Making Sense of Messy Medical Data Through Constructed Charts of Functional Distances

Abstract

This chapter explains a technique called constructed cartography that makes maps or diagrams of events that you might not initially consider appropriate for mapping, including various real or imagined manifestations of diseases and our successful and unsuccessful attempts to mitigate those manifestations. The goal is to reveal underlying patterns in messy data. Starting with data in a form where underlying patterns are elusive, a visual summary is constructed. Information is neither created nor destroyed, just transformed into a more useful form. Two fundamental ideas that allow these charts to be constructed – functional distances and multidimensional scaling – are reviewed. This chapter then describes in detail the data that are needed for constructed cartography. A simulation shows that a sample of 35 cases is sufficient to extract 90% of the pattern in a much larger population. Finally, this chapter reviews several diverse instances where the technique has worked well: administrative funding versus successful vaccinations, migrations, metastases, and symptoms of disease. Finally the potential of aligning our attempts to fight diseases as well as our knowledge of how the disease spreads with ways that the disease actually spreads is outlined. Such a test of congruence of efforts can occur at the individual or at population levels. Places where these constructed charts do not line up are places for closer inspection, consideration, effort or training.