Embrace the change: ICPs should take lead on RCAs
Sentinel events linked to nosocomial infections
Rather than a dreaded, unnecessary task, a new requirement to conduct a root-cause analysis (RCA) for fatal and impairing infections is an idea whose time has come in the new era of patient safety, said Denise Murphy, RN, MPH, CIC, chief of patient safety at Barnes Jewish Hospital at Washington University Medical Center in St. Louis.
"Stop tying this stuff to anything regulatory," she said recently in Phoenix at the annual conference of the Association for Professionals in Infection Control and Epidemiology (APIC). "This is so about doing the right thing for the patient. Infection control will always be at the patient safety table. That’s why I say take a leadership role in this."
The Joint Commission on Accreditation of Healthcare Organizations caused quite a stir in the infection control community when it set a 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 infection." Sentinel events require an RCA, a process that more typically falls to risk managers or other performance improvement professionals.
Murphy concedes she was not thrilled with the new requirements when they first came out, in part because she didn’t know how to conduct an RCA and feared the process would be undercut by lack of resources. Having gotten thoroughly involved in the process, she urges other ICPs to embrace the change.
"You do not have to operate in a silo," she said. "Your organizations know how to do an RCA. There are experts in the risk management department, in the performance improvement department, who do this all the time. What you bring to the table is a unique, wonderful skill set in hospital epidemiology. You bring your brain to the table and let other people facilitate you through the process."
Murphy shared an eye-opening anecdote, saying she recently heard a Joint Commission surveyor tell an audience at another conference that many sentinel events could be prevented if clinicians simply spoke up. As Murphy explained it, a Joint Commission analysis of a data set of wrong-site surgery RCAs revealed that in "100%" of cases at least one health care worker knew something was wrong but said nothing.
"That is so frightening," she told APIC attendees. "If you remember nothing [else] from today remember the words of that Joint Commission surveyor who shared that with us. People in our organizations are afraid to call out, they’re afraid to speak up [for] fear of being punished, retaliation. . . . You are a leader in your organization in patient safety. Go back to your organization and share that information. One hundred percent of the time someone in the room said, I saw that something wrong was about to happen, and I was too intimated to call out.’"
Candidly detailing her own hospital’s experience, Murphy said two separate RCAs had to be conducted within a six-month period after wrong-site procedures occurred.
"This is in at the No. 2 medical school in the country and the No. 8 hospital in the United States," she said. "Human error is going to happen in every hospital in this country. We are not talking to each other. What we are finding out after we sit through a painful RCA where health care workers are in tears, sharing their stories: We are not telling other people [about mistakes] because we are embarrassed. So what happens? Another patient has [a wrong-site procedure] and ends up in an ICU because — in the same department — they were too humiliated to share their story."
Identifying a sentinel event
Murphy walked through a step-by-step breakdown of the process of identifying a sentinel event and then conducting an RCA, citing resources that include scenarios created by her and co-authors for APIC. (See Take-home points on sentinel events and RCAs.) A sentinel event is "an unexpected occurrence involving death or serious physical or psychological injury or risk thereof," she said.
The two key words in that definition that determine whether you do an RCA investigation are "unexpected occurrence," she emphasized.
"We know that health care-associated infections are very strongly tied to the morbidity and mortality for our patients," Murphy said.
"The area of gray is the part of the definition that says the risk thereof.’ Now, if we had to do a root-cause analysis for every patent at risk for infection, we would do nothing else. So that is not really what I believe the Joint Commission’s interpretation of this definition is."
In considering a decision to call something a sentinel event, return time and again to the "unexpected occurrence" concept. For example, bloodstream infections (BSIs) are serious infections that occur in 250,000 patients annually in U.S. hospitals.
"Of course, we are not going to do sentinel event [investigations] on all of them," Murphy continued. "But don’t panic. Think about this practically, saying, Well, which ones will we do it on?’ [Think about] healthy patients who have had easy procedures, procedures that are not usually related to the acquisition of a BSI. Or [look for] very rapid onset of BSI and sepsis. Something is definitely wrong. You guys are the best people at following your gut. . . . So try not to make this too much of a science and really move with what your gut is telling you."
Another factor to consider is the expected mortality rate associated with a type of infection. "Patients having elective surgery with few risk factors for surgical site infection (SSI) are not expected to die of SSI-related complications," she said. "This is something that sounds hard, but it is really very easy if you just take it a step at a time."
To better identify sentinel events, ICPs should work with the medical records department and get a daily list of patient deaths. "Every hospital has a team of individuals who are responsible for reviewing all mortality in your organization," she said. "They have a list every day of every patient who has died. Make the connection and get that list."
The ICP should then compare the mortality data with a list of people with known hospital infections, she advised.
If the lists intersect, work with the hospital epidemiologist or other physician leaders in this area to try and determine if the death was an unanticipated occurrence.
Overall, the root causes of sentinel events include problems with communication, training and orientation, patient assessment, and staffing levels.
"Staffing levels are not just tied to hospital-acquired infections but also to patient deaths, very serious medication errors, falls, just about every type of medical error," she said. In addition, staffing patterns or workflow, technical failures, inadequate policies, and procedures all can contribute to a sentinel event.
Conducting an RCA
Once an ICP identifies a sentinel event, an RCA must be done within 45 days of the event occurring.
"This is important because this is a time-sensitive issue," Murphy said. "If you go through a debriefing, it sometimes takes a couple of days to get people around the table. It should happen as soon as possible after the event and include the people who were involved in it — while things are still fresh in their minds. Remember, you have to act pretty quickly on this. It’s not like you have six months."
An RCA investigation is not unlike an outbreak investigation, which ICPs are very familiar with, she reminded. (See comparative lists.) First, organize the RCA team, clarifying at the outset that identification and reduction of risks is the aim. "Focus on processes, not individuals," Murphy said. "This is not a blame game."
Indeed, you certainly need to talk to the people directly involved in the incident, but staff members may be too upset about the patient outcome to actually participate on the RCA team, she added. "You need to get their information, but sometimes, they can’t sit through the RCAs," she said. "It is just too upsetting to them to sit and go through this process."
The committee should be multidisciplinary, but can’t be too big or you’ll never make the 45-day deadline, Murphy warned.
"Keep the key decision makers in the loop," she said. "They’re the owners of the resources. So many times in organizations we develop, develop — generate so many wonderful ideas. But where we fall down is on execution and getting stuff implemented."
Eventually, ICPs may want to serve as facilitators in the committee process, but for now, Murphy suggests deferring to more experienced RCA investigators in your risk management or performance improvement department.
"Focus on what happened initially, not why it happened," Murphy added. "You’re going to get to that. Verbalize it accurately and succinctly: Mrs. Jones was a 55-year-old who underwent an elective [arterial bypass surgery]. She had a cardiac arrest and died on her third post-op day. There were no signs of SSI. Sepsis was confirmed by blood cultures and autopsy. Central-line sepsis is suspected."
Study the problem and collect information that is relevant to the event. If available, physical evidence should be secured, including documents, and photographs of equipment or settings (but not people).
"This is not punitive, keep reinforcing that," Murphy said. "We are trying to track the process from beginning to end and see if there is any place where we introduce risk to our patients."
Compare the event against hospital policies and procedures, looking for gaps, she said.
"Unfortunately, you’re going to find them. Sentinel events are not happening for no reason. It’s incredible to go through the process. When you keep asking why’ you get down to the causes. You keep drilling down," Murphy noted.
Go the next step and decide whether the solution is going to create another problem, she added. "As improvements or actions come up by lots of team members, your role as the ICP is going to be to make sure they are evidence-based."
The RCA report must clarify to the Joint Commission who is responsible to enact follow-ups, and its best to hardwire this aspect into a committee to ensure the actions outlined are being completed, she said.