Study design

A multi-center, transversal, descriptive study was conducted that included PHC medical and nonmedical personnel from the Spanish National Healthcare System. The study was conducted between June 2016 and March 2017.

Sample size

The sample (n=3994 participants) was made up of 1842 physicians (general practitioners and paediatricians), 1076 nurses (nurses and midwives), and 1076 non-medical personnel. A smoking prevalence of 20% for the first group, and 30% for the other two groups was assumed. A 20% dropout rate was taken into consideration for calculations.

Cluster sampling was carried out in which, initially, by simple random sampling, 5% of the health centers of each Autonomous Community were selected, which constituted the primary sampling units. Subsequently, as secondary sampling units, all members of the primary care teams of the centers selected in the first phase were selected.

Study survey

A questionnaire designed by the members of the Approach to Smoking Group (GAT) from the Spanish Family and Community Medical Society was validated during the first phase in a pilot study. The final version was anonymous and online. The voluntary collaboration of individuals in charge of the selected PHC was requested by email, and in order to improve participation they were reminded by telephone.

Variables

Ethical considerations

Prior to the study, participants were informed about the aim of the research and that the questionnaire was voluntary and anonymous. The study was exempt from the Institutional Review Board. No individual identifiers were used at any analytical step.

Data analysis

Descriptive statistics were used to summarize overall information. A high dropout rate and greater response from participants who were more aware of smoking issues were assumed. Categorical variables were expressed as percentages and continuous ones as means and standard deviation (±SD). Comparative analysis was performed with the χ² test for the former group and Student’s t-test and ANOVA for the latter, as applicable. Statistical significance was established by p<0.05. All analyses were performed with the SPSS software program.

Opinion about tobacco

Whilst no significant differences in opinion about tobacco were observed between men and women, location of the PHC (urban/rural), and professional status, there was, however, a difference between those who had been employed for more or fewer than 10 years. The former considered tobacco a chronic illness (p=0.003) that increased the risk of ischemic cardiopathy (p=0.036). Moreover, they believed that healthcare personnel represented a role model, their advice was effective in quitting tobacco, and that electronic cigarettes were harmful to one’s health (p=0.000 for all points). Regarding PHC medical personnel and their opinion about anti-smoking legislation changes, there were significant differences with respect to the efficacy of counselling (p<0.001), and whether or not electronic cigarettes were harmful (p=0.002), again in favour of those employed for more than 10 years (Table 2).

Opinion about anti-smoking legislation

Concerning the current legislation, 92% of the participants agreed with it. Nevertheless, 25% were of the view that it did not encourage individuals to quit. Almost 100% considered that healthcare centers and closed public areas should be smoke-free, although almost 1 in 5 thought that electronic cigarettes need not be prohibited in closed areas (1 in 4 in the case of non-medical personnel).

We did not observe differences in opinion about anti-smoking legislation, whether healthcare centers and closed public spaces should be smoke-free, and prohibition of electronic cigarettes in closed public areas with respect to gender, PHC location, professional status, and years of PHC employment (Table 3).

Comparison of consumption data in Spain

Research carried out prior to the first legislation of 2006 shows that healthcare personnel smoked far more than we found in our present study. In 1997, Gordo et al.²³ reported that 38.1% were smokers although data were only obtained from one region (Guadalajara). In 2002, Casas et al.²² found 26.5% in Barcelona, however, the study included other medical specialists and differences in consumption were found among administrative staff (35.9%), nurses (31.2%), and physicians (15.8%). A follow-up study by Fernández Ruiz and Sanchez²¹ from 1998 to 2001 of 1235 female medical personnel representing 31.65% (1998) and 39.1% (2001) of the personnel in their PHC, reported a smoking prevalence of 43.1% in 1998 and 43% in 2001. In both years, tobacco consumption was more frequent among the nurses (47.6% and 47%, respectively) than the physicians (34.7% and 37%, respectively), and among those employed in hospital care (46.6% and 46.7%, respectively) compared to PHC (35.3% and 37.3%, respectively). The 2005 study by Cerrada et al.²⁰ with healthcare personnel in Madrid observed a 21.6% prevalence of tobacco consumption.

Following the initial 2006 legislation, Tenas et al.¹⁹, in 2008, reported that 28.7% of PHC personnel in the region of Murcia declared themselves smokers.

In 2015, Jiménez-Ruiz et al.¹⁶ published consumption data from a sample of 612 healthcare personnel (155 PHC physicians). They reported a prevalence of 11.7% with differences between physicians (8.99%) and nurses (11.2%).

In conclusion, our findings show a reduction in the prevalence of smokers among PHC medical personnel. There was a 13% decrease compared to 2006 and even greater when compared with 1998.

Comparison with other countries

Studies by Ravara et al.^24,25 with 605 physicians (196 family doctors) in northern Mediterranean countries and Portugal observed that 20.4% of the smokers were family doctors. In Italy, Nobile et al.²⁶ reported that there were 13.4% smokers from a sample of 722 PHC physicians. Sonmez et al.²⁷, in Turkey, with a sample of 1182 physicians and 1063 nurses, found a smoking prevalence of 34.4% and 30.7%, respectively. In Croatia, Juranic et al.²⁸ reported that 35.1% of healthcare personnel were habitual smokers whilst Azuri and Nashef²⁹, in Israel, with a sample of 302 physicians, observed a prevalence of 13.5%. Stamatopoulou et al.³⁰, in a study undertaken in 40 health centers in rural mainland and island Greece, reported that 32% of the nurse respondents were smokers. Data from a study carried out by Huddlestone et al.³¹ in the UK with a sample of 171 PHC personnel revealed that 11% consumed tobacco.

A possible explanation for the differences found between these countries may be a different application of the laws, if we take as reference the Smoke Free Partnership and the data of The Tobacco Control Scale 2016 (Table 4). In conclusion, our data show a low level of consumption, compared to some of the studies in the Mediterranean area, which have been carried out in different years but all in primary healthcare and similar to the study in England.

Comparison with other healthcare personnel in Spain

In a four-year follow-up (2001, 2004, 2008, and 2011) carried out with hospital employees in Spain, Reyes et al.¹⁸ observed an increase in smoking prevalence from 30.00% in 2001 to 36.21% in 2008 followed by a decrease to 29.42% in 2011. In the group of physicians, it decreased progressively from 25.97% in 2001 to 8.88% in 2011 (p=0.007) whilst in that of nurses it went from 35.15% in 2001 to 25.61% in 2011 (p=0.007). With respect to administrative staff, the highest figure for this group was 43.93% in 2008, decreasing to 35.71% in 2011. Perez-Rios et al.⁹ also analyzed data before and after the 2010 legislation and found a non-significant reduction of 23.4% in 2006 and 20.7% in 2011. Our data partially coincide with the reduction found in other health professionals in Spain. A possible explanation can be that primary healthcare professionals attend to a higher number of patients with this problem and that several initiatives have been developed in primary healthcare to reduce tobacco consumption such as the ‘smoke-free week’ in primary care.

Comparison with other data from the general population

According to data from the 2017 Eurobarometer³⁴, 24% of the population in the European Union (EU) consider themselves daily smokers. A figure that rises in some southern European countries (36% in Bulgaria, 35% in Greece, and 33% in Croatia) and France (33%). In contrast, most of the countries in northern Europe, particularly Sweden (5%) but also the UK, Holland and Denmark (all 16%), Belgium (17%), and Finland (28%) report a lower prevalence. Whilst at 26% Spain is little higher than the EU mean²⁹, our findings show that PHC personnel (15.5%) are more than 10% below the European value. Moreover, if we take into consideration prevalence for physicians (11.8%) and nurses (12.8%), the difference is even more marked.

According to the same 2017 survey, electronic cigarette use stands at 2% throughout the EU³¹ and 1% in Spain. We found a slightly lower value of 0.8%.

Comments with respect to opinions about tobacco and anti-smoking legislation

In 1997, Alonso et al.²³ reported that 91.3% of nonsmoking healthcare personnel, compared to 48.2% of the smokers, were in favor of a normative that would prohibit smoking in the PHC. In 2002, Casas et al.²² found that 68.1% of the healthcare personnel considered that the non-smoking signs in the centers were inadequate, 72.8% did not respect the legislation, 77.2% thought there should be reserved areas for smokers, and 61.2% objected to other people smoking in the center. Fernández Ruiz and Sanchez²¹ observed similar opinions regarding legislation. Such findings contrast with those of our study, where 99.6% thought there should be smoke-free areas and 92.2% agreed with the normative, although there were some differences between non-smokers (94.2%) and smokers (79.9%). We are thus led to conclude that the 2010 legislation has had a very positive effect on the opinions and behavior of PHC healthcare personnel towards tobacco consumption and associated legislation. It has had an influence on their acceptance of the normative and resulted in a decrease in tobacco consumption.

Limitations and strengths

Our study has some limitations. Due to the possible self-selection of the participants and the questionnaire that was based on self-reporting, bias may be implied. There could have been an underestimation of the real rate of smoking prevalence and misrepresentation of attitudes towards smoking and legislation. We are aware that such questionnaires are open to respondees’ prejudices, particularly regarding a theme such as smoking behavior.

Participation was voluntary and some smokers may have avoided answering the questionnaire or changed their response. Whilst our results are not completely representative of Spanish healthcare personnel, they do reflect those who habitually work in PHC³⁵. Moreover, our sample is the largest to date and has included the whole of Spain. Some sociodemographic data that could have been of interest, such as marital status and socioeconomic level, have, however, not been included.