Handoffs from inpatient to outpatient care are a known problem area in hospital medicine. One of the biggest opportunities for improvement may be documenting and communicating the results of tests still pending at discharge.
“All pending test results have the potential to be problematic in terms of communicating them to the right people, but at discharge, they can really fall into the gap,” said Christopher L. Roy, ACP Member, director of the hospitalist service at Brigham & Women's Hospital in Boston. “There really is a lack of clarity about who is ‘catching the ball’ as patients move from the high-intensity care environment of the hospital into an outpatient setting.”
A study published in the September 2009 Journal of General Internal Medicine helped quantify how bad the problem is. It found that of 696 patients released from two academic medical centers with pending test results, only 25% of their discharge summaries mentioned the tests, and only 13% documented all pending tests. In addition, only 67% of the discharge summaries made it possible to discern which clinician should receive the results.
These outcomes were not found to be associated with level of clinician experience, patient age or race, length of hospitalization or amount of time it took to return results.
Martin C. Were, MD, lead author of the study, said he and his co-authors were not terribly surprised by these findings.
“We knew there were problems with discharge documentation, and we wanted to put the information out there and start to work on ways to improve the process,” said Dr. Were, a hospitalist and medical informatics researcher at Regenstrief Institute, Inc., and the Indiana University School of Medicine in Indianapolis.
In a related study that appeared in Annals of Internal Medicine in 2005, Dr. Roy and his colleagues found that of 2,644 consecutive patients discharged from hospitalist services from February to June 2004 at two academic medical centers, 40% had pending test results, 40% of which were abnormal. Of those, 9% potentially required relatively urgent clinical action.
Dr. Were pointed out that even when hospitalists do get results, they don't always know what to do with them. “It may be as basic as not having the outpatient provider's phone number,” he said.
Chaim Bell, MD, a hospitalist at St. Michael's Hospital in Toronto who researches continuity-of-care issues, said part of the problem with results pending at discharge is “separating the wheat from the chaff.”
“We have to decide what we need to know before sending a patient home,” Dr. Bell said. “Uncommon procedures or those for uncommon pathologies that are not run daily really can be problematic.”
Dr. Were stressed that there is no magic bullet for this problem, and noted that it reflects a larger systems problem. He believes that greater awareness, education, institutional support, enforcement and system-based improvement measures are essential.
“The best of intentions aren't always enough,” he said. “There are many systemic hazards to getting these done right. We need to identify ways to make it easy to know what is pending and get that information into the discharge summary.”
He sees electronic medical records (EMRs) as key to making that happen. “The future lies here—in having an EMR that can easily show what is pending and then letting the discharging physician identify which tests the outpatient follow-up provider needs to see,” he said. “We wouldn't want to populate the summary with all pending tests, because many are daily information such as routine CBCs that have little effect on the ongoing management of the patient.” He added that EMRs could potentially be used to identify results that come back after discharge.
Having this capability sounds easy, he said, but most hospital systems are not designed to communicate with other systems in this manner.
“Some systems don't even tell the discharge provider that there are pending results,” he said. “Physicians need to demand this functionality as future systems are designed. Too often they get no say in what functionality the system should have before these systems are purchased.”
Dr. Roy is working to add a function to his facility's EMR that lets the discharging physician query what results are pending at any given time. “This is a new view, as opposed to just seeing what is finalized,” he explained.
Automatic population of arriving results or email alerts are useful too, he said, but conceded that receiving too many such emails could trigger “alert fatigue” because many would not be relevant to care decisions. “It is important to filter actionable alerts from nonactionable ones, but that is not easy from an IS standpoint,” he said.
Dr. Bell, who holds a chair in patient safety and continuity of care funded by the Canadian Institutes of Health Research and the Canadian Patient Safety Institute, has developed an electronic tool that functions as a reminder to discharging physicians about which pending results to flag.
“It encourages physicians to follow a template with a series of headings that serve as placeholders instead of the free flow of thoughts that often characterize dictated discharge summaries,” he said. “It makes it explicit what items need to be reviewed by the primary care physician later.”
The tool prints patients two copies of a discharge letter, one for them to keep and one for them to give their primary care physician personally, helping ensure it makes it into the right hands. A copy is sent directly to the primary care physician as well, but Dr. Bell said that using the patient as a courier is a proven method of improving timely information continuity.
“Using both mechanisms for information continuity makes it more likely that the PCP gets the information and also helps empower patients,” he said.
Dr. Were said he is studying using a health information exchange to gather results returning after discharge from multiple sources such as laboratories and forward them to the appropriate follow-up clinicians.
He described his institution's exchange as one of the best in the country, but said that since such technology is not available everywhere, some hospitalists need to use more traditional approaches to improve follow-up.
“Every practice is different, but hospitalists know their setting and are best placed to come up with their own solutions,” such as designating a staff member to monitor pending tests, he said.
Dr. Bell said that scheduling follow-up clinics to review pending test results is another possible solution to fill in the gaps that occur at handoff. Dr. Roy agreed that such clinics can be useful, especially in areas where access to primary care physicians is limited. However, he cautioned that it can backfire if local primary care physicians feel that hospitalists are encroaching on their turf.
Like Dr. Were, Dr. Roy believes that hospitalists who lack high-technology solutions should use the technology they do have available to be responsible for results. “Put it in your task list on Outlook to remind you to follow up. That is at least the fail-safe, particularly if there is a high potential for harm if something is missed,” he said.
“The bottom line is you need to work on this if you want to avoid missing tests that could affect patient management,” Dr. Were said.