Quit Evaluation Studies: Volume 7

CHAPTER 5

Smoking behaviour, knowledge and intention: Change 1983­1993

Robyn Mullins
Ron Borland


Introduction

The Victorian Smoking and Health Program (VSHP) began in 1985, after a Ministerial Review established smoking as the single biggest cause of disease in Victoria. Before its establishment, smoking prevention activity in Victoria had been very limited, restricted largely to activism from the Anti-Cancer Council of Victoria (ACCV) and a No Butts day in 1984, instigated by the Health Department Victoria (HDV).

In its first three years Quit, as the VSHP is usually known, was funded by the ACCV, HDV and the National Heart Foundation (Victorian Branch), and had a budget of approximately one million dollars per year. In 1987, Victoria's Tobacco Act established the Victorian Health Promotion Foundation (VHPF), funded by a tax on tobacco products. Each year, Quit receives 10% of this budget. In 1988 this led to an increase in budget to around four million dollars, and the figure is currently approximately $2.5 million.

Since its inception, Quit has run an annual media campaign in the electronic and print media, produced written materials about smoking, and offered courses and advice for those who want to give up. The extra funding in 1988 enabled the scale of these activities to be enlarged, and allowed Quit to sponsor some sporting and cultural events, most significantly the Australian Rules Football club, Fitzroy, a sponsorship that started in 1987 with a special grant from ACCV.

Although the VSHP did not begin until 1985, it was in 1983 that the ACCV began collecting annual data on smoking in Victoria. The first annual survey in 1983 was a very brief one compared to current surveys and, apart from monitoring the prevalence of smoking, focused almost entirely on current smokers. The issues covered were:
This report looks at the response to these questions between 1983 and 1993. It is important to note that Quit does not operate in isolation, but as part of a wider social context. The Federal Government has run anti-smoking campaigns during the same period, and there have been some important court rulings affecting smoking - in particular the Morling judgement which ruled that passive smoking was harmful to non-smokers. Quit is one of many influences. Clearly, other factors, some of which Quit has influenced, have also played a role in determining current smoking prevalence and attitudes in Victoria.

Method

Since 1983, the same methodology has been used to collect the smoking data - a random survey of people interviewed in their own homes by a market research company. The method is described in full in Hill and Gray (1984). Some questions are asked every year, and others are introduced or excluded each year according to the demands of research or evaluation. In this report only those questions which were first included in 1983 have been included. Table 1 lists the questions and the years they have been asked from 1983.

Table 1 Years in which each of the original questions were asked
'83 '84 '85 '86 '87 '88 '89 '90 '91 '92 '93
Prevalence x x x x x x x x x x x
Likelihood of quitting x x x x x x x x x x x
People in favour / against quitting x x x x x x x
Risk of fatal illness x x x x x x x x
Advantages and disadvantages if quit x x x x x
Illnesses caused / helped x x x x x x x x x
When quit x x x x x x x x x x


Only the prevalence question, and the question about the smokers' likelihood of quitting in the next three months have been asked every year since 1983. The other questions have been asked in at least five years, and all but the advantages and disadvantages of quitting and people in favour and against quitting were included in the 1993 survey.

From 1985 onwards, data from each survey has been reported in the annual Quit Evaluation Studies volumes. In cases where data is not reported for specific years in this report, further detail can be found in the relevant volume. In most years the sample was around 2500 people, but in 1983 it was far smaller, and for some questions there were only about 200 smokers asked. In some cases, when it is desirable to make socio-demographic comparisons, 1985 data has been compared with 1993 data, as the 1985 sample size was bigger and there were few changes between 1983 and 1985.

Statistical analysis

To test for the significance of relationships between variables, various statistics have been used, mainly chi-square tests, t-tests and analyses of variance. Details of statistical tests are not given in the text. Where relationships between variables are reported, the probability level of significance was less than 0.05, indicating a less than 1 in 20 chance that the finding is due solely to chance.

To look at demographic differences in the data, age has been divided into three categories: 16 to 29 years old; 30 to 49 years old; and 50 years and over. Educational attainment was divided into four categories: Year 9 or less; Year 10 or 11; completed Year 12 but not university; and completed tertiary education. Unless otherwise stated, occupational status is based on that of the main income earner in the household, classified as upper or lower white collar or upper or lower blue collar. Only 41% of the sample worked full time, making classification by the respondent's own occupational status inappropriate.

Table 2 Household survey respondents by year and smoking status
Current Past Never Can't say Total
1983 516 315 643 38 1512
1984 381 251 536 35 1203
1985 808 558 1221 26 2613
1986 790 585 1093 31 2499


Total		 2495 1709 3493 130 7827
1988 689 542 998 22 2251
1989 669 535 1117 13 2334
1990 658 626 1064 9 2357


Total		 2016 1703 3179 44 6942
1991 633 654 1146 7 2440
1992 606 641 1094 14 2355
1993 601 656 1129 1 2387


Total		 1840 1951 3369 22 7182

Prevalence of smoking

Between 1983 and 1993 there was a slow but steady decline in the proportion of the population who described themselves as smokers. In 1983, 33.1% of people said they currently smoked at least one tobacco product. By 1993 this had declined to 25%, a drop over the years of 8.1% (95%CI of diff= 5.2 to 11.0).

In 1983, 54.4% of respondents had smoked at some stage, and of these 39.1% had quit. By 1993, a very similar proportion of respondents (52.7%) had smoked, but more (53.1%) had quit.

In the figures below, smoking prevalence is presented by sex, age and occupational status. The data has been smoothed by averaging each three years of data, to reduce any random fluctuations. For example, the 1987 figure is calculated by averaging the measured prevalence for 1986, 1987 and 1988. The end points, 1983 and 1993, are based on only two years of data, as the third year either was not collected or has not yet been collected.
Figure 1 Smoking prevalence by sex
Fig 1:Smoking prevalence by sex
It is apparent form Figure 1 that there has been a decrease in the prevalence of smoking amongst both men and women over the past ten years. The difference between men's and women's smoking prevalence has closed.
Figure 2 Smoking prevalence by age
Fig 2: Smoking prevalence by age
As indicated in Figure 2, the relationship between age and smoking is somewhat more complicated than that with sex. For both those over 50 and those under 30 years old, there has been a steady decline in smoking prevalence. However, since 1991, smoking prevalence in the 16-29 year age group appears to be increasing. For those aged 30 to 49, the overall decline has been slight, and in the late 1980s there was a small increase in smoking prevalence in this age group.
Figure 3 Smoking prevalence by occupation of main income earner
Fig 3: Smoking prevalence by occupation
As indicated in Figure 3, there is certainly a difference in the proportion of blue collar and white collar workers who smoke, with more blue collar smokers. There has been a downward trend for both groups over the last ten years, but the difference between the two still exists. In fact, approximately the same proportion of blue collar workers smoked in 1993, as the proportion of white collar workers who smoked in 1983. It is reassuring to see that the campaign had roughly equal effects on the two groups, and not just an effect on white collar occupational classifications.

Likelihood of quitting

In each year since 1983, smokers have been asked to indicate how likely they thought it was that they would give up smoking in the next three months. They indicated their response on a seven-point scale, ranging from extremely unlikely, to extremely likely. The figure below presents "likely" and "unlikely" data for each year. In each year, around 10% of respondents said they were "neither likely nor unlikely to quit" in the next three months.

As the data in Figure 4 indicates, there has been an increase in the percentage of smokers intending to quit since 1983, and a corresponding decrease who say it is unlikely they will quit. The increase mainly occurred over the period 1984 and 1987. From 1988 until 1993, the levels have remained fairly stable with total likely between 40% and 45%, and total unlikely between 45% and 50%.
Figure 4 Intention of quitting in the next three months
Fig 4:Intention of quitting

Chance of smoking causing a fatal illness

The perception that smoking will cause the smoker personally a fatal illness may affect the possibility that he or she will quit. To measure this perception, all smokers were asked "If you continue to smoke, what do you think is the percentage chance that smoking will cause you a fatal illness?". Respondents were shown a card with options ranging from 0 to 100%, in increments of 10%, and asked to indicate what they believed the chance was.

Between 1983 and 1993, there was a 14% decrease in the proportion of people who believed they had less than a 50/50 chance of getting a fatal illness if they continued to smoke (95%CI of diff= 6.5 to 21.5). The proportion who completely denied any risk dropped from 15% to 8% (95%CI of diff= 1.9 to 12.2).

Table 3 Percentage chance of smoking causing a fatal illness
1983 1993


n=		 224 601
0 15% 46% 8% 32%
10% 11% 8%
20% 10% 6%
30% 7% 7%
40% 3% 3%
50% 31% 27%
60% 1% 21% 7% 39%
70% 5% 6%
80% 6% 7%
90% 2% 5%
100% 7% 14%
Can't say 2% 4%


In some ways, it is a good thing that smokers generally rate their risks from smoking quite highly, as it indicates that they are aware of the risks associated with smoking. However, it is disturbing that, despite this perception of risk, they continue to smoke. The proportion of people who said there was a 100% chance they would contract a fatal illness from smoking had increased from 7% to 14%, a rise of 7% (95%CI of diff= 2.6 to 11.4). Belief that a fatal illness is a certainty is not due to older people reaching a stage where they are already suffering smoking-related illnesses, but believe that it is too late for them to quit, as in 1993 only 10% of those over the age of 60 rated their chances as 100%, yet 16% of those aged 16 to 29 years old said their chance was 100% and yet continued to smoke.

Why then do people continue to smoke when they consider the risks to be so high, higher in fact than the risk actually is? It is likely that they do not actually respond in terms of probability, but are really saying that they think smoking is very dangerous. The change over time suggests that while the Quit Campaign has been successful in making smokers aware that they personally face serious health consequences from smoking, some smokers are capable of holding that belief, but still continuing to smoke.

Barriers to, and benefits of, quitting

Clearly, knowledge of the health risks associated with smoking is not sufficient to lead smokers to give up. There are other factors which affect the decision, including beliefs about the benefits of quitting and the level of support a person expects to receive from others if they attempt to quit. Some of these additional factors are discussed below.

People in favour/against

A supportive social environment may be an important factor in influencing a smoker to quit, or at least to make an attempt to do so. In 1983, and in six other surveys since then, smokers have been asked to identify first the person who was most in favour of them quitting, and then every person who was in favour of it. In Table 4 data from 1985 is compared with the most recent data (1992). The reason for this is that the 1983 sample was very small, which makes comparisons among subgroups difficult.

As the table below indicates, there has been no real change in the pattern of response from 1985 to 1992, with the same people being named as being in favour of the respondent quitting in each year. The only exception to this was a slight increase in the proportion who named either their son or daughter as the person most in favour of them quitting. In each year, most people could name someone who wanted to quit. The very small number who named anyone as being against them quitting made it impossible to draw conclusion about any change in data.

Table 4 People most in favour of, and most against, respondent quitting
In favour Against
1985 1992 1985 1992


n=		 792 606 792 606
Spouse/partner 34% 35% 2% 2%
Mother 18% 16% * *
Father 4% 3% * *
Sister 2% 2% * *
Brother * - * *
Son 8% 10% 1% *
Daughter 5% 10% * *
Other relative 2% 2% 1% 1%
Workmate(s) 3% 2% 1% 1%
Friend(s) 5% 5% 2% 4%
Own doctor 4% 2% * *
At least one 86% 87% 8% 11%
No-one 14% 12% 92% 89%
* <1%

Of more importance than the person most in favour of the respondent quitting is the overall support for them quitting, that is, whether there is support for them quitting among all their family and friends. Naturally, not everyone will have all of the social reference groups described above - for example, many respondents will not have children - so it is important to compare relevant subgroups.

For people who were married, there was a trend towards an increase (47.5% to 52%) in the number who named their spouse as being in favour of them quitting; however, this difference was not significant (95%CI of diff= -1.7 to 10.7; ns).

For respondents aged between 16 and 19 years of age, whose parents may feel very strongly about them smoking, there has been a big increase in the percentage naming a parent as in favour of them quitting. In 1985, 70% of respondents had said their mother was in favour of them quitting, and in 1992 this had increased to 89% (95%CI of diff= 4.2 to 33.8). Fathers were named less frequently than mothers, but there was still a trend for them to be named more frequently in 1992 than 1985 (73% to 58%, 95%CI of diff= -3.1 to 33.1).

For people with children under 16, pressure was mounting from their sons and daughters, with more than half saying their sons (51%) and daughters (53%) would like them to quit in 1992. In 1985, only one-third of sons (32%) and daughters (33%) had said this. This shows a significant increase for both sons (95%CI of diff= 10.4 to 27.6) and daughters (95%CI of diff= 11.4 to 28.6).

In both years, 13% of those in full time employment said their workmates were in favour of them quitting.

Overall, the vast majority of smokers had at least one person in favour of them quitting. The pattern of response of those most in favour had not changed since 1985, but there had been changes when the total people in favour were considered. Increased social support came mainly from relatives, not the work environment.

 



 Home


 more Chapter 5