Abstract
Objectives:
Investigate the association between perceptions of health misinformation on social media and trust in the healthcare system among U.S. adults, and to assess whether this association varies by frequency of health care visits, perceived healthcare quality and experiences of medical care discrimination.
Methods:
Cross-sectional survey study using data from the 2022 Health Information National Trends Survey 6 (HINTS 6). Analysis was conducted on data collected from March to November 2022. Participants included 3,805 adults who reported using social media and had at least one healthcare visit in the past year. Survey weighted, multivariable logistic regression models were used to assess associations.
Results:
Among those who reported high trust in the healthcare system, 65.1% perceived less than substantial health misinformation on social media, while 34.9% perceived substantial misinformation. In multivariable models, participants who perceived substantial health misinformation on social media had higher odds of reporting low trust in the healthcare system (OR 1.66; 95% CI 1.11–2.48). This association between misinformation and trust varied by perceived healthcare quality and experiences of discrimination. Among those perceiving less than substantial misinformation, the probability of low trust was 11% (95% CI: 9–13) for individuals without medical care discrimination and 33% (95% CI: 20–45) for those reporting discrimination. The interactions between misinformation and healthcare visit frequency and quality were not statistically significant.
Conclusions:
Perceptions of substantial social media health misinformation were associated with lower trust in the U.S. healthcare system, particularly among individuals reporting experiences of medical care discrimination.
Keywords: Health misinformation, social media, healthcare trust, healthcare quality, healthcare discrimination, HINTS
INTRODUCTION
The rise of social media as a primary source of health information has reshaped public engagement with the healthcare system.1,2 While it expands access to health resources, the widespread presence of misinformation raises concerns about its effects on trust in healthcare.3–5 Misinformation, ranging from misleading medical claims to falsehoods about healthcare policies, poses significant public health risks.6 The COVID-19 pandemic highlighted these risks, as misinformation contributed to increased skepticism toward health authorities and lower adherence to preventive measures.7–12 This study extends prior research by examining how social media misinformation relates to healthcare trust beyond the pandemic era.
Trust in the healthcare system is not only a determinant of healthcare utilization, patient adherence, and overall health outcomes, but it also acts as a critical filter for information credibility.13–17 Trust, defined here as the confidence individuals have in the competence, reliability, and integrity of healthcare providers and institutions, shapes how individuals process and act on health information.18,19 People with high trust are more likely to view information from established medical sources as credible, while those with low trust may be more susceptible to misinformation.20,21 Consequently, trust in the healthcare system influences not only the willingness to seek care but also how health information is interpreted and utilized.22–25 Thus, understanding the relationship between perception of health misinformation and trust in the healthcare system has become imperative in public health research.
To better understand the factors influencing healthcare system trust, we can draw on the Health Belief Model as a conceptual framework.26 The HBM posits that health-related decision-making is influenced by an individual’s perceptions of susceptibility to health issues, the severity of those issues, perceived benefits of action, and perceived barriers to taking that action. In the context of healthcare trust, misinformation on social media could lower individuals’ perceived susceptibilities to adverse health outcomes by undermining their trust in healthcare professionals or modern medical advancements.27 Additionally, misinformation may also reduce perceived benefits of engaging with the healthcare system, which exacerbates distrust.22
Trust in the healthcare system varies by social and demographic factors, with racial and ethnic minorities, lower-income or less-educated individuals, and those with fewer social ties often report lower trust.28–33 Individual healthcare experiences also play a role—frequent healthcare interactions may strengthen trust, while experiences of low-quality care or discrimination can contribute to distrust, which may be further reinforced by misinformation.25,33,34 Identifying how these factors shape responses to misinformation can help pinpoint populations most vulnerable to its effects.32 This study examines the role of healthcare visits, perceived care quality, and medical discrimination in shaping trust amid health misinformation, providing insights to inform public health efforts that address misinformation while improving patient experiences.
This study examined whether perceived health misinformation on social media was associated with lower trust in the healthcare system and whether this relationship varied by healthcare experiences, including visit frequency, perceived care quality, and medical care discrimination. Grounded in the Health Belief Model, we hypothesized that misinformation was linked to lower trust, particularly among those with infrequent visits, lower perceived care quality, or experiences of healthcare discrimination. By exploring these factors, this study provided insights into healthcare trust in the era of digital misinformation.
METHODS
Data
This study utilized data from the 2022 Health Information National Trends Survey 6 (HINTS 6), a cross-sectional survey that assesses the U.S. public’s access to and use of health information, including trust. Data for 6,252 respondents were collected via mail and online surveys from March to November 2022 with a response rate of 28.1%.35 The survey is nationally representative of adults in the United States and uses a complex sampling design to ensure generalizability. Detailed survey methodology is available from the National Cancer Institute.35
For this analysis, we first excluded 134 respondents missing data on the outcome variable. Next, we excluded 1,237 participants who indicated they had not used social media in the survey item about perceptions of social media health information. Additionally, we excluded 565 participants who reported no healthcare visits in the past year because healthcare quality is only asked of those that visited a healthcare provider. Finally, we excluded 511 respondents missing data on any covariates using listwise deletion, resulting in a final analytical sample of 3,805 respondents. This sample is representative of the adult U.S. population that used social media and had at least one healthcare visit in the past year. To assess potential bias from exclusions, we compared excluded and included respondents (Appendix eTable 1) and found no significant predictors of exclusion. Multiple imputation analysis (Appendix eTable 2) confirmed the robustness of findings, yielding results consistent with the primary analysis and reinforcing the reliability of the observed associations between misinformation, healthcare utilization, and trust.
We conducted post-hoc power calculations to evaluate whether our sample size provided sufficient power to detect significant associations, particularly with the primary predictor variable: perception of health misinformation on social media. Based on the observed effect sizes, a sample size of 3,805 respondents offered over 80% power to detect significant associations at the 0.05 significance level. This sample size was adequate to examine the relationships between healthcare trust and the key predictors, including health misinformation, frequency of healthcare visits, perceived quality of care, and perceived discrimination in healthcare.
Measures
The primary outcome variable was trust in the healthcare system, measured by asking, “How much do you trust the healthcare system (for example, hospitals, pharmacies, and other organizations involved in health care)?” Responses were dichotomized as high (“very” and “somewhat”) versus low (“not at all” and “a little”), based on prior literature.17,18,33,36 The primary predictor variable assessed perceptions of health misinformation and disinformation on social media with the question, “How much of the health information that you see on social media do you think is false or misleading?” Responses were dichotomized as “substantial” (“a lot”) versus “less than substantial” (“none,” “a little,” or “some”), consistent with past studies.6,21,22 In addition, we explored several secondary predictors. Frequency of healthcare visits in the past year was categorized as infrequent (1–3 visits) and frequent (4+ visits). Perceived quality of healthcare was categorized as high quality (“excellent” and “very good”) vs. low quality (“good”, “fair”, “poor”). Perceived medical care discrimination was measured as yes versus no. A series of demographic and socioeconomic control variables were included in the analysis to adjust for potential confounders, including age, gender, marital status, urban vs. rural residence, race and ethnicity (Non-Latino White, Non-Latino Black, Latino, Non-Latino Other), education (college graduate or not), full-time employment status, self-reported income satisfaction (finding it difficult, getting by, or living comfortably on present income), health insurance status (insured or uninsured), self-reported health status (excellent, very good, good, fair, or poor), and frequency of social media visits (daily vs. less than daily).
Statistical Analyses
All analyses were survey-weighted to account for the complex sampling design of HINTS and to ensure that the results were nationally representative. Descriptive statistics were generated to provide an overview of the study sample, including both raw sample sizes and survey-weighted percentages for all key variables. The bivariable association between the outcome and main predictor variable was assessed using column percentages and design-based Wald chi-squared test. We ran multivariable logistic regression models to assess the association between perceptions of health misinformation on social media and trust in the healthcare system. Odds ratios (ORs), 95% confidence intervals (CIs), and p-values were reported for each predictor. Interaction terms were included to assess the association between perceptions of health misinformation and healthcare trust varied by frequency of healthcare visits, perceived quality of healthcare, or experiences of medical care discrimination. Predictive margins were calculated to estimate the probability of low trust in the healthcare system across combinations of misinformation perceptions and each moderating variable. To evaluate the statistical significance of interaction effects, we conducted joint significance tests using the Wald test for all interaction terms combined, followed by individual significance tests for each interaction term separately. All multivariable estimates were adjusted for the covariates listed above. All analyses were survey-weighted to account for the complex sampling design of HINTS 6 and to produce nationally representative estimates. We followed the STROBE checklist for cross sectional studies.37
RESULTS
Among the 3,805 adult social media users in the past year included in the analytical sample (reported as unadjusted sample size and survey-weighted percentages in Table 1), most participants reported high trust in the healthcare system, with 84.8% (n=3,276) reporting high trust and 15.2% (n=529) reporting low trust. Approximately 63.2% (n=2,446) of participants perceived less than substantial health misinformation on social media, while 36.8% (n=1,359) perceived substantial misinformation. Most participants reported infrequent healthcare visits in the past year (60.1%), perceived the quality of healthcare they received as high (69.1%), and did not report experiencing racial or ethnic discrimination in medical care (92.8%). The age distribution was skewed towards younger adults, with 55.2% ages 18–49 years, 27.9% aged 50–64 years, and 16.9% ages 65 years or older. The sample was slightly more female (54.3%). Most participants (87.8%) lived in metro areas. The sample was predominantly non-Latino White (63.9%). Nearly two-thirds of the participants (63.1) had not graduated from college. More than half of the participants (57.9%) were employed full-time. Most participants (92.9%) had health insurance and rated their health as excellent, very good, or good (84.5%). Lastly, most participants (69.2%) visited social media daily.
Table 1:
Unadjusted sample size and survey-weighted percentages for study variables from the 2022 Health Information National Trends Survey 6 (N=3,805)
| Raw N | Weighted % | |
|---|---|---|
|
| ||
| N | 3,805 | |
| Trust of Healthcare System | ||
| High | 3,276 | 84.8% |
| Low | 529 | 15.2% |
| Perception of Social Media Health Misinformation | ||
| Less than Substantial | 2,446 | 63.2% |
| Substantial | 1,359 | 36.8% |
| Frequency of Healthcare Visits | ||
| Infrequent visits | 2,104 | 60.1% |
| Frequent visits | 1,701 | 39.9% |
| Perceptions of Healthcare Quality | ||
| Low | 1,183 | 30.9% |
| High | 2,622 | 69.1% |
| Medical Care Discrimination | ||
| No | 3,495 | 92.8% |
| Yes | 310 | 7.2% |
| Age | ||
| 18–49 | 1,540 | 55.2% |
| 50–64 | 1,139 | 27.9% |
| 65+ | 1,126 | 16.9% |
| Gender | ||
| Male | 1,402 | 45.7% |
| Female | 2,403 | 54.3% |
| Marital Status | ||
| Married/Cohabiting | 2,102 | 58.3% |
| Formerly Married | 959 | 10.6% |
| Never Married | 744 | 31.1% |
| Rural/Urban Designation | ||
| Nonmetro | 490 | 12.2% |
| Metro | 3,315 | 87.8% |
| Race and Ethnicity | ||
| Non-Latino White | 2,241 | 63.9% |
| Non-Latino Black | 595 | 10.9% |
| Latino | 643 | 14.7% |
| Non-Latino Other | 326 | 10.4% |
| Education | ||
| Not College Graduate | 1,835 | 63.1% |
| College graduate or higher | 1,970 | 36.9% |
| Full time Employment | ||
| No | 1,804 | 42.1% |
| Yes | 2,001 | 57.9% |
| Feelings about Household Income | ||
| Finding it very/difficult on present income | 717 | 18.4% |
| Getting by on present income | 1,382 | 36.6% |
| Living comfortably on present income | 1,706 | 45.0% |
| Health Insurance | ||
| Uninsured | 244 | 7.1% |
| Insured | 3,561 | 92.9% |
| General Health | ||
| Fair/Poor | 638 | 15.5% |
| Excellent/Very Good/Good | 3,167 | 84.5% |
| Frequency of Social Media Visits | ||
| Less than Daily | 1,291 | 30.8% |
| Daily | 2,514 | 69.2% |
Note: Healthcare system trust was dichotomized as high (“very” and “somewhat”) versus low (“not at all” and “a little”). Perceptions of health misinformation on social media were dichotomized as “substantial” (“a lot”) versus “less than substantial” (“none,” “a little,” or “some”). Frequency of healthcare visits in the past year was categorized as infrequent (1–3 visits) and frequent (4+ visits). Perceived quality of healthcare was categorized as high quality (“excellent” and “very good”) vs. low quality (“good”, “fair”, “poor”). Perceived medical care discrimination was measured as yes versus no.
As shown in Table 2, there was a significant association between healthcare system trust and the perception of health misinformation on social media among adult social media users in the past year (p-value = 0.01). Specifically, among those who reported high trust in the healthcare system, 65.1% perceived less than substantial social media health misinformation, while 34.9% perceived substantial misinformation. Conversely, among those who reported low trust, 52.7% perceived less than substantial health misinformation and 47.3%, perceived substantial health misinformation.
Table 2:
Survey-weighted bivariable column percentages for healthcare system trust and perception of health misinformation on social media among adult social media users in the past year from the 2022 Health Information National Trends Survey 6 (N=3,805)
| Healthcare system trust | ||||
|---|---|---|---|---|
| High | Low | Total | ||
|
|
||||
| Social Media Health Misinformation | ||||
| Less than Substantial | Weighted % | 65.1 | 52.7 | 63.2 |
| Raw N | 2137 | 309 | 2446 | |
| Substantial | Weighted % | 34.9 | 47.3 | 36.8 |
| Raw N | 1139 | 220 | 1359 | |
| Total | Weighted % | 100 | 100 | 100 |
| Raw N | 3276 | 529 | 3805 | |
Note: Healthcare system trust was dichotomized as high (“very” and “somewhat”) versus low (“not at all” and “a little”). Perceptions of health misinformation and disinformation on social media were dichotomized as “substantial” (“a lot”) versus “less than substantial” (“none,” “a little,” or “some”). Adjusted Wald chi-square p-value = 0.01 for this bivariable relationship.
The multivariable logistic regression analysis in Table 3 shows several significant predictors of low trust in the healthcare system among adult social media users. Individuals who perceived substantial social media health misinformation (compared to those who perceived less) were significantly more likely to report low trust in the healthcare system, with an odds ratio (OR) of 1.66 (95% CI: 1.12, 2.48). Frequency of healthcare visits was not significantly associated with lower trust in the healthcare system. Participants who rated the quality of healthcare as high were significantly less likely to report low trust in the system (OR = 0.44; 95% CI: 0.28, 0.69). Participants who experienced medical care discrimination were significantly more likely to report low healthcare system trust (OR = 2.86; 95% CI: 1.41, 5.77). Participants who reported living comfortably on their present income had significantly lower odds of low trust in the healthcare system (OR = 0.48; 95% CI: 0.26, 0.87). The other covariates in the model were not statistically significant.
Table 3:
Multivariable logistic regression (odds ratios (ORs), 95% Confidence Intervals, p-values) predicting low trust of the healthcare system: 2022 Health Information National Trends Survey 6 (N=3,805)
| Variable | OR | 95% CI | p-value |
|---|---|---|---|
| Misinformation | |||
| Less than Substantial | Reference | ||
| Substantial | 1.66 | 1.12, 2.48 | 0.01 |
| Frequency of Healthcare Visits | |||
| Infrequent | Reference | ||
| Frequent | 0.72 | 0.48, 1.10 | 0.13 |
| Perceived Quality of Healthcare | |||
| Low Quality | Reference | ||
| High Quality | 0.44 | 0.28, 0.69 | <0.01 |
| Experienced Discrimination in Healthcare | |||
| No | Reference | ||
| Yes | 2.86 | 1.41, 5.77 | <0.01 |
| Age Group | |||
| 18–49 | Reference | ||
| 50–64 | 0.87 | 0.55, 1.36 | 0.52 |
| 65+ | 0.71 | 0.43, 1.19 | 0.19 |
| Gender | |||
| Male | Reference | ||
| Female | 0.81 | 0.55, 1.19 | 0.27 |
| Marital Status | |||
| Married/Cohabiting | Reference | ||
| Formerly Married | 0.93 | 0.63, 1.38 | 0.73 |
| Never Married | 1.37 | 0.90, 2.09 | 0.14 |
| Rural/Urban Designation | |||
| Nonmetro | Reference | ||
| Metro | 0.73 | 0.50, 1.07 | 0.11 |
| Race and Ethnicity | |||
| Non-Latino White | Reference | ||
| Non-Latino Black | 0.66 | 0.39, 1.13 | 0.13 |
| Latino | 0.82 | 0.56, 1.21 | 0.32 |
| Non-Latino Other | 1.03 | 0.39, 2.75 | 0.95 |
| Education | |||
| Not College Graduate | Reference | ||
| College Graduate or Higher | 1.05 | 0.72, 1.53 | 0.80 |
| Full-Time Employment | |||
| No | Reference | ||
| Yes | 0.97 | 0.60, 1.55 | 0.89 |
| Feelings About Household Income | |||
| Finding it Difficult on Present Income | Reference | ||
| Getting by on Present Income | 0.65 | 0.36, 1.17 | 0.15 |
| Living Comfortably on Present Income | 0.48 | 0.26, 0.87 | 0.02 |
| Health Insurance | |||
| Uninsured | Reference | ||
| Insured | 0.86 | 0.47, 1.57 | 0.62 |
| General Health | |||
| Fair/Poor | Reference | ||
| Excellent/Very Good/Good | 0.70 | 0.42, 1.16 | 0.16 |
| Frequency of Social Media Visits | |||
| Less than Daily | Reference | ||
| Daily | 0.85 | 0.58, 1.25 | 0.40 |
Note: Healthcare system trust was dichotomized as high (“very” and “somewhat”) versus low (“not at all” and “a little”). Perceptions of health misinformation and disinformation on social media were dichotomized as “substantial” (“a lot”) versus “less than substantial” (“none,” “a little,” or “some”). Frequency of healthcare visits in the past year was categorized as infrequent (1–3 visits) and frequent (4+ visits). Perceived quality of healthcare was categorized as high quality (“excellent” and “very good”) vs. low quality (“good”, “fair”, “poor”). Perceived medical care discrimination was measured as yes versus no.
The marginal effects for low healthcare system trust across different interaction terms between perceived health misinformation and key factors were calculated. Full marginal effect estimates and statistical significance tests for all interaction models are available in Appendix eTable3. The interaction between perceived social media health misinformation and frequency of healthcare visits was not statistically significant (p-value = 0.09). The interaction between perceived misinformation and perceived quality of healthcare was jointly significant (p < 0.001), indicating a combined effect of these variables on trust in the healthcare system. As shown in Figure 1, among those perceiving less than substantial misinformation, individuals reporting high-quality healthcare (9%, 95% CI: 6–11) had a significantly lower probability of low trust compared to those reporting low-quality care (21%, 95% CI: 16–26). No significant differences were observed among those perceiving substantial misinformation.
Figure 1: Predicted Probability (95% CIs) of Low Healthcare System Trust by Perceived Health Misinformation and Perceived Quality of Healthcare.
Note: Predicted probabilities of low trust in the healthcare system (with 95% confidence intervals) are shown for different levels of perceived health misinformation on social media (“Substantial” vs. “Less than Substantial”) and perceived quality of healthcare (“High” vs. “Low”). Estimates are derived from survey-weighted logistic regression models controlling for age, gender, marital status, urban/rural residence, race/ethnicity, education, employment, income satisfaction, health insurance status, general health, and frequency of social media use. The joint significance test (p < 0.001) confirmed that perceived healthcare quality moderates the association between misinformation and trust. While the interaction term alone was not statistically significant (p = 0.1175), the marginal effects suggest that those perceiving less than substantial misinformation and low-quality healthcare were more likely to report low trust of the healthcare system compared to those perceiving less than substantial misinformation and high-quality healthcare.
The interaction between perceived misinformation and medical care discrimination was significant in the joint test (p < 0.001) and in the specific interaction term (p = 0.03). As shown in Figure 2, those perceiving less than substantial misinformation and no discrimination had a 11% (95% CI: 9–13) probability of low trust, while those perceiving less than substantial misinformation and discrimination exhibited a much higher probability of low trust at 33% (95% CI: 20–45). Among those perceiving substantial misinformation, the difference in low trust between individuals with and without experiences of medical care discrimination was not statistically significant.
Figure 2: Predicted Probability (95% CIs) of Low Healthcare System Trust by Perceived Health Misinformation and Experiences of Healthcare Discrimination.
Note: Predicted probabilities of low trust in the healthcare system (with 95% confidence intervals) are plotted for different levels of perceived health misinformation on social media (“A Lot” vs. “< A Lot”) and self-reported experiences of medical care discrimination (“Yes” vs. “No”). Estimates are derived from survey-weighted logistic regression models adjusting for age, gender, marital status, urban/rural residence, race/ethnicity, education, employment, income satisfaction, health insurance status, general health, and frequency of social media use. The joint significance test (p < 0.001) and the interaction term (p = 0.03) were both statistically significant. Individuals perceiving less than substantial misinformation and no discrimination had a much lower probability of low trust compared to those who experienced discrimination. Among those perceiving substantial misinformation, trust levels were lower for individuals reporting discrimination, though the differences were not statistically significant.
DISCUSSION
In a cross-sectional analysis of U.S. adults who used social media and had at least one healthcare visit in the past year, we observed significant associations between perceptions of social media health misinformation and trust in the healthcare system. Specifically, individuals who perceived substantial misinformation on social media reported significantly lower levels of trust, even after adjusting for sociodemographic, health-related, and behavioral factors. This finding aligns with previous research indicating that misinformation can erode public trust in healthcare systems and professionals, potentially leading to poorer health outcomes.22–24,27 The growing reliance on social media for health information highlights the potential for misinformation to reach a wide audience and negatively shape public perceptions.3–6
Both perceived healthcare quality and experiences of discrimination were associated with trust in the healthcare system.34 Individuals who reported higher-quality healthcare had greater trust in the healthcare system, while those perceiving lower-quality care were more likely to report lower trust. Additionally, among individuals who perceived less than substantial misinformation, those with lower-quality healthcare had lower trust compared to those with high-quality care. While this pattern suggests that positive healthcare experiences may correspond with higher trust, the interaction between perceived misinformation and healthcare quality was not statistically significant.
Discrimination was strongly associated with trust in the healthcare system. Individuals who reported discrimination during healthcare encounters were more likely to report lower trust. Additionally, the association between misinformation and trust differed by discrimination status, which aligns with prior research suggesting that experiences of discrimination in healthcare and exposure to misinformation may coincide with lower trust in the healthcare system.21,22 Discriminatory practices in healthcare have been linked to lower patient satisfaction, reduced healthcare engagement, and poorer health outcomes, all of which can further exacerbate distrust in healthcare institutions.25,28,38 Addressing discrimination in healthcare settings may be a candidate for interventions aimed at improving trust and mitigating the potential effects of misinformation.32
The findings of this study support the application of the Health Belief Model (HBM) in understanding the dynamics of healthcare trust. According to the HBM, individuals’ health-related decisions are shaped by their perceived susceptibility to health issues, perceived severity, perceived benefits of action, and perceived barriers.26,27 Misinformation can distort these perceptions, lowering perceived susceptibility or severity and thus undermining trust in healthcare systems. The interaction effects observed—such as the higher likelihood of trust among individuals with positive healthcare experiences—demonstrate that personal experiences significantly shape health beliefs. These insights highlight the value of using the HBM framework to explore how individual and contextual factors interact to shape trust in the healthcare system.13–15
Public Health Implications
This study highlights the need for multifaceted strategies to address health misinformation and build trust in the healthcare system.1,39,40 While perceived healthcare quality and absence of discrimination were associated with higher trust, these associations were most evident among individuals perceiving less than substantial misinformation. These findings suggest that positive care experiences may help maintain trust in the healthcare system but may not fully mitigate the effects of perceived misinformation.41,42 Public health interventions should address both the content of misinformation and structural factors such as care quality and equity in treatment to support trust.
Healthcare providers play a key role in countering misinformation and thereby mitigate its impact on healthcare system trust. Strengthening provider communication through strategies like the Teach-Back Method, Motivational Interviewing, and the AHRQ Health Literacy Universal Precautions Toolkit have been shown to improve patient trust and adherence.43–47 Additionally, cultural humility and implicit bias training can address disparities that contribute to mistrust, particularly among populations subject to discrimination.48,49 Expanding the role of community health workers (CHWs) as trusted messengers may further bridge gaps in health literacy and provide culturally tailored, effective responses to misinformation.50
Limitations
This study used cross-sectional data, which limits the ability to infer causality. While we hypothesized that perception of health misinformation on social media was associated with trust in the healthcare system, it is also possible that individuals with lower trust were more susceptible to misinformation. Second, the reliance on self-reported data may be subject to recall bias and social desirability bias. Participants may have misreported their perceptions of misinformation or trust in healthcare, which could lead to an over- or underestimation of these variables. The survey question measuring perceived misinformation did not capture the specific types of misinformation encountered or its sources, which could provide deeper insights into the mechanisms behind its influence on trust. Participants who did not report healthcare visits in the past year were excluded from the analyses, which limits the generalizability of the findings for individuals with no interactions with the healthcare system. Our measure of healthcare trust was limited to a single item, which may not fully capture the multidimensional nature of trust in healthcare systems, providers, and institutions.
However, this item was derived from the Hall Trust Scale, a widely used and validated measure of trust that has been tested across diverse populations.15–17,36 Despite these limitations, the study has notable strengths. The use of a nationally representative sample of U.S. adults enhances the generalizability of the findings. Furthermore, the study’s sensitivity analyses add robustness to the results by addressing potential biases arising from missing data. The application of the Health Belief Model as a guiding framework provides a theoretical basis for interpreting the findings and understanding the mechanisms underlying healthcare system trust in the context of social media misinformation.
Future Directions
Further research is needed to better understand the relationship between health misinformation on social media and trust in the healthcare system. Longitudinal or experimental studies could help clarify the direction of these associations, while qualitative methods such as interviews or focus groups may provide deeper insight into how individuals interpret misinformation and its impact on trust. Community-engaged research could examine how misinformation disproportionately affects marginalized communities and identify culturally appropriate strategies to rebuild trust. Additionally, future studies should explore the role of trusted messengers, such as healthcare professionals and community health workers, in countering misinformation and reinforcing patient-provider trust.
Conclusion
This study observed that perception of substantial health misinformation on social media was associated with lower trust in the U.S. healthcare system. Trust also varied by individual experiences with medical care discrimination. These findings highlight the need for public health interventions that not only counter health misinformation but also address inequities in healthcare experiences to support trust.
Funding:
ANO and JPS were supported by the National Institute on Minority Health and Health Disparities (NIMHD) at the National Institutes of Health (NIH) under award number R01MD018727. The content is solely the responsibility of the authors and does not necessarily represent the official views of NIMHD. The funders had no role in the study design, data analysis, decision to publish, or preparation of the manuscript.
Online Appendix
eTable1:
Multivariable firth logit regression predicting exclusion from the analytical sample among U.S. adults: 2022 Health Information National Trends Survey 6 (N=3,805)
| Variable | Odds Ratio | P-value |
|---|---|---|
|
| ||
| Healthcare System Trust | 1.97 | 0.729 |
| Misinformation | 1.41 | 0.839 |
| Frequency of Healthcare Visits | 1.43 | 0.858 |
| Perceived Quality of Healthcare | 0.82 | 0.923 |
| Perceived Racism in Healthcare | 2.62 | 0.619 |
| Age Group | ||
| 50–64 | 1.94 | 0.763 |
| 65+ | 2.25 | 0.793 |
| Gender | 0.82 | 0.907 |
| Female | ||
| Marital Status | ||
| Formerly Married | 1.00 | 0.999 |
| Never Married | 1.79 | 0.781 |
| Rural/Urban Designation | 0.25 | 0.432 |
| Metro | ||
| Race/Ethnicity | ||
| NH Black | 3.03 | 0.628 |
| Hispanic | 3.22 | 0.599 |
| NH Asian & Other | 4.67 | 0.469 |
| Education | 1.24 | 0.904 |
| Full-Time Employment | 1.31 | 0.911 |
| Feelings About Household Income | ||
| Getting by on Present Income | 0.90 | 0.963 |
| Living Comfortably on Present Income | 0.84 | 0.947 |
| Health Insurance | 0.20 | 0.434 |
| General Health | 0.50 | 0.762 |
| Frequency of Social Media Visits | 0.75 | 0.860 |
| Constant | 0.01 | 0.228 |
| Overall Model | 0.999 | |
Note: The outcome variable is exclusion from the analytic sample (1 = excluded, 0 = included). Odds ratios (ORs) represent the likelihood of exclusion relative to inclusion for each predictor. Estimates were obtained using a Firth logistic regression model, which reduces small-sample bias and prevents complete separation in logistic regression. A p-value < 0.05 was considered statistically significant.
The very high p-value (0.9992) suggests that the overall model does not significantly explain variation in exclusion from the analytic sample. Results indicate none of the study variables were statistically significant. Specifically, missing data for individuals with no healthcare appointments was not significantly associated with exclusion (OR = 0.82, p = 0.923). These findings support the robustness of our results, suggesting that our estimates of the relationship between misinformation and trust are unlikely to be biased due to missing data patterns.
eTable 2:
Multivariable logistic regression (odds ratios (ORs), 95% Cis, p-values) predicting low trust of the healthcare system using multiple imputation: 2022 Health Information National Trends Survey 6 (N=4,269)
| OR | 95% CI | p-value | |
|---|---|---|---|
|
|
|||
| Social Media Health Misinformation | |||
| Less than Substantial | Reference | ||
| Substantial | 1.50 | 1.07, 2.11 | 0.02 |
| Frequency of Healthcare Visits | |||
| Infrequent visits | Reference | ||
| Frequent visits | 0.62 | 0.44, 0.87 | 0.01 |
| Perceived Quality of Healthcare | |||
| Low Quality | Reference | ||
| High Quality | 0.42 | 0.28, 0.62 | 0.00 |
| Perceived Medical Care Discrimination | |||
| No | Reference | ||
| Yes | 2.40 | 1.42, 4.06 | 0.00 |
Note: The multiple imputation (MI) procedure was conducted using a chained equations approach with 20 imputations. The “quality of care” variable was imputed using logistic regression, incorporating all available covariates to predict missing values. The analysis was conducted on the imputed datasets using survey weights to account for the complex survey design, and the final estimates were combined using Rubin’s rules. This method allows for the inclusion of all cases, reducing potential bias due to missing data and improving the robustness of the results. All estimates were adjusted for age, gender, marital status, urban vs. rural residence, race & ethnicity, education, employment status, self-reported income satisfaction, health insurance status, self-reported health status, and frequency of social media visits.
Social media misinformation was significantly associated with increased odds of low trust in the healthcare system (OR = 1.50, 95% CI = 1.07, 2.11). Additionally, high perceived quality of care (OR = 0.42, 95% CI = 0.28, 0.62) and more frequent healthcare visits (OR = 0.62, 95% CI = 0.44, 0.87) were both associated with reduced odds of low trust.
eTable 3:
Marginal Effects of Low Healthcare System Trust Across Interaction Terms Between Perception of Health Misinformation and Key Factors Among U.S. Adult Social Media Users: 2022 Health Information National Trends Survey 6, N=3,805
| Interaction Effect | Margin (%) | 95% CI | Joint Significance p-value | Interaction Term Only p-value |
|---|---|---|---|---|
| Misinformation × Frequency of Healthcare Visits | 0.09 | 0.14 | ||
| Less than Substantial and Infrequent Visits | 13 | 11 – 15 | ||
| Less than Substantial and Frequent Visits | 13 | 9 – 16 | ||
| Substantial and Infrequent Visits | 23 | 15 – 30 | ||
| Substantial and Frequent Visits | 14 | 10 – 18 | ||
| Misinformation × Perceived Quality of Healthcare | < 0.001 | 0.12 | ||
| Less than Substantial and Low Quality | 21 | 16 – 26 | ||
| Less than Substantial and High Quality | 9 | 6 – 11 | ||
| Substantial and Low Quality | 24 | 17 – 31 | ||
| Substantial and High Quality | 17 | 10 – 23 | ||
| Misinformation × Medical Care Discrimination | < 0.001 | 0.03 | ||
| Less than Substantial and No Discrimination | 11 | 9 – 13 | ||
| Less than Substantial and Discrimination | 33 | 20 – 45 | ||
| Substantial and No Discrimination | 18 | 14 – 23 | ||
| Substantial and Discrimination | 26 | 10 – 43 |
Note: Healthcare system trust was dichotomized as high (“very” and “somewhat”) versus low (“not at all” and “a little”). Perceptions of health misinformation and disinformation on social media were dichotomized as “substantial” (“a lot”) versus “less than substantial” (“none,” “a little,” or “some”). Frequency of healthcare visits in the past year was categorized as infrequent (1–3 visits) and frequent (4+ visits). Perceived quality of healthcare was categorized as high quality (“excellent” and “very good”) vs. low quality (“good”, “fair”, “poor”). Perceived medical care discrimination was measured as yes versus no. All estimates were adjusted for age, gender, marital status, urban vs. rural residence, race & ethnicity, education, employment status, self-reported income satisfaction, health insurance status, self-reported health status, and frequency of social media visits. To assess the significance of interaction effects, a joint significance test examined whether the main effects and interaction term collectively contributed to the model. Additionally, a separate test evaluated whether the interaction term alone was statistically significant.
Footnotes
Conflicts of Interest: The authors have no relevant financial or non-financial conflicts of interest to disclose.
CRediT Statement:
JPS (Conceptualization, Formal Analysis, Methodology, Writing – Original Draft, Writing – Review & Editing)
SP (Conceptualization, Methodology, Writing – Review & Editing)
EHA (Methodology, Writing – Review & Editing)
DBN (Methodology, Writing – Review & Editing)
ANO (Funding Acquisition, Writing – Review & Editing)
Ethical Considerations:
These publicly available, de-identified data did not require institutional review board approval, as determined by the University of Texas Southwestern Medical Center human research protection program.
Data Availability:
The data sets analyzed during this study are available through the National Cancer Institute: https://hints.cancer.gov/.
REFERENCES
- 1.Borges do Nascimento IJ, Pizarro AB, Almeida JM, et al. Infodemics and health misinformation: a systematic review of reviews. Bull World Health Organ. 2022;100(9):544–561. doi: 10.2471/BLT.21.287654 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Laranjo L, Arguel A, Neves AL, et al. The influence of social networking sites on health behavior change: a systematic review and meta-analysis. J Am Med Inform Assoc. 2015;22(1):243–256. doi: 10.1136/amiajnl-2014-002841 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 3.Moorhead SA, Hazlett DE, Harrison L, Carroll JK, Irwin A, Hoving C. A new dimension of health care: systematic review of the uses, benefits, and limitations of social media for health communication. J Med Internet Res. 2013;15(4):e85. doi: 10.2196/jmir.1933 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 4.Gatewood J, Monks SL, Singletary CR, Vidrascu E, Moore JB. Social Media in Public Health: Strategies to Distill, Package, and Disseminate Public Health Research. Journal of Public Health Management and Practice. 2020;26(5):489–492. doi: 10.1097/PHH.0000000000001096 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Xu S, Markson C, Costello KL, Xing CY, Demissie K, Llanos AA. Leveraging Social Media to Promote Public Health Knowledge: Example of Cancer Awareness via Twitter. JMIR Public Health Surveill. 2016;2(1):e17. doi: 10.2196/publichealth.5205 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6.Stimpson JP, Ortega AN. Social media users’ perceptions about health mis-and disinformation on social media. Health affairs scholar. 2023;1(4):qxad050. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7.Moran P Social Media: A Pandemic of Misinformation. Am J Med. 2020;133(11):1247–1248. doi: 10.1016/j.amjmed.2020.05.021 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8.Liu PL, Zhao X, Wan B. COVID-19 information exposure and vaccine hesitancy: The influence of trust in government and vaccine confidence. Psychol Health Med. 2023;28(1):27–36. doi: 10.1080/13548506.2021.2014910 [DOI] [PubMed] [Google Scholar]
- 9.Amara PS, Platt JE, Raj M, Nong P. Learning about COVID-19: sources of information, public trust, and contact tracing during the pandemic. BMC Public Health. 2022;22(1):1348. doi: 10.1186/s12889-022-13731-7 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10.Lanier HD, Diaz MI, Saleh SN, Lehmann CU, Medford RJ. Analyzing COVID-19 disinformation on Twitter using the hashtags #scamdemic and #plandemic: Retrospective study. De Silva D, ed. PLoS ONE. 2022;17(6):e0268409. doi: 10.1371/journal.pone.0268409 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 11.Suran M, Bucher K. False Health Claims Abound, but Physicians Are Still the Most Trusted Source for Health Information. JAMA. 2024;331(19):1612. doi: 10.1001/jama.2024.6837 [DOI] [PubMed] [Google Scholar]
- 12.Jamison AM, Broniatowski DA, Quinn SC. Malicious Actors on Twitter: A Guide for Public Health Researchers. Am J Public Health. 2019;109(5):688–692. doi: 10.2105/AJPH.2019.304969 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 13.Mechanic D, Meyer S. Concepts of trust among patients with serious illness. Soc Sci Med. 2000;51(5):657–668. doi: 10.1016/s0277-9536(00)00014-9 [DOI] [PubMed] [Google Scholar]
- 14.Gilson L. Trust and the development of health care as a social institution. Soc Sci Med. 2003;56(7):1453–1468. doi: 10.1016/s0277-9536(02)00142-9 [DOI] [PubMed] [Google Scholar]
- 15.Taylor LA, Nong P, Platt J. Fifty Years of Trust Research in Health Care: A Synthetic Review. Milbank Q. 2023;101(1):126–178. doi: 10.1111/1468-0009.12598 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16.Platt J, Taylor L. Measuring Trust: Where Are We and Where Do We Need to Go? AcademyHealth; 2024. https://academyhealth.org/publications/2024-02/assessing-trust-health-care-compendium-trust-measures [Google Scholar]
- 17.Hall MA, Camacho F, Dugan E, Balkrishnan R. Trust in the medical profession: conceptual and measurement issues. Health Serv Res. 2002;37(5):1419–1439. doi: 10.1111/1475-6773.01070 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18.Hall MA, Zheng B, Dugan E, et al. Measuring patients’ trust in their primary care providers. Med Care Res Rev. 2002;59(3):293–318. doi: 10.1177/1077558702059003004 [DOI] [PubMed] [Google Scholar]
- 19.Meyer SB, Brown P, Calnan M, et al. Development and validation of the Trust in Multidimensional Healthcare Systems Scale (TIMHSS). Int J Equity Health. 2024;23(1):94. doi: 10.1186/s12939-024-02162-y [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20.Stimpson JP, Park S, Pruitt SL. Trusting information on cancer varies by source of information and political viewpoint. Cancer Causes & Control. 2024;35(1):177–184. [DOI] [PubMed] [Google Scholar]
- 21.Stimpson JP, Park S, Pruitt SL, Ortega AN. Variation in Trust in Cancer Information Sources by Perceptions of Social Media Health Mis-and Disinformation and by Race and Ethnicity Among Adults in the United States: Cross-Sectional Study. JMIR cancer. 2024;10(1):e54162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22.Stimpson JP, Park S, Wilson FA, Ortega AN. Variations in Unmet Health Care Needs by Perceptions of Social Media Health Mis-and Disinformation, Frequency of Social Media Use, Medical Trust, and Medical Care Discrimination: Cross-Sectional Study. JMIR public health and surveillance. 2024;10(1):e56881. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23.Lee SJ, Lee CJ, Hwang H. The impact of COVID-19 misinformation and trust in institutions on preventive behaviors. Health Education Research. 2023;38(1):95–105. doi: 10.1093/her/cyac038 [DOI] [PubMed] [Google Scholar]
- 24.Tabler J, Snyder JA, White C, Freng A, Thunström L. COVID-19 health practices and attitudes in the United States: the role of trust in healthcare. Z Gesundh Wiss. Published online August 6, 2022:1–14. doi: 10.1007/s10389-022-01737-9 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25.LaVeist TA, Isaac LA, Williams KP. Mistrust of health care organizations is associated with underutilization of health services. Health Serv Res. 2009;44(6):2093–2105. doi: 10.1111/j.1475-6773.2009.01017.x [DOI] [PMC free article] [PubMed] [Google Scholar]
- 26.Glanz K, Rimer BK, Viswanath K, eds. Health Behavior: Theory, Research, and Practice. Fifth edition. Jossey-Bass & Pfeiffer Imprints, Wiley; 2015. [Google Scholar]
- 27.Ruppel EK. Scanning Health Information Sources: Applying and Extending the Comprehensive Model of Information Seeking. Journal of Health Communication. 2016;21(2):208–216. doi: 10.1080/10810730.2015.1058438 [DOI] [PubMed] [Google Scholar]
- 28.Boulware LE, Cooper LA, Ratner LE, LaVeist TA, Powe NR. Race and trust in the health care system. Public Health Rep. 2003;118(4):358–365. doi: 10.1093/phr/118.4.358 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29.Lindström M Marital status and generalized trust in other people: A population-based study. The Social Science Journal. 2012;49(1):20–23. doi: 10.1016/j.soscij.2011.07.002 [DOI] [Google Scholar]
- 30.Ahern MM, Hendryx MS. Social capital and trust in providers. Social Science & Medicine. 2003;57(7):1195–1203. doi: 10.1016/S0277-9536(02)00494-X [DOI] [PubMed] [Google Scholar]
- 31.Stimpson JP, Park S, Srivastava A, Cano MÁ. Belief that Progress has Been Made in Curing Cancer Varies by Perception of Social Media Health Mis-and Disinformation, Education, Frequency of Social Media Use, and Healthcare System Trust: A Cross-Sectional Study. Cancer Control. 2024;31:10732748241289259. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 32.Richmond J, Anderson A, Cunningham-Erves J, Ozawa S, Wilkins CH. Conceptualizing and Measuring Trust, Mistrust, and Distrust: Implications for Advancing Health Equity and Building Trustworthiness. Annu Rev Public Health. 2024;45(1):465–484. doi: 10.1146/annurev-publhealth-061022-044737 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 33.Bazargan M, Cobb S, Assari S. Discrimination and Medical Mistrust in a Racially and Ethnically Diverse Sample of California Adults. Ann Fam Med. 2021;19(1):4–15. doi: 10.1370/afm.2632 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 34.Radević I, Dimovski V, Lojpur A, Colnar S. Quality of Healthcare Services in Focus: The Role of Knowledge Transfer, Hierarchical Organizational Structure and Trust. Knowledge Management Research & Practice. 2023;21(3):525–536. doi: 10.1080/14778238.2021.1932623 [DOI] [Google Scholar]
- 35.National Cancer Institute. Health Information National Trends Survey 6 (HINTS 6) Methodology Report.; 2023. https://hints.cancer.gov/data/methodology-reports.aspx [Google Scholar]
- 36.Merenstein Z, Shuemaker JC, Phillips RL. Measuring Trust in Primary Care. Milbank Quarterly. 2023;101(3):841–880. doi: 10.1111/1468-0009.12654 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 37.von Elm E, Altman DG, Egger M, et al. The Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. Ann Intern Med. 2007;147(8):573–577. doi: 10.7326/0003-4819-147-8-200710160-00010 [DOI] [PubMed] [Google Scholar]
- 38.Benjamins MR, Whitman S. Relationships between discrimination in health care and health care outcomes among four race/ethnic groups. J Behav Med. 2014;37(3):402–413. doi: 10.1007/s10865-013-9496-7 [DOI] [PubMed] [Google Scholar]
- 39.Kozyreva A, Lorenz-Spreen P, Herzog SM, et al. Toolbox of individual-level interventions against online misinformation. Nat Hum Behav. 2024;8(6):1044–1052. doi: 10.1038/s41562-024-01881-0 [DOI] [PubMed] [Google Scholar]
- 40.Tay LQ, Hurlstone MJ, Kurz T, Ecker UKH. A comparison of prebunking and debunking interventions for implied versus explicit misinformation. Br J Psychol. 2022;113(3):591–607. doi: 10.1111/bjop.12551 [DOI] [PubMed] [Google Scholar]
- 41.Timmermans S The Engaged Patient: The Relevance of Patient-Physician Communication for Twenty-First-Century Health. J Health Soc Behav. 2020;61(3):259–273. doi: 10.1177/0022146520943514 [DOI] [PubMed] [Google Scholar]
- 42.Hong H, Oh HJ. The Effects of Patient-Centered Communication: Exploring the Mediating Role of Trust in Healthcare Providers. Health Communication. 2020;35(4):502–511. doi: 10.1080/10410236.2019.1570427 [DOI] [PubMed] [Google Scholar]
- 43.Bidlespacher K, Mulkey DC. The Effect of Teach-Back on Readmission Rates in Rehabilitation Patients. Rehabil Nurs. 2024;49(2):65–72. doi: 10.1097/RNJ.0000000000000452 [DOI] [PubMed] [Google Scholar]
- 44.Ha Dinh TT, Bonner A, Clark R, Ramsbotham J, Hines S. The effectiveness of the teach-back method on adherence and self-management in health education for people with chronic disease: a systematic review. JBI Database System Rev Implement Rep. 2016;14(1):210–247. doi: 10.11124/jbisrir-2016-2296 [DOI] [PubMed] [Google Scholar]
- 45.Papus M, Dima AL, Viprey M, Schott AM, Schneider MP, Novais T. Motivational interviewing to support medication adherence in adults with chronic conditions: Systematic review of randomized controlled trials. Patient Educ Couns. 2022;105(11):3186–3203. doi: 10.1016/j.pec.2022.06.013 [DOI] [PubMed] [Google Scholar]
- 46.Mabachi NM, Cifuentes M, Barnard J, et al. Demonstration of the Health Literacy Universal Precautions Toolkit: Lessons for Quality Improvement. J Ambul Care Manage. 2016;39(3):199–208. doi: 10.1097/JAC.0000000000000102 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 47.Weiss BD, Brega AG, LeBlanc WG, et al. Improving the Effectiveness of Medication Review: Guidance from the Health Literacy Universal Precautions Toolkit. J Am Board Fam Med. 2016;29(1):18–23. doi: 10.3122/jabfm.2016.01.150163 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 48.Okeahialam N, Salami O, Siddiqui F, Thangaratinam S, Khalil A, Thakar R. Effects of strategies to tackle racism experienced by healthcare professionals: a systematic review. BMJ Open. 2025;15(1):e091811. doi: 10.1136/bmjopen-2024-091811 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 49.Vela MB, Erondu AI, Smith NA, Peek ME, Woodruff JN, Chin MH. Eliminating Explicit and Implicit Biases in Health Care: Evidence and Research Needs. Annu Rev Public Health. 2022;43(1):477–501. doi: 10.1146/annurev-publhealth-052620-103528 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 50.Ivantechenko M, Akhter S, Gutnick D. Empowering Trusted Messengers With Motivational Interviewing Skills: Improving Dissemination of Accurate Information to Communities. Community Health Equity Res Policy. Published online November 27, 2024:2752535X241304081. doi: 10.1177/2752535X241304081 [DOI] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Data Availability Statement
The data sets analyzed during this study are available through the National Cancer Institute: https://hints.cancer.gov/.