Critical illness events on wards may be ‘contagious'

Does focusing on one patient who's very, very sick have an impact on other patients on the same ward?


When one hospitalized patient becomes very ill and needs to be transferred to the ICU or treated for cardiac arrest, the impact can be felt by many patients.

Such were the findings of a recent study, published as a research letter in the Dec. 27, 2016, JAMA. After assessing more than 80,000 consecutive admissions to 13 medical-surgical wards at the University of Chicago Medicine from 2009 to 2013, researchers found that an ICU transfer or cardiac arrest increased the likelihood of another transfer or arrest happening on the same ward within the next six hours.

Having multiple critical events occur on a ward increased the likelihood of additional events even more. Events were also associated with a slight decrease in discharges of other patients within the next six hours.

Dr Churpek
Dr. Churpek

These results show evidence of a potential “contagious risk effect” in hospital wards, said senior author Matthew Churpek, MD, MPH, PhD, an assistant professor in the section of pulmonary and critical care medicine at University of Chicago Hospitals. He recently spoke with ACP Hospitalist about the implications of these findings.

Q: What got you interested in studying this issue?

A: I think that we all realized that when somebody in the regular general ward gets even more acutely ill, to the point where they either need to go to the ICU or they're getting close to suffering a cardiac arrest, those events take up a lot of person power, a lot of time, a lot of energy. It seemed like if we're all focusing on this one patient who's very, very sick and needs us all to be there, it is important to ask: What's going on with all the other patients?

Q: What did you think of your results?

A: We didn't know whether this would be a huge effect or a small effect, but our hypothesis was that we would find something, that there would be some increased risk. A few things, to me, suggest that this is a real effect and we're on to something here. I think the first is that when one event happened, the risk went up by 18%, but when more than one event happened, the odds increased by 53%. We also found that when these events happened, that the other patients on the ward are a bit less likely to be discharged from the hospital. And although we weren't powered for this finding, it did seem like at nighttime, this risk may be a bit greater for the patients. That goes along with the [idea] that if you have fewer people around or maybe less staffing at nighttime, then this effect could be greater. I certainly think if a hospital has decreased staffing at nighttime or decreased staffing on weekends or even holidays, those are definitely potential danger zones where this effect could be more pronounced.

Q: What are some possible explanations for these findings?

A: One explanation for our findings is that when you're spending all of this time taking care of this one patient, that some things could potentially slip through the cracks with your other patients, and that may then increase the other patients' risk of having something bad happen. Another potential explanation would be that...maybe [physicians] get sensitized to these events when they happen, so they're more likely to send people [to the ICU] a little bit early. I think there are a few different explanations, and we're doing some further work to see if we can figure out exactly what the most common reason is for these events.

Q: Do you think that these results should make hospitalists care for patients on the wards any differently?

A: I think on the one hand, these events (particularly things like cardiac arrest) are relatively uncommon, which is good...but there are some things for the primary team to think about. When you see a patient who's really sick on your team and you care for them, I think once that event's over, it's important to talk to your team and talk to your nurses and say, “OK, so just to make sure that these other things we were supposed to do didn't slip through the cracks, let's check on Mr. Jones or Mrs. Smith and...a couple other patients on our list who are fairly sick.”

Something that we're doing here at the University of Chicago is that when our rapid response team sees a patient on the wards who is really sick and may need to go to the ICU, once they're done seeing that patient, they also will come out and talk to the charge nurse and say, “Who else on this ward are you worried about?” and see if there are other people on the ward that they can go see who they weren't necessarily called about. We didn't really start doing that until after [collecting] the data from this particular study, so one of the other things that we are interested in doing is to see if practices like that can maybe decrease this effect and improve outcomes.

Q: What's next for your research in the area?

A: One of the most important things, in terms of next steps, is to try to determine “Why is this happening?” We have our own hypothesis, but I think until we know exactly why this is happening and what are all the different contributing factors, we can't really design the best interventions. Once we figure out what the most common causes are, then we can do the next step, which is to design an intervention to try to see if we can prevent it from happening as much as we can. We're certainly trying to do what we can to try to mitigate it but also, at the same time, trying to study it so we can create some generalizable knowledge for other hospitals that could potentially have the same problem.