Elsevier

Environmental Research

Volume 164, July 2018, Pages 340-345
Environmental Research

Validity of self-reported indicators to assess secondhand smoke exposure in the home

https://doi.org/10.1016/j.envres.2018.03.014Get rights and content

Highlights

  • Self-reported indicators showed moderate-strong correlations, low sensitivity and high specificity.

  • Low sensitivities suggest that SHS exposure could be underreported due to other SHS sources.

  • The self-reported exposure indicators studied are good to assess SHS exposure at home.

  • Questions about outdoor sources and third hand smoke should be included in questionnaires.

Abstract

Introduction

Questionnaires are widely used to assess secondhand smoke (SHS) exposure. However, the validity of self-reported SHS exposure indicators has been rarely assessed. We aimed to assess correlations, sensitivity, specificity, and predictive values between self-reported SHS exposure indicators and airborne nicotine concentrations.

Methods

We performed a cross-sectional study with a convenience sample of 175 homes in Barcelona and Santiago de Compostela, Spain. Airborne nicotine samples were collected from participants’ homes and a self-administered questionnaire was completed on SHS exposure in the home. Spearman correlations coefficients and sensitivity, specificity and predictive values were assessed between self-reported SHS exposure indicators and nicotine concentrations in the home.

Results

All self-reported SHS exposure indicators correlated moderately strongly with airborne nicotine concentrations (Spearman correlations coefficient ranging from 0.58 to 0.65). Moreover, sensitivities and negative predictive values between self-reported indicators and the presence of nicotine in the home were below 66.4% while specificities and positive predictive values were over 78.4%. The “number of people usually smoking in the home” showed the best results (rs = 0.65, p < 0.001; sensitivity = 50.4%, specificity = 95.2%, PPV = 95.0, NPV = 51.3).

Conclusions

The self-reported SHS indicators assessed in this study showed moderate and strong correlations, low sensitivities, and high specificities. Among them, the best results were obtained with the “number of people usually smoking in the home”.

Introduction

Secondhand smoke (SHS) is a health hazard with no safe exposure levels. It is considered as carcinogenic to humans by the International Agency for Research on Cancer and it has been causally linked to cancer, cardiovascular and respiratory diseases (IARC Working Group on the Evaluation of Carcinogenic Risks to Humans, 2004; Öberg et al., 2011). In 2011, 45.2% of adult non-smokers in Spain were exposed to SHS (Fernández et al., 2017), causing 1028 attributable deaths (López et al., 2016).

Environmental markers can be used to assess SHS exposure in a particular setting by measuring concentrations of a specific SHS compound. Most common airborne markers are nicotine, particulate matter, carbon monoxide, 3-ethenylpyridine, polycyclic aromatic hydrocarbons, and tobacco-specific nitrosamines (Apelberg et al., 2013). Nicotine is notable among other markers due to its specificity to tobacco smoke, sensitivity at low concentrations, and easy collection (Apelberg et al., 2013; López and Nebot, 2003). Nicotine has often been used to assess SHS in different indoor settings such as homes or working areas such as hospitality venues (Apelberg et al., 2013, Arechavala et al., 2017, Martínez-Sánchez et al., 2014), and it is also considered a reliable environmental gold standard (Avila-Tang et al., 2013).

Questionnaires are an extended and well-accepted tool to assess SHS exposure and their administration is easy and affordable (Avila-Tang et al., 2013). To assess SHS exposure, questionnaires usually include items on the frequency of exposure, collected as the number of days per week that exposure takes place, and/or intensity in terms of the number of smokers, number of hours of exposure, number of cigarettes smoked, or the perceived intensity of the exposure (Avila-Tang et al., 2013; Galán et al., 2014; Nebot et al., 2011). Although most questionnaires on SHS exposure have not been validated, some analyses have been conducted combining reported information and objective measurements. Validity in terms of correlations between airborne nicotine and self-reported intensity indicators have been analysed in various settings with heterogeneous results. For example, reporting the number of cigarettes smoked in the home per day showed strong correlations with nicotine levels (Leaderer and Hammond, 1991, Marbury et al., 1993), while reporting the perceived intensity of the exposure in hospitality venues showed weak to moderate correlations (Galán et al., 2014). However, there are scant formal validity assessments of questionnaires using performance classifications tests (sensitivity and specificity) against a true gold standard.

In 2012 and 2016, as part of an investigation of SHS exposure in homes, airborne nicotine levels were measured in a sample of homes and a questionnaire on self-reported SHS exposure was administered to household members. In the present study, we aimed to assess the validity of self-reported SHS exposure indicators in the home in terms of correlations, sensitivity, specificity, and predictive values using airborne nicotine as the gold standard.

Section snippets

Design and study sample

This is a cross-sectional study based on a convenience sample of Spanish homes. 75 households were recruited in Santiago de Compostela in 2012 and 100 in Barcelona in 2016 through contacts of the researchers. A total of 175 homes took part in the study and there was representation of houses with smokers and houses without smokers.

Airborne nicotine measurement

We collected vapor-phase nicotine using a passive sampling monitor containing a 37-mm diameter filter treated with sodium bisulphate. Monitors were installed in the

Results

Nicotine concentrations were measured in 175 households, of which 88 were inhabited by at least one smoker. Most (73.1%) of the survey respondents were women and 69.8% had university education. Among households without smokers, 80.2% did not allow smoking in the home and 83.7% reported that no one had smoked at home the last week. In homes with smokers, 36.4% did not allow smoking in the house and the same proportion reported to smoke outdoors (33%) or nowhere (3.4%) at home the last week (

Discussion

In this study, we analysed the validity of self-reported SHS indicators often used to assess SHS exposure. These indicators moderate to strongly correlate with airborne nicotine concentrations. Moreover, sensitivities and the NPV between self-reported indicators and the presence or absence of nicotine concentrations in the home were generally low, and specificities and PPV very high. Among the variables analysed, the best results were obtained with the “number of people usually smoking in the

Acknowledgment

We would like to thank Marc Botella Periago for collecting the environmental samples, and all of the participants who were involved in the study.

This work was supported by the Instituto Carlos III (PN I+D+I 2013–2016) and co-funded by the European Regional Development Fund (FEDER) (Grant nos. PI13/02734 and PI10/00400). The study was also partially supported by the Ministry of Universities and Research of Catalonia (AGAUR) (Grant nos. 2014 SGR 1373 and 2014 SGR 999).

Contributors

TA, XC and MJL conceptualized the study. TA analysed the data and lead the manuscript writing and XC and MJL contributed to the drafting of the manuscript. All authors took part in the data interpretation, reviewed the manuscript and accepted the final version before submission.

Declaration of interests

The authors declare they have no actual or potential competing financial interest.

References (24)

  • S.K. Hammond et al.

    A diffusion monitor to measure exposure to passive smoking

    Environ. Sci. Technol.

    (1987)
  • IARC Working Group on the Evaluation of Carcinogenic Risks to Humans

    Tobacco Smoke and Involuntary Smoking, IARC Monographs on the Evaluation of Carcinogenic Risks to Humans

    (2004)
  • Cited by (16)

    • Measurement of airborne nicotine, as a marker of secondhand smoke exposure, in homes with residents who smoke in 9 European countries

      2023, Environmental Research
      Citation Excerpt :

      The highest educational level reported was compared according to each country's compulsory years of schooling. This questionnaire was adapted from a preceding study assessing the validity of self-reported SHS exposure indicators at home (Arechavala et al., 2018). We included national-level tobacco control related covariates in the analyses.

    • Methodological guidelines for the estimation of attributable mortality using a prevalence-based method: the STREAMS-P tool

      2022, Journal of Clinical Epidemiology
      Citation Excerpt :

      In the case of obesity, if the measurements of weight and height are self-declared, the obesity prevalence is underestimated [38]. The same can happen with exposure to second-hand tobacco smoke derived from self-declaration, which is usually an underestimation compared to objective measurements [39,40]. A similar bias is possible for diabetes diagnosis, as the proportion of undiagnosed people is high [34].

    • Burden of respiratory disease attributable to secondhand smoke exposure at home in children in Spain (2015)

      2019, Preventive Medicine
      Citation Excerpt :

      Additionally, SHS exposure was defined on the basis of self-reported information, which may generate information bias. However, self-reported SHS exposure has been proven to have good validity (Arechavala et al., 2018b). Moreover, the indicator used was the number of smokers at home, which should not be that sensitive to this type of bias.

    View all citing articles on Scopus
    View full text