Advanced Search

  • Topic
  • Specialty
  • Content Type


More Episodes

Also Related

< Back To Patient Safety
0 dislikes

< Hide

Comments For

Filling Holes in Electronic Test Result Follow Up


< Shrink

Add Your Voice

All comments are posted anonymously. Your comment will be attributed to: "Anonymous user."

post comment


Are you sure you want to delete this comment?

Filling Holes in Electronic Test Result Follow Up

By Tom A. Augello, CRICO

Related to: Ambulatory, Communication, Diagnosis, Electronic Health Records, Primary Care, Other Specialties

Guest Commentator

  • Susan Cornacchio, RN, JD; CRICO/RMF; Cambridge, MA
  • Hardeep Singh, MD, MPH; Houston Veterans Affairs Medical Center; Houston, TX
  • Dean Sittig, PhD; University of Texas; Houston, TX


In the Harvard medical system, one of the most persistent professional liability risks is the mishandling of abnormal test results. Problems related to physician follow up of test results appeared in one-fifth of out-patient diagnosis-related claims in a recent five-year period. This is more significant because out-patient claims now account for more than half of malpractice claims in the Harvard system. The most common allegations in these ambulatory cases are related to diagnosis.

Will sophisticated electronic health records solve the problem? New research in the Archives of Internal Medicine in late 2009 suggest the answer may be ‘no.’ Or at least, ‘not altogether.’ The findings take on added significance as the medical community prepares for increased pressure to implement electronic systems as part of health care reform in the U.S. The Archives study showed that, for abnormal critical imaging results, timely follow up was lacking nearly 8 percent of the time. This result held, even with advanced integrated electronic health records that sent automated alerts to the ordering physician.

What really surprised the researchers was the finding that doctors who acknowledged receiving the abnormal result by clicking on the electronic notice, did not do much better at acting on the information than their peers who never saw the electronic warning.

“We never expected that, in the acknowledgement group, we would see the same risk of follow up, same rates of lack of timely follow up as the unacknowledged group.”

Dr. Hardeep Singh was the lead author for the study, which looked at 123-thousand imaging results. About one percent of them generated an alert with an abnormality. Tracking software detected whether the alert was clicked on by the ordering clinician. The researchers later called the physician to see what follow up occurred if any. Dr. Singh, a research scientist at Houston Veteran’s Affairs Medical Center, says timely follow up failed for various reasons.

"One reason was that someone was looking at the alerts and got distracted or they were doing it in a hurry, maybe 6:00 in the evening on the way home, and decided ‘oh, I’m going to take care of it next day, let me make a note of it on a piece of paper.’ And of course the piece of paper gets lost or something. So that was one reason, but the second sort of very interesting reason was alerts go to two people many times.”

Sending an abnormal imaging result to two people should be better than one, right? Wrong. In Dr. Singh’s study, results that went to two physicians were twice as likely to lack needed follow up.

“So, for instance, if I’m the lung specialist and I order a CAT scan of a chest x-ray which is abnormal, the alert goes to the primary care doctor and the specialist. So what might happen is the specialist might think ‘oh, I’m the specialist and that’s the primary care doctor, so the primary care doctor might take care of it,’ or I’m like in another hospital, being a subspecialist I often go to other institutions or it might be a fellow who went to another hospital. And the PCP might be thinking ‘hey, the specialist ordered the test so they will take care of it.’ So there was sort of this shared, ambiguous responsibility.”

Dr. Singh says the test results they were tracking had clinical issues that required follow up.

“Yeah, most of them had more tests done. Nearly all abnormal results which did not have follow-up in four weeks but eventually had some sort of measurable clinical impact either in diagnostic testing of treatment.”

Involvement by trainees proved problematic as well. Another predictor for a lack of timely follow up was if a resident had ordered the test. These results often went to another physician when the resident’s rotation changed.

The complexity of the system and the difficulty delineating responsibility for patient follow up is seen often in outpatient evaluations done by Harvard malpractice insurer and patient safety company, CRICO/RMF. Susan Cornacchio is program director of ambulatory risk and office practice evaluation at CRICO. Cornacchio is an attorney and a nurse, and she recognized some familiar themes in the electronic test result study.

“With electronic medical records and with notification systems, a lot of times it is very hard to set up algorithms or design programs to really target the person who is responsible for the follow up. And I think from the liability standpoint, now you’ve just tagged three or four different people and if something goes wrong and there is a malpractice claim later on, then that gives the patient actually additional people to name in any kind of an action.”

Cornacchio’s team at CRICO has conducted evaluations of hundreds of out-patient and office settings since the late 1990s. Chart reviews and interviews help the insurer bring to the surface any process flaws that undermine the provider’s delivery of safe care. According to Cornacchio, most practices struggle in this area.

“What we’re seeing is when we look at test result management—and for our purposes is broader than just imaging studies, it’s also laboratory studies—we’re seeing that practices don’t always have a process in place to make sure that the right tests are ordered, that the tests are performed, that the results are given back to the office and the patient, that there’s documentation in the medical record that the patient got the results, the physician got the results, and then an action plan was developed, and the right follow-up or treatment plan was implemented.”

A factor in the Singh study that led to improved likelihood of timely follow up, was direct communication from the interpreting radiologist to the ordering physician. Cornacchio says that CRICO has learned from the malpractice claims data that a gray area with significant liability and patient safety implications involves incidental findings.

“The patient will be seen in the ER, perhaps for pneumonia or chest complaint. They’re evaluated as part of that work-up, they have a chest x-ray done, they find what looks like a spot on the lung or an area that’s a little bit troubling. The information may or may not make it to the primary care physician’s office and sometimes that information can get just filed with all the other routine paperwork. They would see the pneumonia diagnosis but fail to pick up on the other spot or the incidental finding that really required more aggressive follow up, and very often that’s missed. It can be missed for months or even years until the patient finally does have a chest x-ray or an imaging study that shows that they do have a malignancy, and at that point the patient goes back and retraces their steps and other x-rays that they may have had in the past and realize that they missed the opportunity for an early detection and possibly a cure.”

Practice protocols should guide which findings need direct communication between the radiologist and the ordering physician. She says that electronic systems should be molded by the people who will be asked to work within them.

“I think ideally for the multidisciplinary team, you’re gonna have representatives from the clinical world who can really understand the importance and the significance of different test results. And you would have people who are work-flow engineers and you would also have people in the IS/IT world who really know how to implement the systems here, particularly the electronics systems that need to be put into place to make this all work. I think a lot of times what we see is that the electronic medical records are created by maybe one physician surrounded by a lot of IT people or just IT people and they really miss the mark when it comes to understanding and integrating the clinical component of all this.”

Recommendations in the Singh study follow a similar path. They call for a multidisciplinary approach to creating electronic health record systems that effectively prevent test result management problems. A co-author, Dean Sittig, is an associate professor at the University of Texas at the Houston Health Science Center. Along with Dr. Singh, Sittig wrote a commentary in JAMA that calls for eight “rights” in building a good electronic health record.

“The first one is the right hardware or software and then you have the right user interface, the right personnel, the right work flow and communication, the right organizational characteristics, your right state and federal rules and regulations, and right monitoring of the systems, and of course the right content.”

Sittig says one of the most common significant mistakes people make is thinking they just need to fix one thing to make their electronic record system safe. But it’s often simply the area where that person has the most experience.

“Our biggest recommendation is that you need to sort of monitor that these things are happening so you can figure out if you’re making an improvement in this. One of the things we’ve discovered or used these eight rights for is trying to help people to understand that, you know, that’s a very complex problem that we talked about in the other paper from the Archives and there’s not a simple solution to it. There’s a multitude of things that have to be put in place to get that fixed.”

January 1, 2010
0 dislikes

< Back To Patient Safety