bond

T. Christopher

Merson

Michael H.

Reddy

K. Srinath

Rodriguez

Mario Henry

Sewankambo

Nelson K.

Wasserheit

Judith N.

, et al. 2009. “Towards a Common Definition of Global Health.” Lancet 373(9679):1993–95.

33.

Lambrew

Jeanne M.

Defriese

Gordon H.

Carey

Timothy S.

Ricketts

Thomas C.

Biddle

Andrea K.

1996. “The Effects of Having a Regular Doctor on Access to Primary Care.” Medical Care 34(2):138–51.

34.

Latkin

Carl A.

Knowlton

Amy R.

2015. “Social Network Assessments and Interventions for Health Behavior Change: A Critical Review.” Behavioral Medicine 41(3):90–97.

35.

Levinson

Wendy

Gorawara-Bhat

Rita

Lamb

Jennifer

. 2000. “A Study of Patient Clues and Physician Responses in Primary Care and Surgical Settings.” JAMA 284(8):1021–27.

36.

Lin

Nan

. 2002. Social Capital: A Theory of Social Structure and Action. Cambridge, UK: Cambridge University Press.

37.

Ming-Cheng M.

Stacey

Clare L.

2008. “Beyond Cultural Competency: Bourdieu, Patients and Clinical Encounters.” Sociology of Health & Illness 30(5):741–55.

38.

Lupton

Deborah

. 1997. “Consumerism, Reflexivity and the Medical Encounter.” Social Science & Medicine 45(3):373–81. doi:10.1016/s0277-9536(96)00353-x.

39.

Marmot

Michael

. 2005. “Social Determinants of Health Inequalities.” The Lancet 365(9464):1099–104.

40.

Marsden

Peter V.

Campbell

Karen E.

1984. “Measuring Tie Strength.” Social Forces 63(2):482–501.

41.

McLean

Paul

. 2016. Culture in Networks. New York, NY: John Wiley & Sons.

42.

McPherson

Miller

Smith-Lovin

Lynn

Cook

James M.

2001. “Birds of a Feather: Homophily in Social Networks.” Annual Review of Sociology 27:415–44.

43.

McWilliams

J. Michael

Meara

Ellen

Zaslavsky

Alan M.

Ayanian

John Z.

2007. “Use of Health Services by Previously Uninsured Medicare Beneficiaries.” New England Journal of Medicine 357(2):143–53.

44.

Mechanic

David

Meyer

Sharon

. 2000. “Concepts of Trust among Patients with Serious Illness.” Social Science & Medicine 51(5):657–68.

45.

Möllering

Guido

. 2013. “Process Views of Trusting and Crises.” Pp. 285–306 in Handbook of Advances in Trust Research, edited by Bachmann

Zaheer

Cheltenham, UK: Edward Elgar.

46.

Perry

Brea L.

Pescosolido

Bernice A.

2010. “Functional Specificity in Discussion Networks: The Influence of General and Problem-Specific Networks on Health Outcomes.” Social Networks 32(4):345–57.

47.

Perry

Brea L.

Pescosolido

Bernice A.

2015. “Social Network Activation: The Role of Health Discussion Partners in Recovery from Mental Illness.” Social Science & Medicine 125:116–28.

48.

Perry

Brea L.

Pescosolido

Bernice A.

Borgatti

Stephen P.

2018. Egocentric Network Analysis: Foundations, Methods, and Models. New York, NY: Cambridge University Press.

49.

Perry

Brea L.

Smith

Nicholas C.

Coleman

Max E.

Pescosolido

Bernice A.

2023. “Social Networks, the COVID-19 Pandemic, and Emerging Adults’ Mental Health: Resiliency through Social Bonding and Cohesion.” American Journal of Public Health 114(Suppl. 3):S258–67.

50.

Pescosolido

Bernice A.

1992. “Beyond Rational Choice: The Social Dynamics of How People Seek Help.” American Journal of Sociology 97(4):1096–138.

51.

Pescosolido

Bernice A.

2006. “Of Pride and Prejudice: The Role of Sociology and Social Networks in Integrating the Health Sciences.” Journal of Health and Social Behavior 47(3):189–208.

52.

Pescosolido

Bernice A.

Boyer

Carol A.

1999. “How Do People Come to Use Mental Health Services? Current Knowledge and Changing Perspectives.” Pp. 392–411 in A Handbook for the Study of Mental Health: Social Contexts, Theories, and Systems, edited by Horwitz

A. V.

Scheid

T. L.

New York, NY: Cambridge University Press.

53.

Pescosolido

Bernice A.

Perry

Brea L.

Martin

Jack K.

McLeod

Jane D.

Jensen

Peter S.

2007. “Stigmatizing Attitudes and Beliefs about Treatment and Psychiatric Medications for Children with Mental Illness.” Psychiatric Services 58(5):613–18.

54.

Reich

Jennifer A.

2016. Calling the Shots: Why Parents Reject Vaccines. New York, NY: New York University Press.

55.

Reich

Jennifer A.

2020. “Vaccine Refusal and Pharmaceutical Acquiescence: Parental Control and Ambivalence in Managing Children’s Health.” American Sociological Review 85(1):106–27.

56.

Rosenquist

J. Niels

Fowler

James H.

Christakis

Nicholas A.

2011. “Social Network Determinants of Depression.” Molecular Psychiatry 16(3):273–81.

57.

Rosenquist

J. Niels

Murabito

Joanne

Fowler

James H.

Christakis

Nicholas A.

2010. “The Spread of Alcohol Consumption Behavior in a Large Social Network.” Annals of Internal Medicine 152(7):426–33.

58.

Rosenthal

E. Lee

Brownstein

J. Nell

Rush

Carl H.

Hirsch

Gail R.

Willaert

Anne M.

Scott

Jacqueline R.

Holderby

Lisa R.

, et al. 2010. “Community Health Workers: Part of the Solution.” Health Affairs (Project Hope) 29(7):1338–42.

59.

Sanmartin

Claudia

Ross

Nancy

. 2006. “Experiencing Difficulties Accessing First-Contact Health Services in Canada.” Healthcare Policy 1(2):103–19.

60.

Sarason

Barbara R.

Pierce

Gregory R.

Shearin

Edward N.

Sarason

Irwin G.

Waltz

Jennifer A.

Poppe

Leslie

. 1991. “Perceived Social Support and Working Models of Self and Actual Others.” Journal of Personality and Social Psychology 60(2):273–87.

61.

Scott

John

. 2000. Social Network Analysis: A Handbook. London: Sage.

62.

Sewell

William H.

1992. “A Theory of Structure: Duality, Agency, and Transformation.” American Journal of Sociology 98(1):1–29.

63.

Shami

Elham

Tabrizi

Jafar Sadegh

Nosratnejad

Shirin

. 2019. “The Effect of Health Insurance on the Utilization of Health Services: A Systematic Review and Meta-analysis.” Galen Medical Journal 8:e1411. doi:10.31661/gmj.v8i0.1411.

64.

Shim

Janet K.

2010. “Cultural Health Capital: A Theoretical Approach to Understanding Health Care Interactions and the Dynamics of Unequal Treatment.” Journal of Health and Social Behavior 51(1):1–15.

65.

Starfield

Barbara

Shi

Leiyu

Macinko

James

. 2005. “Contribution of Primary Care to Health Systems and Health.” The Milbank Quarterly 83(3):457–502.

66.

Swidler

Ann

. 1986. “Culture in Action: Symbols and Strategies.” American Sociological Review 51(2): 273–86.

67.

Thanh

Nguyen Xuan

Rapoport

John

. 2017. “Health Services Utilization of People Having and Not Having a Regular Doctor in Canada.” The International Journal of Health Planning and Management 32(2):180–88.

68.

Thom

David H.

Ribisl

Kurt M.

Stewart

Anita L.

Luke

Douglas A.

and The Stanford Trust Study Physicians. 1999. “Further Validation and Reliability Testing of the Trust in Physician Scale.” Medical Care 37(5):510–17.

69.

Umberson

Debra

. 1992. “Gender, Marital Status and the Social Control of Health Behavior.” Social Science & Medicine 34(8):907–17.

70.

Valente

Thomas W.

2010. Social Networks and Health: Models, Methods, and Applications. New York, NY: Oxford University Press.

71.

Washington

Harriet A.

2006. Medical Apartheid: The Dark History of Medical Experimentation on Black Americans from Colonial Times to the Present. New York, NY: Doubleday.

72.

Witmer

Anne

Seifer

Sarena D.

Finocchio

Leonard

Leslie

Jodi

O’Neil

Edward H.

1995. “Community Health Workers: Integral Members of the Health Care Work Force.” American Journal of Public Health 85(8, Pt. 1):1055–58.

73.

K. Tom

. 2002. “Usual Source of Care in Preventive Service Use: A Regular Doctor versus a Regular Site.” Health Services Research 37(6):1509–29.

We,Not Just I,Must Trust Doctors: Network Medical Trust and Health Care Utilization

Abstract

Keywords

Background and Theory

Network Episode Model

Cultural Theory and Hybrid Habitus

Medical Trust as Hybrid Habitus

Health Care Utilization

Data and Methods

Person to Person Health Interview Study

Dependent Variables: Health Care Utilization

Regular doctor (preventive care)

Sought medical services (reactive care)

Key Independent Variables: Medical Trust

Network medical trust (network level)

Personal medical trust (individual level)

Network-Level Variables

Strength of ties

Network size

Network support multiplexity

Individual-Level Variables: Personal Health

Medical problem

Health insurance

Self-rated health

Individual-Level Variables: Demographic Predictors

Gender

Education

Race/ethnicity

Analytic Strategy

Results

Descriptive Statistics

Medical Trust

Network-Level Characteristics

Personal Health Characteristics

Individual-Level Characteristics

Discussion

Limitations

Conclusion

Footnotes

Appendix

Funding

Ethics Approval

ORCID iDs

1

2

3

4

Author Biographies

References