This is an open-access article distributed under the terms of the Creative Commons Attribution License (https://creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided the original work, first published in the Journal of Medical Internet Research, is properly cited. The complete bibliographic information, a link to the original publication on http://www.jmir.org/, as well as this copyright and license information must be included.
There is considerable discussion of risks to health professionals’ reputations and employment from personal social media use, though its impacts on professional credibility and the health professional-client relationship are unknown.
The aim of this study was to test the extent to which workday comments posted to health professionals’ personal Facebook profiles influence their credibility and affect the professional-client relationship.
In a controlled field study, participants (members of the public) reviewed randomly assigned mock Facebook profiles of health professionals. The 2×2×2 factorial design of mock profiles included gender (female/male), health profession (physician/veterinarian), and workday comment type (evident frustration/ambiguous). Participants then rated the profile owner’s credibility on a visual analog scale. An analysis of variance test compared ratings. Mediation analyses tested the importance of credibility ratings on participants’ willingness to become a client of the mock health professional.
Participants (N=357) rated health professionals whose personal Facebook profile showed a comment with evident frustration rather than an ambiguous workday comment as less credible (
This study provides the first evidence of the impact of health professionals’ personal online disclosures on credibility and the health relationship. Public perceptions about professionalism and credibility are integral to developing the evidence base for e-professionalism guidelines and encouraging best practices in social media use.
Professionalism is important to the health professional-client relationship (hereafter, the health relationship) [
An important issue is the need to strike a balance between health professionals’ right to use Facebook as private citizens and their duty to protect professional reputation and health relationships [
Normative beliefs about online or e-professionalism among some practicing health professionals and trainees differ considerably [
We explored the impact of workday comments that blur the line between health professionals’ private and professional lives as they are common and thought to have an impact on perceived professionalism [
The study was approved by the Research Ethics Board at the University of Guelph. Informed consent consisted of a welcome message and a letter of information on the landing page of a Web-based questionnaire. The sample was a convenience sample. Participants were recruited in person in communities across Southwestern Ontario or online, through local social media (ie, advertisements about the study on Facebook and Kijiji, a popular online classifieds site in Canada). Members of the public were approached in public spaces (eg, farmers’ markets, public trailheads, parks, soccer fields, and sports games) to take part in a study described as aiming to understand people’s first impressions of personality from online content. Participants who showed interest in participating were given a card with the researchers’ contact information and URL for the survey. The survey was open and voluntary. To control potential social desirability bias, no mention was made of health professionals [
The first of 3 survey sections presented a randomly assigned mock Facebook profile in PDF format for participants’ review (a mock profile is available to readers from the authors upon request). The second survey section presented an adapted credibility scale on which participants rated the mock profile owner [
Adapted credibility scale showing paired anchor characteristics for competence, caring, and trustworthiness (Cronbach alpha for the scale=.88). Two changes were made to the original validated scalea,b.
Competence | Caring | Trustworthiness |
Novice/Expert | Insensitive/Sensitive | Phoney/Genuine |
Unintelligent/Intelligent | Not understanding/Understanding | Unethical/Ethical |
Incompetent/Competent | Self-centered/Unselfish | Untrustworthy/Trustworthy |
Uninformed/Informed | Unconcerned/Concerned | Dishonest/Honest |
Dima/Bright | Uncaring/Caring | Immoral/Moral |
Untrained/Trained | Doesn’t care about me/Cares about meb | Dishonourable/Honourable |
a
bThis pair was removed from the scale owing to irrelevance to our study scenarios.
Profiles were made to simulate a real Facebook profile, although without working links and with names and headshot photos contributed by members of the research team with permission. Profiles were created using a 2×2×2 full factorial design, with the following attributes: mock profile owner gender (female/male), health discipline (physician/veterinarian), and workday comment type (evident frustration/ambiguous). Profile comments (ie, status updates) were modeled after actual status updates from previous studies of health professionals’ personal use of Facebook to authentically portray workday comments as they may be encountered online [
What is it with some people?? I know I only went through 9 years of university...but really, I know what I’m talking about...yeesh!!!
We also chose the following ambiguous workday comment:
Started with new electronic patient charts today...interesting experience for sure.
Each participant reviewed the content of 1 randomly assigned mock Facebook profile without time constraints. On the next page, they completed the 17 personality ratings, an adapted credibility scale which included 3 subscales: competence (6 items), caring (5 items), and trustworthiness (6 items;
Each item was represented on a visual analog scale, anchored at each end with an adjective describing less (eg, novice) or more (eg, expert) of the trait being assessed (eg, competence). Consistent with first impression research, participants were asked to move fairly quickly through all randomly presented rating scales, sliding the cursor to the place on the line between the 2 adjectives that they felt best represented the profile owner [
Numeric gradations for all visual analog scales (0 to 100) were hidden to access participants’ immediate impressions. Only one response was possible, and all items had a nonresponse option of nonmovement from the middle position on the scale (neutral). Respondents were able to review and change their answers before moving to the next page and could not move backward in the survey so as to ensure collection of their
Univariable linear regression identified unconditional associations between the outcome variable (credibility rating) and covariate or control variables including the following: participant is a Facebook profile owner (no; yes); recruitment (in person; online); participant age (years); participant gender (female; male); annual household income before taxes (low ≤Can $ 74,999; high ≥Can $ 75,000); and participant’s profile viewing and rating times (in seconds). Covariates having a
Sequential linear regression determined the extent to which impressions gleaned from the online content of a personal Facebook profile impact the health relationship [
The total number of people invited to participate in the survey is unknown. A total of 6 surveys were removed owing to insufficient responses, leaving 357 usable surveys. Not all participants answered all questions, thus samples vary as noted below. A missing value analysis revealed that values were missing completely at random [
The following variables were included in the omnibus ANOVA: workday comment type, mock profile owner discipline and gender, participant age and gender, participant profile viewing time, and recruitment venue (in person, online). The ANOVA results yielded only one statistically significant main effect—workday comment type,
Sequential regression supported mediation, meaning that the effect of workday comment type (evident frustration versus ambiguous) on participants’ willingness to become a client of health professional profile owners was mediated by credibility ratings (
Number (n) and proportion (%) of study participants within pretax annual household income and completed education categories.
Demographic characteristic | n (%) | |
<$30,000 | 50 (15) | |
$30,000-$49,999 | 42 (13) | |
$50,000-$74,999 | 75 (22) | |
$75,000-$99,999 | 71 (21) | |
≥$100,000 | 97 (29) | |
Completed or have some high school | 52 (14) | |
College diploma | 66 (18) | |
University degree | 118 (33) | |
Professional degree | 48 (13) | |
Graduate degree | 72 (22) |
Step 1 of the sequential regression examining credibility as a mediator in the relationship between workday comment type and willingness to become a patient of the health professional for 318 study participants (
Variable | Statistics | |
Standardized beta coefficient | ||
Participant age | −.03 | .51 |
Participant income (refa ≤Can $ 74,999) | −.08 | .17 |
Mock profile owner gender (ref female) | −.10 | .05 |
Health discipline (ref physician) | .65 | .21 |
Ambiguous workday comment (ref evident frustration) | .39 | <.001 |
Online recruitment (ref in person) | −.02 | .68 |
Scenario view time (in seconds) | −.08 | .12 |
Credibility rating | —b | — |
aRef: referent.
bVariable not included in the model in Step 1.
Step 2 of the sequential regression examining credibility as a mediator in the relationship between workday comment type and willingness to become a patient of a physician or client of a veterinarian for 318 study participants (
Variable | Statistics | |
Standardized beta coefficient | ||
Participant age | −.02 | .70 |
Participant income (refa ≤Can $74,999) | −.05 | .18 |
Mock profile owner gender (ref female) | −.09 | .02 |
Health discipline (ref physician) | .06 | .11 |
Ambiguous workday comment (ref evident frustration) | .10 | .03 |
Online recruitment (ref in person) | .06 | .10 |
Scenario view time (in seconds) | −.06 | .12 |
Credibility rating | .69 | <.001 |
aRef: referent.
Facebook functions as a space to share information and access social support [
We found a small, though not statistically significant, difference between participant credibility ratings of physicians and veterinarians. Given the different health contexts for physicians and veterinarians, the findings of this study suggest that the public has expectations of online professionalism that warrant further exploration across a range of health professions to broaden our understanding of credibility evaluations in relation to and beyond those studied here.
Our results seem consistent with a conventional view of professionalism and suggest that current cautions around health professionals’ personal use of Facebook are not unreasonable [
In another study, Clyde et al suggest that clients may be motivated to search for health professionals online to supplement information available on a workplace website [
Key strengths of this field study were the random presentation of mock profiles and the realistic context of the survey design, which parallels the way a health professional’s Facebook information may be discovered online. The use of indirect questions to assess judgments of health professionals [
Nevertheless, choosing a health care provider involves more factors than were measured in this study, such as location and word-of-mouth referrals. Facial gestures and physical traits also influence impressions [
In a recent exploration of dentists’ and patients’ attitudes toward social media use, 36% of patients reported searching for their dentist or doctor on social media. Other findings highlighted patients’ desire for more information about their health professionals (eg, by reading online reviews and qualifications) [
We conclude that health professionals, especially early-career professionals who have yet to build their professional reputation, should be mindful of their credibility with the public when using Facebook in their private lives. Adhering to social media guidelines that take a traditional or conventional view of professionalism is a reasonable first step, especially around workday comments. Although confidentiality breeches are more serious and have been a focus of attention for regulators and educators, they are rare [
Although a fuller understanding of e-professionalism is ongoing in this field, given the profound effect of digital technologies on society, social media guidelines should incorporate early evidence to prevent credibility damage because (1) trust, once broken, is difficult to rebuild [
Furthermore, an untapped area of health communication research includes the evaluation of health professionals use of social media skills in the public domain, that is, when, as private citizens, they participate in broader public online discussion about health matters within their professional domain (eg, posting comments in the comments section in news articles about vaccination).
analysis of variance
CW’s stipend was funded by the Canadian Institutes of Health Research under grant number 599547.
None declared.