by Brian Klepper

Last week The Health Care Blog ran two articles about new wiki sites that will develop and continuously update medical information. (A wiki is a “content collaborative” that allows anyone (or anyone authorized by the site) to contribute or modify content; Wikipedia is the best known example.) In Medicine Meets Wiki, Jane Sarasohn-Kahn brought our attention to MedPedia, a collaboration between major academic institutions and governmental agencies to clearly describe the entirety of current medical knowledge.  Then Bob Wachter described Google’s new Wikipedia competitor, Knol, and suggested sites like this could threaten the stranglehold that traditional medical journals have had on emerging information.

As Wikipedia has so effectively demonstrated, Wikis can become go-to authorities on nearly every topic.  However, accuracy and credibility problems may arise when topics are not moderated by subject matter experts. The possibility that information may be less than optimal is a fatal flaw for sites that would presume to offer information that can drive clinical decisions.

So these new health care sites address the accuracy issue with expert panels from every discipline who jury the content creation. Using this model, any professional (and conceivably, non-professional) who brings sufficiently strong arguments, backed by evidence (literature or data), should be allowed to contribute to or modify content.

What makes these efforts particularly interesting is that, through a collaborative Web-based process, they attempt to distill and document the current best knowledge about any topic. In health care, the goal is easily accessible state-of-the-science information, the equivalent of on-going medical/scientific review articles that detail what we know and don’t know about life and care processes. MedPedia and Knol are merely two in a line of health care wikis. The Joint Commission established WikiHealthCare, an interactive forum for health care professionals. Clinfowiki is devoted to clinical informatics. And of course this blog’s readers will be familiar with the Health 2.0 wiki, which has assembled information about that burgeoning sector.

While Health 2.0 generally uses Web-based collaboration to achieve some health care objective, I see it in terms of two broad trends. The first are knowledge, product and service exchanges. Patients Like Me, Sermo, and Carol all fall into this category. They offer platforms that facilitate the organized aggregation of information. People can learn about a topic, or can contribute information that deepens the information readily available. In their present forms, Knol and MedPedia also can be understood as information exchanges.

But the second type of Health 2.0 platform will be stronger and more utilitarian, because it will assemble knowledge or data, and then analyze and reformulate it, presenting the results in ways that support decision-making. For example, Oncology Metrics aggregates clinical and administrative data across oncology practices, feeding the results back to the contributing practices to help them  better understand and manage the clinical, operational and financial decisions that are part of everyday practice. Other organizations are bringing together subject matter experts to describe the rules that
govern specific complex systems, or are aggregating health system quality data to put together a “Travelocity-type” site for medical tourism. The possibilities are endless.

MedPedia, Knol and other health care wikis are exciting developments, because they will help us organize and maintain a unified knowledge base of the best current medical and health care information.  But a related and more thrilling advance is just over the horizon.  We’ll soon be able to use both knowledge and data to understand and accurately describe the rules that define complex systems in health care. The rules and data sets will simply be more action-oriented and experiential versions of the descriptions already contained in the expert-juried wikis. But the rules also will be translatable into readily accessible decision-support tools, built on the best collective knowledge and experience of clinicians and patients and accessible by any clinician.

The trick then will be to encourage clinicians to access these tools. That will require some form of pay-for-performance, with rewards for hitting clinical and financial targets.

All this is a work in progress. But the possibilities and the potential for dramatically improving care and driving down cost, in the US and globally, are clear. And, for those of us who have watched the mechanics of the crisis up close for the entirety of our careers, breathtaking.


Tags: