As many as half of all patients with major traumatic injuries do not receive the recommended care [1-6]. Medical errors are common in critically ill trauma patients and mistakes occur frequently in emergency departments [6,7], where unconscious acts of omission and information overload [8,9] contribute to the problem. Indeed, emergency health care professionals must often make quick decisions, mostly based on intuitive reasoning [9]. Intuitive reasoning is fast, impulsive, effortless, and reflexive: while it serves the emergency health care professionals well, it is also prone to error. Reminders (eg, care protocols, order sets, and treatment algorithms) are knowledge tools that can improve intuitive decision making and help professionals implement best practices [9]. In particular, systematic reviews have indicated that computer reminders to health care professionals at the point of care can be effective at promoting best practices in a variety of clinical areas [10-19], including acute care, where they improve process-of-care outcomes [20]. Such reminders range from simple prescribing alerts to more sophisticated support for decision making. That said, different stakeholders have rejected many reminder systems on the grounds that they are slow, incompatible with work processes, unable to be adapted to local practices, difficult to access, or costly to implement [19]. A wiki, in contrast, stands as a low-cost means of permitting stakeholders from a single or many emergency departments to collaborate asynchronously on creating and updating reminders without duplicating their efforts or spending too much time.

A wiki is a webpage or a collection of webpages whose content can be modified by those who access it [21,22]. Wikis are being used to encourage and make it easier for researchers and clinicians working in medicine to share information and expertise [23-28]. Wikis can also help users adapt knowledge to local contexts within the knowledge-to-action cycle [19,29] and help patients and clinicians collaborate in developing patient support tools [30,31]. Studies have found that clinicians use reminders less than expected [32-35], but by giving clinicians an easy-to-use tool for creating, sharing, and updating reminders for their own and others’ use, wikis could reverse this trend. Wikis’ low cost makes them especially attractive. Clinical decision support systems that are not flexible enough to accommodate regular updating are very expensive [36,37]. For this reason, numerous authors have suggested exploring collaborative Web-based applications to share, create, and update clinical decision support content [19,36,38-40].

Although several wikis exist in health care, a wiki containing reminders to treat trauma patients does not exist yet. These wiki-based reminders for trauma would be created by a multidisciplinary group of clinicians interested in improving the quality of trauma care by implementing care protocols to help standardize trauma care. Using a wiki to share reminders could be a potentially powerful tool to allow a multidisciplinary group of health care professionals within the same hospital or situated in different hospitals to collaborate in the creation of high-quality, evidence-based reminders. The wiki would serve as a shared online repository available for health care professionals at the point of care.

Despite preliminary surveys showing health care professionals’ interest in wikis [28,41,42], very little is known about the determinants that influence professionals’ intention to use the information from wikis in clinical practice. To our knowledge, only one study [43] has explored the intention of physicians to use wikis to share medical information with other physicians. Although this study found that physicians were unsure about sharing medical information with other physicians, this behavior is different from using information from a wiki in clinical practice and has different determinants. Furthermore, trials exploring how eHealth applications—such as wikis—are used by patients have been plagued by high dropout rates, a phenomenon that the founding editor of this journal has termed the law of attrition [44]. No study has yet shown this phenomenon to apply to health professionals, but to minimize dropout in future trials exploring health professionals’ use of wiki-based reminders, we must design wikis that health professionals find helpful in caring for patients. To do this, we must understand what causes emergency physicians (EPs) and allied health professionals (AHPs) to use or to eschew the reminders. To this end, we plan to construct a questionnaire, based on the theory of planned behavior, that will ask health care professionals to quantify the influence of each determinant on their use of a wiki-based reminder in clinical practice. According to the theory of planned behavior, the constructs must be measured using a specific target, action, context, and time, as well as a specific population, to obtain results representative of that population [45]. The choice of trauma care for the setting of this study is linked to the investigators’ research interests and is only the first step in investigating the use of wiki-based reminders in other contexts and with different health care professionals. In particular, wiki-based reminder systems have the potential to support collaboration in clinical settings where multidisciplinary teams work together asynchronously (eg, group-based primary care). The results of this questionnaire will then help construct a theory-based intervention to increase the use of a wiki-based reminder by EPs and AHPs. This research project is the first step in that endeavor. Its goal is to solicit as complete as possible an array of beliefs concerning health professionals’ use of wiki-based reminders that promote best practices in trauma care, and to identify the most salient beliefs for inclusion in the questionnaire.

This study identified EPs’ and AHPs’ beliefs about using a wiki-based reminder to promote best practices in caring for patients with a severe traumatic brain injury. Based on the theory of planned behavior, we categorized these beliefs as behavioral, normative, and control beliefs. After analyzing the beliefs and ranking them in order of frequency of mention, we labeled the 75% most-reported beliefs as salient. We also labeled salient certain beliefs that were not among the 75% most reported. This post hoc decision was based on our knowledge of the literature (eg, “administration” as a normative belief), our experience in implementing care protocols for trauma (eg, “specialists” as a normative belief), or our fear of excluding important negative beliefs (eg, “adds stress” as a behavioral belief).

EPs and AHPs saw many of the same advantages to using a wiki-based reminder: namely, that a reminder gives access to evidence-based data, that it standardizes practices, and that it centralizes protocols. EPs and AHPs also shared similar normative beliefs about parties favorable to the use of a wiki-based reminder (nurses, physicians, respiratory therapists, the trauma team, and the younger generation), and both groups mentioned people less comfortable with computers as a negative referent. Many facilitating factors were common to EPs and AHPs: having a computer at the bedside, accessing information that was peer reviewed and of high scientific quality, ease of use, trialability, and an attractive visual design. The groups also had two obstacles in common: time constraints and undetermined legal responsibility.

Our two groups of respondents also differed in the perceived advantages to using a wiki-based reminder. While both saw the centralization of information and protocols as an advantage, only EPs saw the sharing of information as an advantage and only AHPs saw the promotion of teamwork as an advantage. Similarly, both groups saw easy access to a wiki-based reminder (eg, having a bedside computer) as a characteristic that would make using the reminder simple to use, but only EPs saw rapid access to protocols (“fewer than three clicks”) as a simplifying feature, and only AHPs saw having a workstation for every professional as such a feature. The apparent contradiction between the AHPs’ concern about having a secure website and the EPs’ desire to avoid passwords is worth exploring. Finally, AHPs felt it important to publicize and otherwise promote the wiki-based reminder to make it more visible (or in Rogers’s terms, which we explain below, “observable”). AHPs saw this as important to the innovation’s uptake, recognizing that the more people observe others using a wiki, the more likely they are to use it too.

We noted other differences. Significantly, AHPs named mandatory use as a dominant barrier. At the same time, AHPs often referred to EPs, quality-of-care promoters, and hospital administrators as important decision makers in the care of patients with a severe traumatic brain injury. Thus, if EPs, quality-of-care promoters, and hospital administrators make it mandatory to use a wiki-based reminder, the importance of this barrier might decrease. Research suggests that individuals are more likely to comply with referents’ expectations when the referents in question can reward or punish nonbehavior, as is often the case in a mandatory setting [72,73]. Future work will have to measure and compare the relative importance of these beliefs. These measurements will be used to determine whether implementation strategies should be adapted to different groups of professionals.

Many of the findings in our study confirm the findings of authors who have studied the adoption of other ICTs and of innovations in general. For instance, our participants reported compatibility with work process and trialability as important beliefs. In the diffusion of innovation theory, Rogers identifies these two characteristics in addition to three others—the innovation’s relative advantage, its complexity or simplicity, and its observability (the degree to which it is visible to users and potential users)—as influencing an individual’s decision to adopt or reject an innovation [74]. Authors besides Rogers have also associated an ICT’s lack of compatibility with work process and its poor trialability with the innovation’s unsuccessful implementation [32,39,75,76]. Because wiki-based reminders are designed to facilitate changes and edits by all users, they can be modified to fit different work processes [32,36,39]. Likewise, wikis’ free and open access could facilitate their trialability [41,42,71]. Hence, these seem like important advantages whose influence will need to be measured in future studies. Our participants reported the other three diffusion of innovation characteristics indirectly.

The findings of this survey are also consistent with the factors proposed by Davis [77] as determinants of the adoption of technology in his Technology Acceptance Model. In Davis’s model, the system’s perceived ease of use and its perceived usefulness were among the most frequently mentioned beliefs. Similarly, ease of use was the EPs’ top control belief and the AHPs’ fourth most frequently mentioned belief. As for usefulness, salient beliefs not yet mentioned in this discussion included that a reminder reduces errors (EPs), that it gives users rapid access to protocols (AHPs), that it gives users access to regularly updated protocols (AHPs), and that it is a new teaching tool (AHPs). Prospective measurement of the influence of these beliefs will be important, as high perceptions of system usefulness and ease of use have been associated with cases of ICT adoption [62].

Our study identified additional beliefs (aside from those similar to the beliefs identified by Rogers and by Davis) that were identical to the barriers and facilitators found in a recent systematic review of factors influencing health care professionals’ adoption of ICTs [62]. The beliefs in question were the presence and use of a champion, the participation of end users, and time constraints. Time constraints in particular have been identified as an important control belief in studies on ICT adoption [32,41,62,71,78] and in other contexts as well [79]. This is why we considered it salient, even though it ranked only 13th for EPs and sixth for AHPs. Other salient beliefs concerned the speed with which the user could access the reminder (the fifth-ranked control belief for EPs) and the rapidity with which the reminder gave the user access to protocols (the top-ranked behavioral belief for AHPs). Other control beliefs identified in our study were also similar to those identified in the systematic review mentioned above [62]: the quality of the visual design (salient for both groups), the absence of institutional control (salient for EPs, nonsalient for AHPs), the reminder’s reliance on peer-reviewed information of high scientific quality (salient for both groups), the mandatory use of the reminder (salient for AHPs, nonsalient for EPs), the lack of a reminder to use the wiki (nonsalient for EPs), having a computer at the bedside (salient for both groups), adding stress (salient for EPs), and use by people less comfortable with computers (salient for both groups).

We also found similarities to studies on the adoption of a computerized decision support system. In these studies, clinicians most wanted such a system to remind them of what they already intended to do [19,34,80,81]. Similarly, the EPs’ most frequently reported behavioral belief was that using a wiki-based reminder would refresh their memory. Centralized information (the EPs’ fifth-ranked salient behavioral belief) and access from different areas of the hospital have also been described in the literature as important factors in using a computerized decision support system [38,39]. Respondents also perceived these systems as improving patient care [80]: this is similar to the EPs’ belief that a wiki-based reminder would help reduce errors (sixth salient behavioral belief).

Our study of wikis also confirmed several barriers described in studies of health care professionals’ beliefs about using social media [32,36,39,41-43,78]: concern about the quality of information (EPs and AHPs), undetermined legal responsibilities [36,39,71] (EPs and AHPs), and lack of author identification [32,36] (EPs only). Measuring professionals’ perceptions of the importance of these barriers in our questionnaire will be essential to determine how these barriers might influence the use of a wiki-based reminder.

This overlap between study findings, notwithstanding our rigorous use of a theoretical framework, allowed us to identify new beliefs specific to our target population and related to the adoption of wiki-based reminders, beliefs that studies of the adoption of social media in health care had not identified. The importance of these new beliefs will also be important to measure in a future questionnaire. In terms of behavioral beliefs, both EPs and AHPs reported that using a wiki-based reminder could help standardize practices, promote teamwork (salient only for AHPs), and give users access to regularly updated protocols. EPs did not perceive this last factor as an advantage, but stated that having a system that was updated regularly would be a facilitating factor. Surprisingly, EPs also stated that frequently changing information would be a barrier. The importance of these apparently conflicting beliefs will be important to measure because reminders contained within a wiki could indeed change quite frequently if the literature changes frequently or if an edit war should occur. An edit war arises when a user repeatedly re-edits, undoes, or reverses a prior user’s edits in an attempt to keep visible his or her preferred version of a page [82].

We also identified many influential groups of people who would be favorable to health care professionals’ use of a wiki-based reminder, with EPs and AHPs naming each other as their main influence. So far, interpretations of the role of social influence on the adoption of ICTs have varied. Some authors have argued for the inclusion of normative beliefs (sources of social influence) in models of adoption and use [83,84], while others have excluded them [77]. Furthermore, some work has found that social influence is significant only under certain circumstances: in settings where ICT use is mandatory [71,85], among women in the early stages of their experience [86,87], and among older workers [88]. However this may be, we believe our study to be the first to have rigorously identified health care professionals’ salient normative beliefs concerning the use of any form of social media in health care. This is significant, because understanding the influence of normative beliefs on health professionals’ intentions to use social media such as wikis will be of the essence: social media are hypothesized to operate based on social networking, participation, collaboration, apomediation, and openness between peers [89]—all elements related to social influence.

Although age had been noted as a moderating factor in predicting the adoption of ICTs [87], past studies have not described the “younger generation” as an influential referent group, as AHPs and EPs did in our study. This finding is significant, since members of generation Y (people born between 1977 and 1997) will soon constitute a major part of the health care workforce and have been described as being comfortable with technology [90].

Some EPs working in the level I trauma center expressed another important normative belief. They suggested that clinicians in level II and III centers would be more likely to use a wiki-based reminder for the care of patients with a severe traumatic brain injury than would experienced clinicians working in level I centers, who would not need to refer to a reminder. This suggestion reflects the fact that many clinicians working in level I trauma centers view themselves as leaders and champions who help less-experienced clinicians better manage traumatic brain injuries. The influence of this belief must be measured quantitatively. If the experts do not intend to use wiki-based reminders themselves, future exploration must verify whether they intend to contribute their expertise to a wiki in order to help staff at level II and III centers improve their practices. It seems not unlikely that when recognized experts and strong leaders in a field add material to a wiki of evidence-based reminders promoting best practices, other health care professionals are motivated to take up the material in question. Further exploration of the intention to contribute medical information to a wiki will be particularly important because it seems that the level of sharing of medical information through social media is lower than expected for health care professionals [28,43,78].

In spite of its rigorous methods, our study has limitations. First, we did not perform member checking, even though member checking would have made our results more credible. However, two independent research professionals experienced with using the theory of planned behavior analyzed the contents of the interview transcripts rigorously to interpret our respondents’ beliefs as trustworthily as possible and resolved disagreements by referring scrupulously to the transcripts. Furthermore, to make it possible for readers to interpret the results for themselves, we have presented a sample transcript for each salient belief.

A second limitation is that our survey was conducted with a small group of EPs and AHPs who were recruited from a single region of the province of Quebec. Even though Godin and Kok [50] suggest that a sample of 25 participants is sufficient to elicit salient beliefs, we cannot assume that our results are transferable to all clinicians, especially in the case of AHPs, where no one professional group (nurses, respiratory therapists, or pharmacists) was sufficiently represented. Nonetheless, even if we surveyed only 6% (25/444) of all AHPs, the proportion of each professional group within our sample was very similar to the proportion of those groups within the total population of eligible AHPs in the three trauma centers from which we drew the sample. Finally, we made every attempt to elicit various beliefs by purposefully sampling participants from three levels of trauma centers, with varying levels of experience, and by including clinicians known to be reluctant to use computers and ICTs. We believe that this sampling method resulted in our recruiting AHPs who were more experienced than the norm. Even though it is difficult to predict the influence of this selection bias, we deliberately chose to have more experienced clinicians take part in this survey, as they are known to have more negative views about adopting new technologies [88]. In predicting the adoption of wiki reminders, measuring negative views is just as important as measuring positive ones.

Third, the video we produced to describe the behavior of study portrayed wikis positively and could have influenced interviewees’ beliefs positively as well. Using a mix of animation and actors, the 6-minute video presented wikis as a new and effective way of communicating and collaborating. We produced this video because we needed to demonstrate in a short time how a wiki-based reminder could be used to improve the care of trauma patients and because the use of a wiki-based reminder to promote best practices in trauma remains theoretical for most health care professionals. It also involves many smaller lead-in behaviors (eg, connecting to the Internet, reading the reminder, and applying its suggestions) that are hard to explain during an interview. This said, after realizing that most of the beliefs expressed by participants after viewing the video were positive, we compensated for this possible bias by including negative beliefs that did not meet our 75% most frequently mentioned criteria in our list of salient beliefs.

Finally, our behavior of study was only the use of a wiki-based reminder, not the creation and editing of the reminder. The last two behaviors are important to study in order to understand how a wiki can fulfill its potential to turn a traditionally passive consumer of knowledge into an active producer of knowledge (or prosumer, in Eysenbach’s terms) [89].

A major strength of this study was its rigorous application of the methods proposed by the authors of the theory of planned behavior to assess our target population’s perceived behavioral, normative, and control beliefs related to using a wiki-based reminder. These beliefs will be used to construct a theory-based intervention to increase the use of a wiki-based reminder by EPs and AHPs. Using this theory facilitated our comparison with similar studies and contributed to our elaboration of a theoretical basis for understanding the decision making leading to this behavior. It will also allow researchers to carry out a systematic review in this area. Furthermore, the steps taken to analyze the content of beliefs are rarely described in detail. Our detailed and rigorous description of the content analysis makes it possible for other researchers to reproduce this approach to exploring health care professionals’ salient beliefs about the use of other social media in health care.