Page - 34 - in Applied Interdisciplinary Theory in Health Informatics - Knowledge Base for Practitioners

The paper by Tribus and McIrvine [12] is a good starting point for a deeper study of the conceptualisation of Shannon entropy, and could perhaps be read after the tutorial [11] that was mentioned earlier and the papers by Jaynes [3][4]. We do believe that the time is ripe to propagate a deeper understanding of information theory through practitioners of health informatics. The potential for significant enhancements in the rigour of medical decision making is waiting to be realised. However, some stronger guidelines do need to be developed for its usage. We have described a number of different strategies and there would be real value in documenting a common foundation that would inform all of these. In addition, we would emphasise the need to explicitly model the “noise” that is inherent in the communication model. We have included this in the communication model at the beginning of the chapter, and it is an important but often neglected factor in the risk of misdiagnosis of a patient. Teaching questions for reflection 1. Can you think of a clinical setting from your own experience where information theory might have usefully informed your choices? 2. Do any of the examples provided in this chapter have more mainstream statistical methods of achieving the same result? 3. What do you feel are the barriers to the adoption of information theoretic approaches in the wider community? References [1] D.A. Asch, J.P. Patton and J.C. Hershey, Prognostic information versus accuracy: once more with meaning, Medical Decision Making 11 (1991), 45-47. [2] W.A.Benish, Relative Entropy as a Measure of Diagnostic Information, Medical Decision Making 19 (1999), 202-206. [3] E.T. Jaynes, Information Theory and Statistical Mechanics, Physical Review 106 (1957), 620-630. [4] E.T. Jaynes, Information Theory and Statistical Mechanics, Physical Review 108 (1957), 171-190. [5] S. Kullback and R.A. Leibler, On information and sufficiency, Ann Math Stat 11 (1951), 79-86. [6] J. Lee and D.M. Maslove, Using information theory to identify redundancy in common laboratory tests in the intensive care unit, BMC Medical Informatics and Decision Making 15:59 (2015), 1-8. [7] D.J.C. Mackay, Information Theory, Inference and Learning Algorithms, CUP, Cambridge UK, 2003. [8] L. Pismen, The Swings of Science – From Complexity to Simplicity and Back, Springer Nature, Switzerland, 2018. [9] C.E. Shannon, A Mathematical Theory of Communication, The Bell System Technical Journal 27 (1948), 379-423. [10] W.K. Stadelmann, D.P. Rapaport, S-J. Soong, Prognostic factors that influence melanoma outcome. In: Balch CM, Houghton AN, Sober AJ, et al, eds. Cutaneous Melanoma. 3rd ed. St Louis, MO: Quality Medical Publishing; 1998:11-35. [11] J.V. Stone, Information Theory – A Tutorial Introduction, Sebtel Press, 2015. [12] M. Tribus and E.C. McIrvine, Energy and Information, Scientific American 225 (1971), 179-190. [13] R.T. Vollmer, Entropy and Information Content of Laboratory Test Results, Am J Clin Pathol 127 (2007) 60-65. [14] C. van Walraven and C.D. Naylor, Do we know what inappropriate laboratory utilization is? A systematic review of laboratory clinical audits, JAMA 280 (1998), 550-558. P.Krause / InformationTheoryandMedicalDecisionMaking34