The assumption underlying this approach to defining and classifying vMDTs is that it is not always possible for all the essential members of the team to be present in the same room; they are, in the jargon, not colocated. Table 1a and ​andbb defines and demonstrates the characteristics of conventional (face to face) and completely virtual teams.

An MDT that, for reasons of convenience or geography, chooses to use some form of teleconferencing is not a fully vMDT. The approach is not particularly novel and has been well described in the literature (Axford et al, 2002; Kunkler et al, 2007; Hazin and Qaddoumi, 2010). Asynchronous communication is one of the defining features of a vMDT and has been used in a series of projects (AFIP, iPath, UICC-TPCC, Virtual International Pathology Institute; http://www.diagnomx.eu/vipi/home.php) in diagnostic pathology. This approach, reviewed by Kayser et al (2011), was specific to the discipline of pathology. It provided expert opinion at a distance, but could not be considered multidisciplinary.

A completely vMDT should have all of the characteristics summarised in the second column of Table 1a. Hybrid forms of MDT, between the conventional face-to-face MDT and the fully vMDT, are possible but it is reasonable to regard only those MDT meetings that are asynchronous as truly virtual. This leads to the following definition of a vMDT:

A vMDT meeting involves participants who may, or may not, be part of a permanent team and who interact with each other non-simultaneously using shared clinical data. They may operate at a local or a national level and their remit is not necessarily confined to tumours presenting at a particular anatomical site.

This definition deliberately excludes any specification of the form of communication. Communication could include: text-based comments; images with annotations; short segments of audio or video, any or all of which could be uploaded onto a web-based system used to host the vMDT.

The cardinal principle should be that the technology is subservient to the needs of the team, and not vice versa. There are various ways in which technology can help people to work together. One way to classify the technology is to consider the information that can be transferred and then consider it within the context of a fully vMDT.

• Data-only systems, such as email and messaging services, allow the exchange of text and images. These systems are easy to implement for a vMDT but may not provide the richness of content that is likely to be necessary.
• Decision Support Systems are a particular subtype of data-only systems in which the input (data) are used to provide an output that is usually in the form of a suggestion or recommendation. Decision support systems (Patkar et al, 2011) could be seamlessly integrated into the processes used to support the v MDT. Simple examples could include: the use of such automated systems to screen patients for eligibility for clinical trials and thereby identify which patients might be suitable for enrolment into clinical studies (McNair et al, 2008); linkage between the clinical information entered into the system and clinical guidelines – this could automatically generate a guideline-based recommendation as a point of departure for further discussion (Patkar et al, 2012); and integration of data on molecular profiling of patients' tumours so that specific targeted therapies could be suggested for each individual (Blackhall et al, 2013).
• Audio-only systems, such as telephony and voicemail, transmit the spoken word. They involve, as in a conference call, participants all being available at the same time. Consequently, these systems will have limited utility for a vMDT.
• Video systems allow the transmission of both sound and vision. In the context of a vMDT, a short video segment in which a patient discussed his/her current state of knowledge, concerns, and expectations, would add an important extra dimension to the online discussions.

There is, increasingly, a blurring of the boundaries between the classical route for the electronic delivery of information, telephony, and information transfer via the Internet. Many organisations now used web-based telephone systems that, to the user, seem identical to older, wire-based, systems. Groupware is the generic term used to describe software, usually web-based, that can be used to facilitate meetings and interactions between geographically dispersed individuals.

There are several important factors that need to be considered when choosing the technology to support a virtual team (Table 2).

Virtual microscopy (Fonyad et al, 2012; Kayser, 2012) is an application that has already been used in health care and its conceptual basis is very similar to that of a virtual team. Virtual microscopy could be incorporated into the vMDT, enabling detailed discussions of pathological findings to inform the conclusions reached by the vMDT.

There is a wealth of information available on the role of virtual teams in the industrial and commercial sectors. Standard texts (Lipnack and Stamps, 2000; Duarte and Snyder, 2006) contain accounts of disaster as well as of success. The topic has spawned its own jargon, the result of which is that different terms are often used to describe very similar and very obvious concepts. Recent reviews have, however, pointed out that there is remarkably little empirical fieldwork in this area (Baltes et al, 2002; Kirkman et al, 2002; Axtell et al, 2004; Powell et al, 2004; Hertel et al, 2005; Curseu et al, 2008; Ebrahim et al, 2009). Most studies deal with artificial laboratory simulations, often oversimplified, some using ‘teams' of only two members. What follows is an identification and clarification of the key concepts that are relevant to vMDTs for the management of cancer and its consequences.

When people meet around a table they are aware of each other. They are, whether they like it or not, socially connected. The term social presence is used to describe the extent to which a virtual system facilitates this type of personal connection between team members. When discussions take place, information is exchanged. In a face-to-face meeting, this communication is both verbal and nonverbal. Body language provides additional information about team members' feelings, such as boredom, frustration, anger, or anxiety. Information richness describes the amount and variety of information that a virtual system can handle, including data, images, tone of voice, facial expression, body language, or environmental cues. Sometimes there is simply too much information and participants become distracted by peripherals (that tie is really horrible, why does he make that horrible slurping noise as he drinks his tea?). This is, in the world of virtual teams, referred to as surplus meaning, there is too much information richness or social presence and the result is that team members become distracted and lose concentration.

A vMDT for the management of cancer or its consequences should combine a high degree of social presence with sufficient information richness while avoiding surplus meaning. The process of setting up any virtual team must respect the fact that different teams will have, or will develop, different cultures and that, provided the teams function well, this cultural heterogeneity is to be welcomed rather than feared.

The previous discussion of barriers to implementation has already given some idea of the issues that need to be resolved if a vMDT is to be effective. Any attempt to set up a vMDT has to start with the involvement of the team members themselves. Imposed solutions are unlikely to succeed. Participation, from the beginning of the process, will bring a sense of power and ownership (Kerber and Buono, 2004; Cordery and Soo, 2008). The technology should be appraised and selected after, but not before, team members have agreed about how they would wish to work together. The vMDT should be set up so that it will replace, rather than be an addition to, existing working practices.

The size and composition of a virtual team are crucial to its success. If the team is too large, communication becomes difficult as the number of potential interactions between participants will increase exponentially. If the team is too small, it may lack expertise and breadth of opinion, particularly if it is being asked to deal with complex problems. One solution is to have a core team who are able, as required, to call on the expertise of additional clinicians. The idea of a problem-based MDT, whose membership changes according to the nature of each individual problem that is discussed, is appealing but it might be difficult to maintain the interest of potential participants. The presence of a defined goal for a defined group will improve the sense of social cohesion among group members.

The easiest way to decide upon the membership of a vMDT is to base the virtual team upon an existing MDT. This has several advantages: members know each other and their foibles, and as they have established patterns of working together, a degree of social cohesion already exists; they are likely to regard the approach as supportive – adding to their capabilities, while making it easier for each individual to contribute; and it is reasonably straightforward to establish what is acceptable and what is not. However, there are some disadvantages associated with converting an existing team to a virtual team: dysfunctional relationships and working practices may already be entrenched and the move to a virtual team may exacerbate the problems; the problems that need solutions may not have a team available or an existing team may not have the full repertoire of skills or expertise available for the task; and the approach involves consolidation rather than innovation, when it may be innovation that is required.

Setting up a virtual team from scratch has its difficulties. It is not easy to turn a collection of strangers, who may never meet face to face, into a cohesive social unit. It is particularly difficult when the team members and coordinators are also trying to come to terms with unfamiliar technology. It is naive to assume that such adversity will somehow bring the group together. The effect could be quite the reverse – creating a group of individuals who are resentful of being used as guinea pigs. There are clear advantages to setting up a completely new team: there are no historical grievances and the team members can be selected so that all the expertise that is required will be available. An initial face-to-face meeting, in the form of a retreat or an away day, can help ease some of the problems that arise when a group of geographically dispersed strangers are asked to work effectively together.

Data protection and confidentiality are important issues. There are wide variations in policies between trusts and, if a vMDT is to work across several trusts, the procedures will have to comply with the criteria set out by the most restrictive of the participating units. One way round this potential problem is to strip all distributed data and images of personal and institutional identifiers. This involves extra work initially and at the end of the process where the recommendations are fed back to the referring clinicians, but does mean that the salient features of each problem can be discussed openly and without elaborate security procedures.

Figure 1 illustrates one approach to the organisation of work flow through the v MDT. A key feature is that patients' views can be incorporated into the process from the very beginning. The initial scenarios are prepared using direct input from each patient. At the end of the process, each patient is given a summary of the online discussions that explains, in simple language using lay terms, the nature and provenance of any recommendations.

Any changes to the organisation and delivery of MDT discussions should not be allowed to compromise the effectiveness of the existing process. A document published by the National Cancer Action team (NCAT) outlines the characteristics of an effective MDT (NCAT, 2010), and this provides a framework within which some of the potential advantages and disadvantages of a wholly vMDT might be considered. Table 3 summarises an analysis based on this approach.

Evaluation of the effectiveness of each vMDT should be built in from the very beginning (Lamb et al, 2012b). Those that are successful provide an example to others, and those that fail provide counterexamples and lessons for learning. Table 4 summarises some of the domains that need to be considered when assessing the performance of a vMDT. With an increasing emphasis on ensuring that patients are involved in decisions made about their care (Coulter and Collins, 2011), it is important to incorporate patient-related outcome measures (PROMs) into the evaluation of the vMDT. These measures should go beyond the traditional measures of quality of life (de Haes et al, 2000) and should include patient satisfaction (http://www.nhssurveys.org) (Jean-Pierre et al, 2011; Kamo et al, 2011) and patients' perceptions of the extent to which they felt that they were involved in the decision-making process (Elwyn et al, 2005; Kriston et al, 2010).

The potential role of the vMDT will be to extend the advantages of the MDT approach into areas that are currently underserved. These could include: the treatment of rare or unusual tumours; the assessment and management of patients who have unusually severe late effects following previous treatment; the investigation and management of patients who present with tumours of unknown primary site; the management of recurrent disease in patients previously discussed by a traditional MDT at their original presentation. Current arrangements for re-discussion are highly variable and the breadth of expertise that the vMDT could offer might be of particular benefit to patients with unusual or difficult problems.

Any recommendations that the virtual team makes should be as appropriate and as robust as those made by a traditional team. The process of engagement and participation should be as easy, and certainly no more difficult, than that associated with a conventional team. Convenience should be increased because of lack of need for a fixed time and place of meeting, and costs should be less. There should be evidence of the ability to expand into clinical areas where locally based expertise is unavailable. The vMDT should be judged by all the standards that apply to conventional MDT, but with the following additional outcomes: proof that team members have embraced the technology; proof that team members find the vMDT more convenient than the traditional MDT and that this convenience translates into more frequent attendance and deeper engagement; proof that the approach is affordable.

The truly vMDT, non-colocated and asynchronous, offers a potential means for dealing with some of the limitations and difficulties associated with conventional MDT meetings. However, vMDTs should not be introduced unless and until:

• there is an established framework for their constitution, organisation, and function;
• there is proof that the technology and IT systems are acceptable to team members;
• mechanisms are in place to ensure that technical and IT support is available at all times;
• there are processes to ensure that each individual team is evaluated from its inception.

Data on team organisation and performance should be pooled for all teams and should be available in real time so that we are able to assess not only how well or how badly each team is doing, but whether or not the overall approach is improving the management of patients with cancer. Given all these considerations, it is highly unlikely that existing site-specific MDTs can be effectively replaced by vMDTs.

There is promise here, but clarity and preparation are required in order to maximise the potential that the approach will deliver on its promises. Otherwise, a decade from now, there will be only scattered accounts of a few hopeful experiments. Clinicians will still not know whether or an MDT that is fully virtual can offer significant advantages over traditional methods such as face-to-face meetings or video conferencing. If vMDTs are to make any useful contribution to the management of patients with cancer, then a nationally coordinated and planned programme of research is required. The ad hoc development will not lead to meaningful progress.

The MDT meetings were introduced somewhat precipitously into cancer care in the United Kingdom in the mid to late 1990s. There was little initial planning and structure, and virtually no prospective assessment of performance or outcome. This all came later (Kee et al, 2004; Fleissig et al, 2006; Lanceley et al, 2008; Lamb et al, 2011a, 2011b, 2012a) and, as a result, despite nearly two decades of MDT team working, the extent to which this approach has directly contributed to improvements in cancer care in the United Kingdom is still unclear. We should not allow ourselves to be seduced by the apparent charms of the ‘vMDT' into making the same mistake again.