Feature Story

Getting results from point A to B

September 2004
Karen Lusky

A recent Arizona Supreme Court ruling, Stanley v. McCarver, offers a good news, bad news liability scenario for radiologists, pathologists, and labs in settings where they do work for nonpatients, such as examinees for insurance physicals.

The good news is that the Arizona Supreme Court rejected a lower appellate court ruling that a radiologist had a duty to communicate an abnormal finding directly with the person who had an employment-related physical examination, says CAPgeneral counsel Jack Bierig. "The bad news is that the court left it to the jury to decide whether the radiologist should have contacted the patient."

The ruling doesn’t materially change the duty of the pathologist in traditional clinical settings, legal experts say. "Pathologists are historically viewed as owing a duty to a patient in a situation where a specimen is referred to them for a pathologist examination and report, and this case doesn’t alter that scenario," says Robert Sonne, legal counsel to the Arizona Medical Association. "But the fear of doctors," he adds, "is that the Arizona Supreme Court ruling opened the door to a duty of physicians to nonpatients to let them know about their abnormal testing findings." The full ramifications of the court’s ruling are not completely clear, he says, "and therein lies the concern among physicians."

In the case in question, Stanley v. McCarver, a radiologist discovered a nodule and signs of pneumonia on a nurse’s routine pre-employment chest x-ray. The radiologist followed the established protocol for reporting his findings to the company where he worked as an independent contractor interpreting chest x-rays for pre-employment physical exams. The company, in turn, forwarded the radiologist’s report to the nurse’s employer. But the employer somehow dropped the ball and never let the nurse know about the abnormal x-ray.

The nurse was diagnosed with lung cancer about 10 months later and subsequently sued not only the employer, but also the portable x-ray company that had contracted with the radiologist to read her x-ray—and the radiologist. The plaintiff alleged that the defendants "provided negligent and improper medical care" by failing to "timely and adequately diagnose and/ or communicate to [her] the abnormality evidence on her chest x-ray."

The trial court summarily dismissed the case against the radiologist, reasoning that the plaintiff was not his patient; thus, the radiologist had no legal obligation to protect her from injury or harm, a duty that must exist to establish negligence under tort law. And that might traditionally have been the end of the story in terms of liability for the radiologist.

Yet in a ruling that has implications for other interpreting physicians, including pathologists, the Arizona Court of Appeals and Arizona Supreme Court ruled that the radiologist did owe the plaintiff a "duty to care," even though she wasn’t his patient in any traditional sense. The Arizona Supreme Court remanded the case to the trial court to decide if the radiologist should have personally informed the nurse of the abnormal test results.

Even if the jury finds the radiologist did nothing wrong in this particular case, plaintiff’s attorneys across the nation can cite the Arizona Supreme Court decision in an attempt to keep a similar negligence action alive past summary judgment. And the court’s ruling may apply to at least a couple of relatively common situations in laboratory medicine, Bierig says. "For one, labs examine a person’s blood or urine as part of a pre-employment or insurance exam. Secondly, a laboratory may review a person’s blood or urine as part of a toxicology test for a variety of reasons, such as an athletic event," he says.

"The import of the Arizona Supreme Court ruling," he continues, "is such that a laboratory that detects a serious problem with an individual’s blood or urine in those circumstances but doesn’t report the problem to that individual may not escape liability simply because there is no physician-patient relationship. Rather, the court would leave it to a jury to determine whether, given the specific circumstances, the laboratory should have notified the individual."

In the case reviewed by the Arizona Supreme Court, the employer had a protocol requiring it to notify prospective employees of the results of their physical examinations within 72 hours, which it failed to do, says Leonard Lucey, an attorney with the American College of Radiology. "Yet the court says that regardless of the communication protocol and how it’s supposed to work ... the radiologist possessed information critical to the plaintiff’s care, that is, the nodule seen in the lung ... ," Lucey says. "And this required the interpreting physician to take reasonable steps to convey that information in such a fashion that it would likely reach the patient."

And that, of course, is the million-dollar question: What might be viewed as reasonable on the interpreting physician’s part to ensure the examinee received the information? In Lucey’s view, a process for confirming that the company or a clinician received the abnormal result, such as a receipt, would suffice. He says this practice accomplishes two purposes:

• CA It allows the interpreting physician to know the information got where it was supposed to go.
• CB The radiologist or pathologist, etc., can use that receipt in his or her own defense to show he or she took reasonable steps to ensure transmission of the information to the clinician or third party ordering the testing.

In fact, the Arizona Supreme Court ruling does say doctors can limit their liability exposure through contract agreements that define who is responsible for informing the patient about test results, says Kenneth Gatter, MD, JD, a pathologist with the Oregon Health and Science University, Portland, and an adjunct professor of law at Willamette Law School. "The court writes: ’Doctors may deal with this issue as a matter of contract. They may, for example, require x-ray subjects to consent to having the results reported only to the employers.’

"That’s one of the ’take home messages’ about this case," says Dr. Gatter.

But this message raises concerns for Bierig, who points out that it may be difficult for pathologists to obtain consent from individuals with whom they don’t deal directly. While the pathologist may try to have an employer or insurer get a person to waive rights against the pathologist, "it is questionable whether the consent will be regarded as informed," Bierig says. "And not getting a waiver after trying to do so may be worse than not trying at all," he adds.

A better course, in his view, might be to try to contract with the employer or insurer to have the company indemnify that pathologist against a claim based on the failure of the employer or insurer to report to the tested individual any significant finding that emerges from the lab results. In any event, Bierig says, it’s probably a good idea for pathologists to check with their insurance companies to make sure their malpractice policy covers this sort of liability.

While the Stanley v. McCarver ruling doesn’t affect traditional clinical settings where pathologists and laboratories already clearly owe patients "a duty to care," the case has deepened the ongoing patient safety debate about exactly where the buck does stop in ensuring that patients receive their test results.

At Oregon Health and Science University, for example, "pathology reports patients’ results on the computer system, which provides at least constructive notice to clinicians, so there’s an electronic record," says Dr. Gatter, though the precise legal sufficiency of this kind of notice remains an open issue. "Outside [providers] from different systems receive a fax copy and a hard mail copy [of the report], so we have a record of the fax being sent." Nonetheless, he adds, a telephone call to the clinician for unexpected results is good practice.

Ronald Lepoff, MD, chair of the CAP Commission on Laboratory Accreditation, sees a trend toward health care systems looking for ways to ensure that clinicians not only receive but act on critical test values, particularly in the outpatient setting. "Studies have shown that clinicians do not act on up to 40 percent of reported laboratory results, such as low vitamin B₁₂ levels," says Dr. Lepoff, a pathologist at the University of Colorado Health Sciences Center, Denver.

Dr. Lepoff believes such omissions are getting worse as outpatient clinicians become busier. "The test results get lost or someone in the office drops the ball or the clinician looks at several reports at one time and the results don’t register like they should," he says.

To help address this potential patient safety problem, the University of Colorado Health Sciences Center has implemented a research project to improve communication of critical test values to physicians in its general medicine and family practice clinics. An automated system notifies outpatient physicians of abnormal test results by using e-mail, pagers, and electronic devices. "The clinicians themselves are identifying the tests and critical values about which they want to receive an e-mail or page," Dr. Lepoff says.

The list includes International Normalized Ratios, hCG values (for monitoring testicular tumors in men), hematocrit, white blood cells, and bilirubin in infants.

"The clinicians define the parameters and say, ’Let me know when a bilirubin is greater than X,’" Dr. Lepoff adds. "If we get a result higher than that, the computer sends an automated message to a pager that says, ’Herman Smith has a bilirubin greater than X.’"

The approach used in the research project represents a "push" rather than a "pull" system in the lab information technology world. "In a ’pull’ system, the clinician has to go retrieve the information," says Dr. Lepoff, "but in a ’push’ system, the clinician’s pager goes off or he or she gets an e-mail" about the test result. In other words, "In a push system, the clinician gets told [about a critical test result] in a way that she or he has to hear it."

Yet for a push system to work, labs should limit the number of tests they report on that basis, Dr. Lepoff says, because if a clinician’s pager goes off repeatedly to report an abnormal test, the clinician may stop paying attention. "A very high cholesterol can always wait until the next day, but a patient with an INR in the critical range—say a reading of 8 because he’s getting too much Coumadin—is at risk of a life-threatening hemorrhage. The clinician needs to know that today, not tomorrow," he says.

The ongoing study at the University of Colorado Health Sciences Center will determine whether the push system communicating clinician-defined critical values makes a difference in outpatient care. While the study isn’t set up to look at actual patient outcomes, "researchers will determine whether the clinicians receive and act on the abnormal results for the selected subset of tests," Dr. Lepoff says.

Computerized communication systems that track the "hand-off of test results from the lab to the clinician continue to improve ... and provide the light at the end of the tunnel in preventing missed diagnoses," says Bruce Friedman, MD, professor of pathology and director of pathology data systems at the University of Michigan Medical School and Health System, Ann Arbor. He says there are now Web-based systems with log-on auditing features and reminders to clinicians to check results, as well as mechanisms for the clinicians to indicate that they have reviewed a result before they can move the result out of their personal inbox.

The University of Michigan is, in fact, using such a system on the inpatient side, which has over the past couple of years subtly shifted responsibility almost entirely to the ordering physicians for accessing and reviewing test results, Dr. Friedman says. "The lab information system has electronic inboxes for each physician where the clinical information system copies the vast majority of clinical pathology and anatomical pathology results ordered by the physician, including pending results."

The only test results that don’t routinely go into the inpatient physicians’ electronic inboxes are point-of-care test results "because clinicians have said they don’t want their inbox littered with that data," Dr. Friedman says. "Moreover, we don’t always know the name of the responsible ordering physician of POC tests when the nurses perform the tests at the bedside."

In addition, a "homebrew" software tool called a discharge navigator allows the treating physician to sort a patient’s test results, including pending results, when the patient is being discharged from the facility. "That way, the clinician can make sure there are no loose ends," Dr. Friedman says. That type of computerized capability can head off missed diagnoses that result when hospital lab information systems generate hard copies of patients’ test results, which easily get lost in the paper shuffle on the day that patients are discharged from inpatient care.

Dr. Friedman, in fact, recently heard of a malpractice case against a Pennsylvania hospital where teaching hospital staff missed seeing a patient’s abnormal IgM protein reported on a hard-copy summary of lab results generated late in the day of the patient’s discharge. The patient was diagnosed with multiple myeloma about a year later and sued the hospital for the delay in diagnosis.

The University of Michigan computerized clinical information system can track to see whether clinicians have logged in to their electronic inbox to check test results. Removing a result from the inbox provides positive documentation that the result has been reviewed. "Thus, the system provides a check-off process that didn’t exist when the laboratory issued hard copies of test results," Dr. Friedman says. "Our LIS, because it’s older than our Web-based clinical information system, did not have the capacity to audit by physician which lab results were accessed and which were not."

At what point does the laboratory or pathologist at the University of Michigan have an obligation to personally contact a physician about a test result? "One example would be if the pathologist or lab had made an error which resulted in the posting of an erroneous report," Dr. Friedman says. "The system is set up so that you cannot delete an erroneous report after it has been reported, because if you do so in an electronic system, you’ve lost your audit trail. And a clinician may have already taken action on the basis of the erroneous result."

In cases where pathologists make "a hard, cold mistake—for example, if the pathologist looked at a slide three days later and realized he or she gave the wrong diagnosis"—they must contact the referring physician to let him or her know, says Dr. Friedman. Pathologists who are doing additional special stains to refine a diagnosis may choose to post the result in the ordering physician’s electronic inbox as "tentative with final diagnosis awaiting special stains," he adds. "Or the pathologist may not give a diagnosis until he or she reviews the special stains and renders a final diagnosis."

Some anatomic pathologists are calling for a consensus sensus on what constitutes critical values or diagnoses in surgical pathology requiring immediate communication to clinicians so they can institute emergent treatment. A recent study reported in the American Journal of Clinical Pathology, "Critical Values in Surgical Pathology," retrospectively reviewed 2,659 surgical pathology reports and found 13 cases of critical values, or CVs (0.49 percent). "Four of the 13 reports documented phone calls to the clinician (most at least one day before final sign-out), and in two other reports, the requisition included clinical history that implied previous knowledge of the diagnosis," according to an abstract of the article (Am J Clin Pathol. 2004; 122: 201-205).

The retrospective study also found little agreement among five clinicians and 11 pathologists surveyed about what diagnoses they viewed as requiring a stat phone call to the ordering clinician. Thus, "a consensus conference of leaders in anatomic pathology and key clinical specialties might prove useful for establishing surgical pathology CV guidelines and could represent a practice improvement," concluded study authors Telma Pereira, MD, Yulin Liu, MD, PhD, and Jan F. Silverman, MD.

Pathologist Virginia LiVolsi, MD, of the University of Pennsylvania School of Medicine, who wrote an editorial accompanying the article, agrees. "It seems obvious," she told CAP TODAY in an interview, "that if a pathologist found a tissue-invasive fungus, that finding is just as important as a clinical chemist finding a patient had a blood sugar of 30. Anatomic pathology needs to not only explore and potentially develop a consensus as to what constitutes critical values, but also how to best communicate those in a way that benefits the patient."

Does the pathologist have a duty to inform the patient of serious abnormal results in certain clinical situations? In the view of Stanley Leung, MD, JD, MBA, a pathology resident at the University of Pennsylvania, the Arizona Supreme Court ruling may represent a "harbinger of putting the burden on the diagnostician interpreting tests because that person will be viewed as the last line of defense in a confusing communication system."

Dr. Leung notes that some labs have protocols where the pathologist or lab notifies patients of certain test results when the lab is unable to reach the ordering clinician after repeated attempts. "And from a liability perspective, it’s probably not a good idea to have a protocol like that in place because the pathologist is then responsible for evaluating the patient and directing the person to further care," he adds. "Also, if the lab doesn’t follow the established protocol in every case, or it becomes too burdensome, the lab isn’t following its own best practice or possibly standard of care."

(CAP general counsel Bierig takes a different view. He says the liability risk of not notifying the patient of significant test results after being unable to reach the ordering clinician is so great that, as a risk management matter, pathologists are well advised to have such a protocol.)

From a patient safety and quality-of-care viewpoint, Dr. Leung says, someone does have to make sure patients get their critical or significant test results so they can make decisions about their care. "The buck has to stop with someone," he says. "The question becomes more one of how far down the line the pathologist should be in terms of being that person."

In Dr. Leung’s view, once a pathologist gets notified that the critical value cannot be delivered to the ordering clinician, the pathologist’s next call should be to the patient. "There’s no sense having a system where the pathologist re-pages everyone that the lab staff has already tried to reach. While the pathologist isn’t a direct-care provider, he or she must explain to the patient the test result and advise him or her what to do, which is, in most cases, probably to go to the emergency room," says Dr. Leung.

Some legal experts say pathologists might consider taking additional steps to ensure a patient receives critical findings when they have reason to believe a particular clinician won’t pass the information along to the patient.

"Keep in mind that when things go badly in a pathology case, they usually go very badly," says the Arizona Medical Association’s Sonne. "And a jury will measure the [patient’s] loss of opportunity for treatment of a serious condition ... against what the pathologist knew and could have done to prevent it," he adds. "The jury might ask how tough it would have been for the pathologist to follow up with a clinician or to ensure that the clinician communicated with the patient."

In such instances, for example, Sonne suggests the pathologist could document a call to the clinician saying, "Please confirm that you have notified this patient of the test finding, and if I don’t hear from you in x amount of time, I will notify the patient myself."

What would constitute sufficient grounds for a pathologist to believe or suspect that a clinician wasn’t passing on laboratory information to a patient? According to Bierig, one situation might be where the pathologist is aware of complaints that a clinician was not acting on laboratory findings or not bringing the findings to the attention of the patient. Another possibility: "The pathologist might be aware that the clinician had some psychological or substance abuse problem that might suggest that the clinician wasn’t using the laboratory data responsibly."

Direct-access testing, or DAT, definitely poses a scenario nario where the laboratory or pathologist bears responsibility to inform the patient of any aberrant test results and suggest the need for followup care. Colorado, for example, allows DAT, so patients routinely come to the University of Colorado Health Sciences Center to self-order tests like a CBC or lipids, says Dr. Lepoff. In such cases, the laboratory performs the test for the patient and makes sure he or she gets the report. "Someone talks to the patient about the report to say everything is normal or to say that a test result looks a little worrisome, so please talk to your doctor or clinician about it," he says.

In cases where the lab finds a critical DAT result, such as highly elevated white blood cell count or serious anemia, a staff person contacts the patient immediately, says Dr. Lepoff.

What if a patient self-orders a test that produces a dire result, but he or she never contacts the lab to receive it—and the lab can’t find the patient to let him or her know?

In these circumstances, Dr. Lepoff believes the lab has to take reasonable measures to contact and inform the patient. "The lab can try to mail the result, send the result by certified mail, search national computerized public directories for the person who gave a wrong address or phone number, etc.," he says. "But you don’t have to hire a private detective or try to find the patient in Alaska if that’s where the person chose to go."

The patient bears some responsibility for obtaining results of direct-access testing, in Dr. Lepoff’s view. "It’s not a one-way street. Laboratories and pathologists don’t have the responsibility of picking people up from cradle to grave and make sure they do the right thing to take care of themselves," he says.

In New York, which also allows DAT for FDA-approved over-the-counter tests, labs can provide test results only to the person self-ordering the test, says Thomas Rafalsky, JD, president of the New York State Clinical Laboratory Association. He says the DAT legislation was drafted in that manner because of the potential for a lab to pass along test results to a physician identified by a patient when the physician has no current relationship with a patient—or may not know how to contact him.

In addition, labs in New York may tell the person receiving DAT only that the test is outside the normal reference range and suggest the person consult with his or her health care provider, he says.

How do New York labs handle critical DAT values? "Some labs suggest the person see someone on a list of physicians who have agreed to be consulted in such a situation," Rafalsky says. "But the laboratory cannot give patients advice about what to do" about the abnormal test result.

The patient safety movement may play some role in helping to educate patients to take on more responsibility for following up on their diagnostic tests and to never assume that no news is good news. Dr. Lepoff agrees to some extent: "There’s an implied obligation on the part of patients to take an active part in their care and to act as an advocate for themselves," he says, "but it’s not a legal obligation. For example, does a passenger on an airline have an obligation to ensure that the pilot is knowledgeable about weather conditions during a flight?"

Dr. Gatter says that some defense attorneys in malpractice cases have tried to argue contributory negligence on the part of a patient who didn’t check back on a Pap smear result. And that argument hasn’t gone very far in a court of law because the patient acted reasonably, concluding the Pap must have been normal. "That is not to say," says Dr. Gatter, "that courts never allow contributory or comparative negligence defenses to proceed, because they do, but the standard is whether the patient acted reasonably."

Yet the patient can be a key operator in patient safety if he or she is taught to ask about and know the diagnostic test results and what those mean for health status and followup care, says Dr. Lepoff.

The lab staff can say to the patient: If you don’t hear back about this test result, call this number. In Dr. Lepoff’s view, "That type of effort is especially important in the U.S. health care system, which provides some of the best care in the world, but at the same time is a fragmented and sophisticated system. And it’s easy for information to fall between the cracks."

Karen Lusky is a writer in Brentwood, Tenn.