By John Commins
Surveys of care teams after unexpected patient deaths can provide learning opportunities for hospitals, and catharses for clinicians, a new study in the Joint Commission Journal of Quality and Patient Safety suggests.
"Oftentimes if there is an error, the care team is not going to write in their notes, 'Oh we committed this error!' A lot of times the medical record is not the best source of truth for something like this," says study lead author David Lucier, MD, a hospitalist and director of Quality and Patient Safety at Massachusetts General Hospital.
"What we do know is that when patients pass away unexpectedly there is this pit-of-the-stomach feeling for providers and we were trying to get at that with this project," he says. "It's that gut feeling that maybe something could have been done differently. Is there some truth to that, and if so, how much truth? That's why we did this project, to see if we could augment the existing process that we had in place."
Lucier says that essay opportunities provided to clinicians can often be the most illuminating parts of the survey.
"The idea is to prime them to think about improvements and provide them with a space to write whatever they thought. There was no guidance on what to write. They just wrote it," he says. "In doing so, we found we had a very rich data base of written comments. Some answers were very short. Some were very, very long."
The researchers created a taxonomy from the essays to understand what the clinicians were saying and how they were saying it.
"We were interested in the 'how' in particular. Were people writing blaming comments? Did they feel they needed to be defensive in a patient's death? Were they trying to explain something with emotion because that’s how they were feeling, using descriptive words," Lucier says. "You could say 'a large volume GI bleed,' or you could say ‘the patient had a tremendous amount of bleeding, and it’s shocking how much he bled.' They both describe the same situation, but the words used relay a different provider experience of that patient’s death."
Lucier says many mortality reviews start with the premise that every death might be a preventable death. While laudable, Lucier says that premise is not necessarily congruent with the care goal of the patients, which might be to pass away comfortably at home or in the hospital surrounded by loved ones.
"What we found in the free text comments were repetitive themes about patients' goals of care and whether they had been discussed, and if the care provider was congruent with the understanding of it," Lucier says. "That was the main finding: We are not doing as good a job as we can with understanding patients' goals of care when they are coming in. These are hard conversations to have, and this is at a patient's time of need, so it's a complex issue."
Lucier says the surveys also find that frontline clinicians are not particularly good at identifying preventable deaths.
"Providers overestimate the preventability of death by almost 10 fold," he says. "While something may have happened that didn’t go right, it didn't lead to the death being preventable. If it stopped, the patient likely would have died anyway."
That problem of accurate reporting is compounded by the harried pace of hospital work.
"This is a big problem in safety reporting in general. We rely on the frontline to tell us when something goes on," Lucier says. "The problem is that it's tough to report something when you have a million other tasks at hand, all these patients' demands that are far more important than submitting an adverse event report."
Still, the surveys exposed vulnerabilities in care coordination that would have otherwise gone unnoticed, Lucier says.
"For example, from this we were able to look at the massive GI bleeding pathway, which is a complex dance between multiple services, and look at the drip medications in the critical care units and rework those," he says. "These vulnerabilities were not brought up in any other system that we had within quality and safety to identify vulnerabilities and make improvements."
While the cathartic effect of the survey on clinicians was not examined in his study, Lucier believes that it could also provide an outlet.
"If you have a patient die expectantly, it weighs on you. It makes you question what you did and the care you provided. All of us have this moment where 'did we do something wrong?' Over time that continuous questioning can lead to burnout," he says.
"Providers compartmentalize any emotional distress they might have from an unexpected death. This project was the first step in trying to identify that and provide an outlet to dissipate that emotional distress."