Home Q&A A new look at an old problem: missing clinical information

A new look at an old problem: missing clinical information

May 2005

Published in the May 2005 issue of Today’s Hospitalist.

When it comes to giving primary care physicians the information they need to care for patients, a growing body of literature has shown that in too many instances, critical information is nowhere to be found.

Missing information is a particularly vexing problem for patients who have been hospitalized. Discharge summaries often show up late or without enough detail to be useful, frustrating primary care physicians and patients alike.

While hospitalists find themselves in the hot seat when patient information goes missing, a new study finds that the problem goes beyond hospital medicine and is more complex than many people realize.

The study, which appeared in the Feb. 2 issue of the Journal of the American Medical Association, surveyed 253 Colorado primary care physicians about patient visits between May and December of 2003. Researchers found that clinical information was missing during 13 percent of all visits.

Perhaps not surprisingly, lab reports and dictations/letters topped the list of missing information. Just as unsurprisingly, the physicians who took part in the study estimated that missing information would probably hurt patient care in 44 percent of all patient visits, and produce delays in care or lead to additional services nearly 60 percent of the time.

What surprised researchers, however, was the smaller-than-expected impact that electronic medical record (EMR) systems had on missing information. While physicians with access to a fully functional EMR were 2.5 time less likely to report missing information, their EMRs didn’t completely solve the problem.

Why? In just over half of the instances where clinical information was nowhere to be found, it was outside of the physician’s system because the patient had visited a hospital or another physician outside of the health system or network.

That result is significant, explains lead author Peter C. Smith, MD, assistant professor of family medicine at the University of Colorado Health Sciences Center in Denver.

For one, it points to the difficulty that primary care physicians who need that lab value or test result face in tracking it down. Researchers found that in about a quarter of the instances in which information was missing, about one-third of physicians spent five minutes or more trying to locate the elusive information, while about 10 percent spent more than 10 minutes.

But the fact that so much missing information is outside of the physician’s system or network also raises important questions about how much help EMR systems can provide in eliminating missing information. While physicians can generally find missing information quickly if it’s in an EMR that they can access, the technology is useless if the data are locked away in the EMR of another hospital or health system.

While the research did not focus specifically on patients who had been hospitalized, Dr. Smith talked to Today’s Hospitalist about what the study may mean for hospitalists.

What were some of the common reasons that patient information was missing?

We found that there were relatively few predictors of when information would be missing. Patients with multiple medical problems, however, tended to have information missing more often, which is common sense. We saw a direct correlation between the complexity of their care and the likelihood that they were missing information.

One example from my experience is a patient with multiple medical problems who had extensive diarrhea and lost 20 pounds. The patient was hospitalized, put on parenteral nutrition and underwent an enormous evaluation.

It turns out she was taking a medication that caused a type of colitis. She was put on the drug by a new specialist that no one else at the hospital knew about. The patient was in the hospital for 10 to 14 days. If people had known she was on this medication, the evaluation would have been a lot easier, and the hospitalization probably could have been avoided.

The patient was elderly and taking 10 or 15 medications. In general, she did a good job of keeping up her information, but given her complexity of care, she couldn’t keep track of everything.

It’s not clear whether patients are going to be the best judge of what information is or is not important. As far as the patient in the above example was concerned, the drug in question was just a little arthritis medication. That medication turned out to be very important, but she didn’t think it was essential enough to mention.

Your study found that in about 10 percent of the instances that information was missing, physicians spent more than 10 minutes trying to fill in the information void. What are some of the challenges these physicians faced?

One patient from my practice visited the ER for chest pain at a hospital outside our system. He had a very vague discharge summary. The note said that the patient had atypical chest pain and was given instructions to take medication for reflux disease. The problem was that the discharge summary didn’t specify what kind of evaluation had been done, what kind of lab tests had been done, and so on.

The physician in the office spent at least an hour trying to get the patient’s history. That included time on the phone trying to get records faxed and filling out HIPAA paperwork. If the patient had had a more complete discharge record, the follow-up care would have been a lot easier.

Did you find that patient information was more likely to be missing when patients had been in the hospital?

We didn’t look specifically at hospitalization as a factor in missing information, but if you look at the top three categories of missing information “laboratory results, letters/dictations and radiology results “all can arise from hospitalization.

We also looked at whether or not physicians had electronic access to information from the hospital. Although a high number did seem to have that access through some type of electronic system, it didn’t make any difference in the final number of patients whose information was missing.

We don’t have information from the study to explain that, but you can conjecture a couple of things. One is that if dictations aren’t done in a timely fashion and the patient is visiting the primary care physician a few days after discharge, physicians will not have the information they need, even though it should be available electronically. Another is that those systems are too cumbersome for busy daily practice.

Your research also found that EMR systems often weren’t helpful because the missing information was located outside of the physician’s health system or network. How widespread was this problem?

More than half of the episodes involved care outside of the physicians’ clinical systems. We knew this happens, but we were surprised by just how frequently it happens. It’s like there are lots of cooks in the kitchen.

Our study showed that when physicians work in systems with well-functioning EMR systems, they tend to have better access to patient information. That said, missing information was located outside the physician’s system in about half of all cases. That shows that while EMR systems are necessary, they are going to be an insufficient step to fix the problem of missing information.

The fact that more than half of the episodes involving information that’s outside of the typical clinical system in which outpatient physicians work says one thing: Until we get very good portable and usable information systems, we face an issue that is one of shared responsibility. We need to work together to figure out a way to get information to each other to cope with the cases where information is outside of the system.

Were you surprised by the scope of the problem of missing information?

We were surprised by how accommodated we’ve become to the problem. People tell the story of the frog in boiling water, where if you put a frog in water and turn up the heat gradually, the frog won’t notice how hot it is and will boil to death.

We’ve reached that situation with medical information. We’ve reached the position where people have said of course there is going to be missing information in our system, and they have grown used to it instead of expecting something better.

This is a symptom of a much larger illness in the system, which is that we don’t have a system. We have little clusters of independently acting units that may or may not talk to each other.

We need to work out systems of information exchange that work well, whether it’s encrypted e-mail or fax or phoning people. And while the good old-fashioned thing of talking on the phone is increasingly rare, it is just as effective when it happens.