Download A statistical investigation of the geography of cardiovascular

A statistical investigation of the geography of cardiovascular
mortality in the UK
The UK has witnessed impressive health improvements over the last 50 years or so. People
from every class and region are now healthier and living longer than ever before.
In 2010 cardiovascular diseases were the largest causes of death in England killing around
180,000 people. This accounted for approximately one third of all deaths in the UK.
Nearly half of all deaths from cardiovascular disease are from coronary heart disease
(CHD). In 2012, the NHS reported that CHD was the most common single cause of death
in the UK with around 82,000 deaths each year.
In 2013, it was reported that early mortality rates from cardiovascular disease, i.e. for
those aged under 75 years old, are lower than the national rate in the southeast of England.
In contrast, early mortality rates for cardiovascular disease are higher than the national
rate in the northwest.
This enquiry aims to investigate the possible scale and significance of such inequalities in
health between two regions of the UK. It could be viewed as, ‘Can where you live in the UK
be bad for your health?’ However, it should be remembered that if such an inequality is
confirmed it will not be possible to determine the cause of the inequality.
Statistical investigation
Null Hypothesis: There will be no regional variation in cardiovascular mortality in
England between the NHS Northwest (shown as region 5 on the map below) and the NHS
South East Coast (region 7 below) Strategic Health
Authorities.
Method:
In order to test the above hypothesis you will be
using the following graphical and statistical
techniques:

mean

mode

median

range

interquartile range

dispersion graph

standard deviation
Strategic Health Authorities in England 2013
© 2006 Morwen (talk) (Uploads)
http://en.wikipedia.org/wiki/NHS_strategic_health_authority
© www.teachitgeography.co.uk 2013
19326
Page 1 of 10
A statistical investigation of the geography of cardiovascular
mortality in the UK
Cardiovascular mortality data
The cardiovascular mortality index (CVMI) below uses the number of deaths for those aged
under 75 years old relative to each 100,000 of the population. The index makes
adjustments to take into account local population characteristics and is used to remove, as
much as possible, the effects of differences in age and sex for example. This allows
comparisons that are more direct between different strategic health authorities.
Northwest SHA
CVMI
South East Coast SHA
CVMI
Ashton, Leigh & Wigan
81
Brighton & Hove City
74
Blackburn
98
East Sussex
46
Blackpool
115
Eastern & Coastal Kent
64
Bolton
86
Hastings
65
Bury
83
Medway
64
Central & Eastern Cheshire
58
West Kent
58
Central Lancashire
70
West Sussex
52
Cumbria
69
East Lancashire
81
Heywood, Middleton, Rochdale
81
Knowsley
87
Liverpool
94
Manchester
118
North Lancashire
64
Oldham
86
Salford
102
Sefton
72
Stockport
66
Tameside & Glossop
91
Trafford
73
Warrington
80
Western Cheshire
49
Wirral
82
© www.teachitgeography.co.uk 2013
19326
Page 2 of 10
A statistical investigation of the geography of cardiovascular
mortality in the UK
Measures of central tendency:
In cases where there is a considerable amount of data, it can be useful to calculate an
average in order to summarise the data, as this allows an easy and immediate comparison
between data sets. The mean, mode and median are different methods for finding the
average value of a set of data.
Mean – the total of the data set divided by how many rows of data there are.
Mode – the value that appears the most in a data set.
Median – the middle value of the ordered data set.
Student task 1
a) Copy and calculate to the nearest whole number, the mean, mode and median for the
Northwest and the South East Coast Strategic Health Authorities.
Northwest SHA
South East Coast SHA
Mean
Mean
Mode
Mode
Median
Median
b) Is there a variation in cardiovascular mortality between the two regions? What do
these figures tell you about cardiovascular mortality in the two regions?
Measures of central dispersion
The measures of central tendency (mean, mode & median) do not give a reliable view of the
distribution or spread of the data sets and it is possible for two very different data sets to
have the same mean, mode and median values. It is often useful to show how far figures
differ from the ‘average’. This measure is known as dispersion and there are a number of
ways of showing it:
Range – the difference between the largest and smallest values in the data set.
Interquartile range (IQR) – very similar to the range but refers only to the middle 50% of
the data set.
© www.teachitgeography.co.uk 2013
19326
Page 3 of 10
A statistical investigation of the geography of cardiovascular
mortality in the UK
To calculate the IQR:
 Rank the data in order of size.
 Divide it into 4 equal groups or quartiles.
Note: A dispersion graph is often used to do this.
 Find the upper quartile (UQ) value – this occurs at rank
(n + 1)/4 in the ranked data set.
 Find the lower quartile (LQ) value – this occurs at 3(n + 1)/4 in the ranked data
set.
 IQR = UQ – LQ
Standard deviation – is a measure of the degree of dispersion about the mean value of a
data set.
To calculate the Standard deviation:

 Calculate the mean of the data set.
 Calculate the difference between each value in the data set and the mean.
 Square each difference from the previous step, to eliminate negative values.
 Total the squared differences.
 Divide this by the number of values less one in the data set.
 Calculate the square root.
© www.teachitgeography.co.uk 2013
19326
Page 4 of 10
A statistical investigation of the geography of cardiovascular
mortality in the UK
Student task 2
a) Calculate the range for the Northwest & the South East Coast Strategic Health
Authorities.
b) What does the range tell you about cardiovascular mortality within the two strategic
health authorities?
c) Create a dispersion graph for both sets of data.
d) Using the dispersion graph, calculate the IQR for both the North West and South East
Coast Strategic Health Authorities. Mark the IQRs on the dispersion graph.
e) What does the IQR tell you about cardiovascular mortality within the two strategic
health authorities?
f) Calculate and compare the standard deviation for the Northwest & the South East Coast
Strategic Health Authorities.
Standard deviation for the South East Coast SHA
South East Coast SHA
CVMI
Brighton & Hove City
74
East Sussex
46
Eastern & Coastal Kent
64
Hastings
65
Medway
64
West Kent
58
West Sussex
52
( x – x̄ )
74 – 60 =
14
46 – 60 =
-14
( x – x̄ )2
196
196
4
Mean ( x̄ ) = 60

(x-x)2 =
n=7
n–1=6
© www.teachitgeography.co.uk 2013
19326
Page 5 of 10
A statistical investigation of the geography of cardiovascular
mortality in the UK
Standard deviation for the Northwest SHA
Northwest SHA
CVMI
Ashton, Leigh & Wigan
81
Blackburn
98
Blackpool
115
Bolton
86
Bury
83
Central & Eastern Cheshire
58
Central Lancashire
70
Cumbria
69
East Lancashire
81
Heywood, Middleton, Rochdale
81
Knowsley
87
Liverpool
94
Manchester
118
North Lancashire
64
Oldham
86
Salford
102
Sefton
72
Stockport
66
Tameside & Glossop
91
Trafford
73
Warrington
80
Western Cheshire
49
Wirral
82
( x - x̄ )
( x - x̄ )2
Mean ( x̄ ) =

(x-x)2 =
n=
n–1=
© www.teachitgeography.co.uk 2013
19326
Page 6 of 10
A statistical investigation of the geography of cardiovascular
mortality in the UK
Student task 3
The null Hypothesis was there would be no regional variation in cardiovascular
mortality in England between the NHS Northwest and the NHS South East Coast Strategic
Health Authorities.
a)
Do you now accept or reject the null hypothesis?
b)
What reasons can you suggest for the differences between the cardiovascular
mortality statistics in the two strategic health authorities.
Extension task
Use the data from the East Midlands below to investigate differences in cardiovascular
mortality between the three strategic health authorities.
East Midlands SHA
CVMI
Bassetlaw
53
Derby City
77
Derbyshire County
67
Leicester City
85
Leicestershire County & Rutland
55
Lincolnshire
66
Northamptonshire
61
Nottingham City
93
Nottinghamshire County
63
© www.teachitgeography.co.uk 2013
19326
Page 7 of 10
A statistical investigation of the geography of cardiovascular
mortality in the UK
Teaching notes
Strategic Health Authority cardiovascular profiles are available at
http://www.sepho.org.uk/NationalCVD/NationalCVDProfiles.aspx
Technique
Northwest SHA
South East Coast SHA
Mean
82
60.4
Mode
81
64
Median
81
64
Range
69
28
IQR
21 (UQ=91 / LQ=70)
13 (UQ=65 / LQ=52)
SD
17
9
Standard Deviation table – South East Coast
A rounded mean value of 60 has been used in this calculation.
South East Coast SHA
CVMI
( x – x̄ )
( x – x̄ )2
Brighton & Hove City
74
14
196
East Sussex
46
-14
196
Eastern & Coastal Kent
64
4
16
Hastings
65
5
25
Medway
64
4
16
West Kent
58
-2
4
West Sussex
52
-8
64
Mean ( x̄ ) = 60

(x-x)2 = 517
n=7
n–1=6
Standard deviation for the South East Coast = 9 when rounded.
© www.teachitgeography.co.uk 2013
19326
Page 8 of 10
A statistical investigation of the geography of cardiovascular
mortality in the UK
Standard Deviation table – Northwest
Northwest SHA
CVMI
( x – x̄ )
( x – x̄ )2
Ashton, Leigh & Wigan
81
-1
1
Blackburn
98
16
256
Blackpool
115
33
1089
Bolton
86
4
16
Bury
83
1
1
Central & Eastern Cheshire
58
-24
576
Central Lancashire
70
-12
144
Cumbria
69
-13
169
East Lancashire
81
-1
1
Heywood, Middleton, Rochdale
81
-1
1
Knowsley
87
5
25
Liverpool
94
12
144
Manchester
118
36
1296
North Lancashire
64
-18
324
Oldham
86
4
16
Salford
102
20
400
Sefton
72
-10
100
Stockport
66
-16
256
Tameside & Glossop
91
9
81
Trafford
73
-9
81
Warrington
80
-2
4
Western Cheshire
49
-33
1089
Wirral
82
0
0
Mean ( x̄ ) = 82

(x-x)2
n = 23
n - 1 = 22
= 6070
Standard deviation for the Northwest = 17 when rounded.
© www.teachitgeography.co.uk 2013
19326
Page 9 of 10
A statistical investigation of the geography of cardiovascular
mortality in the UK
Students should reject the null hypothesis. They should identify that cardiovascular
mortality is significantly lower in the southeast of England and higher in the northwest.
Discussion should suggest that the profile of the population e.g. age, socioeconomic status,
lifestyle behaviour and environment will all affect the data. This, and additional
information is available in the strategic health authority cardiovascular profiles at
http://www.sepho.org.uk/NationalCVD/NationalCVDProfiles.aspx
It is also possible to ‘drill down’ within an additional strategic health authority to compare
and contrast differences e.g. between Western Cheshire (49/100000) and Blackpool
(115/100 000) in the Northwest Strategic Health Authority.
Extension task
Standard Deviation table – East Midlands
A rounded mean value of 69 has been used in this calculation.
East Midlands SHA
CVMI
( x – x̄ )
( x – x̄ )2
Bassetlaw
53
-16
256
Derby City
77
8
64
Derbyshire County
67
-2
4
Leicester City
85
16
256
Leicestershire County & Rutland
55
-14
196
Lincolnshire
66
-3
9
Northamptonshire
61
-8
64
Nottingham City
93
24
576
Nottinghamshire County
63
-6
36
Mean ( x̄ ) = 69

n=9
n-1=8
(x-x)2 = 41615
Standard deviation for the East Midlands = 14 when rounded.
© www.teachitgeography.co.uk 2013
19326
Page 10 of 10