Physicians identify many of their own admissions as preventable

Differences in risk assessment and training between ED and medicine were reported to be factors.

To admit or not to admit? Many times, the answer is unclear for hospitalists and ED physicians.

To better understand the rates of and factors influencing potentially preventable admissions, researchers at Mayo Clinic in Rochester, Minn., interviewed 82 physicians involved in 401 admissions from the ED to the general medicine service over a four-week period.

Dr Sawatsky
Dr. Sawatsky

Physicians categorized 22.2% of the admissions as potentially preventable and discussed several factors that affected clinical decision making around those admissions, including differences in risk assessment and training between ED and medicine, with results published in the May 2018 Journal of General Internal Medicine.

Senior author Adam P. Sawatsky, MD, MS, FACP, assistant professor of medicine at Mayo Clinic, recently spoke with ACP Hospitalist about the findings.

Q: What led you to study this issue?

A: The idea really started with the second author [ACP Member Atsushi Sorita, MD] ... wanting to drill down into studying factors that have an impact on health care costs at our institution and nationally. One of the things that he hit on through discussions with a multidisciplinary team was this idea of potentially preventable admissions. ...I think we all had experience, even from our different perspectives, of patients who were unnecessarily admitted—and not just the cost to our system, but the frustrations from the patient perspective, the frustrations from the provider perspective, and the lack of clear communication between the different provider groups.

Q: What were some of the most striking differences you found between the responses of general medicine physicians versus emergency medicine physicians?

A: Right off the bat, it was clear that emergency medicine physicians and general medicine physicians or hospitalists who were admitting these patients have completely different roles and perspectives. Participants talked a lot about how ER doctors' job is to stabilize and triage, whereas general medicine physicians' roles are to diagnose and manage.

Where it got to be more interesting was the discussion around the risks of hospitalization. We labeled this the “gray zone,” which was taken from one of the quotes, where it wasn't clear whether a patient should be admitted or not. ER doctors tend to view the greatest risk being sending this patient home and having them decompensate at home, whereas hospitalists and general medicine physicians saw and viewed the risks of admitting this patient to the hospital, as far as potential hospital infections, risks of hospital delirium, and the cost—not only to the system but also to the patient—of hospitalization.

Q: Were there differences between patients in the “gray zone” and patients who were admitted?

A: We looked at things like the Charlson Comorbidity Index and the [Eastern Cooperative Oncology Group performance status] scores, and there actually was not a significant difference between patients that we had labeled as potentially preventable and patients that we had labeled as control admissions, interestingly. We also looked at factors like time of day they arrived in the emergency department, was it a weekday or a weekend, and other factors to see if there were other patterns that we could find in how these patients present, and [there was] no real difference between them that would predict what would be a potentially preventable admission. This seems surprising, but the interesting thing is other studies have shown the same thing: that there's no great predictor, no model that we can have that would predict potentially preventable admissions, and the standard is still a physician determination about whether or not an admission is or isn't preventable.

Q: How can emergency medicine doctors and hospitalists better work together when making “gray zone” admission decisions?

A: One of the things that I think was clear from both provider groups, from both the ER physician as well as the admitting medicine physician, was that discussions around patients in the gray zone between the two groups are welcomed. An emergency medicine doctor talked about...[how] ER physicians prefer to have the general medicine physicians provide their expertise in real time around these patients within the gray zone. One example would be...if they could consult in real time with general medicine colleagues about what's available and how to get [patients] plugged into the outpatient system.

Q: What might those real-time discussions look like?

A: At an individual level, I think those discussions can happen over the phone. For instance, one of the things that we have here at Mayo is the “doc of the day,” which makes an outpatient physician available for people to call and ask questions, and within that role, I have had ER physicians call me and say, “I have this patient, I think they're stable and I think they don't need to be admitted if I can get them in to see an outpatient general internal medicine physician to coordinate their ongoing care and workup.” And those discussions have led to us being able to do just that and prevent this patient from being admitted to the hospital.

The bigger issue then becomes what do we do from a systems standpoint? The two biggest factors were time pressure—ER doctors have a very distinct time pressure of getting patients through a busy ER—and then their discussion around competing demands. It's hard because discussions and coordination of outpatient care, for instance, for an admission that's potentially preventable, is tough within the system that we've created in our emergency departments. So I think more than just saying there needs to be more discussion ... let's take a look at the system and how we support the ER physicians in actually making some of these decisions, realizing that it may have real cost savings for the patient and for the health system.

Q: Do you have a sense of how often these discussions do happen?

A: I have seen it done very variably across a number of institutions. I have seen everything from little to no communication (meaning ER physicians have the full ability to admit patients and hospitalists just accept those admissions) to places where a general medicine physician or senior-level internal medicine resident is actually stationed in the emergency department there to help with determination of admissions. They're the ones that would do that coordination of care if it's determined that a patient does not need to be admitted but does require further care and follow-up in a timely manner.

Q: Does the latter model seem to work?

A: I think it works when it's utilized. My sense of it is it's not utilized. For instance, the “doc of the day” has a lot of other roles, and my sense is [they] don't get called enough by ER physicians. ... If time is the real issue, are there ways to take those discussions out of the ER physicians' hands, particularly when they have sick and decompensating patients that they need to attend to? Looking at the different institutions, I think a lot of it does come down to culture and how things were set up to begin with and how they've done things over time.

Q: What are your tips on how hospitalists can help prevent unnecessary admissions?

A: I feel like there is a lot of opportunity for discussion around this topic and being able to clearly communicate to our ER colleagues—understanding their assessment of risk is different than ours, but also being able to communicate our own assessment of the risks of hospitalization in certain cases. One of the things that I have seen is that it is easy for hospitalists and myself to become critical of the decisions that are made in the emergency department and frustrated about admissions that could have been preventable. And what I see is that that frustration is felt by the ER physicians when we talk about this with them, and I think it shuts down conversation.

Q: What do you hope happens in response to your findings?

A: Every institution is variable, so what I would hope to see at our institution and at other institutions is that they use this study as a starting point to open up their own conversations between emergency physicians, hospitalists, and outpatient internal medicine physicians about what this can and should look like at their own institution.... Hopefully this [study] provides some evidence and suggestions of ways forward at individual institutions and for people who are doing quality improvement in this area.