Exposing the Test Results No One Sees
Pending Results Often Left Out of Discharge Summaries
By Bill Malone
Laboratorians and clinicians have known for a long time that discharge summaries could do a better job of helping patients transition to their follow-up providers once they leave the hospital. Until now, however, the magnitude of the problem wasn’t clear, especially for pending test results that the follow-up provider might need to know about. A new study described in this issue of Strategies took advantage of a regional health information exchange to explore the true extent of the problem.
Patients are frequently discharged from the hospital with unresolved medical problems requiring outpatient workups. Yet studies of hospital discharge summaries have found that about a third contain errors or inaccuracies. Others have documented that up to 8% have test-related follow-up errors in communicating results that are listed as pending on the discharge summary. But until a recent study, none had been able to accurately examine just how often the discharge summaries missed pending results altogether, sometimes leaving important results out in limbo where no one would come looking for them (J Gen Intern Med 2009;24:1002–1006).
The new study used the Indiana Network for Patient Care (INPC), a health information exchange that serves as a digital vault for patient health records from across the state. INPC is able to integrate data that originates from different databases, including lab test results and hospital discharge summaries. Apart from the health information exchange, these discrete electronic systems are not programmed to communicate—part of the reason that previous researchers who didn’t have the advantage of such a system weren’t able to look for what was going on outside of the discharge summary.
Missing these results on the discharge summary means they might not be appropriately followed up, explained Martin Were, MD, a Regenstrief Institute investigator, assistant professor of medicine at the Indiana University School of Medicine, and an author of the study. "Once a patient is discharged, the presumption is that the outpatient provider is going to take over the patient’s care. But if the outpatient provider is unaware that a test result is pending, they would not even know that they should be looking out for the result."
Were’s retrospective study examined data from 668 patients drawn from two hospitals participating in the INPC. After culling data from the INPC database, the researchers were able to connect patients through their electronic medical records (EMRs) and compare the discharge summaries to lab data. Surprisingly, only 13% of the discharge summaries documented all of the pending tests for the patient, with only 25% mentioning pending tests at all. "It was very striking to find that only 16% of tests with pending results were mentioned in the discharge summaries," said Were.
The researchers were also surprised by how often the discharge summaries failed to mention actionable results. Using a review team of physician investigators and a special algorithm, the researchers considered a test "actionable" if any of the following criteria were met: if the result required starting,
discontinuing, or changing treatment; required a new test or a change in diagnostics testing; or required scheduling of an earlier appointment or a referral to a specialist. Only 10% of the missing tests met these criteria; however, due to the large gap in what appeared on the discharge summaries versus what tests were actually outstanding, only 28% of actionable tests made it into the discharge summaries.
This finding suggested to Were that the missing tests were not just a matter of providers deciding a pending result didn’t merit being included in the discharge summary, but that a systemic error was driving the problem. "The EMR is not directly linked to the admission-discharge-transfer system, and the lab information system is separate from both of these," explained Were. "To determine what’s pending, you have to be able to use all three systems, and more often than not, these systems do not communicate with each other." Though Were and colleagues drew from the health information exchange to compare these sets of data and perform the study, this big picture isn’t readily available to clinicians when they are completing the discharge summary for a patient.
A constellation of factors converge to make it less likely that a pending test result will show up on the discharge summary, said Were. "It’s multiple things that keep all the information from being there—it’s not really the provider that’s the problem, it’s more the system that’s the problem." For one, providers don’t always have the time to figure out which tests are back or not. Second, they might not know what was actually ordered: with people working in teams, someone might have ordered the test at night while the primary provider was away. Finally, sometimes a provider makes a conscious decision not to include a pending result, thinking that it’s not that relevant, Were explained. "For instance, a provider might have a pending chemistry test, but they also have one from the day before and everything looked fine. Still, the main issue is that it’s difficult and time consuming to know every test that was ordered for which a result is still pending at the time a summary is being prepared because of these disconnected systems."
The primary focus of Were’s research was to underscore the sheer size of the discharge summary problem. However, further research that’s currently underway might shed light on some possible solutions. Were highlighted the systemic nature of the problem when, in an experiment separate from the published study, his team changed the discharge summary template, forcing providers to document what they thought was pending. They couldn’t finish the discharge summary unless they documented something. "With this, they at least had to take a moment and include what they thought was pending and important to include in the discharge summary," said Were. "And things did not really improve much. We’re still analyzing that data. So even if discharging providers have the best intentions, they just don’t have the ability to tell what’s truly pending."
As part of an Agency for Healthcare Research and Quality (AHRQ)-funded study Were is involved in, the next step will be to take the value of the comprehensive view of patient data that the INCP INPC provides for the published study and allow providers clinicians access to this complete picture when preparing the discharge summary. This new system will automatically show the physician what which tests are pending and, if possible, automatically send the results of the pending tests to the follow-up care provider listed on the discharge summary.