IMASS v3 User Guide - PN Lee Statistics and Computing Ltd
Contents
1. Country Sex Tobacco Type Start Youngest End Oldest UK HALS v Males v Total Cigarettes TC v 0 Current v 0 Current v Country Ratio denominator None Single Country Y a Prevalence of Smoking Total Cigarettes for Period and Birth cohort 80 0 70 0 60 0 50 0 o o E w 400 2 L a 30 0 20 0 10 0 0 0 lo e wo e wo o ito e 1 o ite o 1 o 1 o ito e wo e x q N N ceo co t rt 1 ite ce oo co Q Q 2 2 2 2 2 2 2 EL 2 2 z 2 EL 2 2 2 2 2 2 m n Nm n NS n zi n n un n x n m n n 2 e e lt N N co e rt Tt ite ite ae ae pe p co o E 2 E 2 2 E EL E 2 2 B 2 2 a 2 2 2 2 2 Period2. 1996 2000 Lung Cancer Mortality Rates for Country and Period Mortality Rate Relative 1946 1950 1951 1955 1956 1960 1961 1965 1966 1970 1971 1975 1976 1980 1981 1985 1986 1990 1991 1995 1996 2000 Period 33 Example 6 Mortality Plots Page 8 Sex Ratio Sex Cohort or Age group Disease M F Sex Ratio Combined 40 69 E Std w Lung Cancer x Relative Scaling Period None v Lung Cancer Mortality Rates for Country and Period Mortality Rate Sex Ratio 1946 1950 1951 1955 1956 1960 1961 1965 1966 1970 1971 1975 1976 1980 1981 1985 1986 1990 1991 1995 1996 2000 Period Example 7 Mortality Rank Plots Page 8 Sex Males Combined 40 69 E Std Disease Lung Cancer Relative Scale Period None Rank Adjustment Normalised bd Country Ranking 1946 1950 Mortality Rate Ranking by Period 1951 1955 1956 1960 1961 1965 1966 1970 1971 1975 Period 1976 1980 1981 1985 1986 1990 1991 1995 1996 2000 a SS SUEDE zu Example 8 Prevalence Plots Page 4 Period or Cohort ea Period for Lagging or Averaging Country Period v Tobacco Type Start Youngest End Oldest France v 1976 1980 v All tobacco products A v 0 Current v 0 Current v Country Ratio denominator None Single Country Y Prevalence of Smoking All Tobacco Products for Sex and Age group Prevale
3. Here unlike the earlier pages any changes to the drop down menus including changes to the Combined Age Group Settings and Relative Scaling Period on page 8 only are not immediately implemented in the plot Instead an Update Country Plots button appears When this is clicked the relevant country workbooks are each opened in turn and the plot is built up See Example 4 generated on page 8 Notice that the Plot Settings were used to restrict the periods on the x axis in Example 3 and this remained set for Example 4 This is a good page on which to demonstrate two features already mentioned The Relative Scaling Period is shown in Example 5 Note that any line with missing data at the chosen period will disappear as happened with Czechoslovakia in the Example Choosing a period in the centre of the range would avoid this This page is also appropriate for plotting the sex ratio see Example 6 Once a country plot has been updated a further option becomes available to convert the mortality rates into country rankings 1 highest 2 next highest and so on These can be either unadjusted or normalised See Example 7 which is a rank version of Example 4 By default the rank plots are presented in Portrait orientation to allow all 30 countries to be displayed but as only a few countries are included in the example this has be altered by changing the y axis limits as described above and changing the maximum to the number of countri
4. i asthma as well as bronchitis unqualified as to whether it is acute or chronic is combined with CB and emphysema in the same WHO broad summary code in the 8 and 9 revisions ii COPD aterm only commonly used in the last 20 years or so is classified with other respiratory disease in the 6 7 and 8 revisions iii emphysema is classified with other respiratory disease in the 6 and 7 revisions and iv asthma is included under allergic disorders and not as a respiratory disease at all in the 6 and 7 revisions and cannot be separately identified from the available data After studying the coding systems at some length it became clear that there was no entirely satisfactory solution The definition of COPD finally selected as detailed in Table 5 is comparable enough for the 9 and 10 revisions but will lead to some discontinuities earlier which may vary by country As an alternative data for respiratory diseases non acute are also included This broader disease grouping has fewer continuity problems but includes various diseases that are not associated with smoking It should also be noted that there are numerous rules used when carrying out the coding and that these may vary between countries and from time to time For instance the death coded is the underlying cause but between 1985 and 1992 England and Wales adopted its own version of the rule to determine the underlying cause resulting in a
5. 15 19 respectively and End to 0 Current e fcreating a Combined age group or using Sexes combined then the simple average lag is calculated first on the original single sex 5 year age group then the population weighted average is calculated In other words only the population at the current time point is used e Some limitations on the validity of this process have been described in section 1 2 Note also that the 85 age group is clearly wider than 5 years so its use in lagged or averaged calculations is not recommended Using the Consumption Plots sheet Consumption data are of two types Sales data are taken from ISS2 and are given as number of cigarettes or grams for all tobacco products per adult per day They start commonly around the 1920s but much earlier for a few countries Tar data mg per cigarette are also available for some countries commonly starting around the 1960s Where original sales data are not available for all 5 years an average from the available years is used and for tar data a value of 35 mg is assumed up to 1955 with a simple linear interpolate used across any missing period Survey data are taken from ISS2 Supplement 1 and have been estimated similarly to the prevalence data They 2 5 13 are given as number of cigarettes or grams for all tobacco products per person per day This sheet is very similar to the Prevalence Plots sheet with some additional options e Either sales based or su
6. 1961 65 include 64 year olds ABBREVIATIONS A CB COPD CPP ICD THD All tobacco products IMASS International Mortality and Smoking Statistics Chronic bronchitis ISS2 International Smoking Statistics second edition Chronic obstructive pulmonary disease MC Manufactured cigarettes Cigarettes per person per day PNLSC P N Lee Statistics and Computing Ltd International Classification of Diseases TC Total cigarettes Ischaemic heart disease WHO World Health Organisation 1 3 1 4 1 5 dying in 1961 born in 1896 and 60 year olds dying in 1965 born in 1905 A more rigorous method taking these overlapping cohorts into account has been used in some other work but is not attempted here Second the approach ignores the fact that the data in the different cells of a diagonal do not relate to the exact same people as immigration emigration and mortality are ignored Differential mortality with more smokers than non smokers dying early may be an important feature for the oldest age groups but is not considered here Countries included 30 countries are included as shown below Coverage is essentially complete for Europe but elsewhere is restricted to some major developed countries One limiting factor is the availability of sufficient smoking data for useful study another is that mortality data submitted to WHO for some other countries are only regional based on inadequate death registration and or cover only a very limited p
7. 1966 70 1991 95 Fernandez Sweden 1961 65 1991 95 1961 65 1991 95 1946 50 1991 95 1961 65 1991 95 1961 65 1991 95 Nordlund Switzerland 1956 60 1986 90 1956 60 1986 90 1956 60 1991 95 1956 60 1981 85 1956 60 1981 85 SOMIPOPS UK 1946 50 1986 90 1956 60 1991 95 1926 30 1991 95 1946 50 1986 90 1971 75 1991 95 HALS AHIP USA 1931 35 1991 95 1931 35 1991 95 1946 50 1991 95 1946 50 1991 95 1946 50 1991 95 Harris USSR 1971 75 1986 90 1971 75 1986 90 1971 75 1986 90 1971 75 1986 90 1971 75 1986 90 x Yugoslavia 1956 60 1986 90 1956 60 1986 90 1956 60 1986 90 1956 60 1986 90 1956 60 1986 90 x x No data d Restricted age range for initial period s for males females start later Data for single period a Restricted age range for females b Restricted age range for initial period s for females c Restricted age range for initial period s e Restricted age range for final period s f Restricted age range for initial period s females start later g Females start later 23 Table 4 Availability of data from retrospective survey based tobacco data Country Name Source Periods Product Notes Start End Australia ABS NHS Australian Bureau of Statistics personal communication Based on 1989 90 National 1906 1910 1986 1990 MC c d e f Health Survey Australia Hyndman Hyndman J Hobbs M Jamrozik K Hockey R Parsons R A retrospective cohort study 1906 1910 1976 1980 A f h i of smoking habits in Australia In Durston B Jamrozi
8. by lines on the plot choosing from age group period birth cohort sex and country Then click on the button labelled with the two chosen factors on the Index sheet to move there directly or scroll down within the sheet to the relevant page The pages are Period and Age group Age group and Birth cohort Period and Birth cohort Sex and Age group Sex and Period Sex and Birth cohort Country and Age group Country and Period O 00 109 tn BP WN Country and Birth cohort Looking at the first page Period and Age group with the default settings Example 1 will be displayed The content and appearance of the plot can be varied in several ways By clicking on the X Axis options buttons e These buttons allow the two factors in the plot to be swapped In this Example if you click on Periods the appropriate labels will appear on the x axis bottom and the age group will appear in the legend side right By using the drop down menus e Data will be for a single country chosen from the Country menu with the Country Ratio denominator set to None Single Country Or for a ratio of two countries choose the numerator country with the Country menu and the denominator in the Country Ratio denominator menu The relevant workbook s will open automatically e Sex can be set to Males Females Males amp Females combined or Male Female Ratio e Disease can be set to Lung Cancer Ischaemic Heart Disease COPD or Respirato
9. estimates from all the survey data presented in ISS2 for standardized age groups and averaged them by sex x 5 year period x 5 year age group These estimates assume that there were no smokers below the age of 15 Prevalence estimates commonly start around the 1950s with consumption estimates commonly starting in the 1960s Fuller details of the availability of the main survey based tobacco data are shown for each country in Table 3 It should be noted that the original surveys on which these estimates are based were drawn from a wide variety of sources as discussed in the Methods section 1 6 of ISS2 and some cautions regarding the estimation process are discussed in the supplement The main survey based data derive from studies in which subjects are asked about their current smoking habits For eleven countries alternative survey based data sets are also available These are based on studies in which subjects are asked about their lifetime smoking history see ISS2 Appendix V and Supplement 2 for a description of this retrospective method and a discussion of some of its advantages and disadvantages These data commonly relate to birth cohorts born after about 1890 and do not involve the assumption of no smoking below age 15 They are taken from various published sources and are not derived on a consistent basis see Table 4 for fuller details of the surveys No attempt has been made to standardize these data The database simply incl
10. missing b Final period based on less than 5 years data see Single years End column i 1979 missing c Data for Czechoslovakia for 1991 for Czech Republic from 1985 and for Slovakia from 1992 are j Data for Switzerland under the 10 revision are available in less detail than for other countries See available but have not been entered Table 5 d 1951 missing k Data for USSR under the 9 revision are available in less detail than for other countries and it is not e W Germany former Federal Republic to 1990 unified Germany from 1991 possible to extract data for an equivalent definition of COPD See Table 5 f Earlier data are available but with insufficient detail 21 Table 2 Availability of sales based tobacco data Country Sales data Tar data MC TC A Start End Start End Start End Start End Australia 1920 1998 1925 1994 1920 1994 1969 1989 Austria 1923 1997 1923 1997 1923 1997 1960 1984 Belgium 1921 1995 1950 1995 1921 1995 1973 1993 Bulgaria 1913 1997 x 1913 1997 x Canada 1920 1995 1955 1995 1920 1955 1968 1981 Czechoslovakia 1923 1992 x 1923 1934 x Denmark 1920 1995 1950 1995 1920 1995 1964 1990 Finland 1920 1995 x 1920 1995 1970 1992 France 1865 1997 1950 1997 1865 1997 1950 1997 Germany 1906 1995 1906 1995 1925 1995 1966 1985 Greece 1920 1995 x 1920 1995 x Hungary 1920 1995 1940 1988 1920 1988 x Iceland 1932 1995 x 1862 1995 x Ireland 1920 1995 1950 1995 1920 1995 x Israel 1
11. substantial drop in the death rate from pneumonia in that period A discussion of other aspects of the validity of mortality data is given by Alderson Tables 1 5 are provided for quick reference to the time span for which the different types of data are available This information is also given in the Country Details sheet of each workbook 2 1 Using the IMASS system Getting started The system works in Excel 97 or later Macros must be enabled some systems default to disallowing macros for virus protection The preferred screen resolution is 1024 x 768 Use View then Full screen or alternatively Zoom 75 for the best effect See also section 2 9 System Requirements When the CD is inserted the install program should auto run If it does not run IMASSinstall exe from the CD Files are installed with read only status to prevent accidental corruption of the data If required the main file can be copied with read write status so that settings for particular analyses can be saved IMASS Version 3 installation will not overwrite any previous IMASS versions and will be identified by new menu entries e g IMASS v3 Excel database The default installation directory is C IMASSv3 Previous IMASS versions can still be used to allow continuity of data in existing projects The program can then be started from the Windows Start Programs menu Alternatively start Excel and then open the workbook C IMASSv3 IMASS xls Althoug
12. 10th revision 1995 1997 were previously coded with special G codes these remain for 1995 1996 but 1997 has been recoded and now uses the ICD 10 Tabulation List 1 as do the new years 1998 1999 This involved minor changes to the definition of COPD but more major problems with continuity in respiratory diseases non acute Consequently data for this disease definition are not now included for Switzerland after 1994 d WHO data codes no longer distinguish between the 6 and 7 ICD revisions but the data are unchanged and the start dates for the 7 revision are still shown in Table 5 and in the Country Details worksheet of each country workbook 20 Table 1 Availability of mortality data as at September 2003 Country Deaths ICD revision year first used Single years Periods 6 7 8 9 10 Start End Start End Australia 1950 1999 1946 1950 1996 2000 1950 1958 1968 1979 1998 Austria 1955 2001 1951 1955 2001 2005 1955 1958 1969 1980 Belgium 1954 1996 1951 1955 1995 2000 1954 1958 1968 1979 Bulgaria 1964 2000 1961 1965 1996 2000 1964 1968 1980 Canada 1950 1998 1946 1950 1996 2000 1950 1958 1969 1979 Czechoslovakia 1953 1990 1951 1955 1986 1990 1953 1958 1968 1979 Denmark 1951 1998 1951 1955 1996 2000 1951 1958 1969 1994 Finland 1952 2000 1951 1955 1996 2000 1952 1958 1969 1987 1996 France 1950 1999 1946 1950 1996 2000 1950 1958 1968 1979 Germany 1952 1999 1951 1955 1996 2000 195
13. 1986 1990 0 02 0 13 0 16 0 83 1991 1995 0 02 0 09 0 22 0 74 1996 2000 0 01 0 07 0 17 0 49 v 35 39 8 60 10 14 9 49 9 44 7 50 5 97 5 49 4 23 3 43 2 69 1 83 Lung Cancer 40 44 25 49 24 99 25 59 22 66 22 13 17 96 14 05 12 00 10 41 8 63 6 85 IY 45 49 50 54 55 59 60 64 58 36 111 73 160 98 207 86 58 68 123 17 200 22 253 19 59 43 125 27 230 74 333 30 56 77 123 57 231 58 369 36 53 45 115 11 222 71 373 28 50 83 107 90 208 53 354 53 39 92 100 98 190 55 334 61 32 10 77 86 171 91 300 44 27 61 60 01 135 37 268 51 22 80 56 14 108 20 207 17 18 46 43 42 95 08 164 23 65 69 208 26 287 30 389 91 484 46 529 66 519 89 498 48 455 35 412 17 365 37 284 61 70 74 177 70 253 70 381 86 494 34 621 39 682 12 674 41 630 10 582 82 504 88 438 58 75 79 136 35 198 64 324 17 450 27 597 39 722 75 804 57 771 18 711 65 627 91 553 00 80 84 98 90 139 08 217 75 332 44 473 07 606 68 769 55 835 75 794 42 716 99 614 69 85 52 49 88 13 140 24 200 81 335 95 433 31 570 43 691 60 713 27 677 03 579 57 29 Mortality Plots Page 2 Example 2 Disease Sex Country v Ischaemic Heart Disease Males IY uk Country Ratio denominator None Single Country Ischaemic Heart Disease Mortality Rates for Age group and Birth cohort 10000 0 e 1000 0 100 0 10 0 1294 Jed 000 001 19d e3e3 Ayez N 0 1 S861L L86
14. 2 1958 1968 1979 1998 Greece 1961 1999 1961 1965 1996 2000 1961 1968 1979 Hungary 1955 2001 1951 1955 2001 2005 1955 1958 1969 1979 1996 Iceland 1951 1997 1951 1955 1996 2000 1951 1958 1971 1981 1996 Ireland 1950 1999 1946 1950 1996 2000 1950 1958 1968 1979 Israel 1950 1998 1946 1950 1996 2000 1950 1958 1969 1979 1998 Italy 1951 1999 1951 1955 1996 2000 1951 1958 1968 1979 Japan 1950 1999 1946 1950 1996 2000 1950 1958 1968 1979 1995 Netherlands 1950 1999 1946 1950 1996 2000 1950 1958 1969 1979 1996 New Zealand 1950 1999 1946 1950 1996 2000 1950 1958 1968 1979 Norway 1951 1999 1951 1955 1996 2000 1951 1958 1969 1986 1996 Poland 1959 2000 1956 1960 1996 2000 1959 1969 1980 1999 Portugal 1955 2000 1951 1955 1996 2000 1955 1958 1971 1980 Romania 1959 2001 1956 1960 2001 2005 1959 1969 1980 1999 Spain 1951 1999 1951 1955 1996 2000 1951 1958 1968 1980 1999 Sweden 1951 1999 1951 1955 1996 2000 1951 1958 1969 1987 1997 Switzerland 1951 1999 1951 1955 1996 2000 1951 1958 1969 1995 UK 1950 1999 1946 1950 1996 2000 1950 1958 1968 1979 USA 1950 1999 1946 1950 1996 2000 1950 1958 1968 1979 1999 USSR 1982 1990 1981 1985 1986 1990 1982 Yugoslavia 1960 1990 1956 1960 1986 1990 1960 1968 1979 Not used g Jewish population only to 1974 then total population a First period based on less than 5 years data see Single years Start column h 1997 and 1998
15. 26 1930 1931 1935 1936 1940 1941 1945 1946 1950 1951 1955 1956 1960 1961 1965 1966 1970 1971 1975 1976 1980 1981 1985 1986 1990 1991 1995 Not Used Not Used NotUsed NotUsed NotUsed NotUsed NotUsed NotUsed NotUsed NotUsed NotUsed NotUsed Not Used UK Sales 2 6 25 5 37 USA Sales i 8 47 UKLC 358 44 USALC 363 45 43 Example 15 Combination Plots Page 5 10 year lag relative scaling Country Country Ratio Sex Age group or Cohort Tobacco Type Lagging or Averaging Start Youngest End Oldest un ae UK None Single Country Y Males Combined 35 54 w Manufactured Cigarettes MC WwW 10 Years previous v 10 Years previous Prev 2 FEE liic NS UK None Single Country v Males Combined 55 74 w Manufactured Cigarettes MC v 10 Years previous M 10 Years previous v Mort 1 P Relative Scaling Period Mo op UK Iv None Single Country Males v Combined 35 54 v Lung Cancer v 9 Mort 2 re Lune w 1971 1975 v Mort 55 74 UK w None Single Country ales Combined 55 74 w Lung Cancer Prevalence amp Mortality for Period 1 4 1 2 1 0 A T 2 08 E o ri gt o 0 4 0 2 0 0 1931 1935 1936 1940 1941 1945 1946 1950 1951 1955 1956 1960 1961 1965 1966 1970 1971 1975 1976 1980 1981 1985 1986 1990 1991 1995 1996 2000 Period 44 Example 16 Prevalence Plots Page 3 E Period for Lagging or Averaging
16. 6 60 1991 95 1956 60 1986 90 1956 60 1986 90 x France 1956 60 1991 95 1951 55 1991 95 1951 55 1991 95 1976 80 1961 65 1991 95 x Germany 1956 60 1981 85 1946 50 1991 95 1946 50 1991 95 1966 70 1981 85 1966 70 1991 95 Brenner Heuer Greece 1956 60 1991 95 1956 60 1991 95 1956 60 1991 95 1956 60 1986 90 1956 60 1986 90 x Hungary 1961 65 1991 95 1961 65 1991 95 1961 65 1991 95 1966 70 1991 95 1966 70 1991 95 HPGS Iceland 1981 85 1991 95 1981 85 1991 95 1966 70 1991 95 1981 85 1981 85 x Ireland 1956 60 1991 95 1956 60 1991 95 1961 65 1991 95 1956 607 1986 90 1956 607 1986 90 x Israel 1961 65 1991 95 1961 65 1991 95 1961 65 1991 95 1961 65 1991 95 1961 65 1991 95 x Italy 1946 50 1991 95 1946 50 1991 95 1961 65 1991 95 1946 50 1991 95 1946 50 1991 95 LaVecchia Japan 1946 50 1991 95 1946 50 1991 95 1946 50 1991 95 1956 60 1991 95 1956 607 1991 95 x Netherlands 1966 70 1981 85 1956 60 1981 85 1946 50 1991 95 1976 80 1956 60 1991 95 x New Zealand x 1956 60 1991 95 1956 60 1991 95 1956 60 1986 90 1956 60 1986 90 x Norway x 1951 55 1991 95 1951 55 1991 95 x 1956 60 1991 95 Ronneberg Poland 1956 60 1991 95 1956 60 1991 95 1956 60 1991 95 1956 60 1986 90 1956 60 1986 90 x Portugal 1971 75 1991 95 1971 75 1991 95 1971 75 1991 95 1986 90 1991 95 1986 90 1991 95 x Romania 1966 70 1991 95 1966 70 1991 95 1966 70 1991 95 1966 70 1966 70 x Spain 1966 70 1991 95 1966 70 1991 95 1966 70 1991 95 1966 70 1991 95
17. 9 Mortality Plots Page 4 Averaged 15 to 5 years previous sses 36 Example 10 Consumption Plots Page 8 sessi enne enne nns 37 Example 11 Consumption Plots Page 8 Tar Adjusted sseseeeeeeene 38 Weare l ANE Flots PA ANEERTTEETM M Hem 39 Example 13 Combination Plots Page 5 10 year lag oo cccecesceceeeeeceseeecseeeeseeessaeereees 40 Example 14 Combination Plots Page 3 20 year average 00 eccceececesececeeeceesseeeesteeessaeenenes 41 Example 14T Combination Plots Page 3 20 year average associated data table 42 Example 15 Combination Plots Page 5 10 year lag relative scaling sss 43 Erample 16 Prevalence Plots Page D ues eavozod mes tx dente puse ek Vote e2E dec eo qua hen en Yn e EN e 44 1 1 1 2 Description of the IMASS system Introduction The system is designed to give ready access to international data on tobacco consumption smoking prevalence and mortality from the major smoking related diseases It consists of a set of linked Excel workbooks containing nationally based data for 30 countries on mortality from four major smoking related diseases and on tobacco and smoking statistics Facilities are provided to help the user explore the relationships in the data via a variety of standard plots and tabulations Data structure Age Period and Coh
18. 960 1961 1965 1966 1970 1971 1975 1976 1980 1981 1985 1986 1990 1991 1995 1996 2000 U 7 o a 38 Example 11 Consumption Plots Page 8 Tar Adjusted T Period for Lagging or Averaging Sex amp Source Cohort or Age group Tobacco Type Start Youngest End Oldest Males amp Females Sales w Age group 70 74 w Manufactured Cigarettes MC bd 0 Current v 0 Current v Tar Adjustment Relative Scaling Period Standard v None v Consumption of Manufactured Cigarettes for Country and Period 12 10 gt amp o o 8 Q c o 5 a 6 i o a o f 2 7 A d S t 9 aa K a i 4 N V N 4 K 1 eo 1861 1865 1866 1870 1871 1875 1876 1880 1881 1885 1886 1890 1891 1895 1896 1900 1901 1905 1906 1910 1911 1915 1916 1920 1921 1925 1926 1930 1931 1935 1936 1940 1941 1945 1946 1950 1951 1955 1956 1960 1961 1965 1966 1970 1971 1975 1976 1980 1981 1985 1986 1990 1991 1995 1996 2000 U 7 o a 39 Example 12 Scatter Plots Page 2 Period or Cohort er Period for Lagging or Averaging Prevalence Sex Prevalence Period v Tobacco Type Prevalence Start Youngest End Oldest Females w 1986 1990 v All tobacco products A hd 10 Years previous v 10 Years previous v Sex Mortality Age Group Disease Mortality Females v Age group 40 44 v Lung Cancer v Preval
19. 967 1995 x 1970 1989 x Italy 1921 1995 1951 1995 1921 1995 1984 1993 Japan 1920 1995 1920 1984 1920 1984 1969 1995 Netherlands 1923 1997 1960 1997 1923 1995 x New Zealand 1920 1995 1950 1995 1920 1995 1990 1995 Norway 1927 1995 1939 1995 1927 1995 1970 1980 Poland 1925 f 1995 x 1925 f 1995 x Portugal 1940 1998 1950 1992 1940 1992 x Romania 1920 1995 x 1920 1995 x Spain 1927 1995 1957 1993 1927 1993 x Sweden 1920 1995 1950 1995 1920 1995 1964 1980 Switzerland 1934 1997 x 1934 1973 1970 1978 UK 1880 1998 1931 1998 1880 1997 1921 1997 USA 1900 1995 1949 1995 1920 1995 1954 1995 USSR 1960 1994 x 1960 1994 x Yugoslavia 1922 1993 x 1922 1935 x x No data c Early data only available at 3 or 5 yearly intervals a This table shows start and end of single year data Missing intermediate years are d Data missing for some intermediate years for details see the lower part of the Tobacco indicated by footnotes c or d Data are presented by 5 year periods using an average of Data worksheet for the relevant country Also Tables 1 2 of the relevant chapter in ISS2 the available years when the full 5 years are not available and intermediate missin e Periods 1936 1955 Missin periods are indicated by other isona ISS2 presented data up to 1995 Later deis have f Period 1941 45 Missing i been entered occasionally but no consistent attempt was made to collect later data g Unified up to 1938 and from 1991 otherwis
20. D Respiratory Diseases non acute Summary Full Summary Full 6 and 7 A093 501 bronchitis unqualified As COPD 502 CB A094 510 hypertrophy of tonsils and adenoids A097 511 peritonsillar abscess A095 518 521 empyema and absess of lung 512 chronic pharyngitis and nasopharyngitis A096 519 pleurisy 513 chronic sinusitis 514 deflected nasal septum 515 nasal polyp 516 chronic laryngitis 517 other diseases of URT 520 spontaneous pneumothorax 522 pulmonary congestion and hypostasis 523 pneumoconiosis due to silica and silicates occupational 524 other specified pneumoconiosis and pulmonary fibrosis of occupational origin 525 other chronic interstitial pneumonia 526 bronchiectasis 527 other diseases of lung and pleural cavity including emphysema g A093 490 bronchitis unqualified As COPD 491 CB A094 500 hypertrophy of tonsils and adenoids 492 emphysema A095 510 empyema 493 asthma 513 abscess of lung A096 501 peritonsillar abscess 502 chronic pharyngitis and nasopharyngitis 503 chronic sinusitis 504 deflected nasal septum 505 nasal polyp 506 chronic laryngitis 507 hay fever 508 other diseases of URT 511 pleurisy 512 spontaneous pneumothorax 514 pulmonary congestion and hypostasis 515 pneumoconiosis due to silica and silicates 516 other pneumoconiosis and related diseases 517 other chronic interstitial pneumonia 518 bronchiectasi
21. International Mortality And Smoking Statistics IMASS Version 3 User s guide and Examples February 2004 Barbara Forey John Hamling and Peter Lee P N Lee Statistics and Computing Ltd 17 Cedar Road Sutton Surrey SM2 5DA UK Tel 00 44 0 20 8642 8265 Fax 00 44 0 20 8642 2135 E mail enquiries pnlee demon co uk 1 Description of the IMASS system 0 cc cesscccssscccssccccssccecssccecsseceessecesssccessceessscesssscesseceesnes 1 1 1 ainean Masa i ENNIO MM 1 12 Datastucture Age Period and LOSDOfE uos edi Noo Ua an asa p R Goa 1 13 Counties 0 oo MCN 2 1 4 Causes of death included TC 2 1 5 DELE BOUES eere rt re e YE dee rr EYE Er vv Ce hes Ce E e C VE EE E E ra DEP 2 16 Detnitons TAS of E 4 PMEMRIUCOPRCUCBIGESEENESUSOEEENITRORIOISU E 7 2 1 Son 0 0 52 SMITH 7 22 Uring the Mortality Flots sheet asocia coprdr e e RUF ex PR Roca i R P Ed edax LE ER UR DR Lue 8 2 3 Using the Prevalence Plots sheet and the Lagging or Averaging option 11 24 Using the Consumption PING BSDEEL 1 iter aesoderobea de euha eod exhibita vidus baa ire Ian kris iaai 12 25 Uaine he Scaner Pe BETREUER 13 26 Using the Combination Plots a uiis sacs cucecnitanesd snstcuse cnbcvaanedaeuiodaatauetdanesbsansiseeanseeedeies 14 2 7 Using alternative retrospective toba co OEIL sse co HORA e hd HL Er e MU naai 15 28 Printin
22. L 0861 9261 GL6L LZ6L 0461 9961 8 961 196L 0961 9861 SS6L LS6L OS6L 9V61 SV61L LV6L OvV6L 9E61 SE6L LE6L O 61 9Z61 GZ6L LZ6L OZ6L OL6L SL6L LL6L 0161 9061 S06L1 L06L 0061 9681 GS681 L68L 0681 9881 S881 188l 0881 978 GL8L LZ8L 0281 9981 Birth cohort 30 Example3 Mortality Plots Page 5 Country Cohort or Age group Disease UK v Combined 35 74 Iw Lung Cancer v Country Ratio denominator Relative Scaling Period None Single Country wY None Lung Cancer Mortality Rates for Sex and Period 250 0 200 0 150 0 100 0 Mortality Rate per 100 000 per year 50 0 0 i 0 1951 1955 1956 1960 1961 1965 1966 1970 1971 1975 1976 1980 1981 1985 1986 1990 1991 1995 1996 2000 Period 31 Example 4 Mortality Plots Page 8 Sex Cohort or Age group Disease Males Combined 40 69 E Std Lung Cancer X Relative Scaling Period None v Lung Cancer Mortality Rates for Country and Period 250 200 150 100 Mortality Rate per 100 000 per year 50 1946 1950 1951 1955 1956 1960 1961 1965 1966 1970 1971 1975 1976 1980 1981 1985 1986 1990 1991 1995 1996 2000 Period 32 Example5 Mortality Plots Page 8 Relative Scaling Sex Cohort or Age group Disease Males w Combined 40 69 E Sta w Lung Cancer v Relative Scaling Period
23. ales 1980 94 the impact of introducing automatic coding and related changes in 1993 Popul Trends 1996 86 29 35 Alderson M International Mortality Statistics London The Macmillan Press 1981 Stephen A Frost C Thompson S Wald N Estimating the extent of compensatory smoking In Wald N Froggatt P editors Nicotine smoking and the low tar programme Proceedings of a symposium Nicotine smoking and the low tar programme London 18 20 November 1986 Oxford New York Tokyo Oxford University Press 1989 100 15 Adlkofer F Scherer G Biber A Heller W D Lee PN Schievelbein H Consistency of nicotine intake in smokers of cigarettes with varying nicotine yields In Wald N Froggatt P editors Nicotine smoking and the low tar programme Proceedings of a symposium Nicotine smoking and the low tar programme London 18 20 November 1986 Oxford New York Tokyo Oxford University Press 1989 116 30 19 Appendix New features in IMASS version 3 1 Country workbooks are now opened automatically when required and Excel 10 11 background calculations are off This gives a considerable improvement in performance and ease of use There is no longer a limit on the number of countries that can be included in a country plot or scatter plot These plots now operate with an Update button which triggers the automatic opening of each relevant country workbook in turn An Index button available throughout to retur
24. cohort among males and females in Spain 1910 1990 Eur J Cancer Prev 2003 12 57 62 Sweden Nordlund Nordlund LA Trends in smoking habits and lung cancer in Sweden Eur J Cancer Prev 1916 1920 1986 1990 A c k m 1998 7 109 16 Switzerland SOMIPOPS Estimates derived at PNLSC from SOMIPOPS Swiss National Health Survey 1981 83 1916 1920 1981 19859 A e f These estimates should be regarded with caution due to some problems in the data set UK HALS ISS2 Appendix V Health and Lifestyle Study 1916 1920 1981 1985 MC A c e f UK AHIP ISS2 Appendix V Alderson Hospital In Patient Study 1916 1920 1976 1980 MC A c e f USA Harris Harris JE Cigarette smoking among successive birth cohorts of men and women inthe 1901 1905 1976 1980 TC e f h i United States during 1900 80 J Natl Cancer Inst 1983 71 473 9 a All data are prevalence except AHIP which has both prevalence and consumption per person data b Females start later c Data relate to a single year d Data given at 10 yearly intervals so alternate periods missing e Based on 10 year cohorts so pairs of adjacent age groups have duplicate data f Retrospective data g Final period based on less than 5 years h Single year data have been averaged to give period i Data extracted from graph j Corrected for differential mortality k Early data retrospective combined with later contemporary surveys Data as given for periods 1920 24 have been entered for 1921 25 etc m Data published as
25. e West b This table shows start and end of single year data Missing intermediate years are h Periods 1916 20 1941 45 Missing indicated by footnote d Data are used for tar adjustment as 5 year averages using for i Periods 1916 1930 1941 45 Missing years with no data available an estimate of 35 mg cigarette up to 1955 and linear j Periods 1941 1955 Missing interpolation between available data points thereafter and if applicable an average based k Periods 1941 1985 Missing on less than 5 years data in the final period So there are no intermediate missing periods 22 Table 3 Availability of survey based tobacco data Country Contemporary surveys ISS2 Alternative retrospective Prevalence CPP surveys see also Table 4 MC TC A MC TC Start End Start End Start End Start End Start End Australia 1976 80 1986 90 1971 75 1991 95 1941 45 1991 95 1976 80 1986 90 1971 75 1991 95 ABS NHS Hyndman Austria 1951 55 1991 95 1951 55 1991 95 1951 55 1991 95 1971 75 1991 95 1971 75 1991 95 x Belgium 1971 75 1986 90 1961 65 1981 85 1951 55 1991 95 1976 80 1976 80 1981 85 x Bulgaria 1986 90 1986 90 1986 90 x x x Canada x 1956 60 1991 95 1966 70 1991 95 x 1961 65 1991 95 Ferrence Czechoslovakia 1961 65 1986 90 1961 65 1986 90 1961 65 1986 90 1981 85 1986 90 1981 85 1986 90 x Denmark 1976 80 1951 55 1991 95 1951 55 1991 95 1971 75 1951 55 1991 95 x Finland 1956 607 1991 95 1956 607 1991 95 195
26. e found in the country workbooks Data type Sheet Start row Numbers of deaths Deaths Male Deaths Female 143 Population in hundreds Pop Male Pop Female 43 Sales and tar Tobacco Data 43 Prevalence and survey based consumption are not available for single years If changes are to be made to a country workbook remove the Read only attribute from the file properties You should always open the main IMASS workbook first before opening the country workbook and save changes to the country workbook before returning to IMASS Additional tobacco data sources can be added to a country workbook by adding the source description to rows 59 or 60 of the Country Details sheet and entering the data in the appropriate Retro or Retro 2 sheet e g Prev MC Male Retro Acknowledgments We thank the tobacco industry for providing generous financial support towards the development of IMASS We are grateful to WHO for making publicly available the mortality data used in IMASS The authors are responsible for the accuracy of transferring the data into IMASS and the implementation of the software 4 10 11 References Lee PN Fry JS Forey BA Estimating age sex and period specific constant tar cigarette consumption in the UK Letter Thorax 1993 48 679 Forey BA Lee PN Fry JS Updating UK estimates of age sex and period specific cumulative constant tar cigarette consumption per adult Thorax 1998 53 875 8 World Health Organizat
27. econd copy rather than by editing in column AB When a change is made to a drop down menu the data will be updated in both copies but labels will be updated only in the original Thus the customized labels will not be destroyed in the second copy Note that you must be careful not to leave erroneous labels e The size of the second copy can be controlled precisely by using Page Setup Margins This will give more accurate sizing than dragging and will avoid the fonts becoming distorted A data table can be transferred to a Word table by using Paste Special Formatted text RTF This is preferable to Paste Cells which also includes any hidden data When transferring results to another Excel file a simple copy and paste will usually create links to the original IMASS files If a stand alone file containing results only is required a data table can be copied using Paste Special Values and Paste Special Formats Any hidden rows or columns will also be copied To copy a plot the following method can be used Right click on the plot area choose Edit Chart Object then Source Data and looking at the Data Range tab make a note of the data range e g sheet Prev Plot Data cells A9 P43 then cancel Copy the chart to the new file Also copy the data from the data range to another sheet in the new file starting at the same start cell e g A9 and using Paste Special Values Hide any rows and or columns that are not required On the new cha
28. ence vs Mortality Mortality Rate per 100 000 per year 0 10 20 30 40 50 60 Prevalence 40 Example 13 Combination Plots Page 5 10 year lag Country Country Ratio Sex Age group or Cohort Tobacco Type Lagging or Averaging Start Youngest End Oldest Prev 1 me YK M None Single Country w Males Age group 60 64 Manufactured Cigarettes MC 10 Years previous vw 10 Years previous b Prev 2 d UK None Single Country Males w Age group 60 64 Total Cigarettes TC Y 10 Years previous v 10 Years previous v Prev 3 g All a UK v None Single Country ww Males w Age group 60 64 All tobacco products A 10 Years previous 10 Years previous v Mort 1 Relati ling Peri Mert LC UK w None Single Country Males w Age group 60 64 w Lung Cancer 798 ative Scaling erod Mort 2 1 None v copp UK None Single Country Males VY Age group 60 64 COPD Prevalence amp Mortality for Period 100 400 90 350 80 300 70 250 z 60 E 2 lt 50 200 gt 5 NM t 40 150 30 100 20 50 10 0 0 1941 1945 1946 1950 1951 1955 1956 1960 1961 1965 1966 1970 1971 1975 1976 1980 1981 1985 1986 1990 1991 1995 1996 2000 Period 41 Example 14 Combination Plots Page 3 20 year average Coun
29. eriod of time Australia Greece Poland Austria Hungary Portugal Belgium Iceland Romania Bulgaria Ireland Spain Canada Israel Sweden Czechoslovakia Italy Switzerland Denmark Japan UK Finland Netherlands USA France New Zealand USSR Germany Norway Yugoslavia Between World War II and reunification data for Germany relate to West Germany only Data for Czechoslovakia USSR and Yugoslavia are only included up to 1990 Causes of death included The system currently contains data for four diseases Lung cancer Ischaemic Heart Disease IHD Chronic Obstructive Pulmonary Disease COPD and Respiratory Diseases non acute See section 1 6 for details of how these diseases are defined Data sources Age sex and cause specific mortality data are made available on the Internet by WHO and commonly start in the 1950s The WHO data are provided as numbers of deaths and populations for single years For all analyses these are converted to rates per 100 000 for the 5 year periods If data are not available for the full 5 years the rate is calculated from the available years Table 1 shows for each country the period of availability of mortality data as at September 2003 For other sources of population data usually before about 1950 see the second edition of International Smoking Statistics ISS2 Smoking and tobacco data are of two types sales based and survey based Sales based tobacco consumption data are mos
30. es plus one the crosses at value will update automatically and dragging the chart area Using the Prevalence Plots sheet and the Lagging or Averaging option Prevalence data represent the percentage of the population who smoke The data are taken from Supplement 1 to ISS2 and are estimates based on all the sex specific surveys in ISS2 standardized to 5 year age groups and 5 year periods They are structured in the same way as the mortality data and commonly start around the 1950s This sheet is arranged into the same 9 pages as the Mortality plots sheet and all the features described above apply Two extra options are available here Firstly the Tobacco Type is chosen as Manufactured cigarettes Total cigarettes 1 e including hand rolled or All tobacco products The second special facility added here is Lagging or Averaging In Example 8 the Lag or Average Start and End settings top right are both If X countries are selected but at a particular point data are available for only Y countries normalised ranks are plotted where Normalised rank Rank X 1 Y 1 This adjustment ensures that the country in the middle stays in the middle and the countries with the highest and lowest ranks stay symmetrically near the top and bottom of the figure 2 4 12 set to the default values of 0 Current so the data relate directly to the period shown exactly as for mortality plots As described in section 1 2 a simple approxi
31. farmer s lung etc CB chronic bronchitis COPD chronic obstructive pulmonary disease DCE diseases classified elsewhere NEC not elsewhere classified RC respiratory conditions RD respiratory disorders URT upper respiratory tract Example 1 Country UK 27 Mortality Plots Page 1 Sex Disease v Males v Lung Cancer v Country Ratio denominator None Single Country Y A Mortality Rate per 100 000 per year 900 0 800 0 700 0 600 0 500 0 400 0 300 0 200 0 100 0 0 0 Lung Cancer Mortality Rates for Period and Age group 15 19 20 24 25 29 30 34 35 39 40 44 45 49 50 54 55 59 60 64 Age group 65 69 70 74 75 79 80 84 85 Example 1T Mortality Plots Page 1 Associated Data Table 28 Lung Cancer Mortality Rate for Period Rows and Age group Columns Country Sex Disease UK v Males Country Ratio denominator None Single Country w 15 19 20 24 25 29 30 34 1851 1855 1856 1860 1861 1865 1866 1870 1871 1875 1876 1880 1881 1885 1886 1890 1891 1895 1896 1900 1901 1905 1906 1910 1911 1915 1916 1920 1921 1925 1926 1930 1931 1935 1936 1940 1941 1945 1946 1950 0 38 0 41 2 03 3 94 1951 1955 0 17 0 72 1 35 3 67 1956 1960 0 17 0 40 1 34 3 56 1961 1965 0 16 0 48 1 17 3 30 1966 1970 0 22 0 47 1 04 2 49 1971 1975 0 07 0 35 0 91 2 33 1976 1980 0 04 0 16 0 53 1 77 1981 1985 0 01 0 11 0 43 1 28
32. fined series of mortality data The actual definitions used for the four disease groupings are detailed in Table 5 As can be seen from Table 5 there is no real difficulty in obtaining a comparable disease definition from the data available for the various ICD revisions for lung cancer the definition used always including cancers of the trachea lung and bronchus For IHD the definitions used in the 8 9 and 10 revisions are also quite comparable However the definitions used in the 6 and 7 revisions are somewhat broader It has been suggested that comparability might be obtained by using codes 420 422 1 in the 6 and 7 revisions but data are not available from WHO for such detailed codes Examination of 1965 data for England and Wales showed that deaths for summary code A081 as used in the database were about 1196 higher for men and 25 higher for women than deaths for 420 422 1 so the discontinuity is not a minor one Given that the alternative was not to have any IHD data before about 1968 when the ICD 8 revision came in it seemed better to include earlier data using a somewhat different definition but making it clear that there was a problem of non comparability when studying time trends For COPD the problem is more complex Ideally for comparison with smoking data one would have liked a disease definition which included the terms chronic bronchitis CB emphysema and COPD However there were a number of problems
33. g abu o 4s Te M 15 2 9 System requirements and CODD sees divse desu ssnstesrannnddusavenuestsvineddsounes Gen n epe Ex en Seats 16 210 Customized analysis RTT 17 EE IA LU rac i rE era eee ee 17 ee RMON E E E E E E 18 Appendix New features in IMASS version 3 oo cc cccccccccccccccececsessnsececececcesceessseceeeceesesessnsaaees 19 Table 1 Availability of mortality data as at September 2003 ssesssss 20 Table 2 Availability of sales based tobacco data sssessssseeeeeeeneeenen 21 Table 3 Availability of survey based tobacco data essseeeeeereree 22 Table 4 Availability of data from retrospective survey based tobacco data 23 Table 5 Definitions of causes of death sseeeessssssseeeeeeeenen nne eene enne 24 Example 1 Mortality Plots 3 03 TE 27 Example 1T Mortality Plots Page 1 Associated Data Table sssseeeeeee 28 Example 2 Mortality Plots Page 2 ba pd ER EXER SEXDHA EEE RE REEERE IGI REDE 29 Example 3 M rtality Plots Pag QT mw 30 Example 4 M rtality Flots Pape S Im 31 Example 5 Mortality Plots Page 8 Relative Scaling 32 Example 6 Mortality Plots Page 8 Sex Ratio sssisiscsscassssienscsasvintsvarivnsessavsanceoersesssanversssssiennseonnee 33 Example 7 M rtality Rank Plots Page 8 o e ar A ARNa 34 Example 8 Pr yalenc Plots xs CANC 35 Example
34. graph table supplied in personal communication from author 24 Table 5 Definitions of causes of death ICD Revision Lung Cancer IHD Summary Full Summary Full 6 and 7 A050 162 malignant neoplasm of bronchus and trachea and of A081 420 arteriosclerotic heart disease including coronary disease lung specified as primary 421 chronic endocarditis not specified as rheumatic 163 malignant neoplasm of lung unspecified 422 other myocardial degeneration g A051 162 malignant neoplasm of trachea lung and bronchus A083 410 AMI 411 other acute and subacute IHD 412 chronic IHD 413 angina pectoris 414 asymptomatic IHD 9 B101 162 malignant neoplasm of trachea lung and bronchus B27 410 AMI 411 other acute and subacute IHD 412 old MI 413 angina pectoris 414 other forms of chronic IHD 10 except C33 malignant neoplasm of trachea 120 7 angina pectoris Switzerland C34 malignant neoplasm of bronchus and lung 121 AMI 122 subsequent MI 123 certain current complications following acute MI 124 other acute IHD 125 chronic IHD 10 G026 Same as other countries G051 G052 Same as other countries Switzerland 1995 1996 10 1034 Same as other countries 1067 Same as other countries Switzerland 1997 on AMI acute myocardial infarction IHD ischaemic heart disease MI myocardial infarction 25 Table 5 Definitions of causes of death continued 1 ICD Revision COP
35. h this workbook contains many sheets the user will carry out the majority of operations using just the first few sheets These are Index Mortality Plots Prevalence Plots Consumption CPP Plots including both sales and survey based consumption Scatter Plots comparison of mortality vs prevalence consumption Combination Plots simultaneous plotting of mortality and either consumption sales and or survey based or prevalence Each of the plot sheets holds a succession of pages each page containing a single plot These are reached directly by simply clicking on the relevant button in the Index sheet or by scrolling down through the plot sheet The user may also want to look at the first sheet Country Details of the country workbooks where information on the exact range of available years of data is shown These too can be reached from the Country Settings section of the Index sheet The raw data can be found in the country workbooks if the user wishes to create their own analyses beyond those provided in IMASS see section 2 10 2 2 Using the Mortality Plots sheet On this sheet the user can plot or tabulate mortality data according to a number of factors The mortality data commonly start in the 1950s In all plots on this sheet the mortality rate forms the y axis First decide which factor is to form the x axis of the plot choosing from age group period and birth cohort Then decide which factor is to be represented
36. ion WHO Mortality Database http www3 who int whosis Forey B Hamling J Lee P Wald N editors International Smoking Statistics A collection of historical data from 30 economically developed countries 2nd edition London and Oxford Wolfson Institute of Preventive Medicine and Oxford University Press 2002 Errata available at www pnlee co uk ISS2 htm Forey B Lee P Estimation of sex specific smoking statistics by standardized age groups and time periods Supplement 1 to International Smoking Statistics a collection of historical data from 30 economically developed countries 2nd edition Sutton Surrey P N Lee Statistics and Computing Ltd 2002 www pnlee co uk Forey B Lee P Estimating past smoking habits by an indirect method An investigation into a method based on recall with application to Great Britain Supplement 2 to International Smoking Statistics a collection of historical data from 30 economically developed countries 2nd edition Sutton Surrey P N Lee Statistics and Computing Ltd 2002 www pnlee co uk Darby SC Doll R Stratton IM Trends in mortality from smoking related diseases in England and Wales In Wald N Froggatt P editors Nicotine smoking and the low tar programme Proceedings of a symposium Nicotine smoking and the low tar programme London 18 20 November 1986 Oxford New York Tokyo Oxford University Press 1989 70 82 Rooney C Devis T Mortality trends by cause of death in England and W
37. k K editors Tobacco and health 1990 The global war Proceedings of the 7th world conference on tobacco and health Perth Western Australia 1st 5th April 1990 189 Royal Street East Perth Western Australia 6004 Organising committee of the Seventh World Conference on Tobacco and Health 1990 264 7 Canada Ferrence Ferrence RG Sex differences in cigarette smoking in Canada 1900 1978 a 1911 1915 1976 1980 TC ef h i reconstructed cohort study Can J Public Health 1988 79 160 5 Germany Brenner Brenner H A birth cohort analysis of the smoking epidemic in West Germany J 1926 1930 1981 1985 TC e f h i j Epidemiol Community Health 1993 47 54 8 Germany Heuer Heuer C Becker N Smoking prevalence and lung cancer mortality in Germany J 1921 1925 1991 1995 TC jk Epidemiol Biostat 1999 4 45 52 Hungary HGPS Estimates derived at PNLSC from Hungarian General Practitioners Study 1936 1940 1991 1995 MC TC e f h Italy LaVecchia La Vecchia C Decarli A Pagano R Prevalence of cigarette smoking among 1906 1910 1976 1980 TC c d e f j subsequent cohorts of Italian males and females Prev Med 1986 15 606 13 Norway Ronneberg Ronneberg A Lund KE Hafstad A Lifetime smoking habits among Norwegian men and 1906 1910 1986 1990 A k women born between 1890 and 1974 Int J Epidemiol 1994 23 267 76 Spain Fernandez Fernandez E Schiaffino A Borras JM Shafey O Villalbi JR La Vecchia C Prevalence 1916 1920 1986 1990 TC b c d e f j of cigarette smoking by birth
38. mation to the lifetime tobacco history relevant to a particular cell in the mortality table can be found by starting at the equivalent current cell in the tobacco table and working backwards up the diagonal of the table The Lag or Average Start and End drop down menus top right allow this to be carried out If both are set to the same time then the tobacco statistic is lagged e g if both are set to 10 years previous then the value used is taken from two steps up the diagonal from the current cell If set to different times then the value is calculated as a simple average e g if the start is set to 15 years previous and the end to 5 years previous then the value is the average of the 3 2 and 1 steps up the diagonal See Example 9 Some points to note e The result will be set as missing unless all the values in the range are available i e are not themselves missing and fall within the subjects lifetime e g consumption 20 years ago will automatically be missing for 15 19 year olds e For the standard ISS2 survey based data it is implicitly assumed that prevalence and consumption were zero below age 15 So if a lag of 10 years is set data for age groups 15 19 and 20 24 will automatically be zero This assumption is made even for periods with no data available so that these zeros are available for lag average calculations e Consumption during lifetime or adulthood can be calculated by setting the Start point to Age 0 4 or Age
39. n to the Index sheet Within the mortality prevalence and consumption sheets data settings made through drop down menus are now shared between pages 1 6 and between pages 7 9 rather than throughout the whole sheet None are shared between sheets Ranking is now available on country plots Colours and symbols have been improved There is improved consistency in the method of selecting Source sales or survey and Sex for consumption data Plot labels and title are now in worksheet cells in column AB to allow for easy editing Previously this had to be done as a chart edit In multi country and combination plots any gap in a data series now shows as a gap on the plot rather than being bridged by a straight line Selection of alternative tobacco data sources is now done in the main workbook accessed via the Index page Alternative data source added for Spain Mortality and population data have been updated using WHO data as at 2 September 2003 a Although WHO has added provision for the 85 age group to be split into 85 89 90 94 and 95 such data are available only for very recent years and they remain combined in IMASS Population data are now available as single persons rather than hundreds but they remain as hundreds in IMASS b This update has provided one more recent year of mortality data for several countries and 2 more years for Switzerland but removed 1951 data for France c Data for Switzerland for the
40. nce 15 19 20 24 25 29 30 34 35 39 40 44 45 49 50 54 55 59 60 64 65 69 70 74 75 79 80 84 85 Age group 36 Example9 Mortality Plots Page 4 Averaged 15 to 5 years previous Period or Cohort ema Period for Lagging or Averaging Country Period v Tobacco Type Start Youngest End Oldest France v 1976 1980 v All tobacco products A v 15 Years previous v 5 Years previous v Country Ratio denominator None Single Country Y Prevalence of Smoking All Tobacco Products for Sex and Age group Prevalence 15 19 20 24 25 29 30 34 35 39 40 44 45 49 50 54 55 59 60 64 65 69 70 74 75 79 80 84 85 Age group 37 Example 10 Consumption Plots Page 8 T Period for Lagging or Averaging Sex amp Source Cohort or Age group Tobacco Type Start Youngest End Oldest Males amp Females Sales Y Age group 70 74 w Manufactured Cigarettes MC v 0 Current v 0 Current v Tar Adjustment Relative Scaling Period Unadjusted v None v Consumption of Manufactured Cigarettes for Country and Period 12 10 gt ke E 8 a c o S 2 6 x o a o 9 4 o 4 N V N 4 N eo 1861 1865 1866 1870 1871 1875 1876 1880 1881 1885 1886 1890 1891 1895 1896 1900 1901 1905 1906 1910 1911 1915 1916 1920 1921 1925 1926 1930 1931 1935 1936 1940 1941 1945 1946 1950 1951 1955 1956 1
41. ntry Country Ratio Sex amp Source Period or Age group Tobacco Tvpe Tar Adiustment Start Youngest End Oldest CPP 1 A Manufactured Cigarettes MC Y 20 Years previous year average UK v None Single Country Y Males amp Females Sales Age group 75 79 v Unadjusted v 5 Years previous CPP 2 Manufactured Cigarettes MC v i year average USA v None Single Country V Males amp Females Sales Age group 75 79 v 9 MC 20 Years previous Unadjusted v 5 Years previous CPP 3 i v 0 Current Not Used UK v None Single Country Y Males amp Females Survey Y Age group 70 74 v Manufactured Cigarettes MG Unadjusted 0 Current Mort 1 UK LC UK v None Single Country Y Males amp Females v Age group 75 79 v Lung Cancer v Mort 2 USA LC USA v None Single Country Y Males amp Females v Age group 75 79 v Lung Cancer v Mort 3 Not Used UK v None Single Country Males amp Females v Age group 70 74 v Lung Cancer v Birth Cohort 1836 1840 1841 1845 1846 1850 1851 1855 1856 1860 1861 1865 1866 1870 1871 1875 1876 1880 1881 1885 1886 1890 1891 1895 1896 1900 1901 1905 1906 1910 UK Sales 2 0 63 1 06 1 60 2 11 2 65 3 20 3 85 4 78 5 47 6 09 6 62 6 87 7 31 7 80 7 96 USA Sales i 0 28 0 63 1 16 1 95 2 70 3 50 4 56 5 87 7 30 8 57 9 59 9 97 10 21 10 24 UK LC 79 55 105 20 153 08 199 17 255 32 307 19 359 31 379 50 USA LC 58 21 69 48 88 19 113 90 152 48 196 95 229 50 274 74 Birth Cohort continued 1921 1925 19
42. on the Update Required button and all the changes to the plot are implemented If you make any changes to the drop down menus from the Data Table part of the sheet then you will be transferred back to the plot in order to carry out the update In the legend the lines are simply labelled as Mort1 Mort2 etc but these can be replaced by more meaningful labels by using Plot Settings then editing in the relevant sections of column K or N If you require a long label split over 2 lines as in Example 14 the format is US sales amp CHAR 10 amp 20 year average If a long label causes the legend to overlap with the axis labels use Edit Chart Object right click in the legend and choose Format Legend Placement Right 2T 2 8 15 On pages 2 and 5 with Period on the x axis there is also a scaling option This applies to all the lines causing the data to be scaled relative to a chosen period In this case it will probably be desirable to change the y axis settings so that both tobacco and mortality have the same maximum value using Edit Chart Object then Format Axis Scale as previously described See Example 15 Note in this case the plot axes are not automatically labelled as relative In the example the labels have been edited manually Note that when two y axis scales are in use it is not possible to have grid lines If all mortality lines are hidden the mortality scale is hidden However if all the tobacco lines are hidden
43. ort In general the data are organized in tables where the rows represent 5 year periods and the columns 5 year age groups 0 4 5 9 10 14 15 19 20 24 80 85 and 85 for males and females separately The diagonals of such a table can be interpreted as representing the lifetime experience of people born around the same time e g people aged 0 4 in 1901 1905 would be aged 5 9 in 1906 1910 and aged 10 14 in 1911 1915 etc Thus the diagonals of the table correspond approximately to the experience of a birth cohort One of the main objectives of this system is to allow mortality from smoking related diseases for a group of people to be related to their tobacco consumption history as recorded at various periods of their life A cell in the mortality table relates to a group of people dying at a specified age and period Their lifetime tobacco history can be found by starting at the equivalent cell in the corresponding tobacco table and working backwards up the diagonal of this table Facilities for this type of analysis are provided In general and subject to availability results are presented by 5 year periods starting at 1851 1855 and continuing to 1996 2000 and in the 5 year age groups 15 19 20 24 80 84 and 85 Some limitations to this birth cohort approach should be noted First people dying within a 5 year period and a 5 year age group actually include those born in a 10 year age range For example those dying at age 60 64 in
44. ours will change To avoid this have all periods cohorts showing when swapping axes 10 e Do not use these methods to change the range of the x axis or to edit its labels Use Plot Settings or edit column AB as described above Then click outside the chart area to de select the plot Click on the Print Plots button to print this plot and its settings Other options for printing and saving the results are discussed below section 2 8 Click on the Data Table button to display the equivalent data table The columns correspond to the x axis factor and the rows to the lines from the plot Normal Excel facilities can be used here e g to change the number of decimal places displayed Those settings which are controlled by the drop down menus can be changed here too in which case the plot will also be updated The Print Table button prints the table see Example 1T When finished click on Go to Mortality Plots which takes you back to the original position in the plot sheet The next few pages of the Mortality Plots sheet for single country or country ratio plots share many of the features already described Example 2 shows a plot generated on page 2 Age group and Birth cohort A restricted age range has been chosen via Plot Settings and a Log Scale set on the y axis via Edit Chart Settings Example 3 generated on page 5 Sex and Period illustrates some features relevant to pages 4 6 e The x axis factor is already fixed since the o
45. prints out the current plot with its settings Use the normal Excel print facilities to change the printer properties When working on drafts for a report a quick way of saving a sequence of plots and their settings is to select and copy the relevant area including one column to the right of the plot and then in Word set the page to Landscape and use Paste Special Picture Enhanced Metafile Each plot will fill naturally one page The Worked example was generated by this method Alternatively Alt PrintScreen in Excel and a simple paste in Word will give a screen image By these methods the drop down settings are stored as part of a graphic so the format and wording cannot be edited Data tables can be copied by the same methods 2 9 16 For plots in a more polished report it may be better to type the headings directly into Word and copy just the plot right click on the plot and select Edit Chart Object then right click in the plot area i e the white area of the plot and select Copy then in Word use Paste Special Picture Enhanced Metafile The plot can be dragged to size either in Excel or in Word but any changes to the labels must be done while still in Excel Another possibility is to create a second copy of a plot within Excel just using simple copy and paste then right click on the new copy of the plot and choose Location As a new sheet This has two purposes e You can create customized labels directly on the s
46. rhinitis chronic rhinitis nasopharyngitis and J42 unspecified CB pharyngitis chronic sinusitis nasal polyp other disorders of nose and nasal J43 emphysema sinuses chronic disease of tonsils and adenoids peritonsillar abscess chronic J44 other COPD laryngitis and laryngotracheitis diseases of vocal chords and larynx NEC other J45 asthma diseases of URT a l J46 status asthmaticus J60 J66 pneumoconiosis coalworker s due to asbestos and other mineral J47 bronchiectasis fibres due to dust containing silica due to other inorganic dust unspecified AE e associated with tuberculosis J67 hypersensitivity pneumonitis due to organic dust farmer s lung etc J68 J70 RC due to inhalation of chemicals gases etc pneumonitis due to solids and liquids RC due to other external agents J80 J82 adult respiratory distress syndrome pulmonary oedema pulmonary eosinophilis NEC J84 J86 other interstitial pulmonary diseases abscess of lung and mediastinum pyothorax J90 J94 other disease of pleura J95 J96 postprocedural RD NEC respiratory failure NEC J98 J99 other RD RD in DCE 10 G063 J40 J46 i e same as other countries except not possible Switzerland excludes J47 bronchiectasis 1995 1996 and J67 hypersensitivity pneumonitis due to organic dust farmer s lung etc 10 1076 J40 J47 i e same as other countries except not possible Switzerland excludes J67 hypersensitivity pneumonitis due to organic 1997 on dust
47. rt go to the Data Range tab as before and change IMASS and the sheet name to the names of the new file and the new data sheet respectively System requirements and configuration The preferred system requirement is a 1200 MHz PC with 512 MB RAM running Excel 2002 which is supplied with Office XP The system can be run with a 256 MB Ram system running Excel 97 although the performance will be slower Macros must be enabled within Excel for IMASS to function correctly In Excel 2000 this is done from menu option Tools Macros Security select the Security Level tab and set to Low 2 10 17 If you see the message requesting virus scan each time Excel opens a workbook it probably means that the Microsoft Office plug in is enabled in your antivirus software The opening of work books can be speeded up by disabling that plug in Consult your antivirus documentation but for example in Norton Internet Security 2003 select menu item Options Norton AntiVirus and the plug in option can be found in the Miscellaneous section Customized analysis The files provide full Excel functionality so that users can create their own customized analyses or add extra data Note that background calculations are switched off so that pressing key F9 Calculate Now is required or change to automatic by Tools Options Calculation Automatic The raw data i e for single years rather than 5 year periods as used in all the provided analyses can b
48. rvey based consumption data can be chosen Since sales based data are not sex specific whereas survey based data are this is combined with the choice of sex on a single drop down menu for pages 1 3 and 7 9 and in Plot Settings for pages 4 6 In Example 10 Males amp Females Sales has been chosen e Sales data are also not age specific and so when current or lagged data have been chosen i e if Start and End are set the same the choice of Age group does not affect the results However if an average e g Start Age 15 19 End 0 Current has been set then the Age group setting will affect the result Setting a lag for sales data simply shifts the data It is possible to choose sales data for e g a Sex and Age group plot but this is not particularly appropriate as the data will all be the same e Tar Adjustment is chosen as None Standard or Sq root Standard In Standard adjustment consumption is expressed in constant tar cigarettes estimated as number of cigarettes x tar 35 where 35mg is the standard tar yield per cigarette assumed before 1955 Sq root Standard 1 e square root of Standard involves an adjustment proposed to model compensation by smokers when smoking lower tar cigarettes Tar adjustment can be applied to any tobacco type although its relevance is primarily to manufactured cigarettes See Example 11 Note that tar adjustment is only possible where tar data are available or have been estimated see
49. ry Disease non acute Definitions of these diseases have been discussed in section 1 6 as has the importance of ICD revision changes the timing of these can be seen in Table 1 and in the Country Details sheet of each country workbook By clicking on the Plot Settings button e By default the x axis is scaled to fit all the standard age groups Use Plot Settings then Age groups to show a restricted range or to choose e g alternate age groups Either just click in the individual tick boxes or highlight all the ones you want to add e g by using ctrl click then click on the Show Selected Age Groups button Note that you cannot combine age groups at this point in the program e All the standard periods are shown in the legend and all that have data available appear as a line on the plot Use Plot Settings then Periods to show only a restricted selection The colours symbols will not be altered e Click on Mortality Plots to get back to your plot e Note that Plot Settings apply to the whole of the Mortality Plots sheet However as we are working in the Age and Period page of the spreadsheet only those settings will affect our current plot By normal Excel facilities e Text can be typed into any of the blank cells outside the chart area e To edit the labels on the plot itself scroll across to the right to column AB where the automatically generated labels will be found Edit as required then press F9 e For other change
50. s 519 other diseases of respiratory system including COPD CB chronic bronchitis COPD chronic obstructive pulmonary disease URT upper respiratory tract 26 Table 5 Definitions of causes of death continued 2 ICD Revision COPD Respiratory Diseases non acute Summary Full Summary Full 9 except B323 490 bronchitis not specified as acute or chronic As COPD USSR 491 CB B313 470 471 deflected nasal septum and nasal polyps 492 emphysema B314 472 473 chronic pharyngitis nasopharyngitis and sinusitis 493 asthma B315 474 chronic diseases of tonsils and adenoids B324 494 bronchiectasis B319 475 478 peritonsillar abscess chronic laryngitis B325 495 extrinsic allergic alveolitis laryngotracheitis allergic rhinitis other diseases of URT 496 chronic airways obstruction NEC including COPD B326 500 508 pneumoconiosis and other lung disease due to external agents B327 511 Pleurisy B329 510 512 519 empyema pneumothorax absess of lung and mediastinum pulmonary congestions and hypostasis post inflammatory pulmonary fibrosis other alveolar and parietoalveolar pneumopathy lung involvement in conditions classified elsewhere other diseases of lung other diseases of respiratory system 9 USSR not possible 329 B323 Same as other countries 10 except J40 bronchitis not specified as acute or chronic As COPD Switzerland J41 simple and mucopurulent CB J30 J39 vasomotor and allergic
51. s to the plot first right click on the plot and choose Edit Chart Object Some of the many possibilities are e To vary the range of the y axis right click on an axis value and select Format Axis Choosing the Scale tab the default settings for the y axis are seen as minimum 0 maximum automatic crosses at 0 These can be changed e g if you want to create a series of plots all using the same fixed range e To seta logarithmic scale on the y axis right click on an axis value choose Format axis then the Scale tab and tick Logarithmic scale You must make sure to avoid having a minimum of zero the easiest way to do this is to fix the Minimum at a low value e g 0 1 and setting Crosses at to the same value This is shown later in Example 2 e Tochange the appearance of the axis labels for instance to change the number of decimal places on the y axis labels use Format Axis then choose the Number tab and alter as required On the x axis you can choose the Alignment tab to change the angle at which the labels are displayed e To change the colours or symbols right click on a line and choose Format Data Series Pattern To display the line label at each point choose Format Data Series Data Labels and tick Series Name There is an Excel problem that if more than 16 colour symbol definitions are in use as is the case for periods and cohorts but not for age groups and if lines are hidden and then the axes are swapped then the col
52. section 1 5 and Table 2 Thus in Example 11 many of the lines stop earlier than in Example 10 Notice that in Examples 10 and 11 a wider range of periods has been set via Plot Settings than in Examples 3 5 because sales data are available earlier than mortality data Notice also that there may be a gap in the data series e g Bulgaria has no sales data for the 1940s or 1950s Example 10 Using the Scatter Plots sheet In this sheet there are two pages with page 1 Consumption and page 2 Prevalence respectively set on the x axis and mortality set on the y axis The countries are shown in the legend and as single points in the plot The Age and Period or Cohort settings apply to both the tobacco statistic and mortality Sex is chosen separately in most cases one will want to choose the same setting for both However if having chosen sales based consumption one might wish to choose a particular sex for the mortality The other settings including the Lagging or Averaging apply only to the tobacco statistic Like multi country plots on the previous sheets any changes to the drop down menus are not immediately implemented in the scatter plots Instead an 2 6 14 Update Scatter Plots button appears When this is clicked the relevant country workbooks are each opened in turn and the plot is built up See Example 12 Using the Combination Plots sheet This sheet allows plots to be constructed showing lines for both mortalit
53. then the plot border the background shading and x axis scale get hidden as well as the tobacco scale this appears to be a deficiency in Excel which can be got round by first right clicking on each mortality line and choosing Format Data Series Axis Primary Axis before deselecting the tobacco lines Then edit the x axis label to Mortality Using alternative retrospective tobacco data As described in section 1 5 alternative survey based data sets are available for 11 countries To choose an alternative data set go to Country Settings from the Index Information on the available sources can be displayed by choosing a country then clicking the Display Survey Sources button The data sets can be chosen from the Survey Source drop down menus Any country scatter or combination plots already using tobacco data for the country in question will immediately be marked as needing update Thereafter whenever the country in question is displayed either in a Country menu or in a legend the abbreviated name of the data set will also be shown Note that the generic labels Source 2 or Source 3 are used in Plot Setting and are initially displayed until a plot is updated The country name is not altered on the Mortality Plots sheets See Example 16 using UK HALS Because this data set relates to 10 year birth cohorts it is appropriate to show only alternate cohorts Printing and saving results As mentioned earlier the Print Plots button
54. ther factor i e sex cannot be used on the x axis and so must appear as the lines e Although the available lines include the M F ratio it is unlikely that this will be required alongside the other Male Female and Combined lines since the scale will be inappropriate It can be hidden via Plot Settings See also Example 6 below e Since age period and birth cohort are interdependent and since one of these must be chosen as the x axis factor a single drop down menu is provided to allow one only of the other two to be chosen The third is thus implicitly fixed e On page 5 Sex and Period a facility is available to allow an age group to be defined as a combination of standard age groups This can be weighted either to the actual population i e unweighted or to a standard population WHO World or European standard populations or a Custom standard population can be defined as any country year Click on Combined Age Group Settings to define the combined group and then choose it from the bottom of the Cohort or Age group drop down menu e Also on page 5 a facility is available to allow the data in all lines to be scaled relative to a chosen period This has the effect of forcing all the lines to pass through a single point at value 1 and the relative changes in the statistic may be more clearly seen This is demonstrated later in Example 5 2 3 11 The final pages 7 9 of the Mortality Plots sheets offer plots for multiple countries
55. tly taken from ISS2 They have been arranged into 5 year periods matching those used for mortality but starting earlier if available They have been entered so far as they are available for three product categories e manufactured cigarettes MC e total cigarettes i e including hand rolled TC and e all tobacco products A They are given as number or grams for all tobacco products per adult per day They commonly start around the 1920s but go back much earlier for a few countries Where original sales data are not available for all 5 years an average from the available years is used Note that sales based consumption data are not age or sex specific Tar data sales weighted average yield mg per cigarette are also available for some countries from ISS2 commonly starting around the 1960s Tar is assumed to be 35 mg cigarette up to 1955 except in the very few instances of actual data before that year with simple linear interpolation used to fill in missing years Tar data are also not age or sex specific The availability of sales based tobacco consumption and tar data is shown for each country in Table 2 The main survey based data consist of prevalence for each of the three product categories as a percentage and consumption of manufactured cigarettes and of total cigarettes as number per person per day adjusted to total sales Both these types of data are taken from Supplement 1 to ISS2 which made
56. try Country Ratio Sex amp Source Period or Age group Tobacco Type Tar Adjustment Lagging or Averaging Start Youngest End Oldest een ae UK v None Single Country Y Males amp Females Sales Y Age group 75 79 w Manufactured Cigarettes MC 20 Years previous Unadjusted 5 Years previous v PP 2 i ho av ND USA v None Single Country Males amp Females Sales Age group 75 79 vy Manufactured Cigarettes MC 20 Years previous v Unadjusted 5 Years previous vj udi p LC UK Y None Single Country Y Males amp Females v Age group 75 79 Lung Cancer v Mort 2 i a Lc USA v None Single Country Iv Males amp Females v Age group 75 79 v Lung Cancer v Consumption amp Mortality for Birth cohort 450 400 350 200 Consumption CPP 150 100 50 1836 1841 1846 1851 1856 1861 1866 1871 1876 1881 1886 1891 1896 1901 1906 1911 1916 1921 1926 1840 1845 1850 1855 1860 1865 1870 1875 1880 1885 1890 1895 1900 1905 1910 1915 1920 1925 1930 Birth Cohort 42 Example 14T Combination Plots Page 3 20 year average associated data table Consumption amp Mortality vs Birth Cohort Period for Lagging or Averaging v v v bd v bd 1911 1915 1916 1920 7 63 7 09 9 91 9 34 381 89 369 23 321 95 354 55 1996 2000 2001 2005 Not Used Not Used Cou
57. udes the data for the nearest relevant standard period and age group either duplicating data or leaving gaps as appropriate to the individual case For instance many of the sources are based on 10 year birth cohorts and these have been entered by repeating the data for the two relevant age groups The user will need to consider how to use these alternative data sets on an individual basis Their advantage is that they often allow lifetime tobacco history to be estimated over a longer time period than is possible with the main survey based data Definitions of causes of death Deaths are coded according to successive revisions of the International Classification of Diseases ICD Countries vary as to when and if they introduce the various revisions see Table 1 In the 1950s the 6 revision was in use and the latest 10 revision published in 1992 is now in use by 16 of the countries Except in the case of the 10 revision which is used in full data are provided to WHO according to various summary coding lists based on the ICD and some countries use their own more limited versions Changes between the ICD revisions reflect differences in the understanding of the disease process and changes in terminology For instance stroke was a disease of the nervous system up to the 7 revision and of the vascular system subsequently while COPD is a term of only relatively recent origin These changes can make it difficult to extract consistently de
58. y and tobacco statistics with each line individually defined The first three pages have Age Period and Cohort on the x axis respectively Both Consumption and Mortality are available on the y axis with the scale relevant to the tobacco statistics shown on the left and the scale relevant to the mortality statistics on the right The next three pages repeat the sequence with Prevalence and Mortality On each plot up to 3 tobacco lines and 3 mortality lines can be drawn If less than the full 6 lines are required use Plot Settings to hide unwanted ones See Example 13 generated on page 5 where the third mortality line has been hidden Then for each line in turn use the drop down menus to choose Age including combined age group Period or Cohort except for the one in use on the x axis Sex including combined and ratio and Country including country ratio Other options applicable to the tobacco lines are Tobacco type Tar adjustment not available for prevalence and Lagging averaging and to the mortality lines Disease Note that in the consumption plots the choice between Sales and Survey based data is made by the Sex amp Source drop down since sales data are not sex specific Changes to the drop down menus are not immediately reflected in the plot Instead a marker Update Req is shown alongside the line in question and an Update Required button is displayed over the plot When all the line definitions are ready click
Download Pdf Manuals
Related Search