Editor,—Worksites represent a promising avenue for reaching smokers who might not otherwise participate in or be exposed to tobacco control activities.1 The need to increase the reach and impact of worksite smoking cessation programmes and smoking policy restrictions has been identified.2 To date, there have been few rigorous, Australian, workplace-based studies involving probability sampling to determine self-reported smoking rates among occupational groups. In particular, there has been no systematic examination of smoking behaviour in Australian healthcare staff.

Healthcare workers are perceived as having a high degree of credibility in health-related matters by the general public3 and are strategically placed to advance the anti-smoking message. Smokers employed in hospital settings set a poor example of health-promoting behaviours and have the potential to unintentionally affect the smoking behaviours of others through modelling.4 5 In addition, there is evidence to suggest that a barrier to health professionals’ use of brief interventions, such as providing opportunistic advice, is smoking behaviour among the health professionals themselves.6 Smoking by nurses and physicians may impede their ability to help patients quit smoking.4 7

In 1997, as part of the Central Sydney Area Health Service (CSAHS) Tobacco Control Plan,8 we randomly selected a sample of 2100 CSAHS employees (from a total of 9501 staff) to receive a single-page, self-administered questionnaire assessing smoking status, smokers’ preferences for cessation strategies, and attendant attitudes. A response rate of 80% was obtained (n = 1457).

Most health service employees (81%) were not smokers. Smokers (n = 276, 19%) were significantly younger than non-smokers, were significantly more likely to be non-professionals or employed in trades or as clerical staff, and were more likely to mainly speak English at home. Smoking rates were lowest among medical staff (7%) and managers (13%). The analysis also revealed that one in five nurses smoke (20%); however, this rate was not significantly different from that among professionals and para-professionals (18%) (χ²= 0.25, 1 df, p<0.62). Smoking was most common among clerical staff and tradespersons (21%) and non-professionals (24%). Smokers in this study reported most interest in written information or self-help booklets (n = 95), personal counselling (n = 80), and group programmes (n = 76) to assist them to quit. There was little interest in telephone support for smoking cessation (n = 19).

Consistent with previous studies,7 9 10 there were clear differences between smokers and non-smokers in their attitudes to smoking issues. Although most respondents (94%) agreed that: “Breathing in other people’s smoke is harmful”, smokers (84%) were significantly less likely than non-smokers (97%) to indicate that passive smoking can be detrimental to health (χ² = 151.63, 4 df, p<0.001). Fewer smokers (37%) than non-smokers (70%) agreed that: “Health care workers who smoke give the impression that smoking is not harmful” (χ² = 124.27, 4 df, p<0.001). Overall, there was clear majority support (83%) for the statement: “Smoking in hospitals and health care centres should be limited to areas not visible to the public”. As might be expected, smokers (67%) were less likely than non-smokers (87%) to agree with such a policy (χ² = 112.48, 4 df, p<0.001). It is encouraging to note, however, that approximately two-thirds of smokers agreed that the policy was a good idea.

Future health service workplace programmes in tobacco control should address deficits in knowledge about the health effects of passive smoking and should raise awareness among healthcare workers who smoke about their exemplary role. The findings suggest that a health service policy that eliminates visible smoking and environmental tobacco smoke would be well received.