Wisdom of the Crowds: Extending the Domain of Medical Information to Case Banks and Online Forums.

Abstract

s Evident in earlier Editorials [1,2,3], we at JOCR are striving for a better understanding at Clinical Decision Making and better A Patient Care. Human deliberations and Single observations have been highlighted as important concepts in building a comprehensive paradigm for individual patient management [2,3]. A re-view of the quality and direction of current body of 'Selective' literature, which is highly biased towards Evidence Based Medicine (EBM), is needed. EBM has its own role and important place in literature but it might be suffering from a 'Central Control' phenomenon in terms of few 'Chosen' groups of experts are digging the Evidence and instructing others on how to interpret and utilize 'Evidence'. Here I will like to introduce a concept from Cybernetics called 'Variety' which represents possible alternatives or possible bits of information. The concept of variety is embodied by Ashby's Law which is stated (and also applicable) in many forms. The most useful form for us is as stated by Steve Hickey and Hillary Roberts: “Ashby's Law stipulates that the minimum amount of information needed to give an accurate answer is exactly the amount needed to specify the problem” [4]. This is interpreted as; if the question has lot of variety the answer too will have same amount of variety. A complicated question will not have a simple answer. Management of a complex fracture in patient with multiple co-morbidites in a resourcefully challenged situation cannot be resolved by 'Cookbook' approach that EBM will like us to believe. Thus if we need an answer to a complex situation, more information will be needed and on a large scale 'Wisdom of Crowds' will be more effective than a controlled guideline. Every one of us holds a part of knowledge about management of these cases and when put together will be much more helpful than the group statistics of large clinical trials. Again as surgeons we need to predict the individual patient outcome and not what will be the outcome in group of such patients. Here if we have had a patient with similar profile, we can model the current patient according to him and apply it to subsequent similar group of patients. However can we get enough information from group statistics to treat Individual patients? This reverse is not found to be true [5] and thus all our Literature which is based on group studies has to be carefully reviewed and interpreted. Again the Best Evidence as defined by a 'central control' will go against the Ashby's law as a group of experts will surely have less variety than an entire profession and their patients. Representation of this individual variety in Patient presentation and management is currently lacking in Literature and the main reason is the group statistical process of approximation and 'rounding off ' at all levels of information, practiced in current evidence based literature.