Keywords

These keywords were added by machine and not by the authors. This process is experimental and the keywords may be updated as the learning algorithm improves.

14.1 Evaluating Team Performance: A Systematic Review

Almost all of us are familiar with teams and/or team observation thanks to sports teams. Whether recreational, school-based, or professional, athletic teams are an example of what a group of people can do to reach a common goal. Assessment of team performance in sports is well established, and the outcomes are obvious (did they win?). For teams in aviation and health care, the outcomes are significantly more important. Did the plane land safely? Is the patient alive? Aviation and health care share a common feature, the potential of tragic consequences when errors occur. Because of the consequences associated with failure, these professions are said to be examples of high-reliability organizations.

The nature of the “hypercomplex environment” in which health care occurs is characterized by several decision makers, whose roles are embedded in an “extreme hierarchical differentiation,” they note that the assurance of patient safety requires interaction and communication in “compressed time” with a “high degree of accountability” (Baker et al. 2006). In identifying the characteristics of high-reliability organizations, Baker, Day, and Salas argue that healthcare providers are often organized in teams, and that their interactions are part of the vital operations in various settings (Baker et al. 2006). The hypercomplexity of the context in which health care occurs is characterized by specialization, where team members have specific roles, responsibilities, and knowledge (Orchard et al. 2012). Because errors, although rare, result in serious consequences, teamwork is essential. Knowledge of their own roles and responsibilities, monitoring of team member performance, and a positive attitude toward teamwork have been shown to relate to team effectiveness (Baker et al. 2006; Driskell and Salas 1992).

Team competencies typically considered for training programs have been identified as leadership, mutual performance monitoring, mutual support, adaptability, team orientation, mutual trust, shared mental models, and communication (Baker et al. 2006; Salas et al. 2005). The extent to which interdependent healthcare professionals are able to communicate and coordinate activities has been shown to relate to failures as well as successes (Healey et al. 2006; Nagpal et al. 2012). While team competencies have been studied extensively in aviation and the military, research on teamwork training and evaluation is gaining prominence in healthcare professions (e.g., Capella et al. 2010; Lurie et al. 2011; Tumerman and Carlson 2012). When considering patient outcomes, patient safety and team performance have been linked to surgery (e.g., Lawton et al. 2012; Nagpal et al. 2012).

Considerable research has been conducted in organizational and cognitive psychology, providing a theoretical basis for competencies associated with effective team performance. While several training programs focusing on team training have been started in health professions, little research has been done to determine whether the theories developed for aviation and military operations are applicable to healthcare settings (Baker et al. 2006). The increasing emphasis on the link between effective teamwork and positive patient outcomes demands that research provide evidence supporting training programs as well as assessment and evaluation of team processes in health care.

Measurement of team processes that lead to successful performance can be challenging. In the health professions, teamwork training introduces new concepts, where autonomy had previously been emphasized (Lerner et al. 2009). Research in other fields has shown that the development of readily observable behavior checklists is likely to be more accurate than self-assessment (Baker and Salas 1992), although more work regarding the extent to which these measures are specialty- or context-specific needs to be done. What is the current state of affairs in assessment of teamwork for these professionals? How has theory from other fields been incorporated in the training and evaluation of health professionals?

Principles meant to guide the development of assessment tools provide a framework for the categorization of the research reviewed in this paper. Baker and Salas (1992) indicated that measures should show clear relation to theory, should present evidence of reliability and validity, should indicate the developmental nature of teams, and should be observation based. In the current investigation, a review of research published that involved healthcare professionals was conducted to determine the extent to which assessments of teamwork had been developed and evaluated between 2006 and 2012. The goal of the study was to summarize the extent to which instruments had been developed, adhering to the principles proposed by Baker and Salas (1992).

14.2 Methods

The National Library of Medicine’s indexed database PubMed was used to identify potential articles for inclusion in the review. Articles published in English between 2006 and 2012 were searched for potential inclusion. Using the search term “assessment OR evaluation AND teamwork,” a total of 549 articles were retrieved. Only those articles that referenced assessment of teamwork were selected for inclusion. Articles were classified by healthcare profession (e.g., pharmacy, medicine, nursing), specialty (e.g., surgery, oncology, anesthesiology), type of research (e.g., program evaluation, quality assurance, psychometric), and factor measured (process, skill, or task).

Of the 549 articles retrieved, 158 articles were selected for inclusion in the study based on review of the article abstracts. Since the current investigation focuses on assessment and evaluation of teamwork amongst health professionals, articles focusing on training, quality assurance, or safety climate were eliminated from further review. Of the 158 articles included for second-level review, 45 were identified that focused specifically on assessment or evaluation of teams, the remainder focused on some other aspect of teamwork (see Fig 14.1).

Fig. 14.1
figure 1

Review and selection process

14.2.1 Results

Nonmeasurement articles on teamwork. Based on the review of 158 articles on teamwork, 113 of the articles reported research focused on topics other than assessment, most of which reported on program evaluation (45 articles, 40 %). Although the majority of these studies focused on medicine (n = 53, 47 %), a number of the studies included various members of the healthcare team (48 articles, 42 %). While surgical teams (n = 12; 11 %) and emergency medicine teams (n = 8; 7 %) were studied, there was an effort to study healthcare teams in a variety of specialties (n = 48; 42 % of articles did so). The primary focus of the articles was on program or training evaluation (n = 45; 40 %), theory specific to healthcare settings (n = 15; 13 %), and program development (n = 15; 13 %). The remaining categories included quality assurance and patient safety (10 articles), review of research on teamwork (9 articles), and attitudes toward teamwork (5 articles), amongst others. Appendix A provides the listing of articles on teamwork that were not focused on assessment.

Review of the 45 articles on assessment and evaluation was primarily focused on psychometric issues (i.e., reliability and/or evidence of validity); 26 of the articles examined psychometric characteristics of the measures (59 %). Other study types included evaluation (n = 11; 22 %), theoretical study of teamwork competencies specific to healthcare professionals (4 articles, 9 %), and review articles on teamwork measures in health professions (n = 2; 4 %). Most of the studies were focused on medicine (n = 27; 60 %), predominately in surgery (n = 17; 37 %), although several of the studies that reported on measures were interdisciplinary (i.e., across professions, but in a particular specialty). Articles classified as general (n = 12; 28 %) included settings that crossed specialties (e.g., Orchard et al. 2012). The investigations primarily focused on measures intended to evaluate teamwork skills (n = 22; 49 %), with eight (18 %) of the measurement themed articles measuring skills and tasks, and eight measuring only tasks. The remaining articles measured attitudes toward teamwork (n = 5; 11 %), climate (n = 1; 2 %), and the relation between teamwork and patient outcomes (1 article).

Because the focus of the article is specifically on assessment, further characterization of articles reporting on measures that examined psychometric qualities were conducted to determine whether the measures studied were self-report, self-assessment, or meant for observation. Of the 26 articles containing information specific to the measure(s) studied, one did not provide sufficient information to determine how the measure would be used (Varkey et al. 2009). Of the remaining 25 articles, 9 (35 %) were self-report or self-assessment measures; and two of these were attitude toward teamwork measures. The remaining 17 articles were observational measures; 8 of those were conducted in surgical settings. Generally, team activities were videotaped, and then the checklists were used to rate performances. Table 14.1 provides information on the articles in which measures were studied.

Table 14.1 Studies focused on assessment of teamwork

When examining which competencies were measured, the work of Salas et al. 2008 was used as the theoretical basis for team training. These included team leadership, mutual performance monitoring, backup behavior, adaptability, team orientation, shared mental models, mutual trust, and closed-loop communication (Salas et al. 2008, p. 1003). Whether these constructs were measured as part of the assessment was examined by review of articles that detailed instrument content. Table 14.2 presents the overlap between the theorized constructs and those measured in the studies included in the review. The work of Patterson et al. (2012) showed that it was possible to design an instrument that measured all of the competencies for effective teamwork, and in addition, they included measures of conflict. Lurie et al. 2011 studied whether the burden of rating using longer checklists could be reduced without loss of information and reliability. In their study, they found that a 29-item checklist could be reduced to as few as five items with similar reliability and factor structure, and that observations could be completed in as few as 3 minutes or less.

Table 14.2 Constructs measured by teamwork assessments

For articles that included the measures as an appendix, authors often found that the items used could be labeled as constructs other than those included in theory focused on training. Team orientation, shared mental models, and mutual trust may have been measured in studies of attitudes more often than in studies which focused on evaluation of teamwork skills. For the six studies included in Table 14.2, most included measures of communication and leadership. These factors are those considered to have the most negative effects when teamwork fails (Nagpal et al. 2012). Two studies (Kenaszchuk et al. 2010; Patterson et al. 2012) considered negative behaviors that could hamper teamwork, isolation, and sources of conflict. In general, instruments described were theory based, and the authors defined constructs in a manner consistent with the theoretical literature.

14.3 Discussion

Research conducted with the military and aviation has informed practices in health care (Baker et al. 2006; Kendall and Salas 2004; Salas et al. 2008). Research on the development of measures that are reliable and based on theory have been conducted and have advanced adaptation of measures of team interaction from other professions. The predominance of articles on psychometric issues is not surprising; work to determine whether measures could be developed or adapted for use with healthcare teams in a variety of settings is essential, and was recommended by those pivotal in the development of teamwork theory in other professions (e.g., Baker and Salas 1992). In addition, study of the factors that can affect teamwork or result in negative performance (e.g., Baker 2010; Capella et al. 2010) support the identification of factors related to the avoidance of negative consequences.

Measurement of teamwork amongst healthcare professionals faces several challenges. First, efforts continue to be specialty-specific (e.g., surgical, emergency medicine, community medicine), although there are studies that have looked to see if the measures can be used across settings (e.g., O’Leary et al. 2012). While several of the measures developed are based in theory, different constructs may be measured. Although there was minimal inconsistency in terminology, papers that do not clearly define the constructs measured can make this effort challenging, particularly if measures are to be used across health professionals and settings.

Interestingly, research has shown that team members are generally not reliable at assessing their level of skill (Baker and Salas 1992; Eva et al. 2004), but practitioners are generally able to self-monitor (Eva and Regehr 2010). Seven of the articles reviewed were self-report or self-assessment measures. Since two of those were attitude measures, the importance of the effect of self-assessment may not have the same significance as it does in measurement of competence.

Although observational measures have been said to be preferable, securing the necessary number of raters to produce reliable measures has been challenging (Morgan et al. 2007), although recent work has shown promise (Russ et al. 2012). Efforts are underway to show that shorter versions of long measures can be used in a fashion that may facilitate recruitment and training of raters, generating more ratings available for the evaluation of teamwork skills (Lurie et al. 2011).

A number of publications that focused on program evaluation highlight a challenge in assessment of teams: finding reliable measurement tools that assess group interaction (Morgan et al. 2007; Murray and Enarson 2007), particularly when targeted training in teamwork skills has been conducted. These studies typically rely on pretest–posttest design (e.g., Aboumatar et al. 2012; Vyas et al. 2012), and often include measures of participant opinion regarding training. While this is legitimate for program evaluation, measures that can be used in practice (i.e., workplace setting) can provide additional evidence of the effect of training in interprofessional teamwork.

Although this review has provided information on the development of assessment tools for use with healthcare professionals it is not without limitations. First, only the author reviewed the abstracts, so no consistency of coding was provided. Other researchers may disagree with the categorization of the studies included in this report. However, the appendices, which include a complete reference list, can be used by others who are interested in the topic. Also, only abstracts were reviewed to determine inclusion/exclusion, and reference lists of included articles were not used to identify other articles for potential inclusion. Additional review and categorization may have examined whether reliability was reported, and the extent to which validity evidence was provided for the measures. Despite these limitations, this review provides preliminary information on the methods used to evaluate teamwork amongst healthcare professionals.

The nature of health care, as typified by Baker et al. 2006, is increasingly complex, and errors have serious consequences. The rigid hierarchical roles that health professionals have traditionally had must change; although knowledge of each person’s role in the team is essential, adaptability and monitoring are important components of successful teamwork. Studies have begun to show the relationship between effective teams and positive patient outcomes (e.g., Mazzocco et al. 2009). Measures developed that can be used in a variety of healthcare settings, in addition to surgery and acute care, will be invaluable as the complexity of providing adequate patient care will increasingly require the coordinated efforts of team members.

Issues/Questions for Reflection

  • Training in use of the assessment may be necessary, particularly if observation of teams will occur.

  • How can work done in psychology on human interaction support the assessments developed for teamwork?

  • The effect of team size, team formation (standing teams vs. dynamic teams), hierarchical structure, professional identity, and more will need to be studied in compiling evidence of validity of team measures.