pub_ajrCover

American Journal of Roentgenology (AJR)

AJR InBrief Banner 


AJR August 2018





Elizabeth H. Dibble
Alpert Medical School of Brown University
Providence, RI

“Automated Delivery of Clinical Follow-Up to the Radiologist via E-Mail: Feasibility Study of a New Information Technology Algorithm”

Using an algorithm to generate high-yield follow-up data for radiologists has potential to improve patient care by facilitating radiologist engagement and self-assessment, according to a study in the August 2018 issue of the American Journal of Roentgenology (AJR).

The authors developed an automated process for radiologists to obtain clinical follow-up on radiology reports via HIPAA-compliant e-mail and determined what follow-up data were collected and whether they were relevant to the radiology report. Researchers led by Elizabeth H. Dibble of Alpert Medical School of Brown University developed an interface engine that allows radiologists to flag reports for follow-up at the time of dictation and receive a HIPAA-compliant e-mail with the radiology report and associated pathology, cytology, or endoscopy reports.

During one month of the study, more than 91 percent of all embedded reports generated an email containing the requested follow-up information. More than 90 percent of the emails sent included relevant follow-up information. More than 93 percent of pathology reports included with emails were deemed relevant. More than 90 percent of cytology reports, 88 percent of surgical reports and 75 percent of endoscopy reports included with emails were deemed relevant.

The authors noted that the algorithm provides personalized self-assessment that relates to specific imaging studies chosen by the reading radiologists, both for routine and unusual or perplexing cases. Radiologists can also create contemporary teaching files to educate medical students, residents, and colleagues. Our model allows correlation across diagnostic fields, monitoring of report accuracy, and monitoring of system usage. In addition to evaluating diagnostic accuracy, proceduralists can use the system to assess yield and concordance with imaging findings and can prompt addition of concordance addenda to biopsy reports.

In this Q&A, Dibble discusses the authors’ ideas behind the study and the outlook for further research.

How did you and your co-authors come up with the idea for this project?

Dibble: It is challenging for providers who are not at the center of a patient's care to get follow up on suspected diagnoses, and radiologist fall squarely into this group. Many of us have developed methods to follow-up interesting cases, but most systems are time consuming and inefficient. One of the authors (Jonathan S. Movson) is the Director of Imaging Informatics in our department and became aware of the capabilities of our enterprise interface engine (Cloverleaf) to potentially automate the process.

We were fortunate that our IT department was prepared to dedicate the resources to develop this capability. We then designed the algorithm described in the paper based on our own experience of the most commonly requested follow up data for radiologists. We were constrained by the formatting of the information that crossed the Cloverleaf interface, but fortunately the most commonly sought-after information (Pathology, Cytology, OR reports, Endoscopy reports) were accessible, and the algorithm was designed around those data points.

What should readers take away from your article?

Dibble: The EMR is often seen by providers as a hindrance to patient care because of the time required to enter information into the system. This is an example of how it is possible to leverage an existing enterprise interface engine and the EMR to save the radiologist time and enhance his or her ongoing education by providing automated follow up to diagnostic questions.

What recommendations do you have for future research as a result of this article?

Dibble: We are in the process of building new algorithms for our radiologists so that they can request automated follow up for additional types of information. Examples we are considering are MRI reports for imaging studies on which follow-up MRI was recommended, diagnostic breast imaging results for BIRADS 0 screening mammograms, and discharge summaries for perplexing cases.

Is there anything else AJR readers should know?

Dibble: The algorithm we describe is very popular among our radiologists, and there is no reason why similar follow-up algorithms cannot be created for other provider specialties. For example, emergency room providers should be able to request a copy of a discharge summary on an interesting case that required admission, and cardiologists should be able to request the surgical findings and pathology reports of patients with abnormal diagnostic studies who subsequently went to the operating room for definitive treatment.

The follow-up algorithm that we described is a great educational tool, and we hope that other institutions with the resources are encouraged to build similar systems for follow-up. As more and more medical applications are integrated into one EMR database, the need to build this capability at the interface level will be less desirable, and it will become increasingly advantageous to make the capabilities native to the EMR.


 

Do you have a story idea for InPractice?
Contact Mike Mason, Communications Manager, at mmason@arrs.org for more information.