VisualDx is proud of its staff of dedicated medical librarians, who help research medical content and verify the accuracy of many kinds of data in the product. They work collaboratively with a team of physician contributors and software engineers to develop and review medical content.
Patricia Gogniat, MLS, AHIP, is our lead medical research librarian. In this guest blog, she writes about the value of diagnostic decision support for medical librarians. Patricia and other members of the VisualDx team will be participating at MLA '16 in Toronto.
Here's Patricia's blog:
Greetings, medical librarians! In this blog post I’d like to share information about VisualDx and explain how it can be used as a point-of-care resource, either by clinical librarians or library patrons such as nurses, doctors, and other clinicians.
As lead medical research librarian at VisualDx, I work collaboratively with our physician contributors, our team of researchers and editors, and the software engineers to build, review, and maintain content in VisualDx. Our editorial team consists of 2 medical editors, 4 medical librarian researchers, 3 in-house physicians, and many consulting physicians. The content that we create and manage includes texts on our diagnosis topics, finding-diagnosis relationships that power the ddx builder (more on that below), references, codes, and our reference interface terminology, which is mapped to standards, along with all sorts of other metadata for each of our diagnosis concepts.
Ever wondered how the ddx search builder works in VisualDx?
When you enter findings to build a differential of matching diagnoses, you’re running a structured query of finding-diagnosis relationships. Let me explain.
On the back end, all of the data is stored in a relational database as concept-concept relationships. These relationships are evidence based, each being documented with a textbook citation, journal article, or other source of medical expertise. The librarian research team, along with our physician team, is heavily involved in creating these relationships. We review the best available literature evidence for each topic and then, using controlled terminology, add the relevant findings to every diagnosis. Relevant findings may include signs and symptoms, drug associations, travel history, exposures, and more.
It’s very similar to an indexing process where relevant index terms are added to an article. Though users can’t see our database of literature evidence (proprietary), we basically perform literature reviews for all 2,700 (and growing!) diagnosis concepts in VisualDx. Physicians then review all of the finding-diagnosis relationships, which currently number over 110,000!
At the point of care, as clinical librarians, in what scenarios would you use VisualDx?
I’ll summarize a few major points here, but if you’re attending MLA, please consider joining our Sunrise Seminar on Monday morning at 7am in room 203D, where I’ll cover scenarios in much greater detail. You can register here!
- Building a differential. When cases are presented at morning reports, you can use VisualDx to enter case findings and build a differential diagnosis. Findings usually include signs and symptoms, but you can also search labs, medications, medical history, social history, and other risk factors such as recent air travel. We have guided workups available for certain chief complaints to prompt additional patient questions to help narrow the differential. Whether you pull up the differential yourself or teach students/residents to pull up the differential, it’s an important step to always consider “what else could this be?” – VisualDx helps clinicians avoid cognitive errors such as premature closure and incomplete data gathering.
- Confirm a diagnosis. While rounding and discussing a specific case, if there is a leading diagnosis, you can pull it up in VisualDx to look at related images and text and help confirm the diagnosis. The Look For and Best Tests texts in particular aid with diagnosis confirmation, as does visual comparison of the patient to our images.
- Patient engagement. Many clinicians use VisualDx to educate the patient about their diagnosis, which helps build confidence in the provider-patient relationship. It can be difficult for a patient to absorb all of the spoken information being conveyed by the clinician; showing the patient words describing their diagnosis or clinical images resembling their own condition can enhance understanding. Clinicians can also print a patient information sheet containing a specific selected image.
- “What’s that rash?” Though this falls under the building a ddx category, our rash workups deserve a special mention. The most traditional use for VisualDx, given our history as a company founded by dermatologists, is when the chief complaint is a skin lesion. You can start by entering “single skin lesion” or “rash” and use the guided workup to build a differential. For any patient with a skin complaint, VisualDx can help you select the lesion type and enter other related findings. The large collection of medical images is essential for comparing rash patterns among diagnoses and is unrivaled by any other resource.
- Therapy information. As of May 2016, we offer therapy recommendations for 1,500 out of 2,700 diagnoses. Many clinicians use VisualDx to look up the therapy recommendations for dermatologic diseases because they are concise and heavily reviewed by our dermatology editorial board.
I hope to see you at MLA! Consider attending the sunrise seminar on Monday if you’re curious about what VisualDx can offer medical librarians and how to use it at the point of care or how to teach it. If you’d like to learn more about the bones of VisualDx, come with your informatics questions! Register here to attend, or stop by room 203D on Monday May 16th at 7am.