Ghost Maps

A few weeks ago, I delivered the keynote address at the First Annual Information and Quality Services Center (IQSC) Educational Event, put on by the Dallas Fort Worth Healthcare Foundation (DFWHC) in Dallas, Texas. It was a great meeting, and afforded me the chance to reconnect with folks I’d collaborated with at Baylor Scott & White, and to meet new and interesting people, one of whom introduced me to the book The Ghost Map by Steven Johnson.

Hearing about this book brought me up short — how in the world had I missed it, and what the hell else had I been missing in my busyness? Clearly it was time for me to slow down, turn my head to the left and right, and take in the bigger world around me.

Anyone who has ever joined one of our trainings or read other newsletters will recall that we always talk about Dr. John Snow and his famous work plotting cholera deaths on a map, thereby helping to make the visual connection between where people were getting their water, and deaths attributable to this lethal disease.

That’s pretty much where the story ends in most descriptions of Snow’s work (including, sadly, mine) leaving out many interesting and instructive details. Fortunately, it’s these missing bits that Johnson writes about in a thoroughly engaging and compelling way.

First, he goes beyond the most immediate details of the 1854 epidemic to describe in vivid detail related subjects like the history of toilets, the upgrading of London’s sewer system, the importance of population density for a disease that travels in human excrement, and the positive as well as the negative aspects of urbanization itself.

Never before Victorian London existed, Johnson teaches the reader, had 2.4 million primates of any species lived together within a 30-mile perimeter. The conditions he describes are quite simply staggering to one’s imagination — this is really the stuff of a Charles Dickens novel.

Next, Johnson describes how Snow used some of the earliest Geographic Information System (GIS) methods to support his arguments.

The map below is one that many of us are familiar with, but the analysis that Snow did was far more complex than this simple rendering conveys. Snow drew Thiessen (Voronoi) polygons around the wells, defining straight-line least-distance service areas for each. That is, the polygons defined individual areas of influence around each of a set of points, and also displayed the area closest to each point relative to all other points. (I know: you thought this would be an easy holiday read. Not so much.)

By analyzing the data in this way, Snow was able to understand that a large majority of the cholera deaths fell within the Thiessen polygon surrounding the Broad Street pump, and a large portion of the remaining deaths were on the Broad Street side of the polygon surrounding the bad-tasting Carnaby Street well.

He then re-drew the service area polygons (using nothing more than a pencil and string) to reflect the shortest routes along streets to wells, thereby revealing that an even larger proportion of the cholera deaths fell within the shortest-travel-distance area around the Broad Street pump. A modern-day version of this technique is displayed in the following version of Snow’s original map:

Source: GIS Analyses of Snow’s Map

Although I found the entire book fascinating, given the work I do, I thought that this additional detail about how Snow analyzed his map using Thiessen (Voronoi) polygons was perhaps the most interesting and instructive. We live in a world where technology allows us to see satellite images of our planet and download way-finding applications to our mobile phones in a matter of seconds.

The possibilities of how to use that data in a proactive and positive manner are pretty staggering.

For example, in the case of public health, it is of paramount importance to obtain a picture of mobility patterns and fluctuations in a continuous manner, particularly during emergencies (such as an outbreak of a potential pandemic or disasters) in order to support decision-making or assess the impact of government measures and restrictions to maximize the effect of interventions.

This type of data has historically been collected via manual health surveys; but with the worldwide use of mobile phones and the potential ability to capture and track infections and diseases in populations, along with their movements in the world, and then display that same data and patterns, using advanced GIS technology, the opportunities to see and understand and prevent human suffering are nothing short of revolutionary.

Here’s the bottom line: if you have a bit of down time over the holidays (or anytime), I highly recommend this interesting and easy read. I feel certain it will broaden and deepen your thinking as well.

This entry was posted in Data Analysis, Data Visualization, Newsletters. Bookmark the permalink.

One Response to Ghost Maps

  1. Meic Goodyear says:

    I bought this on Kathy’s recommendation, and it arrived 2 days before New Year. As a (now retired) public health specialist and health geography researcher working in London, I don’t understand how I can have missed it. What a book! It reads like a thriller, and I devoured it in two sittings. Quite the best thing I read in 2017. As well as the points Kathy draws out, there are fascinating side lights on matters such as experimental method as seen at the time: randomisation had not been developed as a technique, but the need for closely comparably control and study populations was understood; and the irony of the increased private water supplies and WCs contributing to evolution of the epidemic in a particularly vicious example of the law of unintended consequences. Some more recent outbreaks are covered also, but the book was written before the outbreak of cholera in Haiti that arose from Nepalese (Gurkha) soldiers in the British army deployed as aid workers after the earthquake.
    Everyone working in public health should own a copy of this book.

    The book does not include the polygon overlay map – that, I think, must be Kathy’s own work.

    The saddest irony in my view is that in the UK “information governance” and data privacy regulation have become so ridiculously strict that John Snow would not legally be able to publish his analysis, as residential details of people who die is considered person confidential data. Although it’s possible to buy individual death certificates for any deceased person, the computerised Public Health Mortality Files can only be viewed by Directors of Public Health for the local authority or their direct nominees, who have to sign confidentiality agreements every year.

Leave a Reply

Your email address will not be published. Required fields are marked *

8 − = three