From Clinfowiki
Jump to: navigation, search

Free and open source software (FOSS) are software that are free and open source. Free refers to the liberty to use and or distribute, not the price or value. [1] Proprietary software licenses usually take away these rights. Open Source refers to the source code of the software being available. [2]


Some believe that FOSS software licensing is an especially good fit for medicine and medical practice because of its fast-changing nature, relatively iphone photography small market and ability by 3rd parties to readily and directly scrutinize software for privacy, security and interoperability problems. In addition the community building and sharing aspect is akin to the publishing and sharing of medical knowledge for the greater good.

There is a popular though very incorrect impression that FOSS developers and users do not take the idea of intellectual property seriously. Nothing could be further from the truth. FOSS advocates believe very strongly in intellectual property rights and aim to protect them using appropriate licenses based on international copyright law.

Free and Open Source Enabling Technologies for Patient-Centric, Guideline-Based Clinical Decision Support: A Survey.

Turchin A, Kolatkar NS, Pendergrass ML, Kohane IS. Medical Informatics on the Internet in Medicine. 2007 Jun;32(2):93-102. | PDF of Article


What is the current state and future trend of Free and Open Source Software modeling tools to create clinical guidelines suitable for clinical decision support software?


The authors searched PubMed, "Major biomedical informatics websites", the Internet, as well as relied on the their own knowledge to find suitable systems.


Clinical Practice Guidelines (CPG) can improve the quality of healthcare. CPG is a particularly important component of computerized medical decision support. However, computerized is CPG not currently in widespread use. What is needed are tools to help integrate it into clinical work flow. The tools to do this need to be "affordable, adaptable, and interoperable." These qualities are not found in proprietary software, but are found in Free and Open Source software. The authors then provide a brief explanation of Free and Open Source (FOS), its development community, business model, and proponents.

The majority of the article discusses the main concerns with representing CPG's in software. It lists various FOS tools that can help and briefly describes their fundamentals and use. Some of those concerns and the tools that address them include:

  • Terminology Standards that are FOS, or that should be considered for use are: SNOMED-CT, LOINC (free but not open source), UMLS, and OpenGALEN.
  • Data Exchange Standards that are FOS, or that have FOS tools are: Hl7 and DICOM, and ISO/IEEE 11073. (While ISO/IEEE 11073 is not FOS, it has some FOS tools to check for compliance.) The authors also include a list of FOS tools to help with these data exchange standards.
  • Health Record Standards that are FOS are: HL7's RIM, CDA, EHRcom, OHF, OpenEHR, and EGADSS. The authors also include additional information about FOS tools that can help with these.
  • FOS Guideline Modeling Tools:
    • Model Centric tools, that help domain experts create CPG's, include: Protègè is discussed at some length along with its fundamental approach and projects. AsbruView allows those with little training to compose in the Asbru language. Tallis tool is based on the PROforma language.
    • Document Centric tools, which help create CPG's from documents, include: GEM Cutter helps translate text into Guideline Elements Models. DELTA/A allows for CPG's documents placement into XML. Stepper CPG's creation into XML or even code. Uruz is a web based guideline tool based on Degel.

Main Result:

There are many sophisticated FOS tools for creating CPG models for decision support. Some are successful because of the developer community, but are not widely implemented because of the few clinicians using the tools. While there was a lack of standards in the past, there is now a move toward standard terminologies, communication, and architectures. Natural language and machine learning techniques are beginning to help extract information from text. FOS is growing in this area, has learned from its mistakes and successes, and will likely help provide for future CPG use in computerized decision support.


While I can not vouch for the veracity of the information, it is nice to see these sorts of lists as they can really help one begin to research these tools and standards. However, this article goes much further by helping to show one how each tool fits into the over all framework of the process. I find the authors' thoughts on FOS's strengths over proprietary software as well as future prospects interesting, but perhaps not well supported in the article. As a survey I found it quite helpful. It is brief, and somewhat dense to someone new to the field, but to the point.

--John Norris 23:26, 30 January 2008 (CST)