Reduce infections with root-cause analysis: responding to infection-related sentinel events. (The Quality--Cost Connection).

Infection control practitioners play a vital role in reducing nosocomial infections. Collecting and analyzing surveillance data can identify patterns of occurrence so that steps can be taken to eliminate or reduce the factors that contribute to nosocomial infections. The Joint Commission on Accreditation of Healthcare Organization's recent expansion of the sentinel event policy to include nosocomial infections could represent a change in what infection control practitioners view as their traditional role. In addition to discovering the root cause of undesirable infection patterns, practitioners now may be called upon to investigate an unexpected death or patient injury.

A knee-jerk response to a possible infection-related sentinel event would be to explain how difficult it is to establish a clear and concise relationship between a patient's infection and the adverse outcome in question. Often patients who develop nosocomial infections have a host of other medical problems and contributing factors. For example, sepsis from gram-positive pneumonia is a common cause of death among patients hospitalized with severe burns. Is it possible for practitioners to determine if the severely burned patient would have recovered if not for the sepsis? The Joint Commission's revised sentinel event definition suggests that this question must be answered to determine if a root-cause analysis (RCA) is required.

The difficulty in determining whether an infection-related event is, in fact, a sentinel event may be related to our fundamental reluctance to place blame. If the answer is "yes" then it appears we are admitting that the patient care experience was flawed in some way. But nosocomial infections are common in patients with compromised immune systems (such as the severely burned patient).

Thus it is conceivable to answer "no" and blame the patient's physical condition for the death. The danger in answering "no" is that the singular event--sepsis due to pneumonia--might be the symptom of a larger system problem that contributes to the development of various types of nosocomial infections in other patients. The infection surveillance data may not suggest the presence of a system problem; nonetheless, it still could be present. The sentinel event RCA can uncover problems requiring corrective actions. A significant adverse event involving an infection represents a "pattern of one" that deserves more in-depth investigation.

Consider the following situation: A 72-year old patient with severe congestive heart failure is admitted to the hospital with a stroke. On the fourth day, the patient develops aspiration pneumonia and is started on IV antibiotics. On the sixth hospital day, the patient has a sudden cardiac arrest and expires. The patient's attending physician documents the cause of death as cerebrovascular accident and congestive heart failure. Pneumonia is listed as a secondary diagnosis but not labeled as one of the causes of the patient's death.

Does this mean that the patient's death should not be treated as a sentinel event?

It really doesn't matter whether the infection is a known complication or whether it was preventable. Does the event meet the Joint Commission's definition of a sentinel event?