The Joint Commission Update for Infection Control

'The numbers don't match': Joint Commission urges more infection-related sentinel event reporting

Leading IP cites caveats that may explain discrepancy

Citing a dramatic disconnect between the tens of thousands of patients dying annually with health care-associated infections (HAIs) and the paltry number that actually are being reported as sentinel events, The Joint Commission is urging hospitals to file the voluntary reports to help improve patient safety.

"The question here is — we've got a lot of data, we've got lot of sentinel events — where are the infections?" said Louise M. Kuhny, RN, MPH, MBA, CIC, senior associate director of standards interpretation at the Joint Commission. "The CDC is saying we are having 90,000 infection [deaths] a year, the mortality is high, we have all these problems; where are the infections? We would like to know because it doesn't seem that they are being reported through the sentinel event database."

At a recent Joint Commission meeting in Chicago, Kuhny revealed that of the 4,977 sentinel events reported from 1995 through March 2008, only 104 were infection-related. "We certainly know that there were way more than 104 infection-related events in 12-13 years," she said. "We are encouraging this reporting."

The most recent data only underscore the trend of underreporting, as The Joint Commission received fewer than 15 sentinel event reports related to infection in 2007.

"We're concerned . . . because the numbers don't match," Kuhny said. "We're not getting the reporting. We're not getting [reports of] noncompliance with this, either. [Hospitals] probably have all the systems in place to report, yet we know that all of this morbidity and mortality is [occurring]. In 2007, we didn't even have 15, and we know that there are tens of thousands out there."

Indeed, according to the most recent published data, the Centers for Disease Control and Prevention estimates that 5%-10% of hospitalized patients develop an HAI, corresponding to approximately 2 million HAIs associated with nearly 100,000 deaths each year in U.S. hospitals.1

Patient safety goal established in 2004

To be fair, it was not until issuance of its 2004 patient safety goals that the Joint Commission officially called for unanticipated patient deaths and serious injuries related to HAIs to be investigated as sentinel events requiring a root-cause analysis (RCA). NPSG.07.02.01 — which remains in place for 2009 — calls for hospitals to:

"Manage as sentinel events all identified cases of unanticipated death or major permanent loss of function related to a health care-associated infection.

Rationale: A significant percentage of patients who unexpectedly die or suffer major permanent loss of function have health care-associated infections. These unanticipated deaths and injuries meet the definition of a sentinel event and, therefore, are required to undergo an RCA. The RCA should attempt to answer the following questions: Why did the patient acquire an infection? Why did the patient die or suffer permanent loss of function?

Elements of Performance: The hospital manages all identified cases of unanticipated death or major permanent loss of function associated with a health care-associated infection as sentinel events (that is, the hospital conducts an RCA). The root-cause analysis addresses the management of the patient before and after the identification of infection."

How often are HAI events unanticipated?

The "unanticipated" aspect of the definition may be part of the problem, as patients being kept alive by invasive devices may certainly have an HAI among their end-term sequela. In addition, reporting the RCA results is voluntary, but strongly encouraged to identify trends and improve patient safety. "It is a voluntary reporting system, so not all sentinel events of any kind must be reported," said Denise Murphy, RN, MPH, CIC, vice president and chief safety and quality officer at Barnes-Jewish Hospital in St. Louis. "I think hospitals report those that are most serious in terms of how they happened. All sentinel events mean harm, but those that can educate us most in terms of process breakdowns to look out for are what most executive teams would demand be reported."

A frequent lecturer at infection prevention meetings, Murphy has urged IPs to embrace the Joint Commission initiative and conduct RCAs as warranted. "HAIs that are unexpected or 'unanticipated' are what we should be counting," she said in a separate interview after The Joint Commission meeting. "We do that here and do a sentinel event investigation — including a debriefing as soon as we learn of the event — followed by a root-cause analysis. The 'unanticipated' aspect is why you don't see 100,000 HAIs being reported to The Joint Commission."

For example, the death or injury of an ICU patient who develops ventilator-associated pneumonia (VAP) despite the use of infection prevention "bundles" and other cutting-edge interventions is tragic but not completely unanticipated. "[However,] if you have a 30-year-old patient undergo elective surgery, not wake up after anesthesia, require mechanical ventilation and then get a VAP, this was not anticipated or expected and should be debriefed and followed up with an RCA," Murphy explains. "The RCA would look into all aspects of patient safety and why such an adverse event occurred. In fact though, it might be written up as an unanticipated surgical outcome [instead of] the VAP."

By the same token, if a 30-year-old patient undergoes an elective knee replacement due to a sports injury — as opposed to a serious underlying illness such as juvenile diabetes — and then develops a surgical-site infection (SSI), that could be considered a sentinel event, she notes.

"The SSI was not anticipated in a healthy 30-year-old," Murphy says. "Now, even there, a superficial SSI would not be counted. But if the 30-year-old came into the hospital for three incision-and-drainage surgeries, had months of antibiotic therapy, then lost the prosthetic knee due to infection — there is your sentinel event."

Caveats and cost benefits

Thus, given Murphy's points, there is more to the picture than the jarring juxtaposition of the numbers.

"I am not trying to make excuses — HAIs are always horrible outcomes," Murphy says. "But they are not always unanticipated. The patient's underlying conditions at the time of admission or time of a procedure will often help dictate whether or not this was a totally unanticipated outcome that led to death or permanent disability — or risk thereof. The 'risk thereof' is where you could end up investigating every thing as a sentinel event."

In that sense the "cost benefit" of doing an RCA must be considered. "I don't mean dollar cost," she says. "I mean that infection prevention and control programs have finite resources, and we have to decide every day how to best 'spend them.' Taking the time to educate health care teams, assess the safety of our patient care processes, help teams redesign patient care, and build in prevention is where I'd put most of our resources. And I'd use those IP resources to debrief and help risk management do a root-cause analysis every time an HAI was unanticipated and led to death or serious disability."


  1. Yokoe DS, Classen D. Improving patient safety through infection control: A new health care imperative. Infect Control Hosp Epidemiol 2008; 29:S3-S11.