Medicine 2.0 focuses strongly on the use of Web 2.0 tools. However, as a term only four years old and constantly evolving as new tools emerge, academic literature is unlikely to have achieved consensus on its scope as quickly as 2008. For this reason, we used Google’s PageRank system to identify the tools or benefits most important to Web 2.0. Google’s PageRank relies on the democratic nature of the Web’s vast link structure to indicate an individual page’s value. Google interprets a link from page A to B as a vote by page A for page B. Google looks at more than the sheer volume of votes; if the page that casts the vote also has many links to it, this vote cast by that page weighs more heavily [11].

To refine the approach, and to enable a contrast to Medicine 2.0’s salient themes, we started with Web 2.0. We searched with Google for “Web 2.0” to identify the most linked pages with the term. These pages were coded using thematic analysis [12] to identify the terms describing both tool types and the purpose or benefits of Web 2.0. Articles coded included not only the Wikipedia entry and O’Reilly’s definitions, but also a series of blog threads, including over 50 contributions from users attempting to define Web 2.0. The full results of this approach are available in a separate publication [13], but we adapted the approach for this paper and used the result, which summarizes Web 2.0 as:

Democratized Collaborations; a collaboration enabled by web technology that promotes learning and innovation. Democratized collaborations work by connecting participants to harness network effects and knowledge in an open and interactive manner.

A similar approach was used to delineate Medicine 2.0, but no assumption was made that identifying terms such as Medicine 2.0 should take precedence over Physician 2.0 as the researchers had a priori knowledge that numerous terms are associated with the field. Hence, in the first two steps in the method below, we try to determine the identifying terms that describe the field. Step 3 identifies the most popular online discussions relevant to the field, and, finally, steps 4-6 use the comparative method for thematic analysis as described by Techniques and Procedures for Developing Grounded Theory [12] to understand the salient themes. Steps 3-6 were effectively completed three times in order to obtain intercoder reliability of 82% agreement for exact phrases across all of the pages analyzed.

In a second phase, carried out in step 7, the original sample of 2405 academic papers identified as being potentially related to Web 2.0 and health was reduced to 56 papers after excluding those not directly addressing Medicine 2.0, duplicate search results, or papers not available in English. The initial number of papers and those selected for the review are shown in brackets in step 1 of the methodology outlined in Table 1. 

The abstracts and titles of the 2405 papers indicated that “2.0” was associated with Health 2.0, Medicine 2.0, Physician 2.0, Nursing Education 2.0, Medical Librarian 2.0, and Physician Learning 2.0.

Table 2 shows how often the terms used by academics are replicated in the community itself (via Google search). The results show that “Health 2.0” or “Health” and “Web 2.0” are the most commonly discussed terms. The prominence of Health 2.0 and Medicine 2.0 meant only these terms were examined for more precise definitions as detailed by steps 3-6 in the Methods.

For both terms, open coding of the top online descriptions quickly lead to saturation, in the case of Medicine 2.0, after seven articles (articles coded: [4,21-25]) and after nine for Health 2.0 (articles coded: [26-34]). In the early axial and selective coding stages, four core terms were identified: participants or actors, tools, methods, and purpose or objectives. The salient themes or grouping applied to both identifying terms, and there was almost no difference with the ranking (in terms of counts) of exact phrases associated with these themes. Overall, there were few differences between Health 2.0 and Medicine 2.0 in terms of participants, and Table 3 and Table 4 show the individual counts for each term. It is worth noting than one exact phrase, “Privacy,” was identified by both researchers but was not possible to rank using search engines. Different search engines provided widely different rankings for this term (from first to last within the methods and tools grouping), which we believe reflected the fact that some search engines perform key word searchers through the footers of cached pages (see Discussion).

Given the similar definitions of Health 2.0 and Medicine 2.0, and as suggested by other authors to encapsulate research [35,36], we decided to use the term Medicine 2.0. However, choosing either term would not have highly impacted the results of the literature review. The ranking of the terms and the context of use in the pages that we analyzed suggested the following definition for Medicine 2.0:

Supporting this are five salient or structuring themes that we more accurately define as follows:

Can a clear definition of Medicine 2.0 be established across practitioner and academic literature that distinguishes this field from eHealth? Examining this question, we found common salient themes for both Health 2.0 and Medicine 2.0 that describe Web 2.0’s application to health. Its application to health and medicine is not as straightforward as the rule “Medicine 2.0 = Medicine + Web 2.0,” particularly in its emphasis on personalized health care and its participants (not observed in the Web 2.0’s democratized collaborations [13]). In addition, while we did not complete a systematic review of eHealth, previous publications have shown that the field emphasizes the “communicative foundations of eHealth and specif[ies] the use of networked digital technologies, primarily the Internet...for all stakeholder groups” [37]. As such, neither the stakeholders nor the principal tool used (the Internet) distinguishes Medicine 2.0 from eHealth. However, the principles of open source, generation of content by users, the power of networks, personalized health care, and the focus on collaboration across all stakeholders are not always highlighted by eHealth and suggest that these fields have different emphasis.

In addition, earlier in this paper we highlighted the issue that the technology based view of Medicine 2.0 (ie, the use of Web 2.0–like tools) could not clearly distinguish eHealth from Medicine 2.0. For example, we could conclude that every Internet health search using Google becomes a Medicine 2.0 search as the search algorithm is based on user-generated links. However, our definition implies that this cannot be taken for granted as Google does not meet many criteria of the Medicine 2.0 definition. First, it is not open; users do not have transparency on the algorithm or the ability to change it. Second, users do not have an intention to collaborate using Google or to help Google when assigning a link within a page. Rather, Google has commercialized a feature of Internet collaboration for its search and has not created a Medicine 2.0 collaborative platform. Despite this, other authors have argued that Google is the quintessential Web 2.0 company [38] and its use of network effects and user generated content will mean it will probably remain across the eHealth and Medicine 2.0 gray boundary.

Applying this definition to the original set of articles identified via key word searches on health and Web 2.0, we found that fewer papers were associated with the field. One main driver was the fact that the search terms (eg, “Medicine 2.0”) often found identified papers that had no relevance to the subject, though we did not bottom at the root cause of this effect. Others were relevant to eHealth in general, but not Medicine 2.0. For example, the study “Influences, usage, and outcomes of Internet health information searching: multivariate results from the Pew surveys” by Rice [39] provides detailed analysis on the use of the Internet in relation to health, but it does not address Medicine 2.0 issues specifically. A few papers ended on a similar gray boundary to that demonstrated by Google, such as Tse et al [40], and these were excluded from our review. This does not mean their findings are not relevant, but rather we found that the overall paper was not specific to Medicine 2.0 and should consequently be treated as an eHealth paper with potential implications for Medicine 2.0. 

Finally, we noted that rapid saturation in coding was achieved to obtain the salient themes used online. And while we believe this reflects a certain amount of common language used by the Health 2.0 or Medicine 2.0 online community, this does not mean all relevant themes were identified. For instance, social networking is only encapsulated in the “power of networks,” even though some authors would identify this as a very important separate trend and term. As such, this definition only identifies core or salient themes, not excluding other concepts, as being part of Medicine 2.0. We believe any compact definition will have difficulty in precisely delineating its complete scope.

Is there agreement between online discussions and academic communities in their use of the term Medicine 2.0? If not, what does such divergence imply for future research? Regarding research question 2, we identified 56 articles in the research literature that covered four of the five major themes underpinning the Medicine 2.0 definition. In reviewing the literature and comparing it to prominent online themes, we found limited research into personalized health but did find extensive literature on methods and tools. We also noted potential overlaps with a separate body of research into open source health and a general call by authors for further research in specific areas.

Despite the fact that personalized or customized health is a key objective or benefit of Medicine 2.0 (based on online discussions such as those typified by [41]), no academic publications were found that focused extensively on this theme. Specifically, we believe researchers may need to look at how personalized online health care can evolve, such as the trade-offs between an active global site (with rich, regular but fairly uncustomized updates) versus a local site with very specific information to a regional context (but with less contributors and, hence, the risk of inaccuracy or less information).

By contrast, research responding to the tools and methods is the most extensive. In this theme, papers looked at the implications of a particular tool or method, such as the errors in user-generated content or the implications of open-source methods. For example, Deshpande and Jadad [42] offer an overview of the methods or drivers of Medicine 2.0, providing some support for our identified themes and definition. In addition to information inaccuracy and privacy, open-source methods have been widely studied within this theme relating to medical research. Examples can be categorized into two types: those that address the issues and benefits of a common license for the output of research (eg, [10,43]), and those that look at open-source methods to develop information technology tools for medical research (eg, [44,45]). However, open-source health is not subsumed by Medicine 2.0. The extensive literature on open source, such as the 3864 articles in PubMed as of February 2008, covers topics outside Medicine 2.0. For example, Hope [10] explores technology licensing not connected with Internet use, as opposed to Yang et al [46], who do consider a Medicine 2.0 open-source collaboration. As such, while Medicine 2.0 relies on open-source methods in health, and the topics overlap in areas, we believe care should be taken to view them as distinct research topics. 

Overall, there is a call for research in many areas, and Potts [47] suggests that researchers are significantly behind trends in eHealth and, more specifically, in Medicine 2.0 tools such as the use of blogs and wikis. Potts argues that extensive research is required to close this gap, which is supported by other researchers’ calls for more evidence to understand best practice models in using Medicine 2.0 for medical education and practice [48-51]. 

In addition to this call for research, we would expect publications on Medicine 2.0 to continue to grow in this theme for two further reasons. First, Web 2.0 tools are constantly evolving, and hence the impact of new tools will continue to require assessment. Second, two major tensions or research discussions exist that will also require investigation: information inaccuracy, and information privacy and ownership. We return to these tensions in the discussion on research question 3, and detail them in Table 5 along with how papers responded to salient themes. Note that some papers investigate Medicine 2.0’s impact for various stakeholders (indicated by “various” in Table 5), while others either touch on multiple themes or are difficult to classify (indicated by “over-arching or unclassified” in Table 5).

What are the major tensions between the main stakeholders in Medicine 2.0 communities as identified by research? In relation to research question 3, four key areas of debate or tension between stakeholders were identified by our literature review:

The first main area of debate, an overarching theme, is related to the lack of agreement on what Web 2.0 is, and if it really exists [2]. Studies have generated justification for the study of Web 2.0 by the sheer size of its participants and the number of people who recognize it as a concept [92]. This debate has trickled into the Medicine 2.0 domain in discussions by people such as Skiba [52] and is continued by speculation that terms such as Health 2.0 may be a fake “gold rush” [55]. The situation is complicated further by authors introducing Web 3.0 for medicine, speculating that some Web 2.0 tools such as social bookmarking will become redundant [56]. However, we believe people will continue to use Medicine 2.0 tools, and some researchers have argued that Medicine 2.0 and Health 2.0 may evolve into terms with relevance for different audiences, such as Medicine 2.0 as an academic and international focus, versus a business or consumer audience for Health 2.0 [35,36]. Our results neither confirm nor reject this hypothesis, but they do provide support for the idea that the terms currently have a high degree of overlap and that both are more complex than simply applying Web 2.0 to a health care context.

The second main debate surrounds collaboration and practice by doctors and patients. Separate to the issues of information inaccuracy, it encompasses resistance by some doctors to their patients’ use of Medicine 2.0. Their concerns arise from Medicine 2.0 causing unwanted behaviors in patients, such as not consulting a physician, consulting a physician too late, or coming to wrong conclusions about their disease management even if the information available to them online is accurate. The issue is not new and arose with eHealth. Ferguson [54] calls these doctors “e-Patient resistant clinicians” and suggests a sense of loss of control (and risk of being sued), paternalism, or lack of training driving these doctors’ behaviors. However, the issue is distinct in Medicine 2.0, where amplifying effects to this behavior are identified by certain authors, such as lack of training for doctors [5] or the difficulty of advising patients on use of Medicine 2.0 tools [81]. Overall, authors claim that doctors will need to recognize the emergence of Medicine 2.0 and that current training may not be sufficient to do so.

The third main discussion, based on the methods used to generate Medicine 2.0 information, is the risk of inaccurate online information. Misinformation has long been identified as a hazard of eHealth. However, studies have found little support for this concern [93]. These studies pre-date the rapid expansion in Medicine 2.0 use. Looking more closely at Medicine 2.0–specific information, Esquivel [65] notes the error and correction rate on an Internet-based cancer support group. The study found that most information was accurate and most false or misleading statements were rapidly corrected. Eysenbach [74] also examined the impact of information accuracy and credibility in relation to eHealth and noted that that patients will tend to use both intermediated (experts, authorities) and distributed (ie, Medicine 2.0) information to make their health decisions, thereby reducing any risk from inaccurate online information generated by users. In addition, apomediaries or gatekeepers acting at the network or group level work as collaborative filtering processes for distributed information that help users navigate through the onslaught of information afforded by networked digital media, reducing information risk further [35,36]. However, despite this early evidence of low risk, many practitioners and researchers remain to be convinced. This is demonstrated by responses to articles on Medicine 2.0’s potential, such as “the consequences could be disastrous for any inexperienced trainee following the advice” [4], or the need for authors to post a clarification after suggesting that Google could be used as a diagnostic tool [38].

The fourth and final debate is related to the consequences of the methods used to generate Medicine 2.0 information. Authors note that in addition to accuracy of information, privacy, ethical, legal, and ownership issues are also critical due to the nature of health information [49,69]. This applies not only to patients but to doctors who may use social networking sites for medical education and debate [91]. They suggest that potential models of identity management and authorization schemes should also be investigated in the context of Medicine 2.0 research. Once again, this tends to accentuate eHealth trends such as noted by Ferguson [54], who also highlights that those patient groups who run specific sites claim ownership over this data and are increasingly using it to influence the research agenda. Overall, new sources of health information are emerging via these methods, which will impact not only doctors who carry out research but could have potential implications for scientists working in the wider industry, such as pharmaceutical companies.

Our study has several limitations that warrant attention. Clear risks arise from using Google and other search engines to define Medicine 2.0. First, in step 3 of the method, Google’s PageRank system may only identify popular self-referencing communities, which as noted by some researchers has bias against newer online content [16]. Second, search algorithms are rarely published, and hence we cannot be sure of the consistency of the counts, which has been subject to criticism at low levels for both Google and MSN searches [19,20].   

We mitigated the first risk via the iterative manner in which the definitions and themes were identified by comparing academic and online definitions for inconsistencies and by searching for theme rankings across all pages online to reduce the bias toward the popular pages. We did not find any major inconsistencies, even though the small differences in Medicine 2.0’s scope online and in academic publications were established. Examples include the online focus on personalized health and the lack of online focus on social networking, which has been identified as an important trend by other authors [35,36]. Other exact phrases that we anticipated but did not see included “semantic Web.” To mitigate their potential omission, we determined their ranking anyway, but due to lower rankings, they would not have emerged in the defining text of top salient themes used in our paper. This does not indicate that these are not very important themes, but rather that they are covered in more general concepts such as the power of networks.

We also examined the specific criticisms, such as Google returning inconsistent results below 1000 counts or Microsoft Live Search being inconsistent below 8000 counts and hence only ranked exact phrases above these levels. However, the use of different search engines further emphasizes that only the top exact phrases (eg, Blog or Podcast rather than Mashup) can be used with confidence to identify the salient themes as there was good agreement between search engines. Hence, we mitigated this risk by only using the top two to three and commonly ranked phrases, avoiding the bias that a term has been ranked highly only due to a particular search engine’s internal mechanisms.

Following the updated definition of Medicine 2.0, the literature describes five major themes: (1) the participants involved; (2) the impact on different collaborations and practice; (3) the ability to provide personalized health care; (4) the use in medical education; (5) its associated methods and tools.

There is now an emerging body of research into Medicine 2.0; in addition to the 56 papers we identified that address it directly, there are also many eHealth papers that have indirect implications for Medicine 2.0. Overall, they suggest that Medicine 2.0 will have a large impact on all areas of medical practice. Most of these publications are recent, since 2004, and call for more empirical research on various topics.

We expect research to continue to focus on the four major tensions between stakeholders that were found in the literature: the scope of the field including its definition and existence, the patient-doctor relationships impacted by Medicine 2.0, the methods and tools relating to information accuracy, and the methods and tools related to ownership and privacy. These issues are also found in eHealth; however, Medicine 2.0 is accentuating their impact. While touched on by some overarching publications, the lack of research into personalized health does not indicate that its importance is overstated by online discussions. Rather, we concur with other researchers who suggest that research currently lags behind practice in understanding the implications of Medicine 2.0.

None declared.