Twitter is one the fastest growing social media networks and the third most employed social media platform by hospitals in the United States. Pharmacists’ participation on Twitter is increasing, but little information is known about the content or the quality of their communication and interconnections. Therefore, researchers at NSU’s Center for Consumer Health Informatics Research have extended their earlier work on this topic with a deeper exploration of the professional and social use of Twitter by pharmacists, including initial social network analysis (SNA). This work is being presented at the 47th ASHP Midyear Clinical Meeting (Poster 5-257, Tuesday, 2:00-4:30 ).
A cross sectional, mixed method approach was used for this study. A list of self-identified pharmacists on Twitter was generated based on a keyword search of Twitter biographies. Keywords such as “PharmD”, “RPh”, “Pharmacist” were used to identify accounts by employing Followerwonk tools and other online services. Active US pharmacist accounts having tweeted in 2012 and greater than 100 English tweets were included. Student, technicians, and commercially focused accounts were excluded. One-third of the initially identified pharmacists were randomly selected for analysis (N=204). The most recent 30 tweets from each account were extracted using SEOTools. Investigators evaluated the content of extracted tweets in parallel and classified each account as: exclusively professional, predominantly professional, exclusively social, or predominantly social. Accounts containing tweets inconsistent with the Oath of a Pharmacist or Code of Ethics were also recorded. Cohen’s Kappa (k) was used to assess inter-rater classification concordance. Descriptive statistics were used to characterize results. Additionally, social network analysis (SNA) using NodeXL was performed to visually represent pharmacists’ internal cohesion and centrality on Twitter; phenomenological and narrative frameworks were used to determine emerging contextual themes within each category.
A total of 623 pharmacists were located on Twitter in this study, representing a considerable increase from the 102 pharmacists reported in a previous analysis from 2009 data. Pharmacist accounts identified in this study have generated 1.3 million tweets since March 2007 (~2.8/day). From the subset of 204 accounts, 57% (n=115) were classified as exclusively social, 33% (n=68) as predominantly social, 9% (n=18) as predominantly professional and 1% (n=3) as exclusively professional. Additionally, 26% (n=54) of accounts contained tweets of unprofessional nature. Notably, members of this category were highly centric and influential on other pharmacists during SNA. The observed inter-rater reliability demonstrated good agreement (k=0.67, 95% CI 0.59-0.75). Major themes that emerged from the qualitative analysis for the aggregated professional category (i.e., purely and predominately professional) were exchange of health information, job concerns, and insurance coverage. The primary theme that emerged from the aggregated social category (i.e., purely and predominately social) was entertainment.
Although only a small percentage of pharmacists use Twitter, their activity has grown tremendously since 2009. Current use of Twitter by pharmacists remains primarily socially driven, but results of this study indicate a presence of professionally driven exchange of health information as well. This may represent an opportune time for pharmacists to engage in the social media and healthcare dialogue. Reinforcement of online socio-ethical principles may be warranted.