Cyberincivility: Nurses and Nursing Students on Twitter

Sunday, 22 July 2018: 4:05 PM

Jennie C. De Gagne, PhD, DNP, RN-BC, CNE, ANEF, FAAN
School of Nursing, Duke University, Durham, NC, USA
Jamie L. Conklin, MSLIS
Duke Medical Center Library & Archives, Duke University, Durham, NC, USA
Katherine M. Hall, MSN, RN, ONC
School of Nursing, Alderson Broaddus University, Philippi, WV, USA
Sandra S. Yamane, MSN, MS, RN
Duke University, Winston Salem, NC, USA
Sang Suk Kim, PhD
Red Cross College of Nursing, Chung-Ang University, Seoul, Korea, Republic of (South)

Background: A growing number of nurses and nursing students are using Twitter for various purposes, including searching for clinical resources, looking for job openings, and sharing or exchanging ideas (Kung, & Oh, 2014; Piscotty, Martindell, & Karim, 2016). Although such technology has the potential for positive professional development, when used inappropriately, it can degrade the working and learning environment, harm reputations, and negatively affect patient safety. As such, many of the major nursing associations worldwide have offered support and guidance for the use of social media (De Gagne, Yamane, Conklin, Chang, & Kang, 2017). Despite these guidelines and their widespread dissemination, research is uncovering misbehavior in cyberspace among healthcare professionals and students (Chretien, Tuck, Simon, Singh, & Kind, 2015; De Gagne, Choi, Ledbetter, Kang, & Clark, 2016).

Purpose: The aim of this study was to describe the characteristics of tweets posted by nurses and nursing students on Twitter and how they used Twitter with a focus on cyberincivility. The objectives were to (a) examine the prevalence of tweets manifested as disrespectful, insensitive or disruptive, and potentially harmful to patients and/or organizations and (b) to describe patterns and differences in types of uncivil tweets.

Method: A cross-sectional, twitter data-mining method was used to elicit both quantitative and qualitative information. The sample was from self-identified nurses and nursing students on Twitter worldwide. Only tweets in English were included in the sample. Using a desktop application that searched Twitter in bulk for tweets matching a set of predefined hashtags, we extracted 5,680 tweets posted in January 2017. After 207 duplicate tweets were removed, 5,473 were reviewed based on the following inclusion and exclusion criteria. The inclusion criteria were: nurse, nursing student to become a nurse, nurse practitioner, and nursing student to become a nurse practitioner. The exclusion criteria were: those who did not identify themselves as nurses or nursing students in their profiles, private accounts, deleted accounts, and ambiguous accounts. This manual data sorting process resulted in a list of 163 user accounts generating 8,934 tweets after 106 duplicate values had been removed. The framework of the Preferred Reporting Items for Systematic Reviews and Meta-Analysis statement (Moher et al., 2009) was used to guide the data collection procedure. The analysis of 8,934 tweets was performed by a combination of SAS 9.4 for descriptive and inferential statistics including logistic regression and NVivo 11 to derive descriptive patterns of unstructured textual data. The retrieved data were checked and validated in collaboration with the two consultants in the research group and also with a combination of data analysis methods. The institutional review board of a university deemed this study to be exempt.

Results: A total of 163 study samples were analyzed. These came from 143 nurses (87.7%) and 20 nursing students (12.3%), in 14 countries—74 from the United States (45.4%), 29 from the United Kingdom (17.8%), 15 from Canada (9.2%), 26 from unidentified 16.0%), and 19 from other 10 countries (11.7%). The study sample was primarily female (n=133, 81.6%), 16.6% male (n=27), and 1.8% gender unknown (n=3). These users had been tweeting an average of 4.7 years (SD=2.43), ranging from 79 days to 8.9 years. Up to the end of the data extraction period, an average of 6,501.6 (SD=10,457.0) tweets and retweets were issued, with a minimum of 12 tweets/retweets and a maximum of 57,252 tweets/retweets. The number of tweets the user “liked” in his/her account’s lifetime averaged 4,489.5 (SD=14,074.1). The mean number of followers each account had and the mean number of users each account followed were 1,024.1 (SD=3,719.8) and 877.1 (SD=2645.6), respectively. Of the 163 users, 60 (36.8%) had tweeted inappropriately at least once and issued 6.93 inappropriate tweets on average (SD=14.03), with a maximum of 97 inappropriate tweets and a median of 3 inappropriate tweets. Two (1.2%) users had inappropriate descriptions in their profiles. Findings from the logistic regression analysis revealed that Twitter users were more likely to issue inappropriate tweets if they tweeted more often (odds ratio [OR], 1.026; 95% confidence interval [CI], 1.008-1.045). How long the user had been tweeting, the total number of tweets liked, the number of followers the account had, and the number of users this account followed were not associated with the presence of inappropriate tweets.

The 60 users who tweeted inappropriately generated 413 such tweets over 6 weeks. Of those, 240 tweets were related to the nursing profession, whether from nursing students tweeting about their school life or by nurses about their professional work (employment, on-the-job duties, or other areas in the nursing profession). The other 173 inappropriate tweets related to personal life. Major categories of inappropriate content were profanity, product promotion, sexually explicit or suggestive content, and demeaning patients. Others included name-calling, rude comments, interprofessional aggression, alcohol, and drugs, HIPAA (Health Insurance Portability and Accountability Act) violations, racial and ethnic references, intraprofessional aggression, and risky behaviors or violence. Some tweets had multiple instances of incivility with overlapping content categories.

Discussions: Our findings are of importance considering the ethical and professional standards to which nurses take an oath to uphold. While neither the length of time that users had been using Twitter nor the influence of their followers affected the prevalence of inappropriate tweets, the incidence of inappropriate tweets did increase based on the number of posts tweeted by users. This raises questions regarding the effect of extended time on social networking sites (SNS) and the unique culture of cyberspace. Other questions arise regarding how the public image of nursing is affected by the content of these inappropriate tweets.

Conclusion: By socializing students to the organizational culture, faculty and mentors can convey the importance of cybercivility to ensure high standards in patient privacy and confidentiality, appropriate boundaries between the patient and the provider, academic honesty in the online learning environment, and responsible use of SNS. Findings of this study could inform decisions about the design of educational innovations and practice changes in the technology-mediated environment.