The internet has become a prominent source of health information, with social media platforms playing an increasingly influential part in shaping public understanding of health and illness.1 As patients increasingly turn to these platforms for guidance, there has been growing concern regarding the variability in the accuracy, quality and accessibility of the information. Existing research2 indicates that while some content shared online is evidence based and beneficial for patients, health misinformation remains pervasive and, in many cases, can affect health behaviours. A recent systematic review3 highlights the complex relationship between social media use and health. For instance, some social media influencers have been shown to increase awareness of preventive health measures and encourage proactive health behaviours, while others disseminate misleading or harmful content, thereby contributing to confusion, mistrust and suboptimal clinical decision making.
In the Australian healthcare setting, general practitioners (GPs) have expressed longstanding concerns about the impact of online health information on patient expectations and behaviours, with GPs consistently ranking health misinformation and self-diagnosis via the internet as more pressing issues than climate change or health system pressures.4 Despite the proliferation of social media in health-related communication, the peer-reviewed literature to date has primarily examined the utility of the internet, including social media, as a tool for professional communication, either in enhancing doctor–patient interactions or fostering collaboration among healthcare professionals.5–7 There remains a notable gap regarding how patient engagement with social media–based health information influences the nature and quality of the clinical consultation. This is particularly relevant given evidence showing that patient exposure to social media health information can affect decision making directly, with patients presenting to general practice with preconceived ideas or concerns shaped by online sources.8–11
The aim of this study was therefore to assess whether GPs in Australia report that their consultations are being affected by use of social media health information. Specifically, GPs were asked to report how frequently they experience this in their practice, how the problem manifests (ie what patients are requesting on the basis of their social media health exposure) and how social media health information affects the doctor– patient relationship and health outcomes for patients themselves.
Methods
Study design
A cross-sectional online survey of Australian GPs was administered using the online survey software SurveyMonkey (SurveyMonkey Inc.). Ethics was approved by The University of Sydney Human Research Ethics Committee (ref: 2024/HE000524).
Participants and setting
Participants were GPs undertaking a free-of- charge webcast with HealthEd, an online Australian-based GP education provider.12 HealthEd webcasts are offered to a database of approximately 30,000 healthcare professionals, the majority of which are GPs. Following a webcast delivered on 25 June 2025, GPs were invited to participate in a survey unrelated to the webcast topic. The dataset was de-identified prior to analyses.
Measures
The survey consisted of six purposely designed Likert scale and multiple-choice questions on GPs’ experiences with patients who bring health information from social media into their consultations. Social media was defined as information from platforms such as Twitter/X (X.AI Corp.), Facebook (Meta Platforms, Inc.), YouTube (Alphabet Inc.), Instagram (Meta Platforms, Inc.), TikTok (ByteDance) and Snapchat (Snap Inc.). Instructions to the participants stated that this included health-related content or advice shared by influencers, as well as information from online support groups, companies and similar sources (Appendix 1; available online only). Two questions had an ‘Other’ free-text option. Questions were designed by a multidisciplinary team of researchers, clinicians and a clinician- trainee with expertise in general practice, evidence- based medicine, public health and social media, and were piloted with the research team to assess clarity, relevance and usability. Specifically, the survey questions were informed by the team’s experience and interests. Feedback from the pilot informed revisions to wording, response options and survey flow. Demographic variables collected included country of medical training, hours worked per week, gender, age, location (state), rurality and ownership of practice.
Analysis
To test representation, respondent characteristics (state, rurality [practice in metropolitan, rural or remote areas classified according to the Modified Monash Model13], gender, age and initial qualification location) were compared with those of Australian GPs presented in the National Health Workforce Dataset using Chi-squared (χ2) tests. Each question was reported descriptively, including the number of responses for each option of every question recorded and displayed. Analyses of the questions by age and rurality (whether the GP reported that they practised in metropolitan, rural or remote areas) were undertaken in R (R-Project). All missing data were removed. Free-text responses will be reported elsewhere.
Results
Of the 3490 participants (eg GPs, registrars, nurses and pharmacists) attending the HealthEd webcast, 2751 (78.8%) were GPs. We received 2220 completed surveys, of which 1889 (85.0%) were valid GP responses used in the final dataset. Responses were excluded for 331 (14.9%) ineligible respondents (eg not GPs, not practising in Australia or currently not working). The GP survey response rate was at least 68.7%, although the exact rate could not be determined because of the lack of work status data for the 799 participant non-responders.
Most GPs initially trained outside of Australia (51.2%), practised in metropolitan locations (72.9%) and worked >30-hour weeks (59.1%). Many were aged ≥65 years (36.1%). Participants were most commonly from NSW (34.6%). Male (47.5%) and female (52.5%) GPs were equally proportionate. Demographic and practice characteristics were reported and compared with national GP data sourced from the Australian Government.14 The distribution of GPs by state/territory closely corresponded to the nationwide data of GP numbers, as did the distribution of gender.14 However, there was an overrepresentation of metropolitan and older GPs when compared with nationwide data (Table 1).
Most GPs (n = 1607, 85.4%) reported that patients bring up social media information in their consultations. The largest proportion (n = 1113, 59.2%) reported this as at least once a week, with half (n = 570, 30.9%) of these GPs reporting patients bring up social media health information more than once a week. Social media requests stratified by hours worked by GPs are shown in Appendix 2 (available online only), Table A1.
Different types of requests (tests, diagnoses, medications, devices or alternative medicine) made by patients because of the information seen on social media are shown in Table 2 and stratified by GP hours worked in Appendix 2, Table A2. The most frequent social media– influenced intervention GPs noted were requests for specific medications, with 72.0% (n = 1343) of GPs reporting that they receive these requests ‘sometimes’ (once a week) to ‘very often’ (more than once a week). This was followed by requests to investigate specific diagnoses, being received ‘sometimes’ to ‘very often’ (56.5% of GPs, n = 1057). Table 3 shows that 60.1% (n = 1120) of GPs reported that these requests are ‘rarely’ or ‘never’ based on scientific evidence. Most (n = 1416, 76.0%) GPs ‘rarely’ or ‘never’ fulfil unscientific social media–based health requests, although 24.0% (n = 447) reported doing so at least once a week. Responses to questions about requests and scientific evidence stratified by GP hours worked are shown in Appendix 2, Table A3.
Nearly half of GPs (n = 860, 45.5%) reported that there was no change in their professional relationships with their patients due to the impact of patient requests based on social media health information (Table 4). The next most common responses were that patients were less likely to act on the medical advice GPs suggest (n = 330, 17.5%) and more likely to make requests based on social media health advice (n = 286, 15.1%).
The proportion of GPs who perceived that health information from social media is having a ‘somewhat negative’ impact on their patients’ health outcomes (n = 694, 37.6%) was similar to that of those who considered it to be ‘neither positive nor negative’ (n = 668, 36.2%). Around one-fifth (n = 376, 20.4%) considered the impact ‘somewhat positive’ or ‘very positive’ and 5.8% (n = 107) ‘very negative’.
Results by age and rurality of the GP are presented in the Appendix 2, Tables A4 and A5. There were differences by GP age group across most questions. Younger GPs (aged <45 years) were more likely than older GPs (aged ≥65 years) to report patients ‘often’ bringing up social media information during consultations (40.4% [95% confidence interval (CI): 34.0, 47.0] vs 24.6% [95% CI: 21.0, 28.0]) and to receive social media–influenced requests. They were also more likely to perceive negative impacts on the doctor–patient relationship, including loss of trust (13.9% [95% CI: 9.6, 19.0] vs 6.0% [95% CI: 4.4, 8.1]) and reduced patient adherence to advice (24.4% [95% CI: 19.0, 31.0] vs 12.9% [95% CI: 11.0, 16.0]). In contrast, comparisons by rurality showed minimal variation, with a significant difference only for test requests based on social media information, which were more frequent among metropolitan GPs (12.3% [95% CI: 11.0, 14.0]), compared with regional (8.4% [95% CI: 6.0, 12.0]) and remote (10.7% [95% CI: 5.0, 20.0]) GPs.
| Table 1. Characteristics of respondents compared with the Australian workforce |
| |
Number (%) of responses (n = 1889) |
Australian data (%) 2024 (n = 40,375)* |
|
Number (%) of responses (n = 1889) |
Australian data (%) 2024 (n = 40,375)* |
| State |
|
|
Age (years)B |
| ACT |
35 (1.9) |
1.9 |
0–39 |
88 (4.7) |
25.4 |
| NSW |
653 (34.6) |
31.9 |
40–54 |
510 (27.0) |
38.7 |
| NT |
18 (1.0) |
1.8 |
55–64 |
603 (31.9) |
20.3 |
| Qld |
330 (17.5) |
23.7 |
≥65 |
681 (36.1) |
15.6 |
| SA |
111 (5.9) |
7.6 |
Missing |
7 (0.4) |
|
| Tas |
36 (1.9) |
3.2 |
Initial training location |
| Vic |
564 (29.9) |
25.6 |
Australia |
904 (47.9) |
56.2B |
| WA |
142 (7.5) |
11.0 |
Outside Australia |
968 (51.2) |
43.8 |
| Rurality |
Missing |
17 (0.9) |
|
| Metropolitan |
1378 (72.9) |
67.4 |
Hours worked |
| Regional |
430 (22.8) |
29.5 |
≤12 |
104 (5.5) |
|
| Remote |
76 (4.0) |
3.1 |
13–20 |
283 (15.0) |
|
| Missing |
5 (0.3) |
|
21–30 |
386 (20.4) |
|
| GenderA |
31–40 |
711 (37.6) |
|
| Male |
881 (46.6) |
49.8 |
>40 |
405 (21.4) |
|
| Female |
972 (51.5) |
50.2 |
Practice ownershipB |
| Prefer not to say |
25 (1.3) |
N/A |
Yes |
444 (23.5) |
|
| Missing |
11 (0.6) |
|
No |
1424 (75.4) |
|
| |
|
|
Missing |
21 (1.1) |
|
AData collected from the Australian Government Department of Health, Disability and Ageing.13
BIncludes New Zealand. |
| Table 2. Options requested by patients because of the information they are seeing on social mediaA |
| Requests |
Never (less than once per month) |
Rarely (once or twice per month) |
Sometimes (once a week) |
Often (once or twice a week) |
Very often (more than once a week) |
Total number of responsesB |
| Tests (eg anti-Müllerian hormone, follicle- stimulating hormone, gut microbiome, thyroid function tests) |
428 (22.9) |
674 (36.0) |
555 (29.7) |
156 (8.3) |
57 (3.0) |
1870 (100) |
| DiagnosesC (eg endometriosis, perimenopause, adult attention deficit/hyperactivity disorder, anxiety disorders) |
293 (15.7) |
521 (27.8) |
664 (35.5) |
303 (16.2) |
90 (4.8) |
1871 (100) |
| Medication (eg semaglutide, medicinal cannabis) |
156 (8.4) |
367 (19.7) |
646 (34.6) |
508 (27.2) |
189 (10.1) |
1866 (100) |
| Devices (eg transcutaneous electrical nerve stimulation machines, glucose monitoring devices) |
675 (36.1) |
725 (38.8) |
374 (20.0) |
79 (4.2) |
17 (0.9) |
1870 (100) |
| Alternative approaches to medicine (eg chiropractic, naturopathy) |
341 (18.4) |
594 (32.0) |
580 (31.2) |
262 (14.1) |
80 (4.3) |
1857 (100) |
AData are presented as n (%).
BSome data missing for each option.
CPatients discussing the possibility of or seeking a diagnosis. |
| Table 3. Responses to questions about requests and scientific evidenceA |
| Question |
Never (less than once a month) |
Rarely (once
or twice a month) |
Sometimes (once a week) |
Often (once or twice
a week) |
Very often (more than twice a week) |
Total
responses |
| How often do you feel that patient requests based on information seen on social media are supported by scientific evidence |
286 (15.3) |
834 (44.7) |
537 (28.8) |
139 (7.5) |
69 (3.7) |
1865 (100) |
| How frequently do you fulfil a request based on information a patient has seen on social media that you believe lacks scientific evidence |
713 (38.3) |
703 (37.7) |
315 (16.9) |
77 (4.1) |
55 (3.0) |
1863 (100) |
| AData are presented as n (%). |
| Table 4. Impact of social media requests on doctor–patient relationship |
| Impact |
n (%) |
| No change to your professional relationship with the patient |
860 (45.5) |
| Patient becomes less likely to act on medical advice that you suggest |
330 (17.5) |
| Patient makes MORE requests based on social media advice |
286 (15.1) |
| Patient becomes more likely to act on medical advice that you suggest |
276 (14.6) |
| Gain in patient trust in your medical advice |
220 (11.6) |
| Patient makes FEWER requests based on social media advice |
209 (11.1) |
| Loss of patient trust in your medical advice |
188 (10.0) |
| Patient leaves your care |
39 (2.1) |
Discussion
These findings, from a large sample of practising GPs in Australia, indicate that patients are being influenced by health information they see on social media. Patients present with requests based on questionable social media health information, but, from the GPs’ perspective, GPs are managing these requests and maintaining their relationships with their patients. Importantly, findings point to younger GPs being more affected by social media–influenced requests.
GPs in this study reported frequently encountering requests influenced by social media health information, with most regarding the information as being non-evidence based. This is supported by previous studies showing that health and medical information on social media is frequently incorrect or misleading.2,15–18 Despite recognising the lack of scientific basis to their patient’s social media requests, some GPs still reported fulfilling these requests. This may lead to additional patient consultations or interventions as well as unnecessary use of financial and workforce resources – further straining an already stretched healthcare system – with arguably little net benefit to the patient.19
It has been shown that when doctors disregard information patients have seen on social media, or if their advice contradicts that from social media sites and communities, the doctor–patient relationship can become damaged or strained.20,21 Despite the challenges that social media information may bring to their practice, overall Australian GPs in this study noted that they are able to manage the pressures and maintain the doctor–patient relationship, with some further indicating that their patients become more likely to trust them. This is promising, as maintenance of the doctor– patient relationship through supportive communication in instances where there is a lack of evidence to support a request may lead to better health outcomes and person- centred care.22,23
Interestingly, rurality had little impact on GP responses, implying that the documented digital divide between rural, remote and metropolitan areas24 is not apparent with the issue of social media. Test requests, perhaps because of accessibility, were the only difference found by rurality. However, perhaps unsurprisingly, GP age tells a different story. Younger GPs reported seeing more social media health information in their consultations, and this is having a more negative impact on their doctor–patient relationships. It may be that older GPs tend to have older patient bases, with GPs aged over 65 years typically having twice the number of patients aged over 65 years than GPs aged 35 years or less,25 and elderly people tend to use social media less.26
This study has several strengths and limitations. Key among the strengths is the number of GPs from across Australia who completed the survey. When compared with other studies of GP opinions on similar topics in Australia,6,27,28 this study has a larger cohort, with a distribution largely analogous to national data. However, the study is limited by an overrepresentation of GPs aged ≥65 years in the sample who, as noted above, have an older patient base that tends to use social media much less than younger patients.26 The answers from this age group about the frequency with which social media health information is encountered during consultations are statistically significantly different from the under-represented <40-year age group. It is therefore likely that, on the whole, social media health information is more commonly encountered in GP consultations in Australia than our study has found. The experience of GPs in Australia may also differ from practitioners in other countries where social media regulations vary. Furthermore, while the survey was unrelated to the webcast or its topic, it may still have been subject to selection bias resulting from the characteristics of GPs who attend, potentially limiting the generalisability of the findings. This includes GPs who do not like to travel or are too busy to attend other educational or training initiatives to meet professional development requirements. Lastly, as response options were designed to capture a spectrum of perceptions, they offer only approximate representations and cannot precisely quantify the experiences of, or the impacts on, GPs, with responses likely to be influenced by recall bias.
While some published studies have investigated patient views of the effects of social media on their healthcare,10,11 or how GPs view social media as a communication tool,5,6 to our knowledge there have been few, if any, studies on how GPs feel about social media health information affecting their practice. This study fills that gap and provides foundation data for further research in this area. Specifically, this study highlights the need for communication approaches that validate patients’ concerns while guiding them toward evidence-based care. Supporting GPs with practical strategies, such as framing discussions around shared decision making,29 using clear explanations to counter misleading claims, and proactively addressing common online myths, may help strengthen the relationship and improve patient understanding in an environment where social media continues to shape health beliefs.30,31
In conclusion, social media health information is having a discernible impact on GP consultations in Australia. GPs report that patients are often presenting with health information sourced from social media that lacks scientific evidence and leads to requests for unnecessary and unproven investigations and medication. Despite this, current findings suggest that most GPs are able to manage the influence of such information within their clinical practice and therefore reduce the potential for adverse effects on health outcomes. Importantly, the doctor–patient relationship appears largely resilient, with the majority of surveyed GPs maintaining trust and open communication despite these challenges. However, as the social media landscape continues to evolve – particularly with the growth of direct-to-consumer health marketing32,33 and other trends in the modern health information space34 – there is an increasing need to support GPs to navigate misinformation while reinforcing their role as trusted, evidence-based gatekeepers within the healthcare system.