Home Patient Safety What should you do for AMA patients?

What should you do for AMA patients?

January 2014

Published in the January 2014 issue of Today’s Hospitalist

THE DILEMMA OF PATIENTS LEAVING THE HOSPITAL against medical advice (AMA) has prompted a lot of research, with good reason: While only between 1% and 2% of hospital patients leave AMA, they are at higher risk of poor outcomes.

But while most studies have focused on why patients leave, few have looked at what physicians do in such situations. A new study published in the October Journal of Hospital Medicine tries to parse out what transpires before and after AMA self-discharges and identify opportunities for hospitalists to improve patient care.

The retrospective study, led by Jason Edwards, MD, chief internal medicine resident at Wright State University in Dayton, Ohio, looked at the nearly 300 medicine patients who over the course of two years left a university-affiliated tertiary care center AMA. Only 56% of the associated charts included a physician note mentioning an AMA discussion with the patient “even though a warning about an impending AMA discharge occurred in 74% of cases.

Only 37% of records documented the patient’s decision-making capacity, while documentation rates for follow-up plans and discharge medications came in at only 31% and 24% respectively. But when physicians did document the AMA discharge, they were more likely to prescribe medications than when the AMA was not documented (36% vs. 10%).

Dr. Edwards and his colleagues did uncover opportunities to improve how hospitalists can manage impending AMA situations. He spoke with Today’s Hospitalist.

Why did you do the study?
A lot of research has been done on the risk factors for patients leaving AMA, but not on what we are actively doing for these patients. I wanted to figure out what kinds of transition-intervention opportunities exist.

What surprised you most about the findings?
I expected to see better documentation from physicians on their reasoning for why they would or would not give medications, for example, or why a follow-up appointment was or wasn’t made. But only 56% of AMA encounters even included a physician note talking about the AMA or the time of the discharge.

I don’t know why the documentation was so poor. Is it because physicians think they don’t have the same responsibility to these AMA patients as they do to typical discharge patients? Do physicians feel they can’t safely prescribe medications to these patients, or are they worried about being held liable when prescribing to people who aren’t completing their course of care?

What are some dynamics that come into play?
It’s interesting that there’s so much variation in what physicians actually do in these situations. That’s especially true when you consider that physicians are better protected from litigation in AMA cases when they document elements such as the patient’s mental status, health literacy and informed consent.

Some physicians decide to go ahead and prescribe medications, yet others don’t. Of course, because an AMA discharge is “against” the provider’s preferred advice, some may think it illogical to offer “second-best” advice. But second-best therapy may be better than no therapy, and some follow-up plan is better than no follow-up plan.

When I presented this paper at a conference, several people mentioned that their organizations prohibit them from prescribing to AMA patients, so that could be a factor. Or maybe some physicians withhold medications as a possible way to cajole patients into staying. It’s also possible that provider perceptions and attitudes about AMA patients influence their interventions.

In fact, it is legally and ethically defensible to do things like potentially allowing patients to have prescriptions that are low-risk and high-benefit, like antibiotics. That’s true even for patients potentially leaving AMA, as long as we have documentation in place saying that we’ve gone over the risks and benefits.

What should physicians document when patients leave AMA?
They should address the basics: Does the patient have decision-making capacity and health literacy? Are they in their right state of mind? Those are the assessments doctors should make “and document “if they want to have the best medical-legal protection.

It’s also important to discuss the central risks the patient is incurring in leaving the hospital and the potential benefits of staying “and to document that discussion. We do know that in almost every AMA case, it comes down to some kind of communication issue: Patients might not understand why they’re still in the hospital or why the physician doesn’t want them to leave. They might not know, for example, that we want to make sure their hemoglobin is stable before they go.

What about the “impending-AMA” window you found?
We were somewhat surprised to find that three-quarters of the patients had warned they would leave. One thing we didn’t do is look into the charts to find out when that first warning occurred. Was it only an hour before the patient left or several hours?

About 25% of patients just left. But with some, there may have been one or more hours between the time they threatened to leave and they actually left. If patients were willing to stay around and sign paperwork, there definitely is that time to do some kind of intervention.

Of course, not every physician can drop everything and go deal with just that patient. And physicians may feel that “if the patient doesn’t want to listen to me anyway, why do I want to spend my time on this?” Partly because physicians have large patient loads and because they must address this issue in a time-sensitive manner, I think many physicians find such cases irritating.

What kinds of interventions and protocols could hospitalists pursue?
Find out why patients want to leave and what’s keeping them from completing the care course, then see if that can be addressed with an intervention. I have had a few patients who cited that “nobody is there to feed my cat” as the reason they wanted to leave. Using social workers, discharge planners or other staff to work through those issues so patients can focus on their own health would definitely be part of this.

And instead of just leaving it up to individual physicians to decide what to do, facilities could develop a protocol for how to respond when a patient threatens to leave AMA. Hospitals might also create an “early-discharge” checklist to document what has been discussed and done.

That extra planning might create an opportunity to convert some AMA cases to a traditional discharge. It might also help us get rid of this “against medical advice” label and get on board with a more patient-centered approach, accepting patients’ decisions to decline inpatient care if they have the capacity to make that decision.

Bonnie Darves is a freelance health care writer based in Seattle.