Q&A: David M. Yousem on updating the peer review process

David M. Yousem, MD, MBA, director of neuroradiology and vice chairman of program development at Johns Hopkins Medicine in Baltimore, grew frustrated with the long delays associated with traditional peer review, so he looked into finding a solution.

Using a platform provided by Analytical Informatics, a Baltimore-based health IT startup, Yousem and other specialists worked with the Johns Hopkins Technology Innovation Center to develop a new peer review system, Advanced Peer Review. The latest version of Advanced Peer Review went live in 2015, and it incorporates numerous features Yousem and his colleagues thought were missing from other peer review systems, including the ability to provide case reviews within 24 hours of when the original exam was performed.

Now, at the SIIM 2016 Annual Meeting in Portland from June 29 to July 1, Analytical Informatics will be officially launching Advanced Peer Review as a software tool available to the public. 

Yousem spoke with Radiology Business about Advanced Peer Review and how it has changed his colleagues’ opinions on peer review in general.

Radiology Business: What is the primary benefit of moving toward a peer review system with quicker turnaround times?

David M. Yousem, MD: When the traditional peer review read was performed, you looked at a previous study of a patient who was returning for evaluation. So by and large, people would look back at a prior study that may be three months old, six months old or two years old. And then you would do the peer review and see whether there were any discrepancies. The disadvantage of doing it that way is, if there was something that was missed, there is a long period of time during which that process may advance—be it a cancer, an infection or a neurodegenerative disorder.

The advantage of the short-term prospective peer review we’re employing here is that it is of cases that have been performed within the previous 24 hours, so if there were any errors that were made, they were picked up within 24 hours instead of allowing a tumor or infection to go undetected or untreated for weeks, months or years. It just makes more sense to pick up discrepancies and errors as soon as possible.

What were your feelings toward peer review before you helped work on Advanced Peer Review?

For the vast majority of practicing radiologists, peer review had been a compliance issue. You had to do it because there was either a hospital regulation, a Joint Commission regulation or a society regulation. The way it was initially proposed and rolled out with the ACR peer review process, it wasn’t very satisfying. You didn’t get much closure around a case, it was anonymized in a way that you could not fix a problem that occurred with a patient, and, frankly, most people looked at it as a chore without any real patient benefit.

What we’ve evolved to is something a little more specific to the patient. We had notes that would be put in as far as what the errors, if there were any, were about, and we could collect data on physicians in a way that gave them feedback about errors they were making. With the latest iteration, we’ve really changed peer review to a process that is patient oriented. We’re really trying to affect patient care, and my faculty bought into it because they see the benefits of potentially finding errors that could really help a patient.

Since implementing Advanced Peer Review at your hospital, numerous patients have had their diagnosis updated as a result of the peer review. What happened in those cases?

The instances that have had the most impact were the ones in which the earliest findings of a stroke were initially missed and then picked up by peer review within a few hours in a timeframe where the clinician could still intervene and reduce the amount of brain tissue lost due to the infarction. This is sometimes a subtle finding. In addition, we had cases of aneurysms that were not detected, but were detected on peer review. These were not aneurysms that had bled, but they are the types of abnormalities in the brain that could lead to sudden death.

And I want to stress that the way the peer review works, it’s not uncommon for us to be reviewing each other’s reports within two, three or four hours. It’s not as if it’s restricted to being at 24 hours, it’s within the previous 24 hours.

One feature of Advanced Peer Review allows radiologists to request second opinions if they disagree with the peer assessment. Can you provide some insight about that feature?

Fortunately, we’re in a reading environment in which there is a large number of attendings in a reading room together doing neuroradiology. So if I need a second opinion on a case, it’s pretty easy for me to tap someone in the room and say, ‘can I get your opinion on this?’ The idea was, the proliferation of neuroradiologists or subspecialty-trained radiologists may not be such so that there is a large number of people who have that expertise, so in the situation where someone is remote, their study could be submitted for a review and you could get another opinion, even if they aren’t sitting in your work environment.

Another feature I want to highlight is the ability to change the percentage of studies that are reviewed by an individual. For example, if you have new people who are starting out and have less experience than older individuals, you could use the slider to adjust the percentage of their cases you would like to have peer reviewed.

And you can also adjust what percentage of the overall cases in your practice you want peer reviewed. So if it’s a marketing strategy, for example, to say all the cases in the neuroradiology group are double read for accuracy, that could potentially be utilized.

Are there any other features you would like to share information on?

One other element that has added to the benefits of this program is the anonymization. In the past, if I received feedback there was discrepancy on one of my prior reports, all I had to do was look back and see who looked at the report and know who reviewed me. There was potential for a back and forth in that way. A benefit of the current system is it is reviewing totally new cases within the previous 24 hours, so it really is anonymous. For some people, you know, that is an advantage.

In our group, we did agree that if there was a major discrepancy, the person reviewing the case would contact the original reader, discuss it with them and allow the original reader to make the decision about making an addendum and contacting the referring physician. So it has led to greater communication between faculty members about cases and some discussion about some of these discrepancies. It’s more of a team approach, and we’re working together.

This text was edited for space and clarity.