JMIR Publications

ABSTRACT

Background:

Telemedicine draws on information technologies in order to enable the delivery of clinical healthcare from distance [1,2]. Telemedicine has been utilised around the world and a recent World Health Organization survey found that 38% of the countries had some kind of telemedicine system and 30% had agencies that managed telemedicine services [1]. Telemedicine is particularly attractive in rural health areas as well as across long distances where it can be difficult to reach patients [2]. There have been positive and unsuccessful implementations of telemedicine around the world [3]. This has likely led members of the public to have views and opinion towards the technology. It is important to gain an understanding of perceptions towards telemedicine before implementing an order to ensure it is received positively. Twitter is a social networking platform which has 316 million month active users with 500 million tweets per day and its potential for real-time monitoring public health has been well documented [4]. Twitter has been used previously as a platform to disseminate guidelines and perform polls by the European Association of Urology [4,5]. However, there is a lack of empirical research which has critically examined the potential of Twitter polls for providing insight into public health. Twitter is a platform with a demographic which is college educated and where users are likely to be aware of new forms of technology. This makes examining opinions towards telemedicine an interesting case. One of the benefits of utilising Twitter polls is that it is possible to gain access to a large audience which can provide instant and real-time feedback. Moreover, Twitter polls are completely anonymised and it is not possible to learn the identify of a user completing a poll nor is it possible for one user to vote on more than one occasion. Our results will be of interest to health authorities, policy makers, and academics interested in tele-health. They are also likely to be of interest to health authorities around the world seeking low-cost real-time survey methods as well as researchers interested in a critical examination of Twitter polls.

Objective:

The overall objective of this study was to better understand opinions related to telemedicine on Twitter and to assess the potential of Twitter polls to validate and test survey questions.

Methods:

In this study, we devised two Twitter polls using questions from previous questionnaires to explore acceptance of telemedicine among Twitter users. We distributed the polls on one of the author’s Twitter timeline and asked the followers of the account to answer the poll and retweet it to reach a larger audience. The Twitter handle, which we used for this study “@jvalaball”, has more than 9 300 followers. Ethical approval was not required because Twitter polls are completely anonymised.

Results:

Our first Twitter poll was distributed in May 2016 and was ‘pinned’ on the top of the Twitter timeline used in the project for 7 days. By pinning a tweet it will permanently place it on the top of a Twitter users account such that any new visitor will see the tweet appear at the top of a user’s timeline. We used a question from the Telemedicine satisfaction questionnaire, a validated questionnaire developed by Yip et al. in 2002 [7]. The question posted is as followed: “I find telemedicine an acceptable way to receive health care services. Do you agree?” For the responses, only two answers were allowed which were, “yes” or “no”. Figure 1 below displays how the tweet was constructed as well as the responses retrieved by Twitter users. The poll was retweeted 51 times and had 6.698 impressions. It received a total of 108 votes, 90% of which were positive and 10% negative. The second poll was posted during November 2017, which was also pinned in the top of the timeline for 7 days. For this Twitter poll we used a question from the Physician questionnaire in the EU project Health Optimum [8] . The question posted as followed: “How do you rate the quality of care delivered by telemedicine when compared to the quality of traditional care?” Four answers were allowed: which were: “better”, “about the same”, “not as good” and “not sure”. The poll was retweeted 49 times and had 4 364 impressions. Figure 2 below provides insight into how the tweet was constructed as well as the responses retrieved by Twitter users. Overall, the poll received a total of 113 votes. 39% of the respondents stated that they rated the quality of care delivered by telemedicine not as good as traditional care, 19% found the quality of care about the same, 22% rated the quality of care as better and 20% were not sure about the level of care.

Conclusions:

Our study provided an overview of a novel experiment of using targeted Twitter polls to assess acceptance of telemedicine amongst Twitter users. We argue that this tool could be used to quickly perform surveys to assess the opinion of users regarding acceptance of telemedicine in order to obtain rapid feedback of new questionnaires before validating them. An advantage of using Twitter polls is that many can be created and disseminated in very little time as opposed to traditional questionnaire and surveys, which can become resource-heavy. One of the key benefits of utilising social media platforms such as Twitter is the very low cost when compared to traditional survey based methods. In certain departments with low-budgets Twitter could be used as a tool to gain initial public opinion feedback before a survey could be devised. The first poll showed an overwhelming support towards telemedicine as an acceptable way to receive health care services. In the second poll, which asked Twitter users to rate the quality of care delivered by telemedicine when compared to the quality of traditional care, the majority of users found that telemedicine was not as good as traditional care. This highlights how the design of a question can potentially influence the results of a survey. Our method could be used to conduct testing on survey questions and to compare the answers to ensure they are consistent. It must be noted that one of the limitations of using Twitter for gauging public opinion through the use of Twitter polls is that its users are not representative of the general population in terms of demographics [8]. However, a growing body of literature is noting the potential of social media data for providing unfiltered public opinion [9,10,11]. One of the potential strengthens of using social media data has been the ability to avoid the risk of interview bias [9]. Furthermore, due to the ability of social media to set agendas in mainstream media [12,13] it can be argued that it has become important to study content and public opinion held by social media users. Our study has demonstrated the potential of Twitter polls for gaining insight into public health topics such as telemedicine and for developing new light on how Twitter polls can be used to validate and test survey questions. Twitter polls could be utilised by health authorities to gain early real-time feedback on public views and opinion on a range of health issues not just limited to telemedicine. A key strength is the speed at which Twitter polls can be formulated as well as their low cost.