Feb. 11, 2008

In Print

Spotlight

Upcoming

Student Scene

StudenTalk

Lab Works

Science Progress

Home

Student Scene

A Look at the Cost of Interrupted Care

Graham Ramsay Erica Seiguer Shenoy

Working in the emergency department one evening I pick up the chart of a man in his mid-50s whose chief complaint is abdominal pain. I scan the intake sheet for the vitals and then briefly look over what is available in his electronic medical record—a long history of polysubstance abuse, HIV, and hepatitis C. I go into the bay where he is sprawled out on the stretcher, his shirt unbuttoned; he is sleeping. I introduce myself, waking him up, and it is clear he is intoxicated. He begins to moan in pain, clutching his abdomen.

I begin to ask questions to fill in the history, but he refuses to answer most of them. He will barely let me examine him. He begins to bargain with me, demanding narcotics before he will participate in the history or the physical examination. I notice that on his wrist is a hospital band from another Boston area hospital, and it appears that he’s recently had a paracentesis, since the marks on his abdomen show. From the little I can garner from the exam he will allow, his vitals are stable and his abdomen, while distended, is still soft.

Back at the computer, I write for a variety of basic labs, including a toxicology screen; speak with the nurse about the plan (get some basic labs, abdominal ultrasound, possible surgery consult); and move on to the next patient who has arrived, as well as check on the patients who are already in progress.

A while later, the nurse comes to tell me that this patient is not only refusing the ultrasound, but any labs. He has told the nurse that he wants narcotics before he will consent to any tests. I have reason to be circumspect about his complaints of pain: when I walk by his room without his knowing it, I see him lying comfortably on the stretcher, yawning. His vitals show a normal heart rate, blood pressure, and respiratory rate. As soon as I enter, he begins to groan in pain, yet his vitals are stable. I explain that I do not feel comfortable prescribing any medication to him, narcotics or otherwise, unless I have a better idea of what is going on—that is, until we have some basic labs and the ultrasound. After some back and forth, he consents, and I let him know the nurse will be right in.

Minutes later, the nurse comes to me again and says that the patient has refused. I again return to speak with him, and we come to the same agreement. Again, the nurse returns, and the patient refuses. We do this dance three or four times. At this point, I speak with my attending, who has seen the patient as well. I’ve also asked the ED surgeons to check on his abdomen. After they clear him, the question facing us is what to do about a patient who refuses all medical intervention (except narcotic medications) but who probably would benefit from admission, possibly a repeat paracentesis, not to mention a determination that he’s on appropriate therapy for his comorbid conditions.

I’m occupied with taking care of other patients, and we decide just to observe our noncompliant patient, checking his vitals, and letting him sleep. Hours later, I’m approached by his nurse who says that he wants to leave. Because he will be leaving before he has had a thorough evaluation, he will need to leave AMA, or “against medical advice.” I discuss the situation with the patient, informing him of our concerns regarding his health, and he still wants to leave. First, however, I use a breathalyzer to ensure that he’s not currently intoxicated, and sure enough, he has sobered up over the many hours he’s occupied one of our ED bays. He is competent to make the decision of whether to stay or go. He signs the paperwork and exits the building.

What happens to patients who leave AMA? This patient certainly wasn’t the first patient I’ve asked to sign the legal paperwork required when a patient leaves against medical advice. A 2003 study tracked patients who left AMA from an urban hospital and found that they were much more likely to be readmitted to the hospital within 15 days than controls (21 percent versus 3 percent). Male sex and a history of alcohol abuse were also independent predictors of readmission. Across the U.S., AMA discharges account for about 1.4 percent of all discharges from medical services. While patients leaving AMA tend to have lower costs due to shorter hospital stays, because they are readmitted at a much greater rate than other patients, their overall costs to the health care system may be much higher.

—Erica Seiguer Shenoy, MD–PhD ’07, is an intern in internal medicine at Massachusetts General Hospital.

top

Copyright 2008 by the President and Fellows of Harvard College