Public Awareness and Perceptions Surrounding Radon Testing in a State With High Radon Emission Potential and Low Smoking Rates.

Author:Ou, Judy Y.


In the U.S., lung cancer is a major cause of cancer death and radon is the leading cause of lung cancer mortality among nonsmokers (Peterson et al., 2013). The lack of recommended lung cancer screenings for low and nonsmoking populations might contribute to late stage diagnosis and low survival rates for lung cancer patients who fall into these groups (Ellis & Vandermeer, 2011). Early detection of lung cancer in nonsmokers is not currently feasible; thus, preventing exposure to carcinogens should be a priority to reduce lung cancer mortality in low and nonsmoking populations. Utah has the lowest smoking rate in the U.S., with less than 10% of the Utah population estimated to be smokers (Centers for Disease Control and Prevention [CDC], 2019a). Yet lung cancer is still the leading cause of cancer related mortality in the state (CDC, 2019a; Utah Department of Health, 2018). Due to the high proportion of nonsmokers in Utah, preventing radon exposure through testing could have a major impact on the state's lung cancer mortality.

Voluntary radon testing in homes is largely driven by knowledge about radon as a carcinogen, information seeking, and the perception that radon poses a threat to health (Davis et al., 2018; Kennedy, Probart, & Dorman, 1991). Meaningful understanding of radon as a health hazard, however, has been declining nationwide (Laflamme & VanDerslice, 2004) and low radon awareness is correlated with lower income, minority status, older age, rural residence, and lower educational attainment (Halpern & Warner, 1994). A survey of New York residents reported that 82% of residents were aware of radon but only 21% of participants knew that radon was a carcinogen. Those with a meaningful awareness, indicating an understanding that radon is linked to lung cancer, had more Radon awareness and testing practices in Western states with large rural populations, such as Utah, are understudied, even though these states have high radon emission potential (U.S. Environmental Protection Agency [U.S. EPA], 2019a). One assessment in Montana reported that 32% of homes have radon levels above the safe health standard of 4 pCi/L (Hill, Butterfield, & Larsson, 2006), which is comparable to Utah where an estimated 30% of homes have radon >4 pCi/L (Leonard, 2011). Rural Montanans severely underestimated the seriousness of radon exposure, with 52% unsure whether radon causes health problems. Similarly, in a small Utah survey, only 22% of participants could link radon exposure to lung cancer and only 8.6% had tested for radon in their homes (Akerley et al., 2011).

A radon level of 4 pCi/L in indoor air is recognized by multiple health agencies as harmful (CDC, 2019b; U.S. EPA, 2019b; World Health Organization [WHO], 2009). All of Utah's soils have the potential to emit moderate (2-4 pCi/L) to high (>4 pCi/L) levels of radon (Utah Department of Health, 2015; Utah Division of State History, 2018; Utah Geological Survey, n.d.), but the state does not mandate radon testing in homes or public spaces (Geltman, 2016). Identifying which populations are less aware of radon and its risks could aid in the development of targeted interventions to improve voluntary testing rates. At a national level, women, racial/ethnic minorities, less educated populations, and lower-income households tend to be the most uninformed about radon. Whether these trends will hold true in Western states with large rural populations, such as Utah, is unclear (Halpern & Warner, 1994).

State-specific assessments give valuable insight into specific populations and provide additional detail about awareness of radon and current radon testing practices in homes (Utah Department of Health, 2015). This state-specific evaluation of radon awareness and testing presents and assesses data collected through the Utah Behavioral Risk Factor Surveillance System (BRFSS). We identify disparities in radon awareness and testing practices by county, focusing on differences between counties with moderate and high radon emission potentials. In our state-specific analysis, it was possible to assess demographic differences by radon emission potential and to examine disparities in knowledge about testing radon in the context of these demographic and geological factors.


Datasets and Participants

Utah's 2013 BRFSS data are part of the Centers for Disease Control and Prevention's BRFSS system (Utah Department of Health, 2019), a complex population-based telephone survey that collects health data on a variety of health factors for U.S. residents [greater than or equal to] 18 years. BRFSS uses random-digit telephone dialing methods to sample more than 400,000 noninstitutionalized adults each year. BRFSS contains core questions addressing demographic characteristics, health-related behaviors, disease prevalence, and optional modules to be included at a state's discretion.

As inclusion of the radon module is optional, the 2013 BRFSS is the most recent survey available that includes information about radon. The radon module was distributed to a subset of 2013 BRFSS respondents whose responses can be extrapolated to the general population using weights and strata provided by BRFSS. The design, sample characteristics, and surveys are available at www. and through the Utah Department of Health (2019).

Radon Testing and Knowledge

The 2013 BRFSS radon module included questions on:

* Radon testing: Participants indicated if their house was tested for radon (yes, no, or never heard of radon/don't know). We grouped responses as either yes or no/ never heard of radon/don't know.

* Identifying lung cancer as an outcome of radon: Participants were asked what health condition was associated with radon. They could select from 10 options. We indicated if answers were identified, not identified, or don't know/unsure.

* Reason for not testing the home: If testing was not carried out, participants could choose 1 of 20 reasons. We aggregated the reasons into five categories:

  1. not aware of radon or testing (don't know what radon is, don't know where to get test, don't know how testing is done/how test works, haven't thought about it);

  2. cost;

  3. not recommended or needed (not at risk/not needed, house tested by previous owner, not recommended);

  4. problems with test or age of the house (too many other problems with house, house is new, house is old, test doesn't work); and

  5. personal reasons/other (don't want to know, too lazy, no time, planning to do it soon, don't own home/renting, other, don't know/not sure).

For category five, the responses were aggregated into two categories: unaware of testing and all other responses for the regression analysis. Refused responses were considered missing.

Demographics and Smoking

BRFSS core questions provided information on demographics, home ownership, and smoking. Questions included sex, age (18-34, 35-54, 55-99 years), race (White, non-White), ethnicity (Hispanic, non-Hispanic), educational attainment (

Radon County Classification and Rural/Urban ZIP Codes

All counties in Utah have either high or moderate radon emission potential. BRFSS core questions asked participants about their county and ZIP code of residence. Based on their responses, we classified respondents as residing in either a moderate or high radon county. We were unable to assign participants who refused to report a county.

As we were also interested in rural/urban differences in radon awareness, we classified respondents as living in either a rural or urban area using ZIP codes based on rural/urban commuting (RUCA) codes (U.S. Department of Agriculture Economic Research Service, 2019). ZIP codes classified as any large rural, small rural, or isolated community were grouped together as rural.

Statistical Analysis

Descriptive analyses present the raw numbers and percentages weighted using the weights and strata supplied by BRFSS. Rao-Scott chi squared tests and the appropriate stratification and weighting strategies were used to compare differences among counties. We assessed each radon outcome individually in sex- and age-adjusted logistic regression models. We also used logistic regression models to select variables for inclusion in the multivariable model. Variables that were significant...

To continue reading