How to write a good discharge summary

Primary care physicians have some feedback for you. It's about those discharge summaries.

A total of 20 physicians who practice outpatient general internal or family medicine in California participated in a recent study. They were interviewed about their experiences with discharge summaries and their perspectives on some commonly used summary templates.

The physicians had a lot of ideas about how to improve discharge summaries, from how long they should be (spoiler alert: shorter) to which details need to be included more often and how trainees' summaries could be improved. The results of the interviews were published by the Journal of General Internal Medicine on Nov. 27.

ACP Hospitalist recently discussed the study with lead study author and ACP Member Brittany Chatterton, MD, MAS, an associate physician in the department of internal medicine at the University of California, Davis, Medical Center.

Q: What motivated this study?

A: I work as a primary care physician and in my own practice have received a variety of different discharge summaries with different templates, different information added. Sometimes things included are quite helpful to caring for the patient, other times information I need is not included. I was interested to see if other primary care physicians have had similar experiences and what literature was available looking at the perspective of the primary care physician as an end user of this document. I found some survey data, but not a lot in the literature really. That's what motivated me to design a study and talk to different practicing primary care physicians to get their lived experience with discharge summaries and how we could maybe improve the quality of them.

Q: What stood out to you about your findings? Were there any surprises?

A: It was interesting how universal the concern was around incidental findings that occur in the hospitalization. With all of my interviewees, there was a big concern around things getting missed that may have been uncovered during the hospitalization but weren't the acute problem that was being managed. That information may or may not get into the discharge summary, but from the perspective of a primary care physician, thinking about the longitudinal care of the patient, those things are very important, and so a mechanism to clearly communicate that is something that all of them really echoed wanting to have.

Q: Your respondents also mentioned not wanting to be overwhelmed with data. How should discharging physicians balance providing enough information with not providing too much?

A: I 100% acknowledge that discharge summary writing is a very difficult skill. It is very challenging to take a hospitalization that can be anywhere from a couple of days to a couple months, in some cases, and distill it down into a single document. In terms of trying to balance the information needed versus what we in primary care would think is more extraneous information, a central theme that we found is imagine yourself as the primary care doctor. Put yourself in their shoes. What information are they going to absolutely need to continue caring for this patient? If you're in that mindset, incidental findings that are going to need follow-up in three or six months would be very important versus more detailed, nuanced things that happened during the hospitalization. I included as one example how the patient's oxygen level changed throughout the whole hospitalization—that, to the primary care physician, is not information that they're going to be able to utilize in caring for the patient. They just want to know, does the patient need oxygen or not?

Q: Trainees' discharge summaries came up as a particular issue in your study. Do you have thoughts on how they could be improved?

A: It is a challenge with trainees because I think sometimes the trainee assumption is, “More is better, so if I just put everything in, that would be best.” How to help train the trainee in thinking about the discharge summary is to provide feedback. That can be challenging in a busy hospital service, and once that patient has been discharged, the attention is on the patients that are currently admitted. But trying to set some time aside to actually look at the discharge summaries and provide some feedback on what could have been edited out or different things that should have been included, as the attendings or more senior trainees, could be an easy first step.

Q: How can electronic health records help or hinder creation of good discharge summaries?

A: Obviously, note templates are something that everybody uses that can be really helpful, because they can template in information that's going to be necessary. I think, though, that sometimes they can template in too much information. For example, if a patient had chest X-rays multiple days in a row, and you asked the template to pull in all radiologic studies, that's going to give you a recurrent chest X-ray every single day. As a primary care doctor, I don't need to know the daily X-ray findings, just what was initially found and how it differs at the time of discharge. So I think templates can help, because you can build [them] to pull in information that you need. But then you also have to be careful that the template doesn't overload the entire document.

Q: Medications came up a lot in your interviews. What advice would you give those writing discharge summaries on providing the right information on medications?

A: Medications are always that sticking point, right? They're so important, and a lot of changes can happen during the hospitalization. An easy thing that a writer of a discharge summary can do is, on medications the patient's started prior to the hospitalization and now are going to be told “don't continue,” very briefly explaining why it was stopped. That can be so powerful and helpful. It seems rather simple but conveys so much information. It gives a cognitive pause to the person writing it—do they really want to stop this medication? And then it helps the primary care doctor to know the reasoning behind it, so that they'll know later down the line was the reason that they stopped this medication because they thought they had an allergy to it? Can I utilize this in my wheelhouse still?

Then the other medication recommendation was around the durations of the medications. That's super helpful for medications that are more high risk like opiate medications. Help that primary care doctor understand what was the expectation that was set with the patient? Is this something we're continuing past the hospitalization that they're going to get a refill from me for? Or is this just a taper and they're off? Just having a little more explanation can help tremendously in primary care.

Q: Do you have any other tips for hospitalists about improving discharge summaries, from the study or your own experience?

A: An agreed-upon template that everyone's using is incredibly helpful. Because you can customize note templates, you can get a variety of different styles that can come out of that. Having one template that everyone's using can be really helpful to make sure that everyone's including the same type of information in the same way so that there's consistency. This is helpful on the primary care side so that you know where you're going to look in this template to get the information you need. I remember one of the interviewees said, “I would love it if you could universally in every hospital have us all doing the same thing [with discharge summaries].” But on the institution level, if they can do it within their own hospitalist group, that would be really helpful.