The first in a series of shorter posts looking at specific bits of data. Today’s data is a look at ambulance callouts for the London Ambulance service 2009-2013.
Chart 1: Looking a the number of ambulance callouts over the period, we can see graphically a broad increase of around 30 incidents a month over the period. However, the data fluctuates quite a bit month on month. R2 on this is 0.26, which gives R, the correlation co-efficient as being 0.5. Given the number of observations, this means that theres is a less than 5% likelihood that the observation that binge drinking callouts increase with time is random.
Chart 2: If we apply a seasonal adjustment factor based on which month of the year a data point occurs in, to control for differences in drinking at different times of the year (people might disproprtionately drink in say December or summer months) the R2 increases to 0.4, and an R of 0.63. The application of the adjustment, removes some of the month-on-month fluctuation in the data and shows more clearly the increase.
Binge drinking related ambulance callouts are on the increase in LBTH. However, we can’t say with any certainty if this reflects a growth in population, an increase in binge drinking or an increase in how likely someone is to call for an ambulance for themselves, a friend or someone comatose in the street.