An intelligent, interactive textbook on medical diagnosis, DXplain serves as a clinical reference/decision support tool and differential diagnosis educational resource. When clinicians enter a patient description—physical exam findings, patient history and demographics, or laboratory and imaging results—DXplain generates a list of diagnoses. The system then guides clinicians in selecting additional diagnostic questions to ask or tests to order and explains the significance of any finding in relation to a disease.
The DXplain knowledge base contains quantitative information about the relationships between diseases and clinical findings. Disease descriptions display in a consistent format that leverages the data by presenting clinical findings in order of the frequency with which they are found in the disease (usually, sometimes, rarely) and also shows which findings would be diagnostically helpful (findings that support a given diagnosis or findings that make a particular diagnosis less likely). DXplain also allows side-by-side comparisons of two or more diseases to quickly see how each is similar or different. This type of information and display in a consistent and organized format is not available in traditional textbooks.
Developed and supported by the Laboratory of Computer Science at the Massachusetts General Hospital, DXplain has a new streamlined easy-to-use web interface that includes links to the medical literature and other educational resources via PubMed® and Google. Previous versions have been used by over 100,000 practicing physicians, house staff, and medical students since 1987.
No-charge access to DXplain is available to all NIH staff
through the NIH Library Databases webpage.
![DXplain Homepage](https://webarchive.library.unt.edu/web/20090813113106im_/http://nihlibrary.nih.gov/NR/rdonlyres/EB86CF84-DBDB-4407-9AB0-779130243E39/18773/dxplain3.JPG)
The DXplain Homepage
January 2009
The developers would like to hear about your experiences with the service.
Click the "Feedback" link on the DXplain homepage to submit comments or questions.
Please direct general questions or requests for additional information to
Reference & Information Services (nihlibrary@nih.gov or 301-496-1080).