Task Complexity in Emergency Medical Care and Its Implications for Team Coordination

Article excerpt


In recent years interest in human performance in medicine has led to a wide range of studies in medical domains in areas such as human error (e.g., Bogner, 1994; Cooper, Newbower, & Kitz, 1984; Donchin et al., 1995; Schwartz, Matthay, & Cohen, 1995; Xiao, Mackenzie, & the LOTAS Group, 1995), equipment interfaces (e.g., Cook, Potter, Woods, & McDonald, 1991; Harper, Mackenzie, & Norman, 1995; Lin et al., 1995; Weinger & Englund, 1990), team member attitude (Helmreich & Schaefer, 1994), organizational factors (e.g., Gaba, Howard & Jump, 1994), and crisis management training (e.g., Howard, Gaba, Fish, Yang, & Sarnquist, 1992). These studies indicate the need for solutions in a number of areas, such as hardware and software design and team training. In particular, as Helmreich and Schaefer (1994) predicted and Donchin et al. (1995) suggested, problems rooted in team coordination and communications contributed to many of the occurrences of human error.

Like most other medical domains, emergency medical care is delivered by a collection of personnel with differing expertise and professional backgrounds. Unlike most other medical domains, team coordination in emergency situations occurs under severe time pressure. As indicated by the notion of the "golden hour" in treating traumatized patients, there is only a brief window of opportunity during which caregivers can significantly improve a patient's chance of survival. (According to Brown, 1987, about 80% of trauma deaths occur in the first 4 h after injury.)

Collecting and examining human performance data and cataloging human errors are certainly important in efforts to improve the quality of care in emergency medicine. An orthogonal approach is to understand potential mismatches between the capacities of individuals and teams and the task demands imposed on the caregivers. One way of predicting task demands is to investigate task complexity, a characterization of tasks that describes why a task may be complex to perform.

As evidenced by Woods's (1988) treatment of the topic of task complexity, one can obtain many valuable insights by examining the nature of task complexity from the viewpoint of its impact on activities. From previous incident reports, simulator investigations, and field studies, Woods synthesized a picture of cognitive demands attributable to task complexity and plausible ways in which individuals cope with task complexity, particularly during crisis situations. Such a picture can help one to predict the types of cognitive problems that are likely to confront practitioners and the types of cognitive support that are likely to be useful.

Various definitions of task complexity have been put forward, mostly in the context of well-defined experimental tasks. Three such definitions are described here as examples. For a multiple-choice task, Payne (1976) used the number of alternatives as a measure of task complexity. For an inspection task, Gallwey and Drury (1986) used the number of different fault types as a measure of task complexity. For a monitoring task, Kennedy and Coulter (1975) used the number of channels to be monitored as a measure of task complexity.

Real-life tasks in most situations are a combination of individual tasks, and thus the definition of task complexity is not clear-cut. Woods (1988) made an attempt by proposing a four-dimensional scheme oriented for tasks in process control environments to define task complexity: dynamic situations, interacting parts, uncertain data, and risk. These dimensions were proposed for the prediction of demands for cognitive activities, not for the prediction of demands for team coordination.

Despite the fact that most workplace tasks are handled by a group of people working as a team, little empirical or theoretical work has been reported that characterizes task complexity for a team environment. No framework has been put forward to understand what makes coordination of a task complex for a team. …