Commentary
Monitoring Tobacco Use in Canada:
The Need for a Surveillance Strategy

Roberta Ferrence and Thomas Stephens

Abstract

Smoking behaviour has been monitored nationally through population surveys for 35 years in Canada, but these surveys have not been as consistent or rigorous as the magnitude of the smoking problem demands. Inconsistent methods and irregular survey intervals are just two of the characteristics that have made it difficult to know exactly how smoking is changing. Further, an absence of routine data on tobacco control policies (other than the price of cigarettes) has hampered understanding of the determinants of changing prevalence. The advent of two survey series—Canadian Tobacco Use Monitoring Survey (CTUMS) and Canadian Community Health Survey (CCHS)—promises to change this situation for the better. We suggest that both are critical elements of a national smoking surveillance system and that, with a commitment to CTUMS in particular, Health Canada could set a new international standard for surveillance.

Key words: Canada; monitoring; smoking; surveillance

Introduction

There should be no need to detail the devastation caused by tobacco in Canada. Estimates of health and economic costs totalling $9.56 billion in 1992 dollars¹ are conservative. Even before the 1994 federal and provincial tax cuts on tobacco, smoking-related costs were double the tax revenue produced by tobacco.²

Although there is considerable knowledge about strategies that can reduce this damage ³ and many such strategies are in place to varying degrees across Canada, routine surveillance and evaluation of their impact are generally absent. Unlike the United States and the province of Ontario, Canada has never implemented a surveillance system for systematically collecting data on various aspects of smoking behaviour and the determinants of smoking. Although Canada has 35 years of experience in surveying adult smoking at the national level, inconsistent methods and irregular frequency of surveys have provided less useful information than the magnitude of the problem deserves. This inconsistency has led to great difficulty in estimating the impact of policy changes, such as the major tax cut introduced with little advance notice in five Canadian provinces in February 1994,⁴ almost three years after the previous national survey of smoking behaviour. The resulting apparent confusion about the effects of government policies, especially taxation, has been exploited routinely by the tobacco industry, most recently in May of this year.⁵

Health Canada's Commitment

Like the health consequences of smoking and their attendant economic costs, the need for national surveillance appears to be well understood and accepted, at least in principle. Almost a decade ago, the first Directional Paper of the National Program to Reduce Tobacco Use in Canada⁶ identified research and knowledge development as a strategic direction, including "ongoing surveys of tobacco use . . . to aid planning at the regional level." Data gaps, including a critical absence of baseline information needed for many national goals, were again recognized when the National Strategy was updated in 1993.⁷ Yet many of these gaps remain today.

Health Canada's report Tobacco Control: A Blueprint to Protect the Health of Canadians⁸ added a potentially valuable element to the discussion by proposing to expand the requirements for the tobacco industry to report sales, product constituents and other proprietary data. However, although the industry has the longest running consistent time series on tobacco use in Canada, these data are still not available for tobacco control efforts.^a

Most recently, the National Strategy to Reduce Tobacco Use in Canada has again recognized the need to "monitor knowledge, attitudes and behaviour about smoking and second hand smoke among children, youth and adults on an ongoing basis," and to "monitor and evaluate intervention activities and outcomes on an ongoing basis."⁹

This official recognition of the need for surveillance has been matched by much survey activity at the national level (Table 1), but without a systematic approach.¹⁰ The interval for most data series is well over two years, much too long to be able to detect changes let alone attribute them to policy developments. For example, there were no definitive national surveys in 1992 and 1993, which, as already noted, has resulted in much speculation as to what was happening to smoking rates during this period.

There have been other important single surveys besides those cited in Table 1, notably the 1994 Youth Smoking Survey,¹¹ which covered ages 10S19. These have produced very useful data, but their infrequency diminishes their value. Moreover, as Table 1 shows, it is relatively rare to have data for the critical years before age 15, a hold-over from the early days when tobacco surveillance was an add-on to a survey of labour force activity.

Since the mid-1990s, there have also been several studies of tobacco control policies, including national surveys of the following.

• Retailer compliance (AC Nielson 1995, 1996, 1997, 1998)¹²
• Smoking restrictions in municipalities in 1991 and 1995 (Health Canada, 1992, 1995)^13,14
• Smoking restrictions in other public settings (Goss Gilroy, 1995)¹⁵
• Attitudes and knowledge regarding smoking restrictions in homes (EKOS, 1995)¹⁶
• School-based smoking prevention programs (Health Canada, 1994)¹⁷

Although not intended for monitoring tobacco control, an important addition to this list is Statistics Canada's quarterly survey of the price of cigarettes.¹⁸

All of these studies have provided useful data regarding tobacco control policy. However, apart from the monitoring of prices, only one survey (retailer compliance) has so far produced more than two data points.

Besides the importance of being able to assess policy and whether it is being implemented as intended (e.g. retailer compliance), these studies can offer important insights into the relative contributions of particular measures to tobacco control when linked to data on smoking behaviour from population surveys.¹⁹

Current Surveillance Tools

More recently, population surveillance appears to have improved markedly. Beginning in 1999, the Canadian Tobacco Use Monitoring Survey (CTUMS) was put in place to monitor tobacco use in Canada.²⁰ It is only the second adult survey in Canada to be devoted to tobacco use, and its topical coverage is broad: use of cigarettes and alternative forms of tobacco, age of initiation, access to cigarettes, cessation (including reasons and incentives), use of cessation aids, readiness to quit, environmental tobacco smoke (ETS) exposure, restrictions on smoking at home, attitudes toward tobacco control policies, beliefs about "light" cigarettes and awareness of tobacco-industry sponsorship activity. Monthly data collection allows for more precise assessment of specific changes and a larger number of data points for more powerful analyses.

Starting in September 2000, Statistics Canada will inaugurate the Canadian Community Health Survey (CCHS). As part of the core survey content this omnibus health survey will include type of smoker, amount smoked, cessation, age of initiation, use of other tobacco products, workplace restrictions and ETS exposure. The CCHS will be an important addition to the surveillance arsenal, but what it will provide in the way of consistency and robustness will probably be offset by its lack of flexibility and timeliness. This is where CTUMS fits in. When we compare the two surveys on a number of parameters, each shows important strengths for tobacco surveillance (Table 2).

The CCHS is designed to provide basic planning data to 130 local area health units on a wide range of topics. Its strength will be its large sample size (every second year) and geographic coverage. With updated small area data from other sources (e.g. retailer compliance or municipal bylaw coverage), there is very strong potential not only for monitoring but also for assessing the impact of certain tobacco control policies on the fundamentals of tobacco use. However, the reliability of youth data in the smaller sample years will be less than that of CTUMS. Another weakness of the CCHS is timeliness. The gap between introducing a new topic and the time when the results are available will be well over two years, judging by performance to date with the National Population Health Survey and the sheer size of this new survey. In contrast, CTUMS has completed one cycle and has shown that it can produce data on new topics in well under a year. This advantage stems from its more focused content and in-house management of the survey. Further, there are few limitations on the kinds of tobacco content that can be included.²⁰

Components of a National Surveillance System

Given the importance of monitoring not only tobacco use but also tobacco policy "inputs,"²¹ we suggest that an effective national surveillance system should have the following components.

• An ongoing population survey to provide small area data on basic tobacco use indicators (i.e. CCHS every second year, starting in 2000/01)
• An ongoing program of surveys and special studies to provide small area data on tobacco policy inputs, to be linked with the CCHS data, for assessing the impact of such policies as taxation, clean air bylaws and public education
• An ongoing population survey to provide reliable provincial data on emerging issues in tobacco control as well as basic tobacco use, with the flexibility to target specified groups such as youth, women of child-bearing age, heavy smokers, etc. (i.e. CTUMS, preferably every year or, at a minimum, every second year after 1999)

The elements of such a surveillance system are known and could be put in place readily. However, there is apparently still no long-term commitment to support CTUMS, nor is there any evident plan for routinely assessing policy inputs. Continued failure to develop this or an equivalent system—with a commitment to consistency over future years—will seriously hamper tobacco control policy and tobacco research in this country. Adoption of such a plan, given the existence of the CCHS and CTUMS as complementary elements, could simultaneously place Canada in a position of international leadership on tobacco surveillance and build a constructive partnership between Health Canada and its provincial counterparts.

References

1. Single E, Robson L, Xie X, Rehm J. The economic costs of alcohol, tobacco and illicit drugs in Canada, 1992. Addiction 1998:93:991S1006.

2. Kaiserman MJ. The cost of smoking in Canada, 1991. Chronic Dis Can 1997;18(1):13S9.

3. Lantz PM, Jacobson PD, Watner KE, Wasserman J, Pollack HA, Berson J, et al. Investing in youth tobacco control: a review of smoking prevention and control strategies. Tobacco Control 2000;9:47S63.

4. Stephens T. Trends in the prevalence of smoking, 1991S1994 [workshop report]. Chronic Dis Can 1995;16(1):27S32.

5. Parker R. Youth smoking increases happened before tax cuts. Toronto Star. 2000 May 16; Letters.

6. Consultation, Planning and Implementation Committee. Directional paper of the National Program to Reduce Tobacco Use in Canada. Ottawa: Health Canada, 1987.

7. Health Canada. A guide for tracking progress on the objectives of the National Strategy to Reduce Tobacco Use in Canada. Ottawa: Health Canada, 1993.

8. Health Canada. Tobacco control: a blueprint to protect the health of Canadians. Ottawa: Health Canada, 1995.

9. Steering Committee of the National Strategy to Reduce Tobacco Use in Canada, in partnership with the Advisory Committee on Population Health. New directions for tobacco control in Canada: a national strategy. Ottawa, 1999.

10. Kendall O, Lipskie T, MacEachern S. Canadian health surveys, 1950S1997. Chronic Dis Can 1997;18(2):70S90.

11. Health Canada (Stephens T, Morin M, editors.). Youth Smoking Survey, 1994: technical report. Ottawa, 1996.

12. Nielsen AC. Final report of findings: measurement of retailer compliance with respect to tobacco sales-to-minors legislation. Prepared for Health Canada, 1995 (also1996, 1997, 1998).

13. Health Canada. Smoking by-laws in Canada 1991. Ottawa, 1992.

14. Health Canada. Smoking by-laws in Canada 1995. Ottawa: Office of Tobacco Control, Health Protection Branch, Health Canada; 1995.

15. Goss Gilroy Inc. and Thomas Stephens & Associates. Study of smoking policies in various settings in Canada. Report prepared for the Health Promotion Directorate, Health Canada. Ottawa: Health Canada, 1995 Aug.

16. EKOS Research Associates Inc. An assessment of knowledge, attitudes, and practices concerning environmental tobacco smoke. Final report. Submitted to Health Canada, 1995 Mar.

17. Health Canada. School smoking prevention programs: a national survey. Ottawa, 1994.

18. Statistics Canada. Consumer prices and price indexes. Ottawa: Statistics Canada; Cat 62-010-C (quarterly).

19. Stephens T, Pederson LL, Koval JJ, Kim C. The relationship of cigarette prices and smoke-free bylaws to the prevalence of smoking in Canada. Am J Public Health 1997;87:1519S21.

20. Health Canada. Canadian Tobacco Use Monitoring Survey, wave 1. Summary of results. Ottawa: Health Canada, 2000 Jan.

21. Wakefield MA, Chaloupka FJ. Improving the measurement and use of tobacco control "inputs" [editorial]. Tobacco Control 1998;78:333S5. 

Roberta Ferrence, Ontario Tobacco Research Unit, University of Toronto; and Department of Public Health Sciences, University of Toronto; and Centre for Addiction and Mental Health, Toronto, Ontario

Thomas Stephens, Ontario Tobacco Research Unit, University of Toronto; and Department of Public Health Sciences, University of Toronto; and Thomas Stephens & Associates, Manotick, Ontario

Correspondence: Dr Roberta Ferrence, Ontario Tobacco Research Unit, 33 Russell Street, Toronto, Ontario  M5S 2S1; Fax: (416) 595-6068;
E-mail: roberta.ferrence@utoronto.ca

[Table of Contents [Next]