Published in the September 2005 issue of Today’s Hospitalist
Do you know which of your patients are discharged from the hospital with unresolved test results? Even more importantly, do you know which of those test results could potentially change those patients’ diagnosis or treatment?
A new study found that while nearly half of patients leave the hospital with test results still pending, relatively few of those test results are important enough to warrant changing a diagnosis or treatment plan.
The problem, however, is that even when a result does require action, physicians are often completely unaware about the result, an outcome that raises important questions about patient safety.
The study, which was published in the July 19 Annals of Internal Medicine, examined 2,600 patients discharged from two large academic teaching hospitals in Boston during a five-month period in 2004. Researchers found that 41 percent were discharged with one or more lab or radiology tests pending. Of those tests, 9 percent were thought to require some sort of clinical action.
Hospitalists might be relieved to hear that the number of test results deemed to be urgent was relatively low. During the five months of the study, for example, only 15 individual test results were rated as urgent, and only 35 results were said to be important enough to change patient care.
But when the study surveyed attending physicians “a group that included hospitalists, primary care physicians and residents “researchers found that physicians were unaware of almost two-thirds of results considered actionable. Even worse, the clinicians were unaware of 8 of the 15 results that were considered urgent, and 24 of the pending results that changed patient care.
Christopher L. Roy, MD, lead author of the study and associate director of the hospitalist program at the Brigham and Women’s Hospital, views the results with ambivalence.
“We were reassured by the data that there were relatively few actionable tests results,” he explains. “Out of 2,000 pending test results, we as reviewers thought that about 10 percent might be clinically significant. When we surveyed the physicians caring for these patients, only about 15 of these results turned out to be urgent, and 35 turned out to be truly actionable based on the clinicians’ impression.”
Dr. Roy is quick to add, however, that the study points to a potential trouble spot. “We were fairly reassured that we were doing a good job in that relatively few important results were lost,” he explains, “but the fact that any clinically significant results fell through the cracks isn’t acceptable to us.”
An awareness problem
Surprisingly, some inpatient physicians were not only unaware of a test result, but they didn’t know that a test had even been ordered. When researchers asked physicians if they were aware that the pending test had been ordered, for example, 25 percent of inpatient physicians said they were not aware.
“Hospitalists are presumably talking to their team every day about what orders are being written on a patient and what tests need to be ordered and are reviewing these orders in the computer,” Dr. Roy explains, “so they should be aware of all the tests that are ordered. But the reality is that with multiple team members writing multiple orders on multiple patients, they are not aware 100 percent of the time.”
When researchers asked physicians if they were aware of the results of tests, 70 percent of inpatient physicians said they did not know. Two weeks later, only 46 percent of primary care physicians were aware of the test result.
Because researchers waited two weeks before asking primary care physicians if they knew the results of tests, Dr. Roy thought more would have been aware of the test results. “The fact that it wasn’t 100 percent two weeks later was a little disturbing,” he explains.
It turns out that housestaff showed the highest awareness of whether a test had been ordered. Because residents write all orders for their teams’ patients, Dr. Roy says, that should come as little surprise. What was puzzling, however, was that those same residents were no more aware of test results than attendings.
And while it would stand to reason that attending physicians who work on teams without residents “and therefore order their own tests “would be more aware of tests that had been ordered, that wasn’t the case. “On nonacademic teams where there were no housestaff,” Dr. Roy says, “awareness was no better.”
A lack of urgency
While awareness about tests and their results was fairly low, so was the urgency of most of those test results.
Of the results that physicians rated as urgent, many were microbiological test results that required starting or changing antibiotic therapy. “These sensitivities frequently come back after the patient has left the hospital,” Dr. Roy says.
Results that required action but were not urgent included incidental findings of a pulmonary nodule, serologic testing that turned up Helicobacter pylori in patients with GI bleeding, and iron deficiency.
While few of the pending test results revealed life-threatening problems, Dr. Roy worries about the potential for problems. The study estimates that in a large hospital with 30,000 discharges a year, 270 tests results requiring some kind of action would be missed each year.
“What we’re worried about is the volume of test results,” he says, “and that with such a volume, some important results will fall through the cracks. The question is what kind of resources will be necessary to prevent the few that can cause patient harm.”
Dr. Roy notes that the study raises another concern. While the results of thousands of tests were pending when patients left the hospital, their care was affected in only a handful of cases.
“One message is that we should possibly be more circumspect about the tests we order for inpatients,” Dr. Roy says. “We should try to be a little more judicious about tests that are ordered, and maybe even cancel tests that were ordered during the admission that are no longer relevant.”
Too much volume
Dr. Roy says the study also makes a strong case for why physicians need a system to make sure that test results don’t fall through the cracks.
“You can’t rely on the human factor here because of the large volume of test results,” Dr. Roy says. “It’s a frequent occurrence that patients go home with pending tests, so we really need to put systems in place as a fail-safe.”
He adds, however, that the physicians in the study had access to a top-notch electronic medical record system that gives all physicians full access to patient information no matter where they’re located. Physicians without access to such high-tech software would potentially have even more difficulty knowing about tests and their results.
“If you’re seeing a patient in the clinic and you have 15 or 20 minutes for a follow-up visit and you’re sifting through all this information on a complicated admission,” Dr. Roy explains, “you can see how it would be very easy to miss a lab result.”
That possibility was clearly on the minds of the inpatient physicians in the study, who overwhelmingly said they were unhappy with the systems in place to follow up on test results. Dr. Roy says it was that lack of confidence that triggered him and his colleagues to undertake the study in the first place.
“This is the kind of thing that keeps us up at night,” he explains, “that we’re going to send somebody home with a pulmonary nodule and the primary care physician is not going to find out about it, and the patient is going to present with cancer a year later.”
Edward Doyle is Editor of Today’s Hospitalist.