The trusted source for
healthcare information and
Joint Commission’s Q&A on infections as sentinel events
Deadly infections always were sentinel events
The Joint Commission provides the following answers to frequently asked questions about its 2004 patient safety goal to manage as sentinel events all identified cases of unanticipated death or major permanent loss of function associated with a health care-acquired infections:
Question: If this is not a new requirement, why make it a National Patient Safety Goal?
Answer: Even though the requirement for root-cause analysis (RCA) in response to an unanticipated death or major permanent loss of function is not new, many cases that meet this definition have not been considered sentinel events — apparently because infection was associated with the outcome. In other words, there has been an assumption that the presence of infection excludes a case from consideration as a sentinel event. This is not, and never has been, an intended exclusion. As a result, there are very few cases of infection-associated sentinel events in the Sentinel Event Database (in relation to other types of sentinel events and to the number of infection-associated cases reported to be occurring annually). JCAHO believes that managing these cases as sentinel events will provide additional information, not so much about the infection itself, but about managing patients at risk for infection and who have acquired an infection. In this manner, the new goal, while not necessarily a new requirement, will contribute to reducing the risk of patient harm from health care-associated infection.
Question: Many patients who die with nosocomial infections are very sick and may have multiple other problems. How do we determine whether the patient’s death was "unanticipated"?
Answer: This determination is based on the condition of the patient at the time of admission to the organization. A death or major permanent loss of function should be considered a sentinel event if the outcome was not the result of the natural course of the patient’s illness or underlying condition(s) that existed at the time of admission. For example, an otherwise healthy patient who is admitted for an elective procedure, develops a wound infection, becomes septic, and dies should be considered a sentinel event. However, cases in which the patient is immunocompromised or elderly with multiple co-morbidities are more difficult to classify.
The knowledge that a certain percentage of patients with a given condition will die does not mean that the death of any one of these patients is "anticipated." If, at the time of admission, the patient’s condition is such that he or she has a high likelihood of not surviving the episode of care (e.g., the hospitalization), then that patient’s death would not be considered a sentinel event. Otherwise, it should be managed as a sentinel event, that is, an RCA should be conducted.
Question: How should I go about doing an RCA on an infection?
Answer: Just as the identification of an occurrence as a sentinel event is not dependent on whether the patient did or did not have an infection, the RCA we are looking for is not just an analysis of the infection (if there was one), but of the event itself; i.e., why did the patient die or suffer major permanent loss of function? It is anticipated that this analysis will identify system and process factors that through appropriate redesign can reduce the risk of serious adverse patient outcomes even as the risk of nosocomial infection remains high.
Question: Where is the evidence that an RCA will help reduce the risk of health care-acquired infections?
Answer: The efficacy of an RCA to identify system failures and thus direct improvement has been convincingly demonstrated over the past several decades in most high-risk fields and, more recently, in health care for the broad array of serious adverse events that occur. While it is true that the effectiveness of an RCA specifically for reducing harm from nosocomial infections has not been proven, that only may be because it hasn’t been given an adequate chance with this specific type of event. Nor has the traditional rate-based approach, by itself, been sufficient. Perhaps, a combined approach might move us further along.