Second-hand smoke exposure in outdoor hospitality venues: Smoking visibility and assessment of airborne markers

Environmental Research 165 (2018) 220–227

Contents lists available at ScienceDirect

Environmental Research journal homepage: www.elsevier.com/locate/envres

Second-hand smoke exposure in outdoor hospitality venues: Smoking visibility and assessment of airborne markers

T

⁎

Xisca Suredaa, , Usama Bilala,b, Esteve Fernándezc,d, Roberto Valientea,e, Francisco J. Escobara,e, Ana Navas-Acienf, Manuel Francoa,g a

Social and Cardiovascular Epidemiology Research Group, School of Medicine, University of Alcalá, Alcalá de Henares, Madrid, Spain Urban Health Collaborative, Drexel Dornsife School of Public Health, Philadelphia, PA, USA c Tobacco Control Unit, Cancer Control and Prevention Programme, Institut Català d′Oncologia-ICO, Hospitalet de Llobregat, Spain d Department of Clinical Sciences, School of Medicine, Campus de Bellvitge, Universitat de Barcelona, L′Hospitalet del Llobregat, Barcelona, Spain e Department of Geology, Geography and Environmental Sciences, University of Alcalá, Alcalá de Henares, Madrid, Spain f Department of Environmental Health Sciences, Mailman School of Public Health, Columbia University, New York, USA g Department of Epidemiology, Johns Hopkins Bloomberg School of Public Health, Baltimore, MD, USA b

A R T I C LE I N FO

A B S T R A C T

Keywords: Second-hand smoke PM2.5 Airborne nicotine Smoke-free policies Hospitality venues

Introduction: After the implementation of smoke-free policies in indoor hospitality venues (including bars, cafeterias, restaurants, and pubs), smokers may have been displaced to their outdoor areas. We aimed to study smoking visibility and second-hand smoke exposure in outdoor hospitality venues. Methods: We collected information on signs of tobacco consumption on entrances and terraces of hospitality venues in 2016 in the city of Madrid, Spain. We further measured airborne nicotine concentrations and particulate matter of less than 2.5 µm in diameter (PM2.5) in terraces with monitors by active sampling during 30 min. We calculated the medians and the interquartile ranges (IQR) of nicotine and PM2.5 concentrations, and fitted multivariate models to characterize their determinants. Results: We found 202 hospitality venues between May and September (summer), and 83 between October and December 2016 (fall) that were opened at the time of observation. We found signs of tobacco consumption on 78.2% of the outdoor main entrances and on 95.1% of outdoor terraces. We measured nicotine and PM2.5 concentrations in 92 outdoor terraces (out of the 123 terraces observed). Overall median nicotine concentration was 0.42 (IQR: 0.14–1.59) μg/m3, and overall PM2.5 concentration was 10.40 (IQR: 6.76–15.47) μg/m3 (statistically significantly higher than the background levels). Multivariable analyses showed that nicotine and PM2.5 concentrations increased when the terraces were completely closed, and when tobacco smell was noticed. Nicotine concentrations increased with the presence of cigarette butts, and when there were more than eight lit cigarettes at a time. Conclusions: Outdoor hospitality venues are areas where non-smokers, both employees and patrons, continue to be exposed to second-hand smoke. These spaces should be further studied and considered in future tobacco control interventions.

1. Introduction Second-hand smoke (SHS) was responsible for 603,000 deaths among children and adult non-smokers in 2004 worldwide and this number has increased over the years (Oberg et al., 2011; Word Health Organization (WHO), 2015). Long-term SHS exposure has been associated with many adverse health effects including low birth weight and increased risk of respiratory diseases in children, lung cancer and coronary heart disease (IARC Working Group on the Evaluation of

Carcinogenic Risks to Humans, 2004; U.S. Department of Health and Human Services, 2006). Short-term exposure to SHS has been linked to the irritation of the eyes and respiratory tract, (Junker et al., 2001) and there is also evidence suggesting that it can cause significant adverse effects on the human respiratory system, (Flouris and Koutedakis, 2011) and contribute to an increased risk of cardiovascular mortality (Pope et al., 2001). To protect people from the harms of SHS exposure the World Health Organization (WHO) encouraged countries to follow Article 8 of the

⁎ Correspondence to: Social and Cardiovascular Epidemiology Research Group, School of Medicine, University of Alcalá, Campus Universitario - Crta. de Madrid-Barcelona, Km. 33,600, Alcalá de Henares, Madrid 28871, Spain. E-mail address: [email protected] (X. Sureda).

https://doi.org/10.1016/j.envres.2018.04.024 Received 16 March 2018; Received in revised form 19 April 2018; Accepted 20 April 2018 0013-9351/ © 2018 Elsevier Inc. All rights reserved.

Environmental Research 165 (2018) 220–227

X. Sureda et al.

2.2. Variables and instruments

WHO Framework Convention on Tobacco Control (FCTC) to create complete smoke-free environments in all indoor workplaces, public places, and on public transport (Word Health Organization (WHO), 2003). By the year 2017, 20% of the world's population, were protected by comprehensive national smoke-free laws (Word Health Organization (WHO), 2017). In recent years, some countries have also extended smoke-free policies to some outdoor public places such as health centers, playgrounds, beaches, sports facilities, entrances to public buildings, dining patios, or public transportation stops (Kaufman et al., 2010; Global Smokefree Partnership, 2009; Repace, 2008). However, outdoor smoke-free policies are not as popular as indoor smoking bans, although in some outdoor places such as hospitality venues people would continue to be exposed to considerable levels of SHS (Fu et al., 2016; Licht et al., 2013; Sureda et al., 2013; Sureda et al., 2015). Outdoors hospitality venues (such as terraces and patios in bars and restaurants or their main entrances outdoors) are places frequented and used by smokers since the implementation of smoking ban inside those venues (Fu et al., 2016; Sureda et al., 2013; Sureda et al., 2015). Previous studies measuring SHS exposure in those areas have found levels above the values recommended not to be exceeded by the WHO Air Quality Guidelines (Sureda et al., 2013; Cameron et al., 2010; Edwards and Wilson, 2011; Klepeis et al., 2007; Lopez et al., 2012; St Helen et al., 2011; Sureda et al., 2012; Travers et al., 2007; Wilson et al., 2007; Wilson et al., 2011). In a study conducted in Barcelona, Spain, non-smokers claimed that most SHS exposure in all types of outdoor settings occurred in outdoor areas of bars and restaurants (Sureda et al., 2015). In the same study, more than half of the smokers reported smoking in those areas. These results suggest that the effectiveness of indoor smoke-free policies is reduced in hospitality venues by the occurrence of smoking in outdoor spaces when smoking is permitted. From January 2nd, 2011 smoking in all enclosed public places and workplaces including hospitality venues was banned in Spain (Sanidad, 2010). It was also the first time in Europe that smoking was also banned in some outdoor places including health care premises campuses, primary schools and high school courtyards, and children's playgrounds. The law also prohibited smoking in terraces in hospitality venues when they had a roof and more than two sidewalls. This study is part of the “Heart Healthy Hoods” (HHH) study which seeks to understand how social and physical urban environment relates to cardiovascular outcomes in European cities (https://hhhproject.eu/) (Bilal et al., 2016; Carreño et al., 2017). The HHH project includes in its objectives the characterization of tobacco urban environment (Franco et al., 2015). The objective of the present study is to describe smoking visibility and SHS exposure in outdoor hospitality venues in the city of Madrid, Spain.

2.2.1. Systematic social observation: tobacco questionnaire in outdoor hospitality venues We adapted a questionnaire that had already been used in previous studies (Fu et al., 2016; Sureda et al., 2012; Navas-Acien et al., 2016) aimed to characterize hospitality venues including outdoor main entrances and terraces. The questionnaire included information on: a. General characteristics of hospitality venue measurements: date of measurement, address and type of hospitality venue. We included bars or similar (including cafeterias, breweries or bodegas (Spanish traditional drinking venues where wine or beer are produced by the owners); restaurants (including take-away and fast food); and pubs/ cocktail bars. b. Presence of terraces and their physical characteristics: we indicated if the hospitality venue had terrace or not. If it did, we collected information on the number of tables, presence of roof (yes/no), and presence (yes/no) and number of sidewalls. We considered roofs as any permanent or temporary structure that impedes upward airflow. Walls were defined as any structure that impedes lateral airflow, regardless their full attachment to the roof. c. Signs of tobacco consumption at the entrance of the venue and on its terrace (if present): number of smokers, presence of ashtrays (yes/ no), presence of cigarette butts (yes/no), and tobacco smell (yes/ no). For the entrances, we also registered if there were tables or barrels (or similar) with signs of tobacco consumption (yes/no). Venues use those elements attached to the entrance to facilitate tobacco consumption outdoor when they do not have permission to have a terrace (Fig. 2). d. Information related to the airborne marker measurements: we indicated if we measured or not airborne markers at the terrace. If yes, we recorded the time of onset and completion of the measurement, number of lit cigarettes every 5 min, and the total number of smokers and lit cigarettes during the measurement. The tobacco questionnaire was integrated in an app called Open Data Kit (ODK) (https://opendatakit.org/use/collect/) that allowed us to collect the data using smartphones. This app allows data collection, including the possibility of taking pictures and geo-locating the data using the smartphone GPS. 2.2.2. Airborne markers a. Vapor-phase nicotine concentrations: We used an active sampling method connecting nicotine sampler's devices through a tub to a pump (flow 3.02 ml/min), as conducted in previous studies (Fu et al., 2016; Sureda et al., 2012). Sampler's devices contained a 37 mm in diameter filter treated with sodium bisulfate. The pump was calibrated every 3–4 measurements using a Defender 510-M Calibrator (Bios International Corp, USA). Nicotine was analyzed in the Laboratory of the Public Health Agency of Barcelona by gas chromatography/mass spectrometry. We estimated the timeweighted average nicotine concentration (μg/m3) by dividing the amount of nicotine extracted by the volume of sampled air multiplied by the total number of minutes the filter was exposed. The limit of quantification was 5 ng per filter, equivalent to 0.06 μg/m3 of nicotine for an exposure time of 30 min. Samples with values below the quantification limit were assigned half of this value (0.03 μg/m3). b. PM2.5 concentrations: We used a hand-held-operated monitor of particle size and mass concentration (TSI SidePak AM510 Personal Aerosol Monitor) as used in previous studies (Sureda et al., 2012; Sureda et al., 2014). The monitor was fitted with a 2.5-μm impactor to measure the concentration of particulate matter with a mass-

2. Methods 2.1. Study design and area of study This is a cross-sectional study conducted in the city of Madrid, Spain. Madrid is divided into 21 districts, which, in turn, are divided into 128 neighborhoods and 2412 census sections. Census sections are the smallest administrative area for the Spanish Census with approximately 1500 residents per census section. We conducted the present study in 42 census sections used in the HHH study to guarantee the representativeness of the demographic characteristics of the city (Fig. 1). Briefly, we used a multistage design to select the area of study. In the first stage, we selected two neighborhoods for each district of the municipality of Madrid, in total 42 neighborhoods. The selection was representative of socio-economic characteristics of Madrid including unemployment, precarious work, occupational class, educational level and immigration. This was a nonprobabilistic sample. In the second stage, we selected the median census section in each neighborhood in terms of population density, business density, educational level, immigration, and aging. 221

Environmental Research 165 (2018) 220–227

X. Sureda et al.

Fig. 1. Areas selected in the Heart Healthy Hoods study. Madrid, Spain. Madrid is divided into 21 districts, which, in turn, are divided into 128 neighborhoods and 2412 census sections. We selected 42 census sections scattered around the city and representative of their demographic characteristics.

The observer completed the tobacco questionnaire walking along all sides of the street located within the chosen census section. The route in each census section was previously defined using a map that the observer followed the day of the data collection. The observer registered all the hospitality venues within each selected census section (indicating if the venue was open or closed at the time the observer visited the premises). We measured vapor-phase nicotine concentrations and airborne PM2.5 in all the terraces where two or more smokers were present, and we also included some terraces with less than two smokers. Airborne markers were measured for a period of 30–45 min in a central point of the terrace. In each census section where we measured airborne markers in at least one terrace, we also measured an outdoor background of PM2.5 concentration. We missed the background level in two census section where we measured PM2.5 in at least a terrace due to technical problems during the fieldwork. The background PM2.5 concentrations were measured at the end of the venue sampling in a place located more than 10 m from any outdoor terrace and where smoking was not observed.

median aerodynamic diameter less than or equal to 2.5 µm. The sample flow rate through the TSI SidePak monitor was set at 1.7 l/ min to ensure proper operation of the attached 2.5-μm impactor. We used a K factor of 0.52 applied to all the measurements calculated for our specific instrument. The equipment was set to a one-second sampling interval and zero-calibrated prior to each use by the attachment of a HEPA filter according to the manufacturer's specifications. All data were registered by the TSI SidePak monitor and downloaded weekly into a personal computer for management and statistical analysis. PM2.5 concentrations are expressed in μg/m3.

2.3. Data collection During April 2016 one data collector was trained for the fieldwork. The data collector piloted the questionnaire and airborne markers measurements under the supervision of the lead researcher. Data collection took place between May and September (summer season) 2016. Between late October and December (fall season) 2016 the observer repeated the measurements in 11 of the 42 census sections measured in summer to compare possible variability in signs of tobacco consumption and SHS exposure (including airborne markers) between seasons. The selection of the repeated census section was made by convenience to ensure that each census had a minimum of three hospitality venues, and at least one of them with a terrace. The observer collected the data on weekdays, and mostly between 5 p.m. and 9 p.m. without notifying or warning the owners, employees, patrons, or pedestrians to avoid bias. No permission was needed since all the places under investigation were public premises, as approved by the Ethics Research Committee of the Madrid Health Care System.

2.4. Statistical analyses We computed the proportion of outdoor main entrances of hospitality venues with signs of tobacco consumption. We calculated the medians and interquartile ranges (IQR) of nicotine and PM2.5 concentrations measured in terraces of hospitality venues. We described medians and their corresponding IQR by selected potential explanatory variables of SHS levels and we used the KruskalWallis non-parametric test for independent samples to compare medians among groups. 222

Environmental Research 165 (2018) 220–227

X. Sureda et al.

Fig. 2. Examples of outdoor main entrances of hospitality venues, Madrid, Spain, 2016. *All photographs are taken by Víctor G. Carreño. A) Hospitality venues that do not have permission to have a terrace and the owners use their main entrances outdoors as spaces where people can smoke. These spaces are not regulated by smoke-free laws. B) Owners make it easier to smoke outside bars and restaurant using barrels, standing bar, and ashtrays. In the window there is a visible sign indicating the prohibition of smoking inside the venue. C) People smoking are sitting in the tables provided by the owners. This main entrance is a quasi-indoor setting covered with a roof and three walls. There is a sign also indicating the possibility of smoking in a heated terrace inside the venue. D) A man is smoking at the entrance of a bar standing in front of a barrel with ashtrays.

We obtained similar results during fall. There were 78.3% of the main entrances with at least one sign of tobacco consumption: 36.1% had ashtrays; 62.7% had cigarette butts; in 28.9%, tobacco smell was perceived; and, in 28.9%, smokers were observed. There were 42.2% of the main entrances that had tables and/or barrels (or similar) with at least one of the signs of tobacco consumption.

A linear regression analysis was used to explore the determinants of nicotine and PM2.5 concentrations in terraces. We used log-transformed nicotine and PM2.5 concentrations given their skewed distribution. Results are presented as exponentiated coefficients that can be interpreted as the Geometric Mean Ratio. We studied correlations between PM2.5 and nicotine concentrations with the Spearman rankcorrelation coefficient. For all analyses, we used SPSS v. 15.00.

3.2. Second-hand smoke exposure in terraces in hospitality venues 3. Results We observed signs of tobacco consumption in 95.6% of the terraces during summer, and in 93.5% during fall. We measured vapor-phase nicotine and PM2.5 concentrations in 92 terraces of hospitality venues (78 terraces in summer, and 18 in fall). We found 13 samples below the nicotine limit of quantification. The overall median (IQR) concentration was 0.42 (0.14–1.59) μg/m3 for nicotine, and 10.40 (6.76–15.47) μg/m3 for PM2.5 (Table 1). Median (IQR) nicotine concentration ranged from 0.03 (0.03–0.91) μg/m3 when no cigarette was lit to 3.83 (0.97–4.70) μg/m3 when more than eight cigarettes were lit during the measurement (p < 0.001). Median (IQR) PM2.5 concentration ranged from 11.96 (7.02–47.32) μg/m3 to 16.64 (6.11–39.78) μg/m3 (p = 0.061). We observed significant differences for both nicotine and PM2.5 concentrations according to the coverage of the terrace (p = 006, and p < 0.001, respectively). When the terrace had a roof and four sidewalls, median (IQR) SHS levels were 2.40 (0.64–13.36) μg/m3 for nicotine and 78.00 (11.70–158.34) μg/m3 for PM2.5. Median (IQR) SHS levels when the terraces were completely open (without roof or sidewalls) were 0.37 (0.15–1.59) μg/m3 for nicotine, and 11.96 (6.50–16.51) μg/m3 for PM2.5. Both nicotine and PM2.5 concentrations increased in outdoor spaces that were not compliant with the law (they had a roof and more than two sidewalls, and there was at least one sign of tobacco consumption), and during fall season (Table 1). We observed that during

We found 202 hospitality venues (including bars, cafeterias, restaurants, and pubs) that were opened within the 42 selected census sections for the whole city of Madrid, Spain between May and September (summer) 2016. Among those hospitality venues, 91 had terrace and we measured airborne markers in 74 of them. There were 83 hospitality venues that were opened within the 11 census sections measured between October and December (fall) 2016. We measured airborne markers in 18 out of the 32 terraces found during fall. Variables regarding signs of tobacco consumption at the main entrances outdoor were introduced in an updated version of the questionnaire, so it was measured in a subset of the venues in summer (174 out of the 202), and in all venues in fall (Fig. 3). 3.1. Smoking visibility in main entrances (outdoors) in hospitality venues We found signs of tobacco consumption on 78.2% of the entrances measured during summer: 38.5% had ashtrays; 66.1% had cigarette butts; in 24.1%, tobacco smell was perceived; and, in 24.7%, smokers were observed. Moreover, 42.5% of the main entrances outdoors had tables and/or barrels (or similar) with at least one of the signs of tobacco consumption mentioned before. Fig. 2 shows different examples of main entrances we found with signs of tobacco consumption. 223

Environmental Research 165 (2018) 220–227

X. Sureda et al.

Fig. 3. Flow diagram describing the study sample within the city of Madrid.

when the terrace was completely closed (model 2 in Table 2); PM2.5 significantly increased when the terrace was completely closed (model 2 in Table 3). The Spearman rank-correlation coefficient between nicotine and PM2.5 concentration was 0.47 (IC95%: 0.29, 0.61).

fall season almost 77.8% of the terraces were not compliant with the law. We found no differences in the nicotine and PM2.5 concentrations according to the presence of ashtrays or cigarette butts. Nicotine and PM2.5 concentrations increased when we perceived tobacco smell, but with significant differences only for nicotine levels. The median (IQR) PM2.5 concentration in the control outdoor (background measurement) was 7.80 (4.55–10.79) μg/m3. In multivariate analysis (see Tables 2 and 3) nicotine concentrations significantly increased when we perceived tobacco smell, with the presence of cigarette butts, when there were more than 8 lit cigarettes during the measurement, and when the terrace was completely closed (with a roof and four sidewalls) (model 1 in Table 2). PM2.5 concentration was significantly increased when we perceived tobacco smell, and when the terrace was completely closed (model 1 in Table 3). When the model only included the number of lit cigarettes and the number of roofs and/or walls, nicotine significantly increased when there were more than 8 lit cigarettes during the measurement, and

4. Discussion Our results indicate that outdoor main entrances and terraces of bars and restaurants are areas where non-smoking population, patrons and workers continue to be exposed to SHS. A previous study conducted in Spain on self-reported smoking and SHS exposure in outdoor hospitality venues found that the majority of the smokers interviewed reported to smoke in bars and restaurants outdoors (Sureda et al., 2015). In that study, 33.5% of the non-smokers claimed to be exposed to SHS exposure in those settings. In our study, we found signs of tobacco consumption in 78.2% of the main entrances

Table 1 Medians and interquartile ranges (IQR) of vapor-phase nicotine and PM2.5 concentrations in 92 terraces of hospitality venues located in Madrid (2016). Nicotine

Overall Tobacco smell No Yes Presence of ashtrays No Yes Presence of butts No Yes No. of lit cigarettes 0 1–4 5–8 >8 No. roofs and/or walls 0 1 2–4 5 Compliance with law Yes No Season Summer Fall Background PM2.5 level a

n

Median (IQR) (µg/m3)

92

0.42 (0.14–1.59)

PM2.5 p-valuea

Median (IQR) (µg/m3) 10.40 (6.76–15.47)

0.016 10 81

0.07 (0.03–0.40) 0.52 (0.17–1.70)

29 63

0.42 (0.15–3.43) 0.42 (0.14–1.57)

19 73

0.26 (0.06–1.58) 0.46 (0.17–1.62)

5 32 33 10

0.03 0.17 0.69 3.83

(0.03–0.91) (0.07–0.63) (0.27–1.85) (0.97–4.70)

36 29 14 13

0.37 0.32 0.14 2.40

(0.15–1.59) (0.15–1.05) (0.05–1.38) (0.64–13.36)

71 21

0.32 (0.14–1.30) 1.20 (0.19–7.63)

74 18 36

0.32 (0.14–1.35) 1.16 (0.20–7.41) –

p-valuea

0.538 9.36 (6.63–13.00) 10.40 (6.50–15.34)

0.733

0.204 11.96 (7.28–17.42) 10.40 (6.24–14.04)

0.211

0.589 11.96 (5.20–36.92) 10.40 (6.76–14.56)

< 0.001

0.061 11.96 (7.02–47.32) 8.06 (5.33–11.96) 11.96 (8.84–20.02) 16.64 (6.11–39.78)

0.006

< 0.001 11.96 (6.50–16.51) 7.28 (5.46–9.88) 10.92 (7.54–12.74) 78.00 (11.70–158.34)

0.054

0.001 9.36 (5.72–13.52) 15.08 (8.84–108.68) 0.012

0.040

–

Kruskal-Wallis test for medians. 224

9.88 (6.24–14.04) 13.78 (9.10–118.69) 7.80 (4.55–10.79)

–

Environmental Research 165 (2018) 220–227

X. Sureda et al.

Table 2 Multivariable analysis of log transformed nicotine concentration in 92 terraces of hospitality venues located in Madrid (2016). Model 1 for nicotine concentrations* n

Geometric mean ratio (95% CI)

Model 2 for nicotine concentrations* R2

p-value

Geometric mean ratio (95% CI)

p-value

0.344 Tobacco smell No Yes Presence of ashtrays No Yes Presence of butts No Yes No. of lit cigarettes 0 1–4 5–8 >8 No. roofs and/or walls 0 1 2–4 5 Season Summer Fall

0.469

10 81

1 2.36 (1.47–3.79)

0.001

29 63

1 1.14 (0.84–1.55)

0.386

19 73

1 1.42 (1.01–1.99)

0.042

5 32 33 10

1 0.83 (0.57–1.22) 1.29 (0.87–1.89) 2.50 (1.491–4.19)

0.342 0.200 0.001

1 0.74 (0.50–1.09) 1.09 (0.72–1.63) 2.24 (1.34–3.74)

0.128 0.684 0.002

36 29 14 13

1 1.15 (0.82–1.59) 0.97 (0.64–1.46) 2.41 (1.59–3.65)

0.413 0.868 < 0.001

1 1.05 (0.76–1.45) 1.00 (0.66–1.52) 2.03 (1.19–3.49)

0.748 0.993 0.010

1 1.50 (0.94–2.39)

0.091

74 18

R2

* Estimates from multiple linear regressions (log transformed nicotine concentrations). Results are expressed as Geometric Mean Ratios (exponentiated coefficients). For example, in Model 2 the concentrations of nicotine are 2.36 times higher in outdoor venues with tobacco smell as compared to those without tobacco smell.

measured SHS exposure in main entrances outdoors in hospitality venues using environmental markers (Fu et al., 2016; Lopez et al., 2012; Brennan et al., 2010). Their results indicated that smoking in main entrances outdoors not only expose people in those spaces but that tobacco smoke drifts from the outside entrances to the indoor areas. When considering only outdoor terraces, 95.1% showed signs of tobacco consumption. We also measured airborne markers in these settings, and we obtained an overall median nicotine concentration of

in hospitality venues. Those places have become popular for smoking, especially when the venue has no terrace. We observed that in 26.1% of the main entrances there were smokers at the time of the measurement. Moreover, the owners put ashtrays and tables and/or barrels that attract people who want to smoke, and may favor the normalization of smoking. The entrance is a particularly important area for both exposure and visibility as both customers and workers need to go through the entrance to go inside and outside the venue. Other studies have

Table 3 Multivariate analysis of log transformed PM2.5 concentration in 92 terraces of hospitality venues located in Madrid (2016). Model 1 for PM2.5 concentrations* n

Geometric mean ratio (95% CI)

Model 2 for PM2.5 concentrations* p-value

R2

Geometric mean ratio (95% CI)

p-value

0.453 Tobacco smell No Yes Presence of ashtrays No Yes Presence of butts No Yes No. of lit cigarettes 0 1–4 5–8 >8 No. roofs and/or walls 0 1 2–4 5 Season Summer Fall

0.474

10 81

1 1.37 (1.07–1.77)

0.015

29 63

1 0.97 (0.82–1.14)

0.723

19 73

1 0.97 (0.81–1.16)

0.703

5 32 33 10

1 0.94 (0.78–1.13) 1.05 (0.86–1.27) 1.19 (0.92–1.54)

0.497 0.649 0.183

1 0.87 (0.71–1.08) 0.96 (0.77–1.20) 1.12 (0.85–1.47)

0.203 0.730 0.413

36 29 14 13

1 0.86 (0.73–1.01) 1.01 (0.82–1.24) 1.95 (1.59–2.40)

0.065 0.919 < 0.001

1 0.83 (0.70–0.99) 1.02 (0.81–1.27) 1.72 (1.29–2.29)

0.038 0.888 < 0.001

1 1.20 (0.93–1.54)

0.156

74 18

R2

* Estimates from multiple linear regressions (log transformed PM2.5 concentrations). Results are expressed as Geometric Mean Ratios (exponentiated coefficients). For example, in Model 2 the concentrations of PM2.5 are 1.37 times higher in outdoor venues with tobacco smell as compared to those without tobacco smell. 225

Environmental Research 165 (2018) 220–227

X. Sureda et al.

0.42 μg/m3, and an overall median PM2.5 concentration of 10.40 μg/ m3 (statistically significantly higher than the background levels). A systematic review including 18 papers measuring SHS exposure in outdoor areas found 12 studies that measured SHS exposure in hospitality venues (Sureda et al., 2013). Most of them measured PM2.5 with concentrations that ranged from 8.32 μg/m3 (Stafford et al., 2010) to 124 μg/m3 (Wilson et al., 2007). Only one study included in the review used airborne nicotine with a median concentration of 1.56 μg/m3 obtained in outdoor hospitality venues (Lopez et al., 2012). Recently, a study conducted in Barcelona, Spain, also measured nicotine concentrations in outdoor terraces in bars and restaurant, and obtained an overall median of 0.54 μg/m3, similar to our results. Nicotine concentrations increased with the number of lit cigarettes during the measurements and when there was tobacco smell. Both nicotine and PM2.5 concentrations considerably increased when the terraces were completely closed (with a roof and four sidewalls). Other studies also showed that high smoker density and/or number of lit cigarettes, and highly enclosed outdoor areas, generate higher outdoor SHS concentrations (Fu et al., 2016; Licht et al., 2013; Sureda et al., 2013). There is no safe level of SHS exposure. WHO Air Quality Guidelines determines an annual guide value for outdoor long-term exposures to PM2.5 of 10 μg/m3 (World Health Organization (WHO), 2000, 2005). This value represents the lower end of the range over which significant effects on survival have been observed. However, even PM2.5 concentrations of 3–5 μg/m3 have been associated with adverse health effects. In our study most of PM2.5 concentrations were higher than 10 μg/m3, and PM2.5 concentrations reach even values of 78 μg/m3 when the terraces were completely closed. This value can be comparable or even higher than those values obtained in other studies for indoor places (Sureda et al., 2013). These results suggest that part of the population, especially hospitality workers, would be exposed to high levels of SHS under certain conditions, especially in those terraces completely or almost closed and when many smokers are present. In Spain, Law 42/2010 prohibited smoking in those terraces that had a roof and more than two sidewalls (Sanidad, 2010). We found that 22.8% of the outdoor terraces where we measured tobacco markers were not compliant with the law (77.8% during fall season, and 9.5% in summer). Both nicotine and PM2.5 concentrations were higher when the terraces were not compliant with the law, and during fall season, when more terraces did not enforce the law. A previous study conducted in Barcelona after the implementation of the Law 42/2010 observed people smoking in 34.2% of the outdoor terraces that were enclosed and where smoking should not be allowed according to the legislation (Fu et al., 2016). In another study conducted in Spain nationwide this percentage was 87.6%, much higher than the results obtained in our study and the study conducted in Barcelona (Organización de Consumidores y Usuarios, 2015). This difference could be due to the period at which observers visited the premises (there are more enclosed terraces during cold seasons that do not comply with the regulation) or that there were differences by regions within the country. Regardless of the study, all the results indicated a lack of compliance with the law that prohibits smoking when there is a roof and more than two sidewalls. A possible explanation is a problem in the operational definition in the Law 42/2010 of what is considered an outdoor space. There are different definitions of what is considered a roof or sidewalls, including different materials or the level of enclosure. In our study, we considered that the law was not enforced when there was any sign of tobacco consumption in a terrace that had four coverages (one of them including the structure forming the upper covering), and that were complete, considering the stricter possibility of non-compliance. Another explanation to the low enforcement of the law in outdoor terraces can be the lack of inspections in these locations (Perez-Rios and Galan, 2017). For instance, in a survey of hospitality venue owners from Turkey, participants who had received a fine following an inspection were more likely to enforce the law although at the same time those

owners were also less likely to have a positive opinion of the smoke-free law (Aherrera et al., 2016). The wording and definition of what is an illegal terrace in Law 42/2010 should be improved to facilitate their interpretation, as well as reinforcing the inspections. Our results suggest that future policy interventions should consider outdoor hospitality venues to become completely smoke-free, and to ensure that these regulations are enforced. A systematic review including studies on the public support for smoke-free outdoor regulations in USA and Canada found that the support for prohibiting smoking in outdoor hospitality venues ranged from 41% to 82% (Thomson et al., 2016). Moreover, they observed an increase in support for outdoor smoke-free regulations over time. Another study conducted in Australia reported 69% support for smoke-free outdoor restaurant patios (Paul et al., 2007). The public support for outdoor hospitality venues suggests that it would be feasible to extend smoking bans to these settings. This study presents some limitations. All data were collected on weekdays, and mostly between 5 p.m. and 9 p.m. These times were chosen to ensure that most hospitality venues (bars, cafeterias, and restaurants) would be open during the data collection. However, some pubs and nightclubs were closed at the time of the observation. Although these settings do not usually have an outdoor terrace, smokers use their main entrances to smoke. Perhaps data collection at other times may have captured more people smoking in outdoor hospitality venues. Moreover, it is possible that during nights and weekends we have observed less compliance with the regulation due to less fear of inspections. Future studies should consider including measurements during weekends and at night to get a real approach of the exposure to SHS and compliance with the legislation in outdoor hospitality venues. We had results from summer and fall finding that there were no differences in signs of tobacco consumption between these two seasons in outdoor main entrances. However, outdoor terraces were more compliant with the smoke-free law in the summer season. We did not conduct measures in spring and winter because weather and temperature in Spain in those seasons do not vary much from summer and fall, respectively. In this study, we recorded the number of people smoking and the number of lit cigarettes during the measurements. However, we did not distinguish between different types of tobacco product, neither include information on electronic cigarettes consumption. A recent study conducted in Spain observed an increase in roll-your-own cigarette users, especially among young people (Sureda et al., 2017). It would have been interesting to include that information and future research may take it into account to have a realistic view of smoking behavior. In this study, we used both airborne nicotine and PM2.5 to measure SHS in outdoor terraces in hospitality venues. Although in other studies, similar to our findings, the correlation between PM2.5 and nicotine in outdoor settings was moderate, (Sureda et al., 2012; Fu et al., 2013) we decided to measure both for different reasons. PM2.5 is the most widely used airborne marker to measure SHS exposure, and allowed us to compare our results with other similar studies (Sureda et al., 2013). Moreover, the WHO Air Quality Guidelines have recommended values for PM2.5, so this marker is useful as an indicator of the potential harmful impact on health. However PM2.5 are not specific to tobacco, and they are susceptible to atmospheric variations including wind conditions and humidity, what can make the interpretation of the findings more complex (Sureda et al., 2013). For that reason, we combined PM2.5 measures with airborne nicotine, which is specific to SHS and very sensitive at low concentrations. Our study included all hospitality venues distributed within 42 census sections scattered around the city of Madrid and representative in terms of sociodemographic and socioeconomic characteristics. One trained observer did the fieldwork avoiding inter-observer variability. The observational nature of our study allow us to reflect smoking behaviors and SHS exposure in outdoor hospitality venues under normal real-life conditions.

226

Environmental Research 165 (2018) 220–227

X. Sureda et al.

5. Conclusions

IARC Working Group on the Evaluation of Carcinogenic Risks to Humans, 2004. Tobacco smoke and involuntary smoking. IARC Monogr. Eval. Carcinog. Risks Hum. 83, 1–1438. Junker, M.H., Danuser, B., Monn, C., Koller, T., 2001. Acute sensory responses of nonsmokers at very low environmental tobacco smoke concentrations in controlled laboratory settings. Environ. Health Perspect. 109 (10), 1045–1052. Kaufman, P., Griffin, K., Cohen, J., Perkins, N., Ferrence, R., 2010. Smoking in urban outdoor public places: behaviour, experiences, and implications for public health. Health Place. 16 (5), 961–968. Klepeis, N., Ott, W., Switzer, P., 2007. Real-Time measurement of outdoor tobacco smoke particles. J. Air Waste Manag. Assoc. 57, 522–534. Licht, A.S., Hyland, A., Travers, M.J., Chapman, S., 2013. Secondhand smoke exposure levels in outdoor hospitality venues: a qualitative and quantitative review of the research literature. Tob. Control 22 (3), 172–179. Lopez, M.J., Fernandez, E., Gorini, G., et al., 2012. Exposure to secondhand smoke in terraces and other outdoor areas of hospitality venues in eight European countries. PLoS One 7 (8), e42130. Navas-Acien, A., Carkoglu, A., Ergor, G., et al., 2016. Compliance with smoke-free legislation within public buildings: a cross-sectional study in Turkey. Bull. World Health Organ. 94 (2), 92–102. Oberg, M., Jaakkola, M.S., Woodward, A., Peruga, A., Pruss-Ustun, A., 2011. Worldwide burden of disease from exposure to second-hand smoke: a retrospective analysis of data from 192 countries. Lancet 377 (9760), 139–146. Organización de Consumidores y Usuarios, 2015. Espacios sin humo. OCU-Salud 121, 19–22. Paul C., Paras L., Walsh R. et al. 2007. Tracking NSW Community Attitudes and Practices in Relation to Tobacco: A Biennial Telephone Survey. Centre for Health Research and Psycho-oncology. Perez-Rios M., Galan I., 2017. (coordinadores). Evaluación de las políticas de control del tabaquismo en España (Leyes 28/2005 y 42/2010). Revisión de la evidencia. Barcelona. Pope 3rd, C.A., Eatough, D.J., Gold, D.R., et al., 2001. Acute exposure to environmental tobacco smoke and heart rate variability. Environ. Health Perspect. 109 (7), 711–716. Repace, J.L., 2008. Benefits of smoke-free regulations in outdoor settings: beaches, golf courses, parks, patios, and in motor vehicles. William Mitchell Law Rev. 34 (4), 1621–1638. Sanidad, Ministeriode, 2010. Política Social e Igualdad. Ley 42/2010, de 30 de diciembre, por la que se modifica la Ley 28/2005, de 2006 de diciembre, de medidas sanitarias frente al tabaquismo y reguladora de la venta, el suministro, el consumo y la publicidad de los productos del tabaco. Madrid. St Helen, G., Hall, D.B., Kudon, L.H., et al., 2011. Particulate matter (PM2.5) and carbon monoxide from secondhand smoke outside bars and restaurants in downtown Athens, Georgia. J. Environ. Health 74 (3), 8–17. Stafford, J., Daube, M., Franklin, P., 2010. Second hand smoke in alfresco areas. Health Promot. J. Austr. 21 (2), 99–105. Sureda, X., Martinez-Sanchez, J.M., Lopez, M.J., et al., 2012. Secondhand smoke levels in public building main entrances: outdoor and indoor PM2.5 assessment. Tob. Control 21 (6), 543–548. Sureda, X., Fernandez, E., Lopez, M.J., Nebot, M., 2013. Secondhand tobacco smoke exposure in open and semi-open settings: a systematic review. Environ. Health Perspect. 121 (7), 766–773. Sureda, X., Ballbe, M., Martinez, C., et al., 2014. Impact of tobacco control policies in hospitals: evaluation of a national smoke-free campus ban in Spain. Prev. Med. Rep. 1, 56–61. Sureda, X., Fernandez, E., Martinez-Sanchez, J.M., et al., 2015. Secondhand smoke in outdoor settings: smokers' consumption, non-smokers' perceptions, and attitudes towards smoke-free legislation in Spain. BMJ Open 5 (4), e007554. Sureda, X., Fu, M., Martinez-Sanchez, J.M., et al., 2017. Manufactured and roll-your-own cigarettes: a changing pattern of smoking in Barcelona, Spain. Environ. Res. 155, 167–174. Thomson, G., Wilson, N., Collins, D., Edwards, R., 2016. Attitudes to smoke-free outdoor regulations in the USA and Canada: a review of 89 surveys. Tob. Control 25 (5), 506–516. Travers M., Higbee C., Hyland A., 2007. Vancouver Island Outdoor Tobacco Smoke Air Monitoring Study 2007, pp. 1–20. 〈http://www.tobaccofreeair.org〉 (Accessed 3 February 2010). U.S. Department of Health and Human Services, 2006. The health consequences of involuntary exposure to tobacco smoke: a report of the Surgeon General. U.S. Department of Health and Human Services, Centers for Disease Control and Prevention, Coordinating Center for Health Promotion, National Center for Chronic Disease Prevention and Health Promotion, Office on Smoking and Health, Atlanta, GA. Wilson, N., Edwards, R., Maher, A., Nathe, J., Jalali, R., 2007. National smokefree law in New Zealand improves air quality inside bars, pubs and restaurants. BMC Public Health 7, 85. Wilson, N., Edwards, R., Parry, R., 2011. A persisting secondhand smoke hazard in urban public places: results from fine particulate (PM2.5) air sampling. N. Z. Med. J. 124 (1330), 34–47. Word Health Organization (WHO), 2003. WHO Framework Convention on Tobacco Control (WHO FCTC). World Health Organization, Geneve. Word Health Organization (WHO), 2015. WHO Report on the Global Tobacco Epidemic 2015: Raising Taxes on Tobacco. World Health Organization, Geneve. Word Health Organization (WHO), 2017. WHO Report on the Global tobacco Epidemic 2017: Monitoring Tobacco Use and Prevention Policies. World Health Organization, Geneve. World Health Organization (WHO), 2000. Air Quality Guidelines for Europe, second edition. Copenhagen. World Health Organization (WHO), 2005. Air Quality Guidelines. Denmark.

Our results indicated that outdoor hospitality venues (including their terraces and main entrances) are places were non-smokers, both workers and patrons, are still exposed to SHS exposure and where smoking behavior is highly visible. Moreover, we observed signs of tobacco consumption in terraces that were enclosed and where it is actually prohibited to smoke according to the law. It is thus necessary to enforce the existing law regarding what constitutes an outdoor terrace. Additional legislation is also needed to reinforce and extend tobacco control measures in outdoor hospitality venues to become completely smoke-free. Acknowledgements We would like to thank the Public Health Agency of Barcelona for her support in the nicotine analysis. We would also want to thank Rocio Santuy her participation in the data analysis. XS, MF, AN, and EF conceive the idea. RV conducted the fieldwork supervised by XS. XS and UB prepared the databases, and analyzed the data. XS drafted the manuscript. All authors contributed substantially to the interpretation of the data and manuscript review, and approved its final version. XS and MF are the guarantors. Declaration of interest None. Funding This work was supported by the Instituto de Salud Carlos III, Subdirección General de Evaluación y Fomento de la Investigación, Government of Spain (PI15/02146). The Heart Healthy Hoods project was funded by the European Research Council under the European Union's Seventh Framework Programme (FP7/2007-2013/ERC Starting Grant Heart Healthy Hoods Agreement no. 623 336893). XS and EF are also funded by the Ministry of Universities and Research, Government of Catalonia (2014SGR999 and 2017SGR319). The funding sources have not any involvement in the study design; in the collection, analysis, or interpretation of data; in the writing of this work; or in the decision to submit the manuscript for publication. References Aherrera, A., Carkoglu, A., Hayran, M., et al., 2016. Factors that influence attitude and enforcement of the smoke-free law in Turkey: a survey of hospitality venue owners and employees. Tob. Control 26 (5), 540–547. Bilal, U., Diez, J., Alfayate, S., et al., 2016. Population cardiovascular health and urban environments: the Heart Healthy Hoods exploratory study in Madrid, Spain. BMC Med. Res. Methodol. 16, 104. Brennan, E., Cameron, M., Warne, C., et al., 2010. Secondhand smoke drift: examining the influence of indoor smoking bans on indoor and outdoor air quality at pubs and bars. Nicotine Tob. Res. 12 (3), 271–277. Cameron, M., Brennan, E., Durkin, S., et al., 2010. Secondhand smoke exposure (PM2.5) in outdoor dining areas and its correlates. Tob. Control 19 (1), 19–23. Carreño, V., Franco, M., Gullón, P., 2017. Studying city life, improving population health. Int. J. Epidemiol. 46 (1), 14–21 (5). Edwards, R., Wilson, N., 2011. Smoking outdoors at pubs and bars: is it a problem? An air quality study. N. Z. Med. J. 124 (1347), 27–37. Flouris, A.D., Koutedakis, Y., 2011. Immediate and short-term consequences of secondhand smoke exposure on the respiratory system. Curr. Opin. Pulm. Med. 17 (2), 110–115. Franco, M., Bilal, U., Diez-Roux, A.V., 2015. Preventing non-communicable diseases through structural changes in urban environments. J. Epidemiol. Commun. Health 69 (6), 509–511. Fu, M., Martinez-Sanchez, J.M., Galan, I., et al., 2013. Variability in the correlation between nicotine and PM2.5 as airborne markers of second-hand smoke exposure. Environ. Res. 127, 49–55. Fu, M., Fernandez, E., Martinez-Sanchez, J.M., et al., 2016. Second-hand smoke exposure in indoor and outdoor areas of cafes and restaurants: need for extending smoking regulation outdoors? Environ. Res. 148, 421–428. Global Smokefree Partnership, 2009. The Trend toward Smokefree Outdoor Areas. Global Smokefree Partnership, Washington, DC.

227