CRICO MDs ONLY: Register to access your facesheet, and more.

Advanced Search

  • Topic
  • Specialty
  • Content Type


Also Related

< Back To Patient Safety
0 dislikes

< Hide

Comments For

Patient Safety Findings: Part 1: Teams


< Shrink

Add Your Voice

All comments are posted anonymously. Your comment will be attributed to: "Anonymous user."

post comment


Are you sure you want to delete this comment?

Patient Safety Findings: Part 1: Teams

By Tom A. Augello, CRICO

Related to: Communication, Medication, Nursing, Surgery, Teamwork Training

Can analysis of how teams fail, including flaws in patient hand-offs, help prioritize interventions?


  • Allen Kachalia, MD; Brigham and Women’s Hospital; Boston, MA
  • Carol Markus, RN; Massachusetts General Hospital; Boston, MA


Whether handing off information about a hospitalized patient to a new shift of clinicians, or keeping track of test results in the office practice in pursuit of a cancer diagnosis, new ideas for patient safety found a receptive audience at the 2nd Annual Patient Safety Grant Symposium sponsored by CRICO/RMF in the Harvard medical system. For three years in a row, the malpractice insurer has disbursed a half-million dollars to 10 research teams to study ways to improve communication, teamwork and other aspects of patient safety. In this special report: some of those researchers share early views of their findings with their clinical colleagues in October 2006 at the Harvard Club in Boston.

“I’m really very happy to present our progress to date on this collaborative study between MIT and Mass General...”

Carol Markus is a nurse at Massachusetts General Hospital in Boston who helped lead a team that looked at handoffs of clinical responsibility for patients: understanding the characteristics—and the adequacy—of the communication.

“I mean, there certainly is a lot in the literature right now about the importance of effective communication and transitions between caregivers around the status of patients and what their active problems are.”

In their study, Markus and her colleagues clipped a microphone to nurses on a general medical unit to record the transitions. Then they compared the information those providers shared to the actual medical record. They observed 30 shift reports, featuring two to three handoffs each that could be recorded.

“What we were looking for when we were looking for the discussion of the active clinical problems, was not only was the problem mentioned, but was there a discussion about the management and the next steps? And that’s where sometimes we found a disconnect, that there might be a mention of an active clinical problem, but maybe not a discussion about what was being done for the management or the next steps around this.”

The Joint Commission now requires hospitals to have a standardized approach to handoffs. At Mass. General, the process includes a mention of every major clinical issue attached to the patient. Of the 263 active issues that the researchers identified by looking at the medical records, a third were not mentioned in the audio recordings. In nearly 25 percent of the problems that were identified during the handoffs, “next steps” were not mentioned.

Markus offered an example:

“From the medical record we were able to see for this particular patient that... take the first problem… he was intubated with sepsis, acute respiratory distress for two days. And the management as discussed from the medical record, it had lung assessment, suction, O2 and the next steps were monitoring and suctioning.  So, those came from the medical record.  In the handoff from the transcript those things were present.  So, we felt that this was an adequate handoff.  If you look at example two and you look at the bottom one in the record it will tell you there is no bowel movement and what the management in the next steps is, but in the handoff there was no mention of the problem.  So, that would be inadequate.  And then if you look at the somnolent, high LFT and you can see what was in the patient record around the management and the next steps the handoff was insufficient because there was missing information about some of it.  There was some information given, but not complete information given.”

Even though the complexity of the observational approach is limiting, the preliminary results suggest that studying handoffs at change of shift does offer valuable insight into behaviors that influence outcomes.

“Well, as we know, healthcare is often delivered by a team and it is this team structure that brings its own requirements, the first of which is communication.”

Dr. Allen Kachalia of Brigham and Women’s Hospital in Boston, gave Harvard’s patient safety and quality improvement leaders a run-down of his team’s research into failures in communication and teamwork. Dr. Kachalia used his CRICO/RMF grant to examine sentinel events and other incidents investigated by risk management offices at three institutions to try to isolate where improvements in team structure could lead to safer care for patients.

“Clearly, when you have multiple people interacting they need to be able to talk to each other. This includes doctors to doctors, doctors to nurses, nurses to techs, lab personnel and families. You name it, everyone needs to talk. But it is not all just about information transfer. It is also in terms of how well they interact with one another. For example, people’s personalities have to be accounted for.  There has to be minimal conflict on the team for a team to operate well and there has to be enough time for folks to build a relationship so they can communicate as effectively as possible.”

The researchers reviewed 452 records of root cause analyses or other investigations by risk management departments from 2001 to 2005, to identify preventable adverse events. By identifying cases where both an error occurred and the patient was harmed by medical management, the researchers found that 37 percent involved adverse events that were preventable.

“In terms of trying to grade the severity of the injury, six percent were insignificant—and  by insignificant this was meant to include things like small burns or blisters or temporary hypoglycemia requiring insulin. 32 percent of them are what we called minor.  These were items like people getting the wrong side biopsy, this would be a minor biopsy obviously or somebody that got a little somnolent secondary to narcotics and needed a little bit more observation.  50 percent are what we found to be major events, and these were major in the sense that someone either could have had a stroke or brain hemorrhage or needed a second surgery as a result of an error.  Four percent were grave, and grave included things like anoxic brain injury obviously; and eight percent was death, which is self-explanatory.”

The study considered specific individual, teamwork, and communication-related factors that contributed to the 168 preventable adverse events. The most common team factor was team design or organization breakdown, such as coordination of responsibilities, redundancy, supervision, etc. The second-highest factor was a breakdown in mental modeling.

“This is quite simply put where we wanted to see if people ever got their signals crossed and to give you a clinical example you can take me for example.  Let’s say I’m rounding on the floor with my resident.  We see a patient who we think has community acquired pneumonia.  I may look at my resident and say okay, well why don’t we go ahead and treat this patient with the appropriate antibiotics and leave it at that.  I figure my resident is either going to start something like ceftriaxone and Zithro or maybe do Levaquin or something like this, but of course I come back later in the day and they have picked a regimen totally different than mine.  The resident and I were clearly not on the same page.  We didn’t have the same mental model.”

Other team factors were ‘conflict,’ and a failure of team members to properly assert their concerns, which accounted for relatively few of the preventable adverse events. Dr. Kachalia said this method of reviewing risk management investigations for team factors can help prioritize interventions. Yet he said it’s important to note that a key finding was that team-level interventions won’t solve everything.

“71 percent of these preventable events actually had an individual factor contribution and 34 percent of these events had no teamwork contribution.  Despite being in a teaching environment, most were felt to be not related to experience or skill and the big question here becomes how much should we adjust our team structure to capture these individual errors.  I think the classic example is with blood transfusions. We often have people do double checks and we have to confirm with somebody before they transfuse blood and that’s been a team fixed to an individual issue.”

Dr. Kachalia concluded that the high percentage of preventable adverse events involving team breakdowns suggests prioritizing interventions that increase coordination of team member responsibilities, improving redundancy, and increased supervision.

“With regard to mental modeling, we can see that we need more emphasis on making sure that the team members communicate because often they weren’t communicating, but there were many cases in which they were communicating.  I think we need to start paying more attention to exactly what they are communicating.  We also need to perhaps encourage improved comfort with voicing common pitfalls and key concerns that arise.”

November 1, 2006
0 dislikes

< Back To Patient Safety