While there is a growing literature on interprofessional teamwork, much of it consists of uncritical, descriptive accounts of health and social care teams. Many of these studies rely on unvalidated, surveys or superficially analysed interviews with team members. These approaches reinforce a normative view of teamworking rather than revealing new insights or providing rich accounts informed by careful observations of actual practice. In addition, although several reviews of interprofessional teamwork have gathered together large bodies of evaluative literature, many have failed to assess the quality of the included studies or the effects of publication bias (e.g. Canadian Health Services Research Foundation, 2006; Xyrichis and Lowton, 2008). Others categorise different teamwork approaches (e.g. Choi and Pak, 2007), but add little to knowledge of their effects on key outcomes such as team functioning, health care processes and patient care. Furthermore, most evaluations rest upon an assumption that teamwork is essentially a positive and functional way to organise the delivery of health and social care. Few studies offer a critical view of teamwork, or reveal dimensions of power, hierarchy and gender. As we discussed in Chapter 4, these are key elements to understanding interprofessional teams.

In discussing these limitations, Sinclair (1992) has argued that this unquestioning approach to teamwork in health and social care has created a powerful ideology which:

Has been supported by researchers who offer the ‘team’ as a tantalizingly simple solution to the intractable problems of organizational life. (pp. 611–612)

Sinclair points out that teamwork may be problematic when simplistic views about its effectiveness prevail. Such views neglect the role of friction and conflict between members which can impede the work of the team. Similarly, Allen and Hecht (2004) have argued that there is an inherent appeal of teams and teamwork which is not based on much robust empirical evidence.

As we noted earlier, teamwork is a concept describing a range of complex and contingent approaches to the organisation of multiple professionals. Consequently, there is no single, clearly defined ‘teamwork template’ that can be applied to all contexts and services. It cannot also be assumed that teamwork will lead to improved health outcomes. Uncritical support for teamwork is therefore unhelpful. Such support asserts emotional alignment to a value rather than to a thoughtful plan of action or to a practical approach to problem-solving or service provision. Supporters of this uncritical stance may reject evaluation of teamwork activities. For us, however, evaluation is central to developing knowledge on the effects and effectiveness of teamwork interventions. Evaluation demands a critical stance – one that has as a starting point the possibility that different teamwork approaches are needed, depending on local needs; and accepts that teamwork interventions may have a range of outcomes (e.g. success, no effect, harm). All teamwork interventions should therefore be evaluated.

Informative and reliable evaluations of teamwork interventions have become more common in recent years, in response to demands from policy makers, funders, managers, professionals and patients for evidence of its effects. Given the complex array of factors shaping interprofessional teamwork, evaluation of its processes, outcomes and impact is difficult. This is no reason, however, to abandon the attempt to improve teamwork in health and social care through informative, critical and empirical evaluation. Indeed, as Baker et al. (2005) pointed out, given the complexities in evaluating teamwork, any approach must be grounded in theory, account for individual and team-level performance, capture team processes and outcomes, and adhere to standards for methodological quality.

Given the concerns outlined above, careful and thoughtful evaluation of interprofessional teamwork activities and initiatives is essential.

For example, evaluation can help to:

• Understand the nature of teamwork, including how and why it works or has no effects
  
• Identify those situations where one form of teamwork is more (or less) effective than others
  
• Provide accountability for health and social care resources to those who provide or control them – regional/national authorities, research and service funders, consumers and foundations
  
• Strengthen the organisations in which evaluations are conducted, through building capacity for improvement and enhancing effectiveness and efficiency
  
• Provide generalisable evidence which can assist in advancing knowledge in the field.

Without the insights and evidence from evaluations, resources may be wasted on ineffective, or even harmful teamwork interventions. As a result, funders and commissioning bodies may in the longer term become less supportive of efforts to promote teamwork. Our goal is therefore to encourage practitioners, managers and researchers to build an informative and reliable evidence base for improving care and patient outcomes through interprofessional teamwork.

Evaluation has been defined as a process of ‘appraising human activities in a formal, systematic way’ (Kelly, 2004, p. 523). Similarly, Patton saw evaluation as ‘any effort to increase human effectiveness through systematic data-based enquiry’ (Patton, 1990, p. 11). A key purpose of evaluation is to make judgements about the usefulness of health and social programmes. Evaluation is therefore focused on assessing change in activities or programmes; addresses pre-specified questions in a transparent way; and is based on the systematic collection and analysis of data.

Evaluation is an empirical activity in which inquiry is made into the causes and effects of actions. Evaluations, as we discuss, may be quantitative or qualitative or utilise both approaches. A number of qualitative and quantitative components may be assembled to provide a complete picture of the effects of a teamwork intervention. Comparisons between different actions (or interventions) aimed at the same goal (or outcome) can show comparative effectiveness in achieving that goal or outcome. For example, in a randomised control trial undertaken by Wild et al. (2004), patients of a ward in a community hospital were randomised to either the intervention medical team, which conducted daily interprofessional rounds, or to the control team, which provided standard care. The effects of these two different actions on length of hospital stay were measured. The study found no difference in length of hospital stay between the intervention and control groups.

Evaluators may also explore the experiences of those who undertake, or are the recipients of, teamwork in order to reveal the meanings and values attributed to it. For example, a qualitative study was undertaken to evaluate the first specialist adolescent cancer unit established in the UK (Mulhall et al., 2004). The evaluation aimed to explore the culture of the unit; the experiences of patients and their parents as well as the views of staff delivering care in the unit, who constituted a cancer team. Both semi-structured interviews with these groups and observations of routine unit activities were undertaken. Mulhall and colleagues reported a range of findings, including that the availability of an expert group of health care providers was key to ensuring an appropriate care environment. The authors concluded that the complex health care needs of adolescents with cancer may best be met by teams working in specialist units.

In addition, evaluators may examine the costs of teamwork actions or programmes so as to inform decisions regarding their cost-effectiveness. For example, an economic analysis was undertaken alongside a trial of assertive community treatment for homeless adults with severe mental illnesses in the US (Lehman et al., 1999). The effects of this treatment, which promoted continuity of care and was delivered by an interprofessional team, were compared with usual care. The main outcome was the number of days of stable community housing enjoyed by participants. Lehman and colleagues concluded that the assertive community treatment intervention was more effective but not more costly than usual care for increasing time spent in stable housing by adults with severe mental illnesses.

Interventions to improve the process of interprofessional teamworking and to maximise its benefits are complex and are often composed of many different elements. Consequently, evaluations of these interventions are frequently multifaceted. To help understand this complexity, we categorise evaluations according to their purpose, the targets of action and the types of evidence that evaluators may gather.

Evaluation purpose

The purpose of an evaluation may be to provide formative information, summative information or both. An evaluation that is undertaken early in the development of a new or modified teamwork intervention may be useful in understanding the context in which an intervention is to be implemented as well as in improving its design. At this stage of development, there is a need to incorporate feedback from a range of stakeholders in order to improve the ongoing design and delivery of the intervention. Evaluation for this purpose is formative in that it contributes to shaping an intervention or its implementation.

Once the final form of the intervention is defined and implemented, the focus of evaluation may shift towards such issues as the overall worth of the new approach, including its effectiveness and efficiency, in comparison with usual care, existing approaches or alternatives. Such evaluations may be conducted at the site where the intervention was implemented initially, or more widely, at other sites. Evaluation for this purpose is summative and may include data on:

• Perceptions of those participating in and experiencing care within the new model of teamwork – often using qualitative approaches
  
• Processes of care that have changed as a consequence of the intervention – often using both qualitative and quantitative approaches
  
• Impacts on practitioners, resource use or patient outcomes – often quantitative and may include economic data.

Box 7.1 provides two examples of evaluation studies – one which employed a formative approach and the other which employed a summative approach.

Target of the evaluation