Abstract
Introduction
To improve the feasibility of remote biochemical verification of smoking status, our team developed “COast,” a mobile app integrated with REDCap that allows a research participant to complete self-report research assessments and provide a breath sample via the iCOQuit Smokerlyzer for the purposes of carbon monoxide (CO) testing. The aims of the present study were to examine (1) the validity of remote CO data capture using COast as compared to gold-standard approaches (salivary cotinine, stand-alone CO monitor) and (2) the feasibility of remote CO data capture using COast as applied to both daily and weekly CO collection schedules.
Methods
Participants (N = 143, 59% Female), including recently quit (n = 36) and current (n = 107) smokers, completed a baseline video session to capture validity data, and then were randomized to daily or weekly CO monitoring for a period of 1 month.
Results
Balancing both sensitivity and specificity, optimal cut-points for defining abstinence using the COast system were <4 parts per million (ppm) with salivary cotinine as the referent (Sensitivity = 100%, Specificity = 92.8%) and <8 ppm with the stand-alone CO monitor as the referent (Sensitivity = 100%, Specificity = 88.9%). Compliance across groups with CO monitoring was high with average compliance of 74% for the daily group and 84% for the weekly group. Self-reported feasibility and acceptability of using the system were strong.
Conclusions
Pairing the iCOQuit with REDCap via the COast app was both valid and feasible among a sample of adults who smoke cigarettes enrolled remotely. This integration may help to improve the rigor of decentralized smoking cessation trials.
Implications
With increasing prevalence of decentralized trial designs, innovative methods are needed to remotely capture biomarkers. Methods that leverage existing widely available research data capture platforms may be particularly useful for promoting adoption. The COast app, which integrates a Bluetooth-enabled CO monitor with REDCap, is a fitting, valid, and feasible solution to remotely biochemically verify smoking status.
Introduction
The coronavirus disease 2019 (COVID-19) pandemic accelerated a methodological shift from clinical trials being conducted entirely within traditional academic medical centers toward decentralization of trial procedures.1 Whereas traditional clinical trials require that enrolled participants present in person for completion of study procedures, decentralized clinical trials (also referred to as “remote” or “virtual” clinical trials) leverage digital tools to remove geographic constraints to study participation. For decentralized cigarette smoking cessation trials specifically, a key methodological limitation remains: the need for biochemical verification of smoking status.2–5 Until recently, most decentralized trials that incorporated remote biochemical verification typically utilized mailed saliva to assess cotinine, a nicotine metabolite.6–9 This approach is limited as cotinine cannot be utilized in trials that administer nicotine replacement therapy (NRT)10 or in those that accept concurrent electronic cigarette (e-cigarette) use.11–13 More recently, decentralized cessation trials have increasingly utilized remote capture of carbon monoxide (CO) to verify smoking status.14 The increase in the number of trials using remote CO to biochemically verify smoking status has largely been related to the release of smartphone-enabled, single-user CO monitors, and the iCO Smokerlyzer by Bedfont Scientific Ltd, in particular. The iCO was first released in 2015, followed by a bluetooth-enabled model (the iCOQuit Smokerlyzer) in 2020. Smartphone-enabled CO monitors like the iCO have the potential to improve the feasibility of remote biochemical verification of smoking status as expired air is a noninvasive biospecimen to collect, iCO monitors are more affordable than other remote tools including stand-alone, multiuser CO monitors and salivary cotinine, and iCOs can be used repeatedly by the same individual without an increase in cost. Numerous published trials have used iCO monitors to remotely biochemically verify smoking status, typically using a mobile app developed by Bedfont Scientific paired with the monitor15 or by integrating the monitor within other mobile cessation apps.16,17 However, a limitation of these existing resources is that they are not directly integrated with widely available research data capture platforms that are typically utilized for other aspects of clinical trials. This lack of integration with existing research data capture systems can cause challenges with processing, temporally aligning, and analyzing multiple different sources of data, as has been previously cited as a translational science bottleneck.18 An integrated platform that pairs existing research data infrastructure with Bluetooth-enabled remote patient monitoring devices for biomarker capture (eg, the iCOQuit), could circumvent many of the data challenges associated with combining multiple data sources.
The overarching goal of the present study was to improve the feasibility of remote biochemical verification of smoking status by integrating the bluetooth-enabled iCOQuit Smokerlyzer with REDCap,19,20 a web-based research data capture system. REDCap is freely available to research institutions and currently has 2.5 million users across more than 6500 institutions in 153 countries. Toward this end, our team developed “COast,” a mobile application (“app”) integrated with REDCap that allows a research participant to complete self-report research assessments and provide a breath sample via the iCOQuit for the purposes of CO testing, with all data captured via the app (self-report assessments, CO data) and stored into the research project’s REDCap database (Figure S1). In addition to capturing CO data, COast captures a photograph of the participant while they are providing their breath sample, which is also stored into the project’s REDCap database and can be used for the purposes of identity confirmation.
To meaningfully improve the feasibility of decentralized tobacco-focused trials, the COast app/iCO Quit/REDCap integration (referred to in the remainder of this manuscript as the “COast system”) must produce valid results while also being feasible for participants to use independently. Regarding validity, prior research has found that different CO monitors may require device-specific cutoffs for determination of abstinence.21,22 Thus, for the COast system to be deployed within future trials, a valid device-specific abstinence cutoff must be defined. Regarding feasibility, assessment burden, including the frequency by which participants are asked to provide CO samples, may impact compliance with remote CO monitoring. As such, the aims of the present study were to examine (1) the validity of remote CO data capture using the COast system as compared to gold-standard remote approaches to biochemically verify smoking status (mailed salivary cotinine, remote CO capture using a stand-alone CO monitor) and (2) the feasibility of remote CO data capture using the COast system as applied to both daily and weekly CO data collection schedules.
Methods
Study Approval and Guidelines
The study and protocol were approved by the Medical University of South Carolina (MUSC) Institutional Review Board (Pro00085109). Participants provided electronic written informed consent before enrollment. Trial information is presented according to the Consolidated Standards of Reporting Trials (CONSORT).23
Participants
Study participants were recruited nationwide via online advertising (March 2021–March 2023). Study advertisements included a link to a brief REDCap20 eligibility screening. Individuals who currently smoked cigarettes as well as those who had quit smoking within the last 30 days were targeted via study advertisements to facilitate establishing an abstinence cutoff while also assessing feasibility across a range of smoking levels. Eligible participants were English-speaking adults (age 18+) who either reported currently smoking cigarettes daily (1+ cigarette per day, for 20+ days out of the last 30, for the last 6+ months) or reported quitting smoking within the last 30 days, prior to which they smoked at least 5 cigarettes per day, for 20+ of the preceding 30 days, for the last 6+ months. Eligible participants were also required to possess a valid e-mail address to receive electronic remuneration, own an iOS or Android smartphone that was compatible with COast, have access to an additional video camera (eg, via computer, tablet, or secondary smartphone) to complete the baseline session, have a valid mailing address for receiving study supplies, and not currently be using NRT or other nicotine-containing products that would confound cotinine data. Regarding e-cigarettes specifically, participants were excluded if they reported daily e-cigarette use in the last month (>15 days out of the last 30) or if they reported nondaily e-cigarette use in the last month and were unwilling to abstain from using an e-cigarette for 72 hours prior to the baseline study session. Similarly, as combusted cannabis use may confound CO and impact validity outcomes, participants were excluded if they reported daily cannabis smoking in the last month (>20 days out of the last 30) or if they reported nondaily cannabis smoking in the last month and were unwilling to abstain for 24 hours prior to the baseline study session. Individuals who self-reported smoking six or more cigarettes per day but provided a CO sample on a stand-alone gold-standard CO monitor inconsistent with current cigarette smoking (ie, <6 parts per million [ppm]) were excluded from study participation. Those who reported being exposed to high levels of environmental CO during the 24 hours prior to the baseline study session, such as from being around open fires/flames or automobile exhaust, were rescheduled until they could avoid such exposure for at least a 24-hour period.
Procedures
All study procedures were delivered fully remotely. Following informed consent, participants were mailed a kit containing (1) an iCOQuit Smokerlyzer, (2) a Micro+ Smokerlyzer (ie, gold-standard stand-alone monitor), and (3) a swab saliva collection kit (SalivaBio Oral Swab from Salimetrics). After receiving the kit, participants were scheduled for and completed a baseline video visit with research staff, which served to collect validity data. During the baseline session, participants were instructed over video to provide breath samples via each CO monitor. Monitor sequence was counterbalanced to avoid order effects. To ensure a valid gold-standard CO value, participants provided two breath samples via the stand-alone monitor, and a third was requested if the initial samples differed by more than 3 ppm. An average CO value was calculated for breath samples captured via the stand-alone monitors. At least 1–3 minutes, but no more than 5 minutes separated each breath sample. The CO reading captured via COast was automatically entered into REDCap, and the CO readings captured via the stand-alone monitor were shown to the staff person over video who then recorded it into REDCap. Regarding the use of the COast system, all instructions regarding how to use the iCOQuit Smokerlyzer with the COast app to provide a CO sample were included in the app (see Figure S1). No additional training was provided by research staff regarding how to use COast. During the baseline session and across all follow-up timepoints, participants were intentionally not provided with feedback regarding their CO value, on the basis that such feedback may impact compliance with using the COast system. After providing each breath sample, participants provided saliva samples for cotinine testing. At the end of validity testing procedures, research staff instructed participants regarding the return of the saliva sample and the stand-alone monitor. All participants were provided with a prepaid USPS return label and packaging that fit in home mailboxes. Participants were compensated $40 via e-mailed Amazon electronic gift codes for validity testing completion upon return of the mailed saliva sample and an additional $50 for return of the stand-alone CO monitor in working condition. Saliva samples subsequently were assayed for cotinine at the MUSC Research Nexus Laboratory.
At the end of the baseline video session, participants were randomized 1:1 to either daily or weekly compliance testing, for a period of 4 weeks total for both groups. Participants were prompted via automated text messages and push notifications to alert them when they were due to provide a CO sample via COast. Those randomized to the daily arm were asked to submit CO samples via COast daily for 28 days, and those in the weekly arm were prompted on days 7, 14, 21, and 28. Thus, the number and frequency of CO submissions differed by group but not the study duration. Participants in the daily arm had up to 16 hours after receiving the notification to provide a CO sample and compensation for submission of the sample, and participants in the weekly arm had up to 48 hours. Total potential compensation for providing CO samples was held constant across study groups and daily participants were compensated $2 for each CO sample submitted while weekly participants were compensated $14 per submitted sample (total potential compensation for CO submissions in both groups was $56).
Thirty days after the completion of the baseline video visit, participants in both groups were prompted to complete a final set of self-report study assessments within the COast app and asked to provide a final CO sample. Compensation for completion of the 30-day follow-up assessments (including providing CO) was $20 for all participants.
Assessments
All participants at baseline completed a general assessment of demographics. Primary outcomes for this study include measures related to (1) validity of the COast system for accurately identifying current smoking status as compared to gold-standard approaches of mailed salivary cotinine and CO capture via a stand-alone monitor and (2) feasibility and acceptability of using the COast system.
Validity
Validity of the COast system was measured separately against the two gold-standard measures for biochemical verification: stand-alone CO monitor (Micro+ Smokerlyzer from Bedfont; abstinence defined as <6 ppm24) and salivary cotinine (abstinence defined as <10 ng/mL24).
Feasibility and Acceptability
Compliance, an indicator of participant-level feasibility and acceptability, with using the COast system to submit CO via the iCOQuit Smokerlyzer was assessed via passively collected REDCap analytics data. Feasibility and acceptability were also assessed via participant self-report during the end-of-study assessment completed at 30 days. Participants responded to items querying their experiences using the system. Items were developed by the investigative team specifically for this study, and each item was scored on a 5-point Likert scale with response options ranging from Strongly Disagree to Strongly Agree.
Sample Size Calculation and Data Analytic Plan
Sample Size Calculation
Sample size determination was based on the number of participants needed to estimate the sensitivity and specificity of the COast system, relative to gold standard, with 95% confidence. Based on a prior trial,22 we predicted that sensitivity and specificity would be at least 90% at most cutoff values used for determining abstinence. With a maximum marginal error of 10% and assuming that approximately one-quarter of baseline CO samples would be abstinent (participants who had quit smoking within the last 30 days), at least 125 participants would be required.25
Data Analytic Plan
General descriptive statistics for demographic characteristics were summarized for both daily and weekly groups, as appropriate (means and standard deviations for continuous measures; frequencies and percentages for categorical measures).
Sensitivity and specificity metrics along with their 95% confidence intervals (95% CI) were calculated for different cutoff values for determining abstinence (whole numbers ranging from 1 to 10) from the COast system with each of the gold-standard measures. Additionally, the Youden Index26 was used to determine an optimal abstinence cut-point, balancing both sensitivity and specificity, when using the COast system. The Youden Index is often used when both sensitivity and specificity are important in determining cut-points and is calculated by summing the sensitivity and specificity for each potential cutoff value for a given diagnostic test, subtracting 1 from the summed value, and using the highest values (closest to 1) for the cut-point.
Compliance was calculated for each individual participant as the number of CO submissions successfully completed divided by the number scheduled based on group (daily vs. weekly). Individuals in the daily group had 30 submissions to provide (baseline, 28 daily, 1-month follow-up), and those in the weekly group had 6 submissions to provide (baseline, 4 weekly, 1-month follow-up). Compliance scores were descriptively summarized for each time point (day or week) as the average percent completed within group. Additionally, self-report measures of feasibility and acceptability were summarized by group (daily vs. weekly) for each assessment. Strongly Agree and Agree responses were combined and reported for each question as a percentage.
Results
Participant Characteristics
See Figure 1 for information on study screening rates, enrollment, allocation, and attrition, and Table 1 for demographics.
Figure 1.
CONSORT diagram.
Table 1.
Participant Demographics
| Full sample (N = 143) | Daily (n = 71) | Weekly (n = 72) | |
|---|---|---|---|
| Age (M(SD)) in years | 42.0 (11.7) | 42.1 (11.4) | 41.9 (12.1) |
| Sex (% Female) | 85 (59%) | 38 (54%) | 47 (65%) |
| Race (%) | |||
| White | 104 (73%) | 54 (76%) | 50 (69%) |
| Black | 27 (19%) | 13 (18%) | 14 (19%) |
| Asian | 6 (4%) | 3 (4%) | 3 (4%) |
| Multiracial | 6 (4%) | 1 (1%) | 5 (7%) |
| Ethnicity (% Hispanic/Latinx) | 12 (8%) | 5 (7%) | 7 (10%) |
| Education (%) | |||
| ≤High school diploma | 31 (22%) | 17 (24%) | 14 (19%) |
| >High school diploma | 112 (78%) | 54 (76%) | 58 (81%) |
| Annual household income (%) | |||
| <$50k | 92 (64%) | 43 (61%) | 49 (68%) |
| ≥$50k | 51 (36%) | 28 (39%) | 23 (32%) |
| Type of smartphone (%) | |||
| iPhone | 58 (41%) | 30 (42%) | 28 (39%) |
| Android | 85 (59%) | 41 (58%) | 44 (61%) |
| Cigarettes per day (%) | |||
| 1–5 | 23 (16%) | 11 (15%) | 12 (17%) |
| 6–10 | 23 (16%) | 11 (15%) | 12 (17%) |
| 11–15 | 30 (21%) | 15 (21%) | 15 (21%) |
| 16+ | 31 (22%) | 16 (23%) | 15 (21%) |
| Quit within the last 30 days | 36 (25%) | 18 (25%) | 18 (25%) |
Validity
When salivary cotinine was used as the gold standard, a CO cutoff value of <4 ppm on the COast system was needed to optimize both sensitivity (correct detection of abstinent samples) and specificity (correct detection of smoking) using Youden’s Index (Sensitivity: 100%, 95% CI: 59% to 100%, Specificity: 92.8%, 95% CI: 85.7% to 97.1%; Youden’s Index = 0.928). When the stand-alone CO monitor was used as the gold standard, a cutoff value of <8 ppm was needed to optimize both sensitivity and specificity (Sensitivity: 100%, 95% CI: 92.0% to 100%, Specificity: 88.9%, 95% CI: 81.0% to 94.3%; Youden’s Index = 0.889). See Table 2 for the sensitivity and specificity of alternative cutoffs ranging from <1 to <10 ppm.
Table 2.
Validity of the COast System
| Cotinine (≤10 ng/mL) | Stand-alone CO Monitor (≤6 ppm) | |||||
|---|---|---|---|---|---|---|
| iCO cutoff value point (ppm) | Sensitivity | Specificity | Youden’s Index (J) | Sensitivity | Specificity | Youden’s index (J) |
| <1 | 14.3% (0.4, 57.9) | 99.0% (94.4, 100) | 0.133 | 22.7% (11.5, 37.8) | 100.0% (96.3, 100) | 0.227 |
| <2 | 42.9% (9.9, 81.6) | 97.9% (92.8, 100) | 0.408 | 56.8% (41.0, 71.7) | 99.0% (94.5, 100) | 0.558 |
| <3 | 85.7% (42.1, 99.6) | 93.8% (87.0, 97.7) | 0.795 | 79.6% (64.7, 90.2) | 95.0% (88.6, 98.3) | 0.745 |
| <4 | 100% (59.0, 100) | 92.8% (85.7, 97.1) | 0.928 | 86.4% (72.7, 94.8) | 93.9% (87.3, 97.7) | 0.803 |
| <5 | 100% (59.0, 100) | 90.7% (83.1, 95.7) | 0.907 | 88.6% (75.4, 96.2) | 92.9% (86.0, 97.1) | 0.816 |
| <6 | 100% (59.0, 100) | 88.7% (80.6, 94.2) | 0.887 | 90.9% (78.3, 97.5) | 90.9% (83.4, 95.8) | 0.818 |
| <7 | 100% (59.0, 100) | 87.6% (79.4, 93.4) | 0.876 | 95.5% (84.5, 99.4) | 90.9% (83.4, 95.8) | 0.864 |
| <8 | 100% (59.0, 100) | 84.5% (75.8, 91.1) | 0.845 | 100% (92.0, 100) | 88.9% (81.0, 94.3) | 0.889 |
| <9 | 100% (59.0, 100) | 83.5% (74.6, 90.3) | 0.835 | 100% (92.0, 100) | 87.9% (79.8, 93.6) | 0.879 |
| <10 | 100% (59.0, 100) | 80.4% (71.1, 87.8) | 0.804 | 100% (92.0, 100) | 88.9% (81.0, 94.3) | 0.848 |
Bolded text highlights the cut-point that optimizes both sensitivity and specificity.
Feasibility and Acceptability
Compliance for both the daily and weekly groups was generally high with average compliance across all timepoints of 74% for the daily group (Range: 60.6%–100%) and 84% (Range: 62.5%–100%) for the weekly group (Figures 1 and S2).
Feedback regarding the use of the COast system was generally positive across participants in both daily and weekly groups (Figure 2). Across groups, greater than 85% of participants either agreed or strongly agreed that the COast system was easy to use, they could complete submissions when prompted, others would be able to use the integration with ease, the device was compact and easily portable, and they understood how to use the system. Agreement was lowest when asked about liking using the COast system, with 71% of daily participants and 70% of weekly participants either agreeing or strongly agreeing that they liked using the system.
Figure 2.
Self-reported feasibility and acceptability.
Discussion
These results support both the validity of the COast system as well as the feasibility and acceptability of using this integrated system among individuals who smoke cigarettes or have recently quit. First, regarding validity, we considered two different biochemical measures of smoking status, salivary cotinine and a stand-alone CO monitor, as gold standards. Cutoffs on the COast system had high sensitivity and specificity with established cutoffs for abstinence on these gold-standard measures, though the cutoff varied between gold standards (4 ppm for cotinine, 8 ppm for CO). Thus, the cut-point for determining abstinence on the COast system depends on which gold-standard measure is considered to be the true indicator of smoking status. Cutoffs for distinguishing between abstinence and nonabstinence within the range of 4–10 ppm have been supported in numerous prior trials of other CO monitors.21,24,27–29
Study results also support participant feasibility and acceptability of the COast system for providing CO. For both participants randomized to daily and weekly CO submissions, compliance with using the system was generally high across the study period but decreased gradually over time. In terms of decreasing compliance, participants randomized to weekly monitoring had overall higher compliance across the assessment period than participants randomized to daily monitoring, suggesting that less frequent CO monitoring may be more acceptable and associated with higher compliance. In a recent meta-analysis of studies that have implemented remote biochemical verification methods to verify smoking status, Thrul and colleagues reported a mean sample return rate across modes of biochemical verification of 70% for smoking cessation intervention studies without contingency management, 77% for contingency management studies, and 65% for other studies.14 As compared to these compliance rates, average compliance with using the COast system for weekly CO monitoring was higher and for daily CO monitoring was consistent with prior research. Though it will be important to examine patterns of compliance across more protracted endpoints (eg, at 3, 6, and 12 months), CO monitoring using the COast system, particularly when participants are prompted weekly to provide a CO sample, may be a feasible method for remote biochemical verification of smoking status. Interestingly, compliance with CO submission was higher for both groups at the 30-day follow-up assessment timepoint as compared to the prior assessment timepoint at 28 days (Daily: 65% vs. 70%; Weekly: 62.5% vs. 74%). Participants were compensated $20 for completion of the 30-day follow-up versus $2 for prior CO submissions for the daily group and $14 for the weekly group. Thus, increasing remuneration for CO submission may be a useful strategy for promoting compliance when it has waned across the course of a trial.
Self-reported feasibility and acceptability of using the COast system were also generally high across groups. In terms of areas for potential improvement, ratings of liking using the COast system were lowest as compared to other areas of feasibility/acceptability. These comparatively lower ratings may be due, at least in part, to participants not receiving feedback regarding their CO value after submission of their breath samples. This decision was intentionally made a priori as self-monitoring can be considered a cessation intervention by itself,30 which could, in turn, impact compliance with using the system. However, providing participants with information about their CO value likely would increase their liking of using the system as it may provide a more complete user experience.
Results of this study should be interpreted with limitations in mind. While we attempted to control for a number of factors that may impact validity results (eg, combustible cannabis use, environmental CO exposure), it is possible that there may be other factors that could impact validity outcomes that were not assessed and thus not controlled for. Participants were not followed beyond 30 days and as such, results may not apply to more protracted endpoints (eg, 3, 6, and 12 months) typically utilized within many clinical trials. It is likely that, if using the same remuneration, compliance with CO monitoring using the COast system would be lower at these more distal timepoints. Similarly, compliance with CO monitoring was likely impacted by the amount of remuneration provided for each CO submission. Herein, we held total potential compensation constant across groups and selected compensation amounts ($2, $14) that likely would be within the budgets of many clinical trials. However, because compliance was higher at the 30-day follow-up when payment was higher, compliance and remuneration are likely linked. Finally, study-specific inclusion/exclusion criteria may impact the generalizability of results.
There are a number of additional important future directions stemming from this work. The iCOQuit/COast integration was built into MUSC’s institutional instance of REDCap and, as such, cannot easily be deployed within other instances of REDCap. With the support of a grant from the National Center for Advancing Translational Science (NCATS), our team is currently expanding this functionality so that the device/REDCap integration will work across instances of REDCap. Moreover, this study provided an initial proof of concept for the integration of a Bluetooth-enabled remote patient monitoring device with REDCap. This system can be further expanded to support additional Bluetooth-enabled remote patient monitoring devices to improve the feasibility of remote biomarker capture across clinical areas (eg, blood pressure monitoring, pulse oximetry), which is also an area of focus for our current NCATS-supported work. Finally, to promote utilization and compliance, it will be important in the future to optimize system feasibility and acceptability among those who had less positive experiences using the COast system as well as among other end-users, particularly research staff.
In conclusion, pairing of the iCOQuit with REDCap via the COast app was both feasible and acceptable among a sample of adults who smoke cigarettes recruited nationwide and enrolled remotely. This integration may help to improve the rigor of decentralized smoking cessation clinical trials by improving the feasibility of remote biomarker capture.
Supplementary Material
Supplementary material is available at Nicotine and Tobacco Research online .
Acknowledgments
The authors would like to thank the Medical University of South Carolina’s Biomedical Informatics Center for their partnership in developing COast and study staff members including Monika Schindwolf and Olivia Levins for assistance in data collection.
Contributor Information
Jennifer Dahne, Department of Psychiatry and Behavioral Sciences, Medical University of South Carolina (MUSC), Charleston, SC, USA; Hollings Cancer Center, MUSC, Charleston, SC, USA.
Amy E Wahlquist, Center for Rural Health Research, East Tennessee State University, Johnson City, TN, USA.
Erin A McClure, Department of Psychiatry and Behavioral Sciences, Medical University of South Carolina (MUSC), Charleston, SC, USA; Hollings Cancer Center, MUSC, Charleston, SC, USA.
Noelle Natale, Department of Psychiatry and Behavioral Sciences, Medical University of South Carolina (MUSC), Charleston, SC, USA; Hollings Cancer Center, MUSC, Charleston, SC, USA.
Matthew J Carpenter, Department of Psychiatry and Behavioral Sciences, Medical University of South Carolina (MUSC), Charleston, SC, USA; Hollings Cancer Center, MUSC, Charleston, SC, USA.
Rachel L Tomko, Department of Psychiatry and Behavioral Sciences, Medical University of South Carolina (MUSC), Charleston, SC, USA.
Funding
Funding for this research was provided by the National Cancer Institute (R21 CA241842), the National Institute on Drug Abuse (K23 DA045766), the American Cancer Society (IRG-19-137-20), and the National Center for Advancing Translational Sciences (R41 TR004224). REDCap at the Medical University of South Carolina is supported by the National Center for Advancing Translational Sciences of the National Institutes of Health under Grant Number UL1 TR001450. The funding source had no role in study design, data collection, data analysis, data interpretation, in writing this report, or in the decision to submit this article for publication.
Declaration of Interests
The authors declare no relevant conflicts of interest that may impact this work. The authors had full access to all of the data in the study and take responsibility for the integrity of the data and the accuracy of the data analysis.
Author Contributions
Jennifer Dahne (Conceptualization [lead], Funding acquisition [lead], Project administration [lead], Writing—original draft [lead]), Amy Wahlquist (Data curation [supporting], Formal analysis [lead], Writing—review & editing [supporting]), Erin McClure (Funding acquisition [supporting], Methodology [supporting], Writing—review & editing [supporting]), Noelle Natale (Project administration [supporting], Writing—review & editing [supporting]), Matthew Carpenter (Conceptualization [supporting], Funding acquisition [supporting], Writing—review & editing [supporting]), and Rachel Tomko (Conceptualization [supporting], Funding acquisition [supporting], Methodology [supporting], Writing—review & editing [supporting])
Data Availability
We will make de-identified data available to users under a data-sharing agreement that provides for (1) a commitment to using the data only for research purposes and not to identify any individual participant; (2) a commitment to securing the data using appropriate technology; and (3) a commitment to destroying or returning the data after analyses are completed. All data sharing will comply with privacy and confidentiality protections such as the NIH Certificate of Confidentiality and applicable laws, regulations, and policies governing data derived from human participants. Data requests should be addressed to the corresponding author at dahne@musc.edu
References
- 1. McDermott MM, Newman AB.. Remote research and clinical trial integrity during and after the coronavirus pandemic. JAMA. 2021;325(19):1935–1936. [DOI] [PubMed] [Google Scholar]
- 2. Gorber SC, Schofield-Hurwitz S, Hardt J, Levasseur G, Tremblay M.. The accuracy of self-reported smoking: a systematic review of the relationship between self-reported and cotinine-assessed smoking status. Nicotine Tob Res. 2009;11(1):12–24. [DOI] [PubMed] [Google Scholar]
- 3. Wong SL, Shields M, Leatherdale S, Malaison E, Hammond D.. Assessment of validity of self-reported smoking status. Health Rep. 2012;23(1):47–53. [PubMed] [Google Scholar]
- 4. West R, Zatonski W, Przewozniak K, Jarvis M.. Can we trust national smoking prevalence figures? Discrepancies between biochemically assessed and self-reported smoking rates in three countries. Cancer Epidemiol Biomark Prev. 2007;16(4):820–822. [DOI] [PubMed] [Google Scholar]
- 5. Dahne J, Tomko RL, McClure EA, Obeid JS, Carpenter MJ.. Remote methods for conducting tobacco-focused clinical trials. Nicotine Tob Res. 2020;22(12):2134–2140. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 6. Abroms LC, Boal AL, Simmens SJ, Mendel JA, Windsor RA.. A randomized trial of Text2Quit: a text messaging program for smoking cessation. Am J Prev Med. 2014;47(3):242–250. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 7. Cha S, Ganz O, Cohn AM, Ehlke SJ, Graham AL.. Feasibility of biochemical verification in a web-based smoking cessation study. Addict Behav. 2017;73:204–208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 8. Brown J, Michie S, Geraghty AW, et al. Internet-based intervention for smoking cessation (StopAdvisor) in people with low and high socioeconomic status: a randomised controlled trial. Lancet Respir Med. 2014;2(12):997–1006. [DOI] [PubMed] [Google Scholar]
- 9. Free C, Knight R, Robertson S, et al. Smoking cessation support delivered via mobile phone text messaging (txt2stop): a single-blind, randomised trial. Lancet. 2011;378(9785):49–55. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 10. Murray RP, Nides MA, Istvan JA, Daniels K.. Levels of cotinine associated with long-term ad-libitum nicotine polacrilex use in a clinical trial. Addict Behav. 1998;23(4):529–535. [DOI] [PubMed] [Google Scholar]
- 11. Etter JF, Bullen C.. Saliva cotinine levels in users of electronic cigarettes. Eur Respir J. 2011;38(5):1219–1220. [DOI] [PubMed] [Google Scholar]
- 12. Etter JF. Levels of saliva cotinine in electronic cigarette users. Addiction. 2014;109(5):825–829. [DOI] [PubMed] [Google Scholar]
- 13. Etter JF. A longitudinal study of cotinine in long-term daily users of e-cigarettes. Drug Alcohol Depend. 2016;160:218–221. [DOI] [PubMed] [Google Scholar]
- 14. Thrul J, Howe CL, Devkota J, et al. A scoping review and meta-analysis of the use of remote biochemical verification methods of smoking status in tobacco research. Nicotine Tob Res. 2022;25(8):1413–1423. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 15. Schwaninger P, Berli C, Scholz U, Luscher J.. Effectiveness of a dyadic buddy app for smoking cessation: randomized controlled trial. J Med Internet Res. 2021;23(9):e27162. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 16. Kendzor DE, Businelle MS, Waring JJC, et al. Automated mobile delivery of financial incentives for smoking cessation among socioeconomically disadvantaged adults: feasibility study. JMIR Mhealth Uhealth. 2020;8(4):e15960. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 17. Bloom EL, Japuntich SJ, Pierro A, et al. Pilot trial of QuitBet: a digital social game that pays you to stop smoking. Exp Clin Psychopharmacol. 2022;30(5):642–652. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 18. Bertz JW, Epstein DH, Preston KL.. Combining ecological momentary assessment with objective, ambulatory measures of behavior and physiology in substance-use research. Addict Behav. 2018;83:5–17. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 19. Harris PA, Taylor R, Minor BL, et al. ; REDCap Consortium. The REDCap consortium: building an international community of software platform partners. J Biomed Inform. 2019;95:103208. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 20. Harris PA, Taylor R, Thielke R, et al. Research electronic data capture (REDCap)—a metadata-driven methodology and workflow process for providing translational research informatics support. J Biomed Inform. 2009;42(2):377–381. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 21. Tuck BM, Karelitz JL, Tomko RL, et al. Mobile, remote, and individual focused: comparing breath carbon monoxide readings and abstinence between smartphone-enabled and stand-alone monitors. Nicotine Tob Res. 2021;23(4):741–747. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 22. Karelitz JL, Michael VC, Perkins KA.. Analysis of agreement between expired-air carbon monoxide monitors. J Smok Cessat. 2017;12(2):105–112. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 23. Butcher NJ, Monsour A, Mew EJ, et al. Guidelines for reporting outcomes in trial reports: the CONSORT-outcomes 2022 extension. JAMA. 2022;328(22):2252–2264. [DOI] [PubMed] [Google Scholar]
- 24. Benowitz NL, Bernert JT, Foulds J, et al. Biochemical verification of tobacco use and abstinence: 2019 update. Nicotine Tob Res. 2020;22(7):1086–1097. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 25. Hajian-Tilaki K. Sample size estimation in diagnostic test studies of biomedical informatics. J Biomed Inform. 2014;48:193–204. [DOI] [PubMed] [Google Scholar]
- 26. Youden WJ. Index for rating diagnostic tests. Cancer. 1950;3(1):32–35. [DOI] [PubMed] [Google Scholar]
- 27. Cropsey KL, Trent LR, Clark CB, et al. How low should you go? Determining the optimal cutoff for exhaled carbon monoxide to confirm smoking abstinence when using cotinine as reference. Nicotine Tob Res. 2014;16(10):1348–1355. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 28. Perkins KA, Karelitz JL, Jao NC.. Optimal carbon monoxide criteria to confirm 24-hr smoking abstinence. Nicotine Tob Res. 2013;15(5):978–982. [DOI] [PMC free article] [PubMed] [Google Scholar]
- 29. Benowitz NL, Jacob P III, Ahijevych K, et al. Biochemical verification of tobacco use and cessation. Nicotine Tob Res. 2002;4(2):149–159. [DOI] [PubMed] [Google Scholar]
- 30. Foxx RM, Brown RA.. Nicotine fading and self-monitoring for cigarette abstinence or controlled smoking. J Appl Behav Anal. 1979;12(1):111–125. [DOI] [PMC free article] [PubMed] [Google Scholar]
Associated Data
This section collects any data citations, data availability statements, or supplementary materials included in this article.
Supplementary Materials
Data Availability Statement
We will make de-identified data available to users under a data-sharing agreement that provides for (1) a commitment to using the data only for research purposes and not to identify any individual participant; (2) a commitment to securing the data using appropriate technology; and (3) a commitment to destroying or returning the data after analyses are completed. All data sharing will comply with privacy and confidentiality protections such as the NIH Certificate of Confidentiality and applicable laws, regulations, and policies governing data derived from human participants. Data requests should be addressed to the corresponding author at dahne@musc.edu