was not to employ a representative sample of hospitals, but to assess how excellent nursing hospitals handle service failures. From these observations, common basic patterns of problem-solving behavior across the nine hospitals were identified.

Researchers distinguished two types of process failures: “errors” and “problems.” “Errors” were defined as the execution of a task that was either unnecessary or carried out incorrectly. “Problems” were defined as disruptions of the nurse’s ability to execute a prescribed task because a resource was unavailable at the needed time, location, or condition or in sufficient quantity (e.g., missing supplies, information, or medications), thus preventing the task from being implemented.

Of the 194 observed failures, 86 percent were problems rather than errors. This finding is significant to improving patient care for several reasons. Problems are relatively frequent and visible, and also carry fewer stigmas than errors; all of these features facilitate an HCO’s taking action on a problem to improve patient care and safety. However, researchers found that nurses tended to practice “first-order” problem solving, that is, fixing the immediate problem without communicating that it occurred, investigating why it occurred, or seeking to change its cause. Thus the problem was isolated so that it did not become visible to the hospital as an opportunity to learn how to be more efficient or effective in patient care. Second-order problem solving, in contrast, occurs when a worker, in addition to fixing the problem so the task at hand can be completed, takes action to address the underlying cause. Researchers used lenient criteria—i.e., encompassing any behavior that called attention to the problem—to assess the extent to which second-order problem solving had occurred. Nonetheless, only 7 percent of nurse responses were second-order.

Researchers identified three human resource practices that explained why so few problems had received second-order attention that would have enabled the organization to learn from the problems and correct systemic weaknesses. First, instilling in nurses a strong sense of responsibility for individual vigilance can, as a side effect, encourage such a strong emphasis on independence and self-sufficiency that they see a failure not as a system problem, but as one that can be overcome or withstood through individual competence. The majority of the nurses interviewed commented that they believed their manager expected them to work through daily disruptions on their own. Speaking up about a problem or asking for help was likely to be viewed as a sign of incompetence. Second, staffing levels were so tight, with so little slack in the system, that nurses did not have the time to eliminate underlying causes of problems. Instead, they were “barely able to keep up with the required responsibilities and [were] in essence forced to quickly patch problems so they [could] complete their immediate responsibilities” (Tucker and Edmondson, 2003:9). Finally, removal of front-line managers



The National Academies | 500 Fifth St. N.W. | Washington, D.C. 20001
Copyright © National Academy of Sciences. All rights reserved.
Terms of Use and Privacy Statement