91ÉçÇø

Dr. James Hanley and Samantha Hajna

Professor & PhD candidate
Department of Epidemiology, Biostatistics and Occupational Health, 91ÉçÇø

Data analysis in the time of cholera

ALL ARE WELCOME

SYNOPSIS:

We revisit the big- (and smaller-) data on the 1849 and 1854 cholera epidemics in England and the analyses carried out by William Farr, John Snow and John Simon. Some of these data, and the way they were obtained, have not been widely used/described in the teaching of epidemiology. Indeed some 23 pages of individual-level data in Snow’s 1855 book were omitted when, in 1936, Wade Hampton Frost (JHSHPH) had it reprinted as ‘Snow on Cholera.’ [According to Vandenbroucke, it is this 1936 book that made Snow a hero: Snow’s self-financed book sold fewer than 60 copies, and his work was all but ignored in the 19th century.]Ìý

We will present some of the data, analyses and graphics in Farr’s 400 page ‘big data’ report on the 1849 epidemic, a report the Lancet called ‘one of the most remarkable productions of type and pen in any age or country.’ We will focus on the ’statistical law’ Farr developed from these data and how it fared when confronted with the 1854 data.

We will describe the 1854 ‘Grand Experiment’ in South London that Snow exploited, the extra work he had to do to use the limited denominator information available to him at the time, how his results were reviewed by the critics, and our efforts to revive the 23 pages of data.Ìý Eighteen months after his book was published, better denominators became available to him through an extensive study overseen by John Simon. Snow published an updated analysis of these South London data in the Journal of Public Health in late 1856. We will relate Snow’s criticisms of Simon’s numerators and his use of Simon’s denominators, and speculate on how these data might be analyzed today.ÌýÌý

We will also relate how Simon seems to have anticipated, by more than 100 years, a now-increasingly-popular technique for dealing with confounding.

OBJECTIVES:

1. To become better acquainted with our epidemiological history;

2. to appreciate and learn from the population-based investigations before Pacini and Koch identified the vibrio cholerae; and

3. statistical practice ‘BC’ (i.e., before [electronic!] computers, and multiple regression models fitted to big data).

BIO:

Jim Hanley -

Samantha Hajna completed a Bachelor of Science (Honours) and a Master of Science degree in Health Sciences at Brock University. She is currently a PhD candidate in the Department of Epidemiology, Biostatistics and Occupational Health at 91ÉçÇø. Broadly, Samantha is interested in the social determinants of cardiometabolic health. While at Brock University, the focus of Samantha’s research was on understanding the role of family eating behaviours on weight outcomes in children. Her current research focuses on the role of the built environment on physical activity in adults with type 2 diabetes and how these factors may be targeted in interventions to facilitate increases in physical activity among adults. Samantha has been an active member of the Canadian Obesity Network’s Student and New Professional Executive since 2010, serving as the Financial Director in 2011/2012 and the Chair in 2012/2013. Ìý

Ìý