There is extensive evidence from many different settings, including healthcare, that effective teamwork improves organizational performance in terms both of efficiency and of quality. Reviews of literally hundreds of studies in industry and financial services for instance, have concluded that improved team working can lead to increases in productivity, quality and financial performance (Paris, Salas and Cannon-Bowers, 2000). We know there are many differences between healthcare and other settings, but at the very least this should make us think there may be more to good teamwork than a sense of camaraderie, shared grievances about working life and occasional, though welcome, alcohol-induced team bonding. As we will see, many quite simple team interventions have as their primary goal the re-introduction of these basic team processes.

In their comprehensive review of team effectiveness in healthcare, Lemieux-Charles and McGuire (2006) found some evidence that structuring work in healthcare teams improved quality. For instance, team interventions in the Veterans Administration geriatric service, improved functional status, mental health and even reduced mortality (Caplan et al., 2004). West and colleagues (2002) found an association between management practices in hospitals and patient mortality; in hospitals where more than 60% of staff worked in formal teams, mortality was around 5% lower than would have been expected (West et al., 2002).In a major study of 44 sites and over 6000 people, Daniel Davenport and colleagues (2007) found that reported levels of communication and collaboration in surgical teams, though not team climate or working conditions, were associated with risk adjusted morbidity and mortality in the Veterans Health System. These are important findings but there is still much to learn both about the nature of team performance in healthcare and where our priorities should be for strengthening teamwork.

Teams, like individuals, may erode or create safety. For instance, in their study of communication in the operating theatre, Lorelei Lingard and colleagues (2004) classified about a quarter of operation relevant communications they observed as communication failures. Events were classified as failures because they were made too late or too early, because essential content was missing, they were addressed to the wrong person or the purpose was simply unclear. The nurses and anaesthetist, for instance, discussed the positioning of the patient for surgery without consulting the surgeon, resulting in wasted time and interpersonal friction later in the case. A team that is not working effectively multiplies the possibility of error. Conversely teams, when working well, have the possibility of being safer than any one individual, because a team can create additional defences against error, by monitoring, double checking and backing each other up; when one is struggling, another assists; when one makes an error, another picks it up.

Patient safety has been particularly influenced by aviation teams and the use of simulation in pilot training; this approach has been particularly important in anaesthesia, and latterly in surgery and emergency medicine (Cooper and Taqueti, 2004). Crew resource management(CRM) is the term commonly used to describe the training of cockpit teams and other aviation teams; the detection and management of error has always been a central component of the more successful training programmes (Helmreich and Merrit, 1998). The CRM training includes instruction in human vulnerability to stressors, the nature of human error and error counter-measures. The objective of the training is to reduce the risk that crews will make a series of important errors, because they failed to foster teamwork, solve problems, communicate and manage their workload effectively (Risser et al., 1999). Notwithstanding the importance of CRM, it is important to realize first that aviation will only be a good parallel for some healthcare teams and second that, as patient safety matures as a discipline, we should be drawing on a wider range of models, research literature and methods of training.

Most teams believe their teamwork is pretty good. But, what do you see when you actually watch a team work? We will look at two examples from surgery, and one from emergency medicine. In the first, my colleagues Shabnam Undre, Andrew Healey and Nick Sevdalis (2007) and others developed a method of assessing the performance of surgical teams, Observational Team Assessment in Surgery (OTAS), in which two observers, usually one clinician and one psychologist, observe team tasks and team behaviours, respectively. For the moment we will just review the team tasks which were divided into three categories:

• Patient tasks – related to actions or information associated directly with the patient;
  
• Equipment and provisions tasks – included items such as checking and counting surgical instruments;
  
• Communication tasks – included confirming consent, patient details and operative site.

In an initial study of general surgery, we found that up to a third of standard team tasks were not completed. In a second study of urology surgery teams, we found that the average level of task completion across the three categories of tasks that we observed was higher, in that 83% of the tasks were completed; 93% of patient tasks, 80% for equipment/provisions tasks and 71% for communication tasks. Many fewer equipment/provisions tasks were completed during the preoperative phase (61%) than during the intraoperative phase (91%) or the postoperative phase; the opposite was found for communication tasks (Box 18.1).

In a second example, Elaine Hart and Harry Owen observed 20 anaesthetists carrying out the difficult, and increasingly uncommon, procedure of general anaesthesia for Caesarean section in a simulated environment. They had first prepared a checklist of 40 items all viewed as important procedural checks by an expert panel. They found that on average the anaesthetists failed to carry out about a third of the recommended checks:

Some items were omitted because staff assumed they had been checked by others. Such assumptions can have disastrous consequences. . . Most participants admitted that they did forget to check an item that they would deem as important and would routinely want to check.
(HART AND OWEN, 2005)

In spite of these findings, only 40% of participants felt that a checklist would be useful in a clinical situation. Many expressed concerns about increased anxiety in the patient associated with its use, in that hearing the list being read out loud might be disconcerting to an awake patient in what is often an already stressful situation for them. But, if you were a patient, what would you think? Hart and Owen appositely remark that airline passengers are reassured by visible routines and checklists as being evidence that everything has been checked and nothing ignored or overlooked by the flight crew; the same may well be true of patients.

The third observational study concerns the critical team handover within emergency medicine. Accurate communication of information at shift change is one of the primary functions of handover to ensure safe transition of shift responsibility from the outgoing to the incoming teams in a healthcare setting. The shift leader is required to have knowledge of the all patients in the department, in order to prioritize them and organize investigations necessary to make decisions. This requires knowing how many patients are in the emergency department, what beds are available and whether side rooms are needed because of infection control. They also need to be able to anticipate problems in the forthcoming shift due to staffing problems and have identified the senior doctors on call for the various specialties. They would expect to know about serious incidents and patients who had died in the department, as well as equipment problems and any other problems or special circumstances (Farhan, personal communication).

When Maisse Farhan began her observations, the need for this information to be transferred was well understood by senior staff, but the handover was informal. Although having worked in the department for some years she was surprised to find, when she actually watched, just how little information was actually transferred. Only about a third of the deaths and serious incidents were discussed, the hospital bed situation was hardly ever mentioned, equipment problems although quite frequent were never mentioned and seriously ill patients often not discussed. In these circumstances, the incoming shift leader has, in effect, to go round and find out everything for themselves; the handover, being inaccurate, serves little purpose and conveys false and dangerous reassurance.

Direct observation of clinical care has not been used to a great extent in studies of safety and quality, but deserves a much higher profile. Compared, for example, to the laborious and expensive analysis of hundreds of incidents, invaluable information can be extracted relatively quickly. This is just a small selection of studies, but fairly typical of those I have reviewed; always, the care actually provided falls far short of what the staff concerned imagine is happening (and indeed the patients). The most remarkable feature of these studies is that the clinical observers themselves, although they know these environments intimately, can still be surprised by what they see. As human beings, we quickly become accustomed to our working environments and in a very real sense fail to actually see the many inefficiencies and lapses. This is perhaps less surprising than it first seems. Clinicians are almost always busy and do not have time to stand about watching. Once you do stand back though and watch attentively, a whole new landscape emerges. Suddenly one is simultaneously deeply impressed by how everyone copes with the rapidly changing and sometimes chaotic environment, while being simultaneously appalled by how far teamwork in practice diverges from the tidy guidelines and protocols mandated by professional organizations. Teamwork can, however, be improved by effective leadership and by a variety of interventions.

Team leadership in all its forms is particularly critical in high risk activities. For example, team leaders influence safety attitudes and behaviour in the workplace, such as compliance with safety related rules and procedures and are key to the effective management of emergencies (Flin and Yule, 2004). Rigid hierarchies in healthcare teams may not be conducive to high quality care; however, leadership, clarity of purpose and roles remain critical.

I once attended a talk on surgical leadership in which Ernest Shackleton, the Antarctic explorer, was held up as a model for surgeons to emulate. The talk was very inspiring, thoughtful and reflective. However, I remember thinking that healthcare was in serious trouble if all leaders were expected to develop the same qualities as Shackleton. This story directs us to one particular vision of leadership, prominent in early research, in which leadership is founded in character and associated with charisma, drive, intelligence and endurance; one might call it the heroic model. Latterly, however, writers on leadership began to see that effective leaders adapt their style and behaviour to the context and demands of the task. For instance, a completely autocratic string of commands might be absolutely appropriate during an emergency; the same leader would be ill advised to adopt the same approach when trying to engage ward staff in an improvement programme.

Consider the range of leadership responsibilities in a maternity unit. Maternity services are organized in different ways in different countries. In Britain, Canada and other countries, normal births are managed by midwives with obstetricians managing complex births and associated medical problems. The leadership roles in this context include:

• The individual midwife leads the team caring for the woman and her baby, which may include a midwifery student and a maternity support worker. In an emergency, leadership passes to an obstetrician, co-ordinating a larger team, including paediatricians, midwives, anaesthetists and more junior obstetricians.
  
• The senior obstetrician on call needs to provide similar leadership to the team of obstetricians on duty and is ultimately responsible for all the care provided on an obstetrician-led labour ward.
  
• Midwifery co-ordinators are the leaders of a labour ward midwifery shift, providing support to all midwives on duty, taking decisions about staff deployment and reviewing professional decisions where appropriate.
  
• Other clinical areas also have their own leaders: antenatal and postnatal wards have a midwifery shift leader; operating theatres have separate structures of leadership, involving anaesthetists, scrub nurses, and recovery nurses.
  
• Specific leadership and support on safety issues may be provided by a dedicated unit safety lead or risk manager.
  
• At a higher level, the unit’s head of midwifery, clinical director or service lead, and general manager lead the entire maternity unit team.

Clearly leadership is very widely distributed in organizations, and not just confined to those in senior positions. Almost everyone in healthcare has some leadership responsibilities.

Leadership requires specific skills, in addition to clinical ability but these skills are often lacking:

Our experience on the ground is that there are a lot of core management skills that people in very key roles are lacking, and that’s to do with managing conflict, getting teams to work effectively together, being able to analyse incidents and drawing out learning from that. When we do development work with people, some of the basic management skills appear to be a revelation.

We had people telling us either that they didn’t know who was in charge or that those in charge never seemed to be around unless there’s a crisis.
(SAFE BIRTHS: EVERYBODY’S BUSINESS, 2008. REPRODUCED BY PERMISSION OF THE KING’S FUND, LONDON)

Leaders must manage not only their own work but also those of the team. They need therefore to be knowledgeable not only about their own speciality, but to also appreciate the work and challenges faced by other members of the team. Ideally they are respected by other team members for their experience in their own field, but also for their willingness to appreciate the skills of others (Flin, Crichton and O’Connor, 2008). Team leaders have three main tasks (Zaccaro, Rittman and Marks, 2009), which are to:

• Create the conditions that enable the team to do its job. This means making sure the team has clear objectives and that the necessary resources are available.
  
• Build and maintain the team as a performing unit. This includes making sure the team includes members with the necessary skills and abilities. The leader must also develop processes that help the team to perform effectively by nurturing good decision making, problem solving and conflict management.
  
• Coach and support the team. The team leader has to be sensitive to the mood of the team and to note how well members are interacting and communicating. A key task is to ensure that everyone is on ‘the same page’ while training and working together, which is why the best teams engage in constant team briefing and exchange of information.

Team leadership is different from traditional hierarchical leadership. Traditional leaders tend to direct rather than facilitate and support, to give rather than seek advice and to determine rather than integrate views. Effective team leaders, on the other hand, encourage members to offer solutions when things are not going well and do not insist on having the final say when decisions need tobe made. Team leaders differ most clearly from traditional leaders in focusing on the team as a whole rather than on its individual members and in sharing responsibility with the team (West, 2004). Leaders must be easily available and visible to junior staff and have a crucial role to play in supporting more junior staff to be confident about asking for help.

Watching teams and teamwork quickly reveals that a group of well intentioned individuals does not make a team and furthermore, that teamwork has to be planned and organized. In this section we will review some apparently simple interventions, which turn out to have quite profound effects. Daily goals, pre-operative and post-operative checklists seem mundane, and this partly accounts for clinicians’ resistance to their use. However, a checklist is not a piece of paper or even a list; it is a team intervention which, used well, can affect the wider team functioning, the relationships across professions and hierarchies and even the values and safety culture of the team. To my mind, the impact of these simple tools on clinical processes and patient outcome suggests that their effect can only be fully understood by appreciating their wider impact on team performance.

Recall the case of David James, who died from a spinal injection of Vincristine. One of the features of this case was that almost everyone involved made assumptions about the knowledge and abilities of those around them. We assume, by default, that other people have the same understanding of a situation as we do and, even worse, that we have correctly communicated our intentions and wishes. Many instructions for patient care are given rapidly, in a hurry, often in a kind of clinical shorthand and with many assumptions about the kind of basic care that will be provided. In a fixed team that works together day in and day out, this generally works pretty well. However, few teams, especially ward teams, are like that; it’s a shifting population of people on a variety of shift patterns, supported to varying degrees by temporary staff.

Pronovost et al. (2003) posed two simple but critical questions to intensive care doctors and nurses after the daily rounds: (1) How well do you understand the goals of care for this patient today? and (2) How well do you understand what work needs to be accomplished to get this patient to the next level of care? These questions seem unnecessary, almost insulting.

BOX 18.2 Daily goals in intensive care