Social Media, AI, and Health Literacy in Rheumatic Diseases
وسائل التواصل الاجتماعي والذكاء الاصطناعي والثقافة الصحية في الأمراض الروماتيزمية
Journal: Journal of Clinical Rheumatology: Practical Reports on Rheumatic & Musculoskeletal Diseases
University: Tertiary outpatient clinic in Monterrey, Mexico
Study Type: cross-sectional
Evidence Level: low
Participants: 95
Published:
30-Second Summary
This cross-sectional study evaluated how 95 patients with rheumatic diseases engaged with social media and artificial intelligence for health information. It found that younger patients demonstrated higher electronic health literacy and were less likely to alter their treatments based on social media content compared to older patients.
1-Minute Summary
Researchers conducted an observational, cross-sectional study in Monterrey, Mexico, involving 95 adult patients with rheumatic conditions. The study assessed their engagement with social media and artificial intelligence regarding health information. Results indicated that while daily social media use was nearly universal, patients aged 45 or younger had significantly higher electronic health literacy. Furthermore, older patients were more likely to change their treatment based on social media content, highlighting the role of digital health literacy in processing online information.
3-Minute Summary
This comprehensive analysis is based exclusively on the provided abstract from the Journal of Clinical Rheumatology: Practical Reports on Rheumatic & Musculoskeletal Diseases, titled 'The Impact of Social Media and Artificial Intelligence on Illness Perception and Treatment Adherence in Patients With Rheumatic Diseases.' It is imperative to state immediately that this analysis relies solely on a truncated summary of the research; full-text verification is absolutely required to fully understand the study's methodology, the specific clinical characteristics of the participants, and the nuanced context of the findings. No clinical decisions, policy changes, or alterations to patient care should be made based on this abstract alone. The abstract outlines an observational, cross-sectional study designed to evaluate how patients with rheumatology conditions engage with and perceive social media (SM) and artificial intelligence (AI) in the context of health information. The researchers contextualize their study within the aftermath of the COVID-19 'infodemic,' a period characterized by an overwhelming amount of information—both accurate and inaccurate—which has led patients to increasingly rely on digital platforms for health-related queries. The study specifically highlights the importance of understanding electronic health literacy (HL) in regions facing unique sociocultural challenges and varying levels of education, though the abstract does not detail what specific sociocultural challenges were present in the study population. Conducted at a tertiary outpatient clinic in Monterrey, Mexico, the study took place over a specific timeframe from August 2024 to November 2024. The sample size consisted of ninety-five (95) adult patients. The demographic breakdown provided in the abstract reveals a median age of 50 years, with a heavily skewed sex distribution: 92.6% of the participants were female. This demographic profile is a critical factor when interpreting the results, as the findings predominantly reflect the attitudes and behaviors of middle-aged women in a specific geographic and clinical setting. To gather data, the researchers utilized a structured questionnaire that incorporated three specific measurement tools: the eHealth Literacy Scale (eHEALS), the Social Media Engagement Questionnaire (SMEQ), and the AI Attitude Scale (AIAS-4). The abstract does not provide the specific questions asked within these scales, nor does it detail their scoring mechanisms, which underscores the need for full-text verification. The statistical analysis involved comparing responses between two distinct age groups: those aged 45 years or younger (≤45) and those older than 45 years (>45). The researchers employed the Chi-square (χ2) test for categorical variables and the Mann-Whitney U test for continuous or ordinal variables, setting the threshold for statistical significance at p < 0.05. The reported findings indicate a highly digitized patient population within this specific sample. Daily use of social media was reported as almost universal, at 94.7%. Furthermore, nearly a third of the participants (30.5%) reported having used artificial intelligence to ask questions related to their disease. When analyzing the data by age group, the researchers found statistically significant differences. Patients aged 45 or younger demonstrated significantly higher electronic health literacy (p = 0.010) and significantly higher engagement with social media (p < 0.001) compared to the older cohort. One of the most notable findings reported in the abstract relates to treatment adherence and the influence of social media. The data suggests that younger patients (≤45 years) were less likely to alter their medical treatment based on content they encountered on social media compared to older patients (>45 years), a finding that reached statistical significance (p = 0.043). This suggests a potential vulnerability among older patients in this cohort regarding the critical evaluation of online health information and its application to their prescribed treatment regimens. Interestingly, the abstract notes that attitudes toward artificial intelligence did not show significant variation by age. Furthermore, the study highlights a strong foundation of trust and communication between the patients and their healthcare providers regarding digital information. A large majority of the patients (84.2%) believed that their doctors were open to discussing information found online. Additionally, an overwhelming 91.6% of the participants stated they would use social media for health purposes if it were explicitly recommended by their physicians. The authors conclude that digital health literacy significantly influences how patients process online information and manage their treatment adherence. However, as this is classified as a low-evidence-level, cross-sectional study, it is crucial to remember that these findings represent a mere snapshot in time. The study design cannot establish causality; it cannot prove that lower health literacy causes older patients to change their treatments, only that an association exists within this specific sample of 95 patients in Monterrey, Mexico. The limitations inherent in the small sample size, the single-center design, the reliance on self-reported questionnaires, and the heavily female demographic must be carefully considered. Full-text verification remains essential to critically evaluate the robustness of these conclusions and the validity of the measurement tools used in this specific cultural context.
Full Analysis
This extensive research-literacy analysis is predicated entirely upon the provided abstract from the Journal of Clinical Rheumatology: Practical Reports on Rheumatic & Musculoskeletal Diseases. The study under review is titled 'The Impact of Social Media and Artificial Intelligence on Illness Perception and Treatment Adherence in Patients With Rheumatic Diseases.' It is of paramount importance to establish at the outset that this document is an analysis of a summary, not the full peer-reviewed article. Consequently, full-text verification is absolutely essential before drawing any definitive conclusions regarding the study's methodology, the validity of its findings, or its applicability to broader patient populations. The classification of this study as possessing a 'low' evidence level further necessitates a highly cautious and critical interpretation of the reported data. This analysis will systematically deconstruct the abstract's reported design, demographic profile, instrumentation, statistical framework, findings, and inherent limitations. ### 1. Introduction and Study Context The abstract introduces a highly relevant contemporary issue: the intersection of digital information-seeking behavior and chronic disease management. The authors frame their investigation within the context of the COVID-19 'infodemic,' a term used by public health organizations to describe the rapid and widespread dissemination of both accurate and inaccurate information during the pandemic. This environment has purportedly accelerated the rate at which patients turn to digital platforms—specifically social media (SM) and artificial intelligence (AI)—to source health information, self-educate, and potentially make decisions regarding their care. The researchers explicitly state that understanding electronic health literacy (HL) is vital in regions characterized by unique sociocultural challenges and varying educational levels. While the abstract does not elaborate on the specific sociocultural challenges of the target region, this framing suggests an awareness that digital literacy is not uniformly distributed and is likely influenced by broader socioeconomic determinants of health. The primary objective of the study, as stated, is to evaluate engagement with and attitudes toward SM and AI among rheumatology patients, with the ultimate goal of informing patient-centered care. It is important to note that while the goal is to inform care, the abstract itself does not provide actionable clinical protocols; it merely provides observational data. ### 2. Methodological Framework: The Observational Cross-Sectional Design The study is explicitly defined as an observational, cross-sectional study. Understanding this specific methodological design is crucial for research literacy and for accurately interpreting the weight of the evidence. An observational study is one in which researchers measure variables of interest without assigning treatments or intervening in the participants' environment. The researchers are merely observing and recording data as it naturally exists. Furthermore, a cross-sectional design means that the data was collected at a single point in time—in this case, between August 2024 and November 2024. Cross-sectional studies are often described as 'snapshots' of a population. Because all variables (e.g., age, electronic health literacy, social media use, treatment adherence behaviors) are measured simultaneously, it is mathematically and logically impossible to establish temporal precedence. Therefore, a cross-sectional study can never establish causality. It can only identify correlations or associations. For instance, if the study finds an association between older age and a higher likelihood of changing treatment based on social media, it cannot prove that older age causes this behavior; it only demonstrates that the two variables co-occur within the observed sample. Full-text verification is required to ensure the authors do not overstate these associations as causal relationships in their full discussion. ### 3. Demographic Profile Analysis The research was conducted at a tertiary outpatient clinic in Monterrey, Mexico. A tertiary clinic typically handles specialized, complex cases referred from primary or secondary care providers. This setting inherently introduces selection bias; the patients in this study may have more severe disease, longer disease duration, or different healthcare-seeking behaviors than rheumatology patients in the general community. The sample size is reported as ninety-five (95) adult patients. In the realm of quantitative research, a sample size of 95 is generally considered small, which limits the statistical power of the study and increases the margin of error. The median age of the cohort is 50 years, indicating that half the participants were older than 50 and half were younger. Crucially, the abstract reports that 92.6% of the participants were female. While many rheumatic diseases (such as Rheumatoid Arthritis or Systemic Lupus Erythematosus) disproportionately affect women, a sample that is nearly 93% female severely restricts the generalizability of the findings to male patients. The attitudes, digital literacy levels, and information-seeking behaviors reported here are overwhelmingly those of middle-aged women in a specific Mexican tertiary care setting. Any extrapolation of these findings to other demographics, regions, or clinical settings must be done with extreme caution. ### 4. Instrumentation and Measurement Tools The researchers utilized a structured questionnaire comprising three specific instruments: 1. **eHealth Literacy Scale (eHEALS):** Generally designed to measure consumers' combined knowledge, comfort, and perceived skills at finding, evaluating, and applying electronic health information to health problems. 2. **Social Media Engagement Questionnaire (SMEQ):** Typically used to quantify the frequency and intensity of a user's interaction with social media platforms. 3. **AI Attitude Scale (AIAS-4):** Presumably a tool to gauge perceptions, trust, and willingness to utilize artificial intelligence. It is vital to recognize that all three of these tools rely on self-reported data. Self-reported questionnaires are highly susceptible to multiple forms of bias, including recall bias (participants misremembering their behaviors) and social desirability bias (participants answering in a way they believe is expected or viewed favorably by the researchers or their physicians). The abstract does not detail whether these specific instruments were validated for use in a Mexican Spanish-speaking population, nor does it provide the specific scoring thresholds used to define 'high' or 'low' literacy or engagement. Full-text verification is absolutely necessary to examine the psychometric properties of these translated or adapted scales. ### 5. Statistical Framework The abstract explicitly mentions the statistical tests utilized to compare the two age groups (defined as ≤45 years versus >45 years). * **Chi-square (χ2) test:** This is a non-parametric test used to determine if there is a significant association between two categorical variables (e.g., age group and a 'yes/no' response to changing treatment). * **Mann-Whitney U test:** This is a non-parametric test used to compare differences between two independent groups when the dependent variable is either ordinal or continuous, but not normally distributed. The use of this test suggests that the scores from the eHEALS, SMEQ, or AIAS-4 scales did not follow a normal bell-curve distribution, which is common in small sample sizes (N=95). The researchers set the threshold for statistical significance at p < 0.05. In statistical terms, a p-value of less than 0.05 indicates that there is a less than 5% probability that the observed differences between the groups occurred by random chance, assuming the null hypothesis is true. However, statistical significance does not equate to clinical significance. A finding can be statistically significant but have a negligible impact in a real-world clinical setting. ### 6. Detailed Breakdown of Reported Findings The abstract reports several key data points: * **High Digital Penetration:** Daily social media use was reported at 94.7%, and 30.5% of patients had used AI for disease-related questions. This indicates a highly connected patient population, though it is unclear if this is representative of the broader region or specific to this clinic's demographic. * **Age-Based Disparities in Literacy and Engagement:** Patients aged 45 or younger showed significantly higher electronic health literacy (p = 0.010) and engagement with social media (p < 0.001). The p-values indicate a strong statistical probability that these differences are real within this sample. It is logically consistent that younger populations, who are often digital natives, would score higher on these specific metrics. * **The Critical Finding on Treatment Alteration:** The abstract states, 'Importantly, younger patients were less likely to change their treatment based on SM content compared with older patients (p = 0.043).' This is perhaps the most clinically relevant finding reported. It suggests a paradox: while older patients (>45) have lower electronic health literacy and lower overall engagement with social media, they may be more susceptible to acting upon the health information (or misinformation) they do encounter on these platforms, to the point of altering their prescribed treatments. This highlights a potential vulnerability in older populations regarding the critical appraisal of digital health content. * **Attitudes Toward AI and Physicians:** Attitudes toward AI did not vary significantly by age. Furthermore, the study found high levels of trust in physicians regarding digital information: 84.2% believed doctors were open to discussing digital information, and 91.6% would use SM if recommended by their physicians. This suggests that despite the high independent use of digital tools, the physician remains a central and trusted figure in the patient's healthcare journey. ### 7. Limitations and the Requirement for Full-Text Verification The classification of this study as 'low evidence level' is accurate and appropriate based on the methodology described in the abstract. The limitations are substantial and must be explicitly acknowledged: 1. **Cross-Sectional Design:** As previously detailed, this design precludes any determination of cause and effect. 2. **Small Sample Size:** An N of 95 limits the statistical power and increases the risk of Type II errors (failing to detect a true difference) or overestimating the effect size of the differences found. 3. **Single-Center Bias:** Being conducted at a single tertiary clinic in Monterrey, Mexico, means the findings may be heavily influenced by the specific culture, socioeconomic status, and healthcare access of that specific locale. 4. **Gender Skew:** The 92.6% female demographic means the results cannot be reliably applied to male patients with rheumatic diseases. 5. **Self-Reporting Bias:** The reliance on questionnaires means the data reflects what patients say they do, which may differ from their actual behaviors. Full-text verification is required to ascertain the specific clinical diagnoses of the patients (e.g., Rheumatoid Arthritis vs. Osteoarthritis), the exact wording and validation of the questionnaires, the raw data scores, and the authors' full discussion of their study's limitations. ### 8. Conclusion The abstract concludes that digital health literacy significantly influences how patients process online information and manage treatment adherence. While the reported data supports an association between age, digital literacy, and self-reported likelihood to alter treatment based on social media within this specific sample, the broader conclusion must be viewed through the lens of the study's severe methodological limitations. This abstract provides a valuable preliminary observation regarding digital information-seeking behaviors in a specific rheumatology cohort, but it does not provide definitive, generalizable evidence. No clinical actions, treatment modifications, or policy implementations should be derived from this abstract without rigorous verification of the full peer-reviewed text and corroboration by higher-level, multi-center, longitudinal studies.Health Implications
This abstract outlines an observational study exploring the correlation between digital health literacy, social media (SM) use, and self-reported treatment adherence among 95 patients at a single clinic in Mexico. It establishes that within this specific, predominantly female (92.6%) cohort, daily SM use was highly prevalent. The statistical analysis indicates an association where patients under 45 demonstrated higher electronic health literacy and were less likely to report altering their prescribed treatments based on SM content compared to patients over 45. The study also highlights a high degree of reported patient trust in physicians regarding digital health discussions. Critically, this abstract does not establish causality. Because it is a cross-sectional study, it cannot prove that lower digital literacy causes older patients to change their treatments, only that the two factors co-occurred in this small sample. It does not evaluate the efficacy, safety, or outcomes of any specific medical treatment, nor does it establish how these findings apply to different demographic groups, geographic locations, or male patients. Full-text verification is required to assess the validity of the questionnaires used. No practical clinical actions or changes to patient care should be initiated based solely on this abstract.
Key Findings
- The study included 95 adult patients with rheumatic diseases, of which 92.6% were female, with a median age of 50.
- Participants were evaluated using the eHealth Literacy Scale, the Social Media Engagement Questionnaire, and the AI Attitude Scale.