The patient experiences problems after discharge. What’s the responsibility of the hospitalist? Even when the patient misses a follow-up visit, you may still be liable if things go wrong by Phyllis Maguire
Published in the September 2006 issue of Today's Hospitalist
A patient admitted with chest pain has a negative work-up for coronary artery disease, but a lung nodule is found on the chest X-ray. The hospitalist makes an appointment for the patient to follow up on the nodule with his primary care physician, but the patient doesn’t keep the appointment.
A diabetic patient hospitalized with pneumonia is started on an ACE inhibitor for her hypertension. But her primary care physician doesn’t receive her discharge summary before the patient comes in for follow-up, so doesn’t know to check her renal function. Two
“What are we supposed to do—go to patients’ homes, handcuff them and drag them to their appointment?”
months later, the patient is hospitalized again, this time for hyperkalemia and acute renal failure.
And after having her CT reviewed by both the radiologist and hospitalist, a 72-year-old woman admitted for abdominal pain is diagnosed with constipation. Two months later, she contacts the hospitalist, demanding to know why she wasn’t informed of her probable renal carcinoma diagnosis. The CT report had been revised after the initial read, but no one told the hospitalist.
What do all three cases have in common, besides being real-world problems that occurred at an academic medical center? They illustrate problems that keep surfacing when care is handed off from hospitalists to primary care physicians.
In a presentation this spring at the American College of Physicians’ annual meeting in Philadelphia, hospitalist Eric Howell, MD, made it clear that this “voltage drop” in communication continues to have major potential consequences for both patient safety and physician liability.
While he had no silver bullets to offer, he did provide data showing a lot of room for improvement. He also suggested several avenues that hospitalists should take to close the biggest gaps.
Great perceptions, shaky reality
When it comes to hospitalist handoffs to primary care, surveys conducted over the years show that the vast majority of patients—upwards of 90 percent—say they are satisfied with communications between physicians.
Studies have similarly found that most primary care physicians—56 percent in one instance—rated the handoff process as satisfactory, and that 63 percent report that they “always or usually” receive patients’ discharge summaries.
Despite such glowing perceptions, “the data show that we’re not doing very well at all,” said Dr. Howell, who is associate director for hospitalist services at the Johns Hopkins Bayview Medical Center in Baltimore.
For instance, studies have shown that hospitalists document their communications with primary care physicians only 8 percent of the time. (Primary care physicians are only marginally better, documenting only 11 percent of their interchanges with hospitalists.) While information is probably communicated in a much higher percentage of cases, Dr. Howell explained, “we all know that if it’s not documented, it’s not done in legal terms.”
Studies also show that office-based physicians are notified of discharge by phone only 31 percent of the time, and that they receive only one out of three discharge summaries by the time patients arrive for their follow-up appointment.
The fallout from such miscommunication can be especially dire for patients at highest risk, particularly when it comes to what Dr. Howell called “incidentomas,” or incidental findings. Other potential problems include patients whose test results have changed, and information that isn’t received by primary care physicians at the time of follow-up.
Case law: Who’s responsible?
While there’s very little case law that refers directly to hospitalists, Dr. Howell said that hospitalists’ risks and responsibilities can be inferred from other specialists—surgeons, cardiologists and emergency physicians—who have a long legal record when it comes to discontinuous care.
According to Dr. Howell, case law places responsibility squarely on the shoulders of both inpatient and outpatient physicians, with each having different obligations to patients. Hospitalists, he explained, generally have a duty to the patient to ensure care upon discharge.
“That doesn’t mean we have to be staffing a clinic for follow-up visits for patients,” he noted, “but we do have to make sure that patients have care available for pending or changed tests.”
Likewise, he said, case law implies that hospitalists are obligated to make sure that medical treatments started in the hospital—and medicines that require further testing after discharge—are adequately addressed by a physician on the outside.
That holds true even in situations where communicating with a primary care physician is especially challenging, either because patients are from out of state or don’t have a primary care provider.
What about the duty of primary care physicians? Dr. Howell said that case law places responsibility with them to obtain hospital records that haven’t been received and to ensure follow-up once the handoff from the hospital is complete.
How does this translate to the patient with the lung module who failed to keep his follow-up appointment? Because of the patient’s actions, follow-up didn’t take place for nine months, he reported. While the nodule was found to be benign, there would have been trouble for the hospitalist if there had been an adverse outcome and a lawsuit.
“The person sued would be the inpatient physician,” Dr. Howell said. “Just because you’ve scheduled that follow-up appointment, if it’s a really significant problem and you think follow-up is important, you need to make sure the patient actually went and saw the doctor.”
But that assessment of hospitalists’ responsibility clearly touched a nerve with physicians in the audience. As one frustrated attendee asked, “What are we supposed to do—go to patients’ homes, handcuff them and drag them to their appointment?”
But another member of the audience pointed out that as hospitalists’ obligations have evolved, they are no different from those of an office-based specialist who must continue to try to report abnormal test results to a patient who won’t come or call in.
“You have to make so many phone calls, you have to send the patient a registered letter,” the physician said. “It’s the same problem.”
Dr. Howell agreed. “You can’t just have some cursory method of trying to ensure follow-up for all patients,” he said. “You have to do something else, particularly for the highest risk patients.”
But what should that “something else” be? That depends in part on the system where you work. One audience member commented on the woman patient whose late discharge summary led to her not having her renal function checked.
The hospitalist in that case, the audience member pointed out, could have scheduled a follow-up lab for the patient when she was leaving the hospital. While Dr. Howell said he liked the idea, he noted that it would work only if the patient is in the same network as the hospital. For instance, he pointed out, “We’re at a large academic center, and most of our patients actually can’t follow up at our institution for labs.”
Along with audience members, Dr. Howell did suggest some possible routes to better communication:
• Tailor your efforts to patient risk. Part of that decision-making process begins in the hospital. The man with the incidental lung-nodule finding, for instance, could have been given a CT scan before he was discharged—even though throughput is an overriding concern at all hospitals, and a CT could delay discharge by several hours or even a day.
But given the liability potential of an incidental finding, Dr. Howell said that ordering the CT may be the better way to go.
“You have to weigh who you’re going to send home,” he concluded. Hospitalists also have to tailor follow-up efforts to be more vigilant for an elderly patient with documented coronary artery disease than for a relatively young person with few risk factors. And “verbal discussion between the hospitalist and the primary care physician should be mandatory” for high-risk patients.
• Build redundancies into the process. Even though discharge summaries are dictated on the day of discharge at Bayview, Dr. Howell said, primary care physicians may not receive them for many days.
For facilities that face similar lag times, one possible solution is to create a redundant communication system. Immediately upon discharge, hospitals can fax primary care physicians a discharge sheet that lists the basics like the diagnosis. A full discharge summary will follow later.
• Use discharge planners. Some studies show that non-physician discharge planners—who can devote more time to follow-up and communication efforts—improve primary care follow-up and cost per case. However, no data yet support the idea that they improve clinical outcomes.
And as the person responsible for his hospitalist group’s budget, Dr. Howell pointed out that the line item can be a sticking point.
“It may be effective in terms of global care to hire discharge planners,” he said, “but I can tell you that management will ask you why you have to spend $60,000-plus to do the job that you can do.”
• Learn from your mistakes. After realizing that no one had informed the hospitalist of the probable renal carcinoma diagnosis, the hospitalists in that facility now review all radiology reports prior to signing off on the discharge summary.
• Go the extra distance. When it comes to particularly challenging communications—patients from out of state or individuals who have no primary care physician—you need to make an extra effort.
One potentially promising service would allow hospitalists to autosearch any primary care physician in the country. Dr. Howell said he’s heard of one company that is developing a service that would search for the primary care physician, confirm that the physician treats the individual patient and link the two physicians up.
But for unassigned patients—which make up most of the patients at Bayview—”we treat them as the highest risk and we take the most conservative approach with follow-up.” That includes making an appointment with a sliding-scale outpatient clinic and documenting that the patient - was emphatically educated about follow-up.
When patients don’t make that follow-up appointment, Dr. Howell said, “We call them and even fax them the results of tests that were problematic.”
All in all, Dr. Howell noted, the area of transition “is ripe for improvement,” even though the public perception of that need hasn’t yet caught up to reality. That, however, continues to put both hospitalists and primary care physicians at a disadvantage.
“There is little pressure to improve to date because patients still think we do a good job,” Dr. Howell said. But that’s going to change in time—in the form of greater patient dissatisfaction, higher error rates and more liability lawsuits—”if we don’t change it ourselves.”
Phyllis Maguire is Executive Editor of Today’s Hospitalist.
Weighing the pros and cons of phones, faxes and e-mail
According to hospitalist Eric Howell, MD, any type of communication between hospitalists and primary care physicians reduces the risk of serious errors and liability. The problem, however, is that no data offer head-to-head comparisons of different types of communications between physicians.
At the same time, significant barriers and a wide range of personal preferences result in key flaws in just about every communication method, said Dr. Howell, who spoke at last spring’s annual meeting of the American College of Physicians about improving patient handoffs between hospital and office internists. He offered a quick review of pros and cons for each of the following:
• In-person vs. on the phone.Everyone wants what Dr. Howell called the “gold standard” of communicating patient information: an in-person exchange. “Everybody wants to talk in person,” he said, “but that doesn’t happen anymore.”
Phone contact is probably the next best route. According to surveys, three-quarters of primary care physicians say they’d like to get a phone call from hospitalists on both admission and discharge. (See “Communicating with office-based physicians: What do they want?”, above)
“But they really want it on their own time, like at the end of the day when they’re not that busy,” pointed out Dr. Howell, who is associate director for hospitalist services at Baltimore’s Johns Hopkins Bayview Medical Center. When you have two really busy doctors who want to receive phone calls when they’re not busy, he said, “you get a lot of unanswered phone calls.”
One remedy was proposed by a physician in the audience: Have dedicated phone lines with primary care providers, either to individual physicians or to an office practice as a whole. That way, the hospitalist’s discharge message isn’t competing with the rest of a generalist’s voicemail.
• Faxes. Faxing has the advantage of being automatic, once you have the primary care physician’s contact information in your system. But there is no way to confirm that a fax was received, which makes it very one way.
“There is no dialogue,” Dr. Howell said, “so we can’t really talk about the patient and discuss options.”
• E-mail. Many people have hoped that e-mail would loosen the communication logjam and solve the problems posed by both phone calls and faxes. E-mail, after all, can be read by physicians on their own time, and the sender can at least verify that a communication has been received and opened.
But many still think that e-mail has too many security issues to be reliable for sensitive patient information. That probably contributes to the dismally low numbers of physicians who use e-mail to communicate, which one study estimated as one in three.
• Information technology. Dr. Howell pointed out that a literature search he conducted didn’t offer up much data on how technology solutions affect medical or legal risk. And the few studies that do exist show that technology creates its own types of errors.
The April 4, 2006, Annals of Internal Medicine, for instance, detailed a case where a diabetic patient admitted to a large hospital was mistakenly given the bar-coded wristband of a nondiabetic patient. The mix-up almost led to the nondiabetic patient receiving a potentially fatal dose of insulin.
“I firmly believe that in the long-term, information technology will be a large portion of the solution, because new services will be more secure and will allow physicians to document their communications,” Dr. Howell said. In the meantime, he cautioned, information technology systems—like so many other potential solutions—are still in their infancy.