Measuring Teamwork
I have previously discussed the critical nature of clinical teams in producing high quality results in healthcare organizations, and I have considered the detrimental effects of turnover on maintaining highly reliable teams. In Strategic Human Capital I suggested one problem facing health system leaders is the difficulty in measuring the value of the team. I want to consider measuring teamwork, I do not mean productivity. Here I mean the question of how well teams accomplish their fundamental work, which I would define as an internal focus on maintaining the team and an external focus on accomplishing the work. In 2007, the Canadian Health Services Research Foundation published a review of the subject of effective teamwork in healthcare.[1] Their definition of healthcare teams is explicit. Teamwork requires an explicit decision by the team members to cooperate in meeting the shared objective. This requires team members sacrifice their autonomy, allowing their activity to be coordinated by the team, either through the decisions of the team leader or through shared decision making. As a result, the responsibilities of professionals working as a team include not only activities they deliver because of their specialized skills or knowledge, but also those resulting from their commitment to monitor the activities performed by their teammates, including managing the conflicts that may result. The authors concluded creating functional teams became an important policy objective for Canadian authorities in 2004. They point out that teamwork and collaboration are often used interchangeably, but collaboration may occur without a formal team structure. Teamwork requires a formal structure, but cannot succeed without collaboration. They pointed out the greatest obstacle to change was the “hierarchical culture of healthcare.” Barriers include “the self-regulation of professions, current malpractice and liability laws and funding and remuneration models.” In a 2005 meeting of experts, barriers to creating teamwork listed, among other things, “the absence of efforts to capture evidence for success and communicate this to key stakeholders, including the public.” A review of interventions to improve team effectiveness was published in 2010.[2] The authors identified 48 articles, 42 of which were published after 2000. 32 of these articles were considered of low quality and often related to the non-technical aspects of teamwork such as communication, cooperation, coordination, and leadership. Eight studies studied the effect of Crew Resource Management (CRM) training on attitudes toward quality and safety. These tended to be more robust studies, but again, show very limited results. This program has been adopted in varying degrees in U. S. healthcare settings. The program comes from the aviation industry, where rigorous “failure” analysis showed that many fatal accidents could have been prevented if the team communicated better. Since aviation is also a hierarchical industry, the parallels with operating rooms and ER’s has made it attractive. At its heart is a structured process for calling potential problems to the attention of the decision maker—the pilot in the case of an airplane, the doctor in the case of a healthcare setting. Lastly, a review of the various instruments available to measure aspects of teamwork was published this year.[3] These authors were interested in the question of whether statistically valid measures exist, what properties of teamwork are measured, and have they been used in empiric studies linked to an outcome of interest? Their definition of teams included both formal teams, as outlined in the Canadian review, and also the informal teams, which they defined as collaborations. They identified 36 scales, only 15 of which were published in health services or medical journals. 12 scales had documented relationships with a non-self-reported outcome, (4 clinical, 6 non-clinical, and 2 a combination.) Seven of these 12 scales included a full set of psychometric properties, and 3 satisfied the four pre-specified criteria. These 12 scales included questions designed to assess two dimensions of teamwork: the quality of the social interaction and the quality of the task-related interaction. The social interactions assessed included social support, respect, psychological safety, active conflict management, and group cohesion. The task-related interactions assessed included communication, shared decision making, use of all members’ relevant expertise, full participation, collaboration, learning orientation, coordination and effort. The authors note, however, The inclusion of items assessing both the quality of the task-related and social interactions between team members in all the scales but one suggests consensus that both of these elements are important components of teamwork. In contrast, the lack of consistency in the dimensions included to assess the quality of these interactions, and in how thoroughly each dimension is explored suggests lack of consensus about the fundamental sub-components of teamwork. The difficulty with this lack of conceptual consistency is that it limits what we can learn from research on teamwork and limits the ability to effectively intervene to improve teamwork. This review concludes that teamwork incorporates both social (internal) and work (external) components. To those who approach healthcare from a financial background, the social component may seem too subjective and even ethereal, and so may be ignored. Those who approach healthcare from a personnel background may focus on the social component and fail to emphasize the work component. My experience in the dialysis unit suggests this split is common. The corporate administrators, who have financial responsibilities, place a lot of emphasis on the work without considering the details. Nurse managers, on the other hand, put a lot of emphasis on trying to make sure everybody is getting along. I usually try to bridge the gap by pointing out to the nurse managers the goal is to be fair. It is not in a manager’s power to “make” someone feel anything, much less happy. I point out to the administrators that the nurses are people who have events in their own lives that may temporarily impair their ability to do the work, but that person is still a valued employee and a valued member of the team. A Working Approach to Measuring Team Effectiveness Since there is no generally accepted measure of teamwork, what should we do while awaiting developments in the field? I suggest measuring the social aspects of teamwork by looking at longevity and turnover of the team members, vacancy rates, and time to fill open slots. We may not be able to measure teamwork, but people “vote with their feet.” We should at least count the votes. While the organizational structure has an important impact on how teams perform, these effects should be fairly uniform within an organization. However, it should be clear that these measures are not directly comparable to data from other organizations. In other words, benchmarking is going to be difficult for the time being. We can measure technical outputs by defining items of interest. These would include patient satisfaction, core measures, and safety measures. These items need to be measured at the unit level, not the organization level. Clearly, some items, say heart failure core measures, are not going to be useful measures of work on the orthopedic ward. On the other hand, falls and hospital acquired infections are likely important on all wards, but probably not useful in the operating room or the emergency room. I suggest establishing a battery of six items for each team of interest—three related to quality and three related to safety. In some cases, those items will be evident to outsiders, but it is equally valid to let the team members choose their own metrics. Items measured must not only be relevant to the individual unit, they must also be compared to the individual unit.[4] The issue is not whether patient satisfaction on the labor and delivery ward is better than on the oncology ward, the issue is whether patient satisfaction improves over time. Is there a role for traditional markers of work, such as occupied beds, number of cases, etc? These markers are commonly used to determine staffing levels, and certainly, staffing levels have an impact on team function. But the issue is to measure the effectiveness of the team function in ways that are meaningful to both the organization and to the individual members of the team. Once the organization has working metrics for each team, the question becomes what can be done to improve results? Here, again, the literature is sparse. As has already been suggested, crew resource management training seems associated with improved communication and reduced errors. Team composition is another area where changes can be made. One research group, for instance, reported a study of 51 teams comprised of 652 employees showing that the teams scoring higher in general mental ability, extraversion, and emotional stability were deemed by supervisors to be both more effective and more viable.[5] Recent data suggest external leadership can also be important, even to empowered teams, although routine “leadership” activities seem to have no effect.[6] Given the absence of proven pathways, each team needs to be given the freedom to make deliberate changes in its work processes that, in the opinion of the team, seem likely to improve performance on one or more of the chosen metrics. Of course some outside review may be needed to make sure the changes are congruent with established legal and clinical standards, but, in general, management needs to be open to the ad hoc, experimental nature of such efforts. The key, as in all process improvement efforts, is frequent measurement of the items of interest and integration of the observed results into further efforts. I appreciate that in these times of economic stress, it is not realistic to exclude money from the conversation, which I have done in this discussion thus far. So how should this be done? Traditional cost accounting methods do not, in my view, provide information actionable by clinical teams. This requires breaking expenses down to the individual unit level, and requires breaking down direct expenses such as personnel and equipment. However, revenue should not be attributed to each unit. Senior management determines the cost structures of the organization, not the clinical teams, and the latter have no control over payer mix, acuity level, or volume. On the other hand, the staff can be responsible for controlling the direct costs associated with each clinical activity. In such a system, each expense results in a “negative” balance for each unit, but clinical people can understand a goal such as “keeping our number less than -$100,000.” Organizations that wait for the perfect system will be getting further behind in the race to develop effective and resilient clinical teams. The scheme I have outlined may have gaps and certainly needs to be adapted to each local situation, but does cover the elements that the research base supports. Why not start now? [1] Clements, D., Dault, M., Priest, A. Effective Teamwork in Healthcare: Research and Reality. HealthcarePapers, 7(SP) January 2007: 26-34.doi:10.12927/hcpap.2013.18669. Accessed 12 September 2014 at http://www.longwoods.com/content/18669. [2] Buljac-Samardzic, M., Dekker-van Doorn, C. M., van Wijngaarden, J. D. H., van Wijk, K. P. Interventions to improve team effectiveness: a systematic review. Health Policy 2010;94:183-195. Accessed 12 September 2014 at www.elsevier.com/locate/healthpol. [3] Valentine, M. A., Nembhard, I. M., Edmondson, A. C. Measuring Teamwork in Health Care Settings: A Review of Survey Instruments. Med Care 2014 (Apr). Accessed 12 September 2014 at http://www.rrsstq.com/stock/fra/p.217. [4] Competition between teams may be useful, but competition within teams impairs knowledge sharing, a key component of teamwork in healthcare. He, H., et al. Modeling Team Knowledge Sharing and Team Flexibility: The Role of Within-Team Competition. Human Relations. Early publication accessed 12 September 2014 at http://hum.sagepub.com/content/early/2014/02/03/0018726713508797. [5] Barrick, M. R., Stewart, G. L., Neubert, M. J., Mount, M. K. Relating Member Ability and Personality to Work-Team Effectiveness. J Appl Psychol 1998 (Jun);83(3):377-391. doi: 10.1037/0021-9010.83.3.377. Accessed 12 September 2014 at www.longwoods.com/content/18669. [6] Luciano, M. M., Mathieu, J. E., Ruddy, T. M. Leading Multiple Teams: Average and Relative External Leadership Influences on Team Empowerment and Effectiveness. J Appl Psychol 2014(Mar);99(2):322-331. doi: 10.1037/a0035025. 22 September 2014 |
Further Reading
A Good Place To Work Is your organization a just one? How do you know? Culture Matters The scandal at the VA shows the importance of choosing performance measures wisely and the need to consider organizational culture in applying standard management techniques. Horizontal Violence and Nursing Staff Turnover A recent study shows horizontal violence - conflict between nurses in a hospital - is common and a major cause of job dissatisfaction and intention to leave. What can be done about it? Organizing for Success - Lessons from Keystone The Keystone Cooperative ICU Project obtained major improvements in safety. The "soft science" lessons need to be applied more widely. |