Background
The Exposure, Location and lung Function (ELF) tool (previously called the Mobile Exposure Device) is an example of an innovative approach to community-defined research needs [
1]. As described previously [
1], the ELF combines a portable spirometer, a customizable app (ELF Tracker) on an Android smart phone [
2] and lightweight silicone wristbands [
3]. This allows the ELF to simultaneously record daily chemical exposure for polycyclic aromatic hydrocarbons (PAHs), geospatial coordinates of participants, and lung function measurements.
There has been an increasing need for personal chemical monitoring devices that are low-cost, easy to use, require minimal maintenance and can generate robust, reliable data [
4‐
7]. Current personal chemical exposure monitors are often hampered by a limited range of chemical substrates detected, a need for power (electrical or battery) and maintenance, and can be bulky,difficult to use and may alter a participant’s behavior due to the weight (~5lbs) [
8‐
10]. In addition, the need to evaluate chemicals as complex mixtures rather than individually, has added a difficult layer for personal monitoring. Studies have shown that passive samplers reflect the bioavailable fraction of lipophilic organic chemicals [
11,
12]. When tested concurrently with an active air monitor backpack (capable of detecting both semi-volatile and particulate matter), the wristband correlated more strongly with urinary PAH metabolites than either the polyurethane foam or filter [
10]. Similarly, other studies report strong significant correlations between concentrations in wristbands and concentrations in urine for flame retardants [
13,
14] and nicotine [
15]. Furthermore, the passive wristband sampler has high temporal and spatial sensitivity [
3], and, to date, has been used to detect and quantify 1530 different organic chemicals, including 62 different PAHs [
3,
10,
16,
17]. PAHs are present in crude oil, tobacco smoke, certain petroleum products and are produced through incomplete combustion, such as the burning of fuel or smoking/charbroiling food [
18]. Exposure to PAHs has been linked with diminished respiratory health [
19‐
26].
Mobile phones have the ability to track geospatial location and applications (apps) can include questionnaires to add personal reporting around environmental monitoring [
27‐
30]. Recently, the Smoke Sense app released by the Environmental Protection Agency demonstrated the integration of self-reported health data with exposure to smoke from wildfires [
31]. The use of apps for disease management has also been explored, such as colorimetric tests for detecting glucose, protein and pH levels in urine [
32], as well as cell phones for collecting basic scientific information [
33]. Other apps have been developed to integrate with external instrumentation, such as a lens to collect digital retinal images [
34]. Similarly, the Air-Smart Spirometer collects respiratory health measures via an external spirometer with results displayed on a smartphone or tablet [
35]. Here, we describe the ELF Tracker which integrates compliance data, collection of location data and collection and transfer of spirometry data via an external spirometer linked via Bluetooth.
Spirometry collects three measures of lung health: forced expiratory volume in 1 s (FEV1), forced vital capacity (FVC) and peak expiratory flow (PEF) [
36]. These measures can reflect respiratory responses to exposure to air pollution [
37]. Acute exposures (< 24 h) can result in changes to respiratory function measureable via spirometry [
37]. Prior studies have successfully used FEV1, FVC and PEF to monitor changes in respiratory function [
37‐
40]. Hand-held spirometers are capable of collecting valid spirometry data outside of a clinical setting [
35,
41,
42], making them ideal for multi-day research studies.
The ELF was developed and refined in collaboration with two community groups with similar air quality concerns [
1]. In each community, researchers worked with established community groups and built off community-led research initiatives. In Eugene, OR, and Carroll County, OH, residents face concerns from industrial air emissions and emissions from unconventional natural gas drilling, respectively [
1]. The ELF was designed to capture the breadth of exposures in a full day (24 h). Community members cited differing schedules and routines as a reason for looking at a full 7-day week, explaining that each day might represent different exposures [
1]. In collaboration with these communities, a multidisciplinary team of chemists, software engineers, toxicologists, and environmental public health scientists developed, refined and tested the ELF. To ensure the ELF was responsive to community needs, members in each community tested the ELF, thereby improving the usability and accessibility [
1]. The study presented herein further evaluates the ELF to determine feasibility as an environmental health tool.
Discussion
This work describes the refinement and feasibility testing of a novel environmental health tool and a description and demonstration of preliminary exposure and health outcome measures. The ELF was designed to address community concerns, while also collecting robust data capable of addressing environmental health research concerns. While the use of cell phones to collect location and health symptoms/diagnostics have been successful [
27‐
30], the ELF leverages multiple technologies to expand the capabilities afforded by a mobile phone alone. Importantly, the ELF combines environmental monitoring (i.e. detection of PAHs) with location as well as quantifiable respiratory health measures.
The data indicate that the ELF is easy to use by research participants, and capable of gathering the type of data needed to begin addressing the larger research questions around exposure and health. The high degree of compliance indicates that the tool is easy to use. The ELF Tracker can meet several of the criteria needed for community-engaged research, specifically the need for a tool that can collect robust and rigorous data that is comparable to other scientific studies [
51,
52].
To our knowledge, this is the first demonstration of PAH levels in wristbands being measured concurrently with FEV
1 levels. Personal exposure to PAHs has been previously associated with several adverse outcomes related to lung function [
19,
21‐
25], and this easy-to-use, integrated ELF tool can be used to further explore these relationships. Using this dataset, the ELF collected chemical exposure data for up to 62 PAHs each day. The exposure data demonstrated unique patterns of exposure between individuals and identified the most common PAH exposures across individuals. Previous work has demonstrated that silicone wristbands sequester volatile and semi-volatile compounds and illustrate spatial differences between individuals [
3,
16,
53‐
57]. Here, we detect PAHs in non-occupational 24-h deployment periods using wristbands. In larger studies with the ELF, we will be able to directly compare PAH exposure profiles from wristbands with FEV
1 measurements from spirometers. While this feasibility study focused on PAHs, currently the wristband can be analyzed for 1530 chemicals, allowing future analysis to evaluate a larger chemical inventory [
17].
The ELF Tracker collected physical location data using the smart-phone GPS capability. Regional built environment and atmospheric exposures are associated with multiple respiratory health outcomes. Acute outdoor fine particulate matter and ozone [
58,
59], industrial air emissions [
60], and traffic-related air pollutants [
59,
61] are associated with increased rates of asthma-related hospital visits. Urban nature is also associated with respiratory health outcomes, both positive and negative. While exposure to parks, tree canopy, and gardens are associated with decreased concentrations of air pollution and decreased rates of asthma [
62] and asthma-related hospitalizations [
63], exposure to pollen is associated with increased rates of allergic airway inflammation [
49,
64]. Evaluating these ambient environmental exposures has significant potential to reduce residual confounding and adjust for competing risks in air pollution epidemiology. The collected GPS data can therefore be used to look at multiple metrics, such as co-occurring exposure to pollutants like PM
2.5, NO
2 and ozone, or determine exposure from nearby emission sources, to include major roads or conversely, to estimate time spent near green spaces. Currently, such correlations are dependent on stationary air monitoring networks, such as the Air Quality Data Mart system run by the Environmental Protection Agency, or the Toxic Release Inventory [
47,
65]. However, in Eugene, OR, there are only two monitors within the city, and a third located 25 miles south of the city. As a result, there are geographic and spatial gaps within the linked air quality monitor and location data. TRI data is limited temporally; the data is representative of total air emissions averaged over one year, and therefore may not be representative of exposure during the times participants were monitored. However, emerging data sources such as Google Street View can monitor the built environment [
66,
67] and collect air quality measurements [
68]. Growing use of the NOAA Hazard Mapping System has allowed evaluation of wildfire smoke exposure [
69‐
71], although the system is limited by cloud cover, which can interfere with identification of smoke and/or fire. Additionally, citizen science efforts have resulted in networked community air monitors to measure urban air quality [
72]. These are examples of publicly available databases that can evaluate relationships between locations and PAH sources. We show here that the ELF Tracker can be integrated with multiple databases, allowing analysis of personal location data to sources of pollution as well as time spent near sources.
The portable spirometer allowed analysis of lung function within and between participants with over 90% of all tests collecting valid spirometry data. Furthermore, over 90% of all readings complied with study protocol and accepted spirometry validity guidelines, indicating that the portable Spirotel® spirometer is easy to use, conforms to standards and collects robust, valid data. Finally, the ELF was well-received by participants and demonstrated an overall compliance rate of over 90% across all components. Furthermore, the various datasets generated by the ELF, in combination with existing datasets (NOAA Hazard Mapping, EPA Data Mart, etc.) demonstrated the feasibility of integrating multiple datasets to identify correlations between health outcomes, chemical exposures and location metrics (TRI sites, NOAA Hazard Mapping, EPA Data Mart). This allows a comprehensive evaluation of exposure and health outcomes beyond measuring a single metric of exposure. The purpose of this study was to evaluate and assess the ELF as an environmental health tool capable of being used by study participants with a high degree of compliance, collecting multiple data types (chemical exposure, physical location and respiratory health outcomes), and generating valid, accurate datasets. This trial resulted in a refinement of the ELF protocol, dropping the afternoon spirometry readings and improving the GPS sensitivity to better calculate time spent indoors vs. outdoors. While the ELF is capable of meeting the defined metrics, limitations include the small sample size (
n = 10), preventing any exposure-health insights. Regarding the use of the ELF as an environmental health tool, we have previously conducted research evaluating the ease-of-use and compliance with the ELF tool [
1], as well as the efficacy of using an online app to collect health and exposure metrics [
2]. Furthermore, we have published on the use of the wristband as a tool that demonstrates high compliance [
10,
56,
57]. Importantly, while the sample size is small, it generated relevant and diverse data, allowing insights into data management and data integration methodologies for use in public health.
The ELF tool is now being used in a panel study to evaluate relationships between chemical exposure, physical location and lung function. Future work will enable download of the ELF Tracker on multiple operating systems (Android, iPhone, Windows, Blackberry). The ability to utilize personal phones may improve compliance and reduce burden on the participant [
30]. Finally, we have collaborated with community liaisons to begin developing an interactive, online report-back format to allow study participants to view their data (manuscript in preparation). Previous studies utilizing the wristband alone have reported data back to participants to enable improved understanding of exposure and provide mechanisms to reduce exposure [
56,
57].
Publisher’s Note
Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.