Clinicians talk shop - |
 |
|
|
September 2003 Karen Titus What physicians want from pathologists is quite simple. Every last one desires something different. “That’s the short answer, unfortunately,” says Robert J. Homer, MD, PhD, associate professor of pathology, Yale University School of Medicine, and director of anatomic pathology, VA Connecticut Healthcare System. Dr. Homer should know. He’s one of three coauthors of “Clinicians Are From Mars and Pathologists Are From Venus” (Powsner, et al. Arch Pathol Lab Med. 2000; 124: 1040–1046), which looked at how well clinicians understood pathologists’ written reports. Which leads to another short answer: Not so well—the investigators found that surgeons misunderstood pathologists’ reports 30 percent of the time. “That really shook me up,” says another coauthor on the paper, José Costa, MD, deputy director of the Yale Comprehensive Cancer Center and director of anatomical pathology and professor of pathology, Department of Pathology, Yale. “A third of my life is gone in misunderstandings,” he says, only partly facetiously. At the same time, he couldn’t have been completely surprised. After all, the study was born from his concern that he and fellow pathologists were overly “pathocentric,” as he calls it. “As pathologists we worry about the report, and we think about the report, and always from the pathologist’s point of view,” he says. “That’s not necessarily what our customers, the physicians, want. And in fact, we found out that it’s very difficult to get information across in a way that is easily apprehended by the consuming physician.” You hardly need a study for that truth to dawn, however. Anyone who’s ever worked in a laboratory knows pathologists and clinicians aren’t always comfy bedfellows. Everyone has a tale. Clinicians who scribble “rush, rush, rush” on every order, assuming the turnaround time will stretch into weeks if they don’t goad the lab into faster action. Pathologists who drop a note about a cancer diagnosis in the mail, nothing e- about it. Clinicians needing tutorials in the finer points of screening versus diagnostic tests. Pathologists who struggle to explain that distinction. The stories go on forever, like a messy divorce, unleashing shock, disbelief, and anger, all giving way eventually to a resigned tiredness. Underneath this sea of woes lies the real conundrum: How can pathologists make sure clinicians get what they truly need? Which brings us back to Dr. Homer’s observation—it depends on the clinician and the pathologist. Clinicians
aren’t reluctant to say what they want. They also emphasized that much of the time, they’re already getting what they want. “Pathologists are not a frustrating part of our experience,” says Gretchen P. Purcell, MD, PhD, of Duke University Medical Center. Still, if Yale surgeons misunderstand pathologists’ reports 30 percent of the time, is it going out on a limb to suggest there might be room for improvement? In that spirit, here’s a sampling of physicians’ responses to the question, What do you need from your pathologists? Show them what you’ve got. Seeing is believing, says Dr. Purcell, chief resident in general surgery and assistant research professor of surgery and clinical informatics, Duke. “Sometimes seeing the pathology is very useful to me,” says Dr. Purcell. Yet unless a case involves cancer, such viewings are not the norm. Dr. Purcell admits seeing slides for every case would overwhelm her. But she’d like to see more than she currently does—unusual cases, for example, or those with equivocal findings. “I’m not a pathologist, nor do I profess to have any expertise in this area,” she says. “But it is useful when someone says, ‘This is unusual,’ and you look at it yourself. The more you look, the better you feel in making a judgment when the findings are equivocal.” Pull them into your world. Harold C. Sox, MD, editor of the Annals of Internal Medicine, says he enjoyed a spirited relationship with pathologists during 30-plus years of clinical practice. “I’ve always appreciated a pathologist who imposed on me to come look at a biopsy slide, who said, ‘The urinalysis on your patient is pretty exciting—can you come up right now and take a look at it?’” he says. “I’m in favor of pathologists crowding clinicians a bit to come see interesting things and broaden their education, to get them out of the clinic and learn something.” Pathologists should welcome being pushed as well, says J. Stephen Jones, MD, staff urologist at the Cleveland Clinic. He likes that at his institution, when he or his clinical colleagues ask to look at a biopsy, the pathologists “bend over backwards to make sure we understand why they’re reading the biopsy the way they are.” “Sometimes they know it’s cancer, they’re just not willing to go out on a limb,” Dr. Jones continues. “And that’s when I like to be able to challenge them. Sometimes they’ll say, ‘No, there just isn’t enough to say it’s cancer,’ and other times they’ll say, ‘You know, you’re right, we’ve looked at it again, and we’re ready to commit to saying it’s cancer.’ Having that dialogue is obviously crucial.” But get out of your own world as well. It’s a delicate matter, one clinicians raise consistently but gingerly. Begged one physician: “Find a way to say this so your readers don’t beat the crap out of me.” As some clinicians see it, pathologists’ professional lives are built around scheduled procedures. Pathologists could be tonsured monks, for all the times they get out of the lab. Yes, the age-old myth persists: They think you chose pathology, in part, because it’s more controlled, more contained, than other specialties. Clinicians on the front lines of patient care lay claim to no such luxury. “It can be infuriating. You can go long periods of time without much happening, and then suddenly it’s a deluge,” says Lee A. Hebert, MD, director of the Division of Nephrology and professor of medicine, Ohio State University. Pathologists who don’t rise to the occasion just make matters worse. A clinician who suspects a patient has TTP, for example, doesn’t want the pathologist to respond with, “Oh, sure, I’ll look at it Monday,” says Dr. Hebert. Nor is “Hmmm, really? Are you sure about that?” a helpful reply. “Pathologists have to learn to not find reasons not to do something,” Dr. Hebert says. “Even if it’s 11 at night, or 1 in the morning, you want to hear, ‘That’s my job; I’ll be right there.’” Certainly pathologists aren’t the only reluctant warriors in medicine, Dr. Hebert acknowledges. But now, more than ever, there’s reason for pathologists who drag their feet to change their ways. “There’s an increasing movement for pathologists to take on clinical activities,” says Dr. Hebert, pointing to pathologist-directed plasmapheresis services as one example of this shift. “So we’re expecting pathologists will become more willing to be there for the patient at times inconvenient to themselves and their families.” Stop speaking in tongues. “It’s very simple what we need from our pathologists,” says Gary Falk, MD. “Expert pathologic consultation, using a standardized, acceptable terminology.” That latter portion is critical, he says. “That’s the single biggest problem—that pathologists don’t use standardized terminology. It creates an incredible amount of extra work for clinicians and pathologists, and it increases health care costs.” Dr. Falk directs the Center for Swallowing and Esophageal Disorders at the Cleveland Clinic’s Department of Gastroenterology and Hepatology; he’s also an associate professor of medicine. Not surprisingly, he picks Barrett’s esophagus to make his point. Though specific terminology exists for the dysplasia classification as well as intestinal metaplasia, “Pathologists get it wrong all the time,” he says. “They don’t use the correct terminology for dysplasia, and they don’t state in the pathology report whether they see intestinal metaplasia or not.” Accordingly, Dr. Falk sees his fair share of patients who come to him in a panic, having been told—mistakenly—that they’ll die of esophageal cancer. “When in fact they don’t have Barrett’s esophagus. They’ve never had Barrett’s esophagus. It’s because the pathologists didn’t read the darn thing right. If they would use the right terminology, a lot of these diagnostic inconsistencies would get better.” John Goldblum, MD, professor and chairman of the Department of Anatomic Pathology at the Cleveland Clinic, says Barrett’s esophagus is defined as esophageal intestinal metaplasia with an endoscopic abnormality. “Our job as pathologists,” he says, “is to recognize intestinal metaplasia in a biopsy that the gastroenterologist tells us is from the esophagus.” If the biopsy has Barrett’s, Dr. Goldblum says, then the pathologist must comment specifically on the presence or absence of dysplasia using the following classification:
Dr. Falk is also irked by the way pathologists frequently report on chronic active gastritis. “There’s no mention oftentimes of whether the pathologist looked for H. pylori or not, or if special stains were done,” says Dr. Falk. “Again, it’s the issue that standard terminology, standard ways of reporting information, isn’t there.” Get to the point. It’s a dead-ringer for a line in a country-western song: Make the bottom line your top line. This request comes courtesy of Dr. Jones. “What I usually want—and I think I speak for most clinicians—is to be able to get the high points of any pathology report in a matter of seconds,” says Dr. Jones. “Then I can go through the text if I need clarification.” Dr. Jones finds that a bulleting system works well. A prostate biopsy report, for example, would contain one line each for Gleason’s score, location, tumor size, and percentage of biopsy that’s positive. “I can go down the list, and I know just where to find each piece of information,” Dr. Jones says. “Whereas if it’s just in paragraph form, I have to sort through someone’s verbiage, hoping that I interpret them correctly.” Too much prose adds confusion. “Help me out here,” Dr. Jones pleads. “I don’t have time to read three paragraphs of a negative report, to make sure that at some point in there it says that there’s no cancer. I want something that says, ‘There’s no cancer. The details are below.’” Equivocal findings fit the format just as well, Dr. Jones says. “I still want that summary at the top: ‘Suspicious but not diagnostic for cancer.’ I need that summary right off the bat, hitting me on the forehead. And then I’ll choose whether I need to sort through why they couldn’t tell.” Have a point. Pinning pathologists down to an answer is a little like watching Donald Rumsfeld spar with the press, some clinicians say. “In medical school we probably would have preferred the true-false test, and the pathologists would have preferred the multiple-choice,” says Dr. Jones. “We can understand if something is equivocal,” he continues. “That is an answer, and we’re willing to accept that. What we have a hard time with is two paragraphs explaining that they’re unwilling to say they don’t have an answer.” Instead, he’d like pathologists to say—up front, of course—“We have done everything we can do, and this is the best diagnosis we can make.” Making a change? Let them know. When changing a test or adding a new one, don’t keep clinicians in the dark, requests Maurie Markman, MD, chairman of the Department of Hematology and Medical Oncology at the Cleveland Clinic. E-mails or a paper trail announcing the change is one way to get the word out. But face it: Both can get lost, or at least not read. A more direct way is literally going to a staff meeting, says Dr. Markman. “That way you’ll be able to have a direct discussion with the doctors involved, and you can have a give-and-take about whatever questions and answers come up.” That also holds true when changing report formats. “In our place the tests change constantly,” says Michael Fischbach, MD, professor of medicine and chief of the Division of Clinical Immunology and Rheumatology, University of Texas Health Science Center, San Antonio. “One week it’s reported in titers, the next month it’s reported in units, and you’re not sure how they correlate.” Unleashing new information in pathology reports will confuse clinicians if it’s not introduced carefully. Says Dr. Markman: “If the pathologists think there’s something new that they want to add—just as when clinicians find something new that they think is relevant—meet together and decide how to modify the reports so everybody understands it.” Remember that not every test is equal, cautions Dr. Fischbach. No amount of e-mails or newsletters can make him remember a test has been changed if it’s one he rarely orders. “You forget about it. When you order it, that’s when you need to deal with the information. Maybe it can be attached to the test results.” Use numbers ... More numeracy in answers would improve care, says Robert Badgett, MD, associate professor of medicine, University of Texas Health Science Center, San Antonio. Important test results should include information on the test’s sensitivity and specificity in typical clinical populations, he suggests. “How many clinicians are inappropriately ordering and applying the BNP assay to diagnose heart failure because they don’t realize its specificity isn’t strong at any cutoff level, and its sensitivity is strong only when the BNP is less than 50 pg/mL?” he asks. ... Or don’t. “On the other hand, I got to admit, docs are real reluctant to have their answers in numeric terms,” Dr. Badgett continues. It’s true. “For one thing, it takes some work to interpret the numbers,” he explains. “And studies show that only three percent of practicing docs and about 50 percent of teaching docs use sensitivity/specificity when they’re trying to interpret tests.” Dr. Badgett takes those numbers to mean that clinicians don’t want numbers. “They see it as a burden.” (But really, you probably should use numbers. Because as Dr. Badgett points out, “There is research that shows if you give them those numbers, they make better clinical decisions. A current example is probably the BNP.”) Pick up the phone. OSU’s Dr. Hebert can recall a case when he was told of a terminal cancer diagnosis by mail. “By mail? Geez!” he says, still amazed at the delivery method. Not that other methods are necessarily more efficient. Dr. Hebert recounts another case involving a pathologist who was performing a kidney biopsy, working in conjunction with a radiologist. After the pathologist had completed his work and left, the radiologist, in reading the films, realized the patient had an aneurysm. The radiologist reacted to this finding by recording a message on his Dictaphone—as Dr. Hebert puts it, “leaving it to his secretary, the mails, and interoffice help” to convey the message. “There’s a 10 percent chance the report never got back to the doctor,” he says. “If it does get back, and you’re going through six inches of reports, and here’s this one talking about a kidney biopsy under fluoroscopy—you’re barely going to look at it. And buried in that is the note about the aneurysm.” Dr. Hebert became involved “when the patient showed up six months later and died in the ER because his aneurysm had ruptured.” The original pathologist was subsequently sued; Dr. Hebert, who served as an expert witness in the case, reports the radiologist defended his actions by saying, “I told you—it was in the report.” “As if the responsibility to his patient ended when the words left his mouth and entered a Dictaphone,” Dr. Hebert says angrily. Though this particular incident was tied to a radiologist, not a pathologist, the point is well-taken: Make sure that clinicians receive critical information. “We’re deeply grateful for calls,” says Dr. Hebert, who now gets all reports of a malignancy by telephone. He also advocates telephoning for panic values. “Our lab is really quite good at that. They call you at 8 PM, after you’ve been in the clinic all day and generated a bunch of labs. And they’ll say, ‘By the way, Dr. Hebert, Mrs. Jones has a potassium of 6.6.’ And they’ll have her phone number so I can follow up—because they’re calling me at home. That’s hugely useful,” Dr. Hebert says. But for such calls to be useful, they can’t be perfunctory. “We see it all the time in the reports—‘Dr. Jones was called’—when in fact, he wasn’t called,” he says. Don’t shy away from e-mail, either. Phone calls may be ideal, says Dr. Falk, “but I don’t think it’s fair to ask them to do that all the time—everyone is so busy.” Not every case requires a call, agrees Dr. Hebert. “If it’s not that urgent, an e-mail to promptly alert us is fine,” he says. “The trouble with e-mail is, you’re never sure if the person reads it. They may be gone for the day.” At the Cleveland Clinic, physicians receive an automatic e-mail notification as soon as laboratory values or pathology reports are complete, says Dr. Jones. “It saves me a lot of effort.” Though he may call a pathologist on more urgent cases, requesting that they look at a slide posthaste, routine matters don’t require constant checking. “If I don’t have that e-mail, I know it’s not ready,” he says. Or the Internet, for that matter. An active researcher in clinical informatics, Dr. Badgett avows a strong interest in understanding how clinical information flows. Specifically, he’s intrigued by the question, How much information do clinicians want in their answers? Research has found, he says, that so-called opinion leaders—those most likely to change medical practices, either formally or informally—are likely to want added information, such as original evidence. Teaching physicians are likely to fall into this category, as are nonteachers who strive to stay abreast of new developments. “But at the point of care, I think they’re just going to want you to give them the answers—the assay results, and its sensitivity and specificity,” says Dr. Badgett. Eventually, says Dr. Badgett, the Internet and electronic medical records will make it possible, perhaps even easy, for pathologists to provide layers of information in their reports. In addition to including a test value, they can also link to studies related to the test, to the disease, or to both. This passive approach makes the information accessible without overwhelming clinicians, he suggests. Making such information useful is another matter. “If you just throw out a couple references that you think are probably pretty good, but you haven’t done a good literature search and formally assessed the articles, or worked out the sensitivity and specificity, then people aren’t going to respect that work and use it,” Dr. Badgett says. Dr. Jones seconds that notion. In fact, he dismisses most added information as useless. “It just clutters up the pathology report when they put that in. I know some labs routinely say, ‘The PSA is positive,’ and then it will give a reference—‘Catalona found that 22 percent of patients with this level of PSA had cancer,’ and things like that.’ Well, if I don’t know that already, I shouldn’t be ordering that test,” Dr. Jones says. Be a consultant. Clinicians use this word freely when they talk about you. To wit:
For example, a report containing a positive antinuclear antibody test result would include guidance on followup testing. “Of course everyone should know that. But everyone doesn’t know that,” says Dr. Fischbach. “Especially when you deal with people who have lots of training in one area or people who are doing family practice, who have to cover so many areas they couldn’t possibly know all those things.” String the above points together, and the result sounds like a personals ad wish list, minus the obligatory “romantic dinners” comment. Unlike those ads, however, there’s nothing toplofty in any of these requests. But for every point raised, pathologists have a response—to which clinicians will often accede. Yale’s Dr. Homer, for example, observes there are subspecialists who want excellent histologic description, then to be left to themselves to make their own clinical-pathologic diagnosis. These generally are older, more experienced clinicians, he’s found. “Younger clinicians who generally have very little, if any, histopathology training want the whole thing tied up neatly with a bow. They want a complete clinical-pathologic correlation provided by you—even though you don’t know the patient,” he says. Dr. Falk concedes the point. “A pathologist has the handicap of us not giving them enough information. That’s a problem, and I don’t doubt that for one minute.” Some pathologists strongly promote the pathologist as consultant, including Michael Laposata, MD, PhD, director of Clinical Laboratories at Massachusetts General Hospital, and a professor at Harvard Medical School. Dr. Laposata and colleagues have been developing expert pathology reports, narrative interpretations that, in simplest terms, focus on two areas: the differential diagnosis and what clinicians should do next to make the diagnosis. Anything beyond that—sensitivity, specificity, pathophysiology—while important, should be related links on a report, not part of the actual narrative, he says. “The fundamental piece of information that these doctors want is, What does my patient have, and what do I do next?” Then there’s the matter of standardization. Consensus criteria make matters much more clear—but only to those who are aware of and understand the criteria, says Dr. Homer. “It makes things completely opaque to the people who don’t,” he says. Dr. Homer takes it a step further, noting that some clinicians don’t understand the fundamental purpose of pathology. He recently worked with a clinician who was reviewing a lung biopsy with cancer in it from a patient with a history of colon cancer. The clinician asked Dr. Homer if he thought the cancer was primary lung or metastatic colon. Dr. Homer said the histology slightly favored lung and suggested that additional stains be done. The clinician responded, “Oh, no, I trust you,” and ended the discussion. “As if this was something magical, as opposed to actual criteria that get used,” says Dr. Homer. “What he seemed to want is the magical statement from a pathologist about some unknowable truth, as opposed to a statement of odds,” he continues. Indeed, he says, he’s encountered more than his fair share of clinicians who wanted him to say what he felt about a biopsy. “Which I’ve never understood,” he says. “Because I don’t feel anything about a biopsy. I can tell them what’s on it, though.” All of which gets back to Dr. Homer’s original point—what clinicians want depends on the clinician, as well as what the pathologist can provide. “I encourage pathologists not to say more than they know,” he says. “I think there’s always the temptation for the clinician to push the pathologist into the corner, and I think the pathologist needs to struggle as best he can to get out of the corner.” This means giving an answer—but not getting “pinned down to things that are not sayable.” At this point, the discussion would seem to have moved to another stage. It’s no longer a matter of looking for desirable traits in another. Now it’s time to dig a little deeper and ask, What do clinicians truly want? The answer to that question might sound (warning: semi-sexist metaphor to follow) like a wife pleading with her husband to please just talk more. Open up. Share. But, as is so often the case in relationships, that message gets lost in a maze of other requests that devolve into inner lives of their own. This explains the befuddled husband who finds himself huddled with a divorce lawyer, and he thinks it’s because he didn’t pick up his socks. Dr. Laposata has long known of the gap between what clinicians say they want and what they really want. He also observes that often pathologists and clinicians aren’t even aware when the wires may be crossed. Take a high potassium, a routine test result hardly in need of one of Dr. Laposata’s expert narrative reports. But a clinician who sees that a patient’s potassiums have spiked wants to know if the change is “real,” and not simply the result of a sample being stored for too long. Most pathologists have the answer to that question, says Dr. Laposata, but they don’t always convey it. Moreover, he says, pathologists may need to prompt the clinicians with questions of their own: What have the potassiums been for the last week? Is there any evidence of cardiac arrhythmia? “Simple results can be confusing for clinicians, and pathologists usually don’t understand that clinicians are going to be confused,” Dr. Laposata says. INR results provide another excellent example. If a pathologist switches to a more sensitive thromboplastin without informing clinicians of the new reagent, a PT result can leap from 20 to 40 seconds for the same patient and same blood sample. That’s an alarm for most physicians, even though the INRs calculated from the different PTs will be the same. “When they start getting into the 30-second range for a PT, they get scared,” says Dr. Laposata. “Communicating to the doctors that this is just due to the reagent, and to ignore the fact that it’s 20 seconds longer, has been a major educational problem, and I’m sure it’s resulted in the infusion of more fresh frozen plasma into people than they should have had.” Can this marriage be saved? A troubled couple might turn to a marriage counselor. The stand-in here is Seth Powsner, MD, associate professor of psychiatry and emergency medicine at Yale and, not coincidentally, the lead author on the aforementioned “Clinicians Are From Mars ...” article. Dr. Powsner says he shared Dr. Costa’s surprise at the error rates they uncovered in their study. The other surprise, for him, was how physicians learned to adjust to poorly designed reports. “It’s like the line from ‘My Fair Lady’—‘I’ve grown accustomed to her face,’” he says. In the study, he and Drs. Costa and Homer found that stylistically improving report formats made matters worse. The reason, says Dr. Powsner, is that physicians eventually accommodated many of the problems with current formats. “Physicians have developed visual calluses. They’ve learned to work a little harder to find that number they need that’s on the second page.” Of course, tolerating problems is hardly a socko point around which to rally. Dr. Powsner echoes the voices of more enlightened pathologists in calling for point-blank dialogue with clinicians. For too long, he says, both sides have relied on formulaic ways of conveying information—the blood-for-numbers/ tissue-for-histology exchanges that keep true understanding under wraps. With Socratic ease, Dr. Powsner nudges along another approach. “Could it be that physicians want the answer to a clinical problem, not a technical question?” Clinicians may ask, What was the patient’s ALT? How many units of enzyme were measured when we sent you the serum specimen? “In truth, some of them are sitting back there saying, ‘I think this patient’s trouble has something to do with their liver, and I would like to know whether we should do more testing on the liver. Or, should we move on to some other part of the digestive tract?’” Dr. Powsner says. The real answer, in turn, would be along the lines of: “‘If you’re worried about the liver, this series of tests doesn’t indicate that the liver is the primary location for the lesion,’” he says. “The challenge then for the pathologist is to figure out a way of offering that kind of information without doing an individual review of every case and then dictating a paragraph.” No one said it would be easy. Indeed, Dr. Laposata’s efforts to provide those types of answers are coming to fruition after nearly a decade of work, and he’s the first to admit there’s plenty of work remaining before all areas of pathology will be covered by clinically focused expert reports. On the other hand, the deeper answers may also be closer, and more simple, than you think, though you’ll first have to endure another depressing blow to your oeuvre: In addition to not always understanding your reports, many physicians don’t even read them, the Yale authors contend. That means clinicians are often getting critical information another way. Oftentimes it’s done in a 30-second conversation in the hallway, or through a quick phone call. Is that so bad? Maybe not. Though these interchanges are rarely talked about, Dr. Powsner observes they usually work. “When the going gets tough, the front-line clinician knows some pathologist to call, and the action is usually resolved over the phone.” Phone calls do get the job done, Dr. Badgett agrees. “We e-mail or telephone someone like John Olson [MD, PhD, pathologist at UTHSC at San Antonio and chair of the CAP’s Coagulation Resource Committee]. Which is probably not the easiest way to get an answer, but it’s always what happens and what works,” he admits. Pathologists might want to keep track of the phone calls they receive from clinicians asking for clarification, Dr. Powsner says. That, combined with a satisfaction survey, would yield interesting data. Twenty calls and high marks from clinicians, for example, would indicate “whatever you’re doing, they like. But it sure seems to involve the phone a lot,” he says. “Now, which part isn’t needed?” No one—not a one—is suggesting pathology reports are ready for a farewell tour. Instead, they will evolve, as Dr. Laposata predicts, to a point where they answer the real questions clinicians are asking and form the bedrock of better clinician-pathologist relationships. Dr. Homer puts a helpful—if somewhat mischievous—spin on matters. Like many clinicians, he agrees that long—that is to say, three-paragraph—notes simply won’t get read. Which is why he encourages his pathologist colleagues to pile on the prose. “Certainly for the complicated specimen, the long note is better than the short note, largely because the long notes mean the clinicians will need to call you,” he says. “Because they’ll read this long note and they’ll say, ‘Geez, what the hell is that?’” Karen Titus is CAP TODAY contributing editor and co-managing editor. |
|||
|
