| Smoking and Health | Statistics In Your World |
|
|
| Student Notes |
|
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Back to Notes | |
|||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
| Brief Description Aims and Objectives Prerequisites Section A
|
|
Brief Description Design Time: 4-5 hours Aims and Objectives On completing this unit pupils will have practised reading tables, plotting graphs, drawing and considering inferences, interpreting tables and data, comparing two or more sets of data, describing and assessing trends, plotting and reading bar charts, using and interpreting death rates, reading time series, calculating percentage changes and using them for comparison, interpreting statements using the word ‘average’ and (optionally) making a critical assessment of other comments based on data. They will be more aware of data sources, the problems of data comparison, the distinction between correlation and causality, the data connecting smoking with various aspects of health and (optionally) the distinction between conclusions shown from cross-sectional and longitudinal studies. Background Reading Royal College of Physicians, Smoking OR Health, Pitman Medical (1977). Prerequisites Pages R1 and R2 are for use in Section A2, and page R3 is for use in sections B2 and D2. Section A introduces the problem of smoking and health and shows the extent of smoking by giving data on manufactured cigarettes smoked in the period 1930 to 1975. Section B considers various aspects of health when compared to smoking. B) considers death rates from coronary heart disease. B2 introduces data on lung cancer and chronic bronchitis. B3 looks at the connection with lung cancer from a different point of view. Section C considers how doctors’ smoking habits have changed since the publication of the first report by the Royal College of Physicians. Section D considers the connection between smoking in pregnancy and the birth weight of babies. Section E reviews all the evidence presented and concludes by considering, in an optional section, some comments from the report Smoking OR Health, to which fairly extensive reference is made throughout the unit. We recommend that a copy of the report be available to the teacher, perhaps through the school library. Detailed Notes Section A Al Some of the reasons for starting smoking given in the report (page 105) include ‘availability of cigarettes’, ‘curiosity’, ‘to appear tough’, ‘social confidence’ and ‘friends smoking’. Four reasons for not ceasing to smoke are quoted in Al. It may be valuable to add to this list from staff colleagues who smoke. Certain statistical implications may be highlighted in the class discussion. 1st quote. Is this sample of size 1 statistically significant? i.e. should one base decisions on the evidence of one person’s experience, or try to see a larger sample of smokers? See also E1d. What about all the grandfathers who might have survived to such an age but did not? 2nd quote. The meaning here is rather obscure. Perhaps it implies ‘Why worry about trying to exercise control over whether you live or die. The hand of fate may strike you dead anyway’. This is a somewhat fatalistic view of life. However, it may be worth discussing the relative probabilities of the two events ‘death by smoking’ and ‘death by road accident’.Although anyone may die of lung cancer, or under a bus, we can reduce the probability of either of these occurrences — by not smoking and taking care on the road, respectively. 3rd quote. This is dealt with in Section C. It may be interesting to establish (i) do pupils believe that many doctors smoke, and (ii) are they likely to be influenced by the smoking habits of doctors? 4th quote. Statistical evidence of the kind available here can never amount to 100% proof. But the evidence is consistent with smoking being a major contributory cause of lung cancer. Of other variables tried by multiple regression none has shown anywhere near so significant a relationship. Other reasons given by pupils in trials of this unit were: ‘to see what it’s like’, ‘because it’s manly’, ‘it’s sociable’, ‘copying adults’, ‘advertisements’, ‘to give your hands something to do’ and ‘it’s a drug’. A2 a This graph could be drawn for homework if required. d 1945. Consumption grew more quickly during the war years(1930-45) but fell off afterwards. e There was a steady increase in consumption to a peak in 1945. After that men’s smoking dropped by a quarter in the years to 1949, with minor fluctuations recovered very gradually up to 1961, then settled at a slightly lower level. Smoking among women was less than a quarter as many cigarettes as men up to 1942. The decline was less pronounced than men’s in 1946-48, whereas the increase since then has continued so that by 1965 women’s consumption was over half the male daily levels and over two-thirds by 1975. Taking men and women together, the 1945 peak was exceeded in 1961 and equalled in 1963. The figures are averaged per member of live population. You cannot tell how far they represent changes in the number of smokers, or in the number of cigarettes each smoker consumed. Section B The tubes (trachea, bronchi and bronchioles) of the respiratory system are lined with ciliated cells which produce a thin mucus. The mucus collects particles of dust, etc., breathed in, and the cilia move this up to the back of the throat from where it trickles into the oesophagus and thence into the stomach. It is believed that smoke causes the cells to produce thicker mucus which cannot be effectively removed by ciliary action, and coughing is necessary to bring up the thicker phlegm as this thickened mucus is colloquially called. Under these circumstances the effectiveness of the filtering action is lessened, and tar and other irritants in the smoke get into the lungs. Smoke contains a relatively high proportion of carbon monoxide, a gas which combines preferentially with haemoglobin, thus reducing the volume of oxygen that the blood can carry. Heart muscle requires a rich blood supply and if the circulation is poor (this may occur for a variety of reasons, including obesity), then the reduced oxygen supply in a smoker’s blood may be critical. Other factors may be involved. B1 Note that Figure 1 refers only to male doctors who are under 45 years old, and the death rates quoted only apply to deaths attributed to C.H.D. The numbers expressed here, as in the original, are annual age specific death rates, per 100000 hypothetical doctors in each smoking category. Actual deaths over a 20-year period were:
B2 a Men and women are listed separately because of the striking differences in mortality rates between the sexes. (It is instructive to compare this with the different smoking habits of the two sexes, in Table 1). i See notes above before B). B3 b Many men have already died by this age. d The levels are higher for men throughout. The increase in cigarettes smoked 1930-45 was about equally shared between women and men, representing a much smaller proportional change among men. Since 1948 there has been fluctuation, but all within the 9-Il a day range, i.e. a fairly level pattern compared with a continuing increase among women. Men now smoke roughly 50% more cigarettes than women, compared to 15 times as many in 1930, nearly four times as many in 1945, and about three times as many in 1949-50. Death rates from lung cancer grew more steeply among men, from possibly double women’s rates in 1930 to about five times as many in 1956. Since then death rates among men have levelled off, but still look about four times as high as among women. One might expect these to continue converging, as the cancer deaths seem to lag some 10 years behind the changes in smoking habits. e The similar shape of the curves is consistent with a connection, but in itself this is not evidence of any direct causal link. It is enough to make one suspect a relationship, and wish to investigate farther, particularly as deaths from other forms of cancer have moved in the opposite direction. f Similarity could be due to a common third cause, coincidental, or due to positive correlation with some genetic effect. Section C C1 b In the period 1950-70, cigarette consumption among doctors fell consistently, while among all men it changed little or rose slightly. Doctors over 60 years old reduced their smoking more gradually, from a lower starting point than younger doctors but a higher one than other men of the same age. These gradually increased their consumption. Doctors aged 35-39 reduced their consumption more sharply to reach levels close to those of the older doctors, while all men aged 35-39 did not alter their consumption very much. d Surveys by random sampling methods among doctors, and among the male population at large, on the dates for which information is required. A question on the age of each respondent would be needed, as well as whether a smoker and the average daily consumption of cigarettes. A list of doctors exists (e.g. general medical register), but the sampling frame for all men is more troublesome and it may be best to use a firm of consultants specializing in this kind of work. Electoral registers give a useful sampling frame but are costly for this kind of work and some sort of clustering seems a probable solution. People do not necessarily remember or report their smoking habits accurately. It might be necessary to sample all smokers (at a lower sampling proportion than the men of 35 and over) so as to estimate total cigarette consumption. This could be checked against sales by the tobacco companies to see that the figures were more or less of the right order. e The figures give average numbers of cigarettes per doctor per day. This is obtained by estimating total consumption of cigarettes by doctors and dividing by the total number of doctors (both those who smoke and those who do not). Hence a drop in consumption could be due either to some people giving up, or to some people smoking fewer cigarettes, or a combination of the two. A specific census of doctors or a sample enquiry properly constructed could be arranged to yield information on how many doctors now smoke, and how many did 20 years ago and have given up. Some of them may have difficulty in remembering exactly when they gave up, and a less specific question like ‘Did you once smoke regularly?’ (to people who already said they do not smoke now) might yield more usable information. In fact the same doctors were questioned at different times. This method is reliable when you can find most of them again. C2 a A lot of the information reproduced here was first published in journals read by doctors. Some of them wanted also to set an example to their patients. They were in a good position to verify the effects of smoking on their own patients. They form a comparatively convenient group to sample, easy to reach and trace (there are specific mailing lists), responding reliably, and enough of them gave up cigarettes to test the effects of it. b It would bring about a small reduction especially among younger doctors. c It fell. d It rose and was the only one of the four death rates to do so. e The reduced smoking by doctors has been associated with a marked decline in death rates from smoking-associated diseases, while among other men stable consumption has been accompanied by an increase in death rates. Cigarette-related diseases have grown to account for more than half all male deaths, but less than half male doctor deaths (despite lower death rates from other diseases among doctors than other men). Section D The data in this section relate to the birth weights of babies of mothers, some of whom smoked during pregnancy. The figures relate only to a sample of 29 babies since the published figures of larger surveys have already been aggregated, and the plotting of weights of individual babies from these figures is not possible. Smaller babies are less likely to survive, as is brought out in the data used in one of the test questions. This is not the only danger to which babies of smoking mothers are exposed. Smoking tends to decrease the maternal blood supply to the foetus, and this may affect the development of the baby’s brain and hence make it less intelligent than it otherwise might have been. D2 d There is a downward tendency so that babies of smoking mothers tend to have a lower birth weight. e (i) Physique of both parents (ii) Nutrition of mother during pregnancy (iii) Duration of pregnancy (premature babies tend to be lighter). Section E Ela In Section B) we found that for male doctors under 45 years of age, the risk of death from C.ll.D. went up with cigarettes smoked. For instance, a doctor who smoked 25 or more cigarettes per day was 15 times as likely to die from C.H.D. as a non-smoker. In Section B3 we saw that the death rate from lung cancer has increased significantly since 1916. However, the pattern of deaths since then shows a marked similarity to the pattern of smoking over the same period. In Section C1 we found that the percentage of doctors who smoke has fallen to less than half of the 54% who smoked in 1951. Section D showed a rather weak negative correlation between smoking and birth weight. The contrast in average birth weight of these babies with a random sample of non-smokers’ babies is quite marked. d The sample size of 1 is too small to give a fair picture. He may be a quite unrepresentative grandad. He is noticed because he is alive — the many people who died earlier are forgotten. Compare with the statement ‘Motor racing does not cause deaths. I know a driver who lived until the age of 82’. E2 In all this section, as an alternative to choosing one subject, several or all of the statements could be reviewed in a wide ranging class discussion. It is notable that all require comparisons collected over quite long periods of time. Other quotations from the same report that could be used are: In 1974, 31000 men and 8000 women between the ages of 35 and 64 died from coronary heart disease in the UK. Over 9 500 of these deaths could be attributed to smoking. In Britain it was found that babies weighing less than 25 kg at birth were nearly twice as common among mothers who smoked than among those who did not. Still-birth and death in the first weeks of life occur nearly 30% more often in the babies of mothers who smoke after the fourth month of pregnancy. Acknowledgements: We thank the following for permission to quote from and use copyright material: The Pitman Medical Publishing Co. Ltd. and the Royal College of Physicians for Figures 1-4 and Tables 2, 3 and 5. All these derive from the publication Smoking OR Health a report of the Royal College of Physicians, published by the Pitman Medical Publishing Co. Answers A2 b 1940 c 1944 d See detailed notes *f l0.2 . 100/47 = 21.7 *g 6.8 x l00/43 =l5.8 3 See detailed notes B1 a 61 per 100 000, about 9 b 104 per 100000, about 15 d see detailed notes. e See detailed notes. B2 a see detailed no9tes b 75+, c 7.6 x 600/319/2 = 2.4% Dl a 61 per 100000, about 9 d See detailed notes. d Table 7 Percentage death rates
e C.H.D. of those causes given f C.H.D. of those causes given *g 45.7% h 26.1% i See detailed notes. B3 a about 98/8.2 = 1.2 b Table 4 Deaths per 100 000 from cancer
c men l60/45=3.6 women30/7.5 = 4 d, e and f See detailed notes. C1 a 21% b See detailed notes. c 1954 d and e See detailed notes. C2 a See detailed notes. b See detailed notes. c (242 — 297)/297 x 100 = —l8.5% d (212 — 282)/282 x 100 = —24.8% e See detailed notes. C3 a 365.25 x 40 = 14610 (365 x 40 = 14600) b 14610 x 20 = 292200 (14600 x 20 = 292000) c 365 x 4 x 24 x 60 + 366 x 24 x 60 = 2629440 d 2629440 + 292200 = 9 minutes (2629440 / 292000 = 9 mins) Dl b See detailed notes. D2 a Table 6, last two lines:
d and e See detailed notes. El a See detailed notes. d See detailed notes. Test Questions
3 Use Graph 2 to answer these questions.
4 The table gives the death rates from lung cancer in British male doctors.
(Source: British Medical Journal 1976, 2. 1527)
Copy and complete the following sentences relating to these doctors:
a Fewer doctors smoke now than 20 years ago. (True/False)
Test Questions: Answers 1 a True b False c True d False 2 Generally declined from about 61% to about 46%. 3 a Men smoke more cigarettes than women.
4b 106 c 10 5 Statement b is true. 6 a True b True c False 7 More likely to die than babies of mothers who don’t smoke (or equivalent). The more cigarettes smoked the more likely the babies are to die (or equivalent). Resource (R) Pages R1
R2
R3 Table 7 Percentage of male and female deaths due to lung cancer, chronic bronchitis and C.H.D. in 1974 (UK all ages)
Table 8
Connections with other Units Other Units at the Same Level (Level 4) Choice or Chance, Testing Testing , Figuring the Future, Retail Price Index, Sampling the Census, Equal Pay Units at Other Levels in the Same or Allied Areas of the Curriculum Level 1 Practice makes Perfect, Tidy Tables Level 2 Seeing is Believing, Getting it Right Level 3 Net Catch, Multiplying People, Leisure for Pleasure, Opinion Matters, Cutting it Fine, Pupil Poll This unit is particularly relevant to: Science, General Knowledge, Health Education, Mathematics. Interconnections between Concepts and Techniques Used in these Units These are detailed in the following table. The code number in the left-hand column refers to the items spelled out in more detail in Chapter 5 of Teaching Statistics 11-16. An item mentioned under Statistical Prerequisites needs to be covered before this unit is taught. Units which introduce this idea or technique are listed alongside. An item mentioned under Idea or Technique Used is not specifically introduced or necessarily pointed out as such in the unit. There may be one or more specific examples of a more general concept. No previous experience is necessary with these items before teaching the unit, but more practice can be obtained before or afterwards by using the other units listed in the two columns alongside. An item mentioned under Idea or Technique Introduced occurs specifically in the unit and, if a technique, there will be specific detailed instruction for carrying it out. Further practice and reinforcement can be carried out by using the other units listed alongside.
|
||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
