E-cigarettes versus combination nicotine replacement therapy following a recent failed quit attempt: a pragmatic randomized trial through state tobacco quitlines

Summary

Background

E-cigarettes have emerged as a potentially more effective and satisfying alternative to nicotine replacement therapy (NRT) for smokers who struggle to quit. Although quitlines are effective platforms for tobacco cessation, they have not incorporated e-cigarettes due to regulatory concerns and limited clinical evidence. We evaluated whether quitline-delivered counseling combined with e-cigarettes was more effective than counseling with NRT among adults who recently failed to quit using standard quitline services.

Methods

We conducted a pragmatic, open-label, parallel-group randomised controlled trial with two U.S. state quitlines between October 2020 and January 2023. Adults [N = 350; 212 (61%) female & 248 (72%) white] who were still smoking after a recent quitline enrollment were randomised (1:1) to receive 8 weeks of either JUUL e-cigarettes or a combination of nicotine patch and lozenge, along with three counseling calls. The primary outcome was biochemically verified 7-day point prevalence smoking abstinence (7-day PPA) at 8 weeks. Analyses used an intent-to-treat approach; secondary outcomes included 12-week abstinence, prolonged abstinence, changes in smoking behavior, dependence, and adverse effects.

Findings

At 8 weeks, 7-day PPA did not differ significantly between e-cigarette and NRT groups [25 (14.3%) of 175 and 17 (9.7%) of 175, respectively; OR 1.56; 95% CI 0.80–3.04; p = 0.19]. Both groups showed similar reductions in cigarette use and dependence. Adherence to counseling and assigned products was high. Adverse events were generally mild; cough and breathing difficulties were more frequently reported in the e-cigarette group and NRT participants reported more dizziness, sleeplessness, and allergies.

Interpretation

Among quitline users with a recent failed quit attempt, e-cigarettes combined with quitline counseling were not more effective than combination NRT in increasing smoking abstinence after 8 weeks’ follow-up.

Funding

U.S. National Institute on Drug Abuse.

Research in context.

Evidence before this study

We searched the existing scientific literature on the effectiveness of e-cigarettes as a smoking cessation aid and on the use of quitlines as delivery platforms for cessation interventions. We searched PubMed, Google Scholar, and the Cochrane Central Register of Controlled Trials from database inception to October 2020. Search terms included combinations of the following: “electronic cigarettes”, “e-cigarettes”, “electronic nicotine delivery systems”, “vaping”, “nicotine replacement therapy”, “NRT”, “quitline”, “telephone quitline”, “smoking cessation”, “randomized trial”, “retreatment”, “re-engagement”, and “tobacco cessation”. We included randomized controlled trials, prospective studies, and systematic reviews comparing e-cigarettes with nicotine replacement therapy or other standard cessation treatments, as well as studies evaluating quitline-delivered treatments or proactive re-engagement of quitline callers. Prior research showed that e-cigarettes can increase quit rates compared with nicotine replacement therapy and that switching from cigarettes to e-cigarettes reduces exposure to harmful toxicants. However, the evidence base was limited by the use of older generation e-cigarettes, minimal behavioral support in many trials, and the absence of studies examining e-cigarettes delivered through a quitline platform. No trials had specifically evaluated e-cigarettes for smokers who had recently failed a standard quitline intervention.

Added value of this study

This study is the first randomized trial to test the delivery of e-cigarettes directly through a U.S. state quitline system and the first to evaluate their use among smokers who had recently failed to quit with standard quitline services. By embedding treatment within a real-world quitline and providing both groups with structured counseling, the study offers evidence on how e-cigarettes perform when implemented within an established public health platform. The findings may also refine expectations regarding the magnitude of benefit that e-cigarettes may provide in this harder-to-treat population and clarify the potential benefit of proactive re-engagement with state quitline services following a recent failed quit attempt.

Implications of all the available evidence

When combined with the broader evidence base, the results suggest that both e-cigarettes and combination NRT can support reductions in cigarette consumption and short-term cessation when delivered with behavioral counseling. Although our findings appear to contrast with the latest Cochrane review showing higher quit rates with e-cigarettes versus NRT, the results are consistent with its estimated effect size (two to six additional quitters per 100 smokers). The trial may have been underpowered due to lower than-expected abstinence rates. However, incorporating e-cigarettes into quitline services does not currently appear to confer an additional advantage over re-engagement with standard quitline-based therapy for individuals who recently failed a quit attempt. Future well-powered investigations should continue to evaluate how e-cigarettes might be optimally integrated, whether specific subgroups may benefit, and whether alternative dosing strategies, device types, or longer-term treatment could improve outcomes.

Introduction

Whereas nicotine replacement therapy (NRT) and counseling can double a person's chance of smoking cessation, most smoking quit attempts involving NRT are not successful. For those smokers, e-cigarettes may offer a more appealing and satisfying option to help them quit. In fact, a recent Cochrane Review concluded there is high-certainty evidence that people are more likely to stop smoking using e-cigarettes than NRT.¹ Reviews have also shown that switching completely from cigarette smoking to e-cigarette use is likely to reduce health risks.²

Telephone-based tobacco quitlines are an effective means for treating cigarette dependence.3, 4, 5 They have the capacity to assist a wide range of tobacco users and have proven effective even among traditionally hard-to-reach populations.6, 7, 8, 9 Standard quitline practices are based on years of rigorous research evaluating the best methods of tobacco cessation and typically include phone-based counseling and provision of free NRT (duration and dose varies by state program offerings).^10,11 To date, quitlines have not incorporated the use of e-cigarettes as a quit strategy, in part due to a lack of U.S. Food and Drug Administration (FDA) approval for these products as a cessation tool.¹⁰ While 8 prior randomised controlled trials have evaluated the efficacy of e-cigarettes compared to current evidence-based treatment, none have examined their efficacy when delivered through a quitline.^1,12 Providing e-cigarettes directly to smokers through a quitline platform could significantly increase the likelihood that e-cigarettes yield a public health benefit, particularly for those who have failed to quit using FDA-approved strategies.

The current trial focused on adult smokers who recently failed to quit with standard quitline services and were still interested in quitting smoking. The trial examined the efficacy of behavioral counseling paired with provision of e-cigarette or NRT (nicotine patch and lozenge) delivered through a quitline. Our primary hypothesis was that quitline counseling paired with e-cigarettes would be more effective at helping smokers quit smoking than counseling paired with NRT.

Methods

Study design

This parallel-arm, open label, randomised controlled trial, with randomisation at the participant level, was conducted remotely by research staff at The Ohio State University (Columbus, Ohio) and Optum (Eden Prairie, Minnesota) from October 2020 to January 2023. Participants were recruited from the Oklahoma Tobacco Helpline and South Carolina Tobacco Quitline and engaged in study procedures over 12 weeks. The trial procedures, including a waiver of signed consent, was approved by the Ohio State University institutional review board.

Participants

Individuals who had enrolled in Oklahoma or South Carolina quitline tobacco treatment services four to seven months prior for smoking cessation were contacted by phone to offer study participation (or reenrollment into their state quitline program if not interested or eligible). This timeframe was selected to allow individuals time to finish their previous quitline enrollment. Individuals were contacted if they were 21 years or older, English-speaking, and had consented to be contacted during their previous enrollment. Those reached were eligible to participate if they had smoked within the past 24 h; currently smoked at least five cigarettes per day; could read, write, and speak English; and reported at least minimal interest in switching to an e-cigarette or other vaping product (responded ≥1 on a 0–10 scale where 0 was “not at all interested” and 10 “very interested”). Individuals were excluded if they reported they were currently enrolled in a related study; cohabitating with someone already enrolled in this trial; currently using varenicline or bupropion; currently using an e-cigarette daily over the last month; had a heart attack or stroke within the last 6 months; diagnosed with bipolar disorder, depression, or schizophrenia, or hospitalized for a psychiatric condition within the past year; known to have a reaction to using patch medication or adhesive tape; allergic to propylene glycol or vegetable glycerin; or pregnant, planning to become pregnant, or breast feeding. Eligible individuals were enrolled after providing verbal consent and were then randomised to receive either an e-cigarette or NRT (combination nicotine lozenge and patch).

Randomisation and masking

After study consent and the baseline assessment, participants were randomised to either the e-cigarette or NRT arms and then completed their first study coaching call. Stratified block-randomisation with small, random-sized blocks was utilized to assign participants to either arm. Randomisation was stratified by sex, state, educational attainment, and having ever used an e-cigarette. Participant screening, enrollment, and randomisation was conducted by Optum Quit Coaches, who also conducted intervention calls throughout the study. Participants and staff were not masked to intervention assignment.

Procedures

Initial participant contact, screening, and enrollment was conducted by a group of Optum Quit Coaches with specialized research training. At the time of this study, Optum was the service provider for both the Oklahoma and South Carolina tobacco cessation quitlines. Quit Coaches must have a bachelor's degree, complete more than 240 h of training, and receive ongoing supervision. Quit Coaches who conducted recruitment and provided coaching in this study received research ethics training and 4 h of training in study-specific procedures and treatment. Study procedures were completed in three phases—Baseline, Product Trial, and Surveillance—over the course of 12 weeks.

Baseline phase

In the Baseline phase (Day 0), participants completed a baseline assessment including sociodemographic, cigarette dependence, and tobacco product use measures. Participants were then randomised to the e-cigarette or NRT arms for the eight-week Product Trial phase.

Product Trial phase (Baseline to week 8)

Participants assigned to the e-cigarette arm were mailed a JUUL device, charger, and e-liquid pods (5% nicotine; tobacco or menthol-flavor depending on participants' preferences). Pods were mailed in two shipments (baseline and four weeks); the number of pods supplied was calculated from the self-reported number of cigarettes smoked per day (CPD). This trial was originally designed to utilize the National Institute on Drug Abuse's Standardized Research Electronic Cigarette (SREC). However, a multi-year delay in the availability of the SREC device necessitated a change in intervention device prior to the start of recruitment. The JUUL device was secondarily chosen due to its market availability, ability to effectively deliver nicotine, and improved toxicant profile compared to other e-cigarettes.13, 14, 15

Participants assigned to the NRT arm were mailed nicotine patches and lozenges, with patch dosing similarly determined by self-reported smoking behavior. Participants were asked to only use their assigned study product and to report any use of products not sent from the study. All participants also received study smartphones to complete surveys and coaching phone calls, as well as a portable exhaled carbon monoxide (eCO) monitor (iCO smokerlyzer monitor; Bedfont Scientific, Medford, NJ) to remotely collect eCO readings. All materials were supplied at no cost to participants.

Participants in both arms received behavioral intervention in the form of three coaching calls from quitline coaches (at baseline and weeks one and four), each lasting 10–15 min. For participants randomised to NRT, coaching calls included addressing five key behaviors for quitting smoking: setting a quit date, using cessation medications effectively, coping with urges, removing tobacco paraphernalia from one's environment, and utilizing social support. For participants randomised to e-cigarettes, coaching calls addressed the same five keys to quitting with the exception that the guidance on cessation medications was repurposed to focus on using e-cigarettes as a device to completely replace smoking. Specifically, education on e-cigarette use and device troubleshooting, developing strategies to help completely switch to e-cigarettes (i.e., using the e-cigarette frequently throughout the day to prevent cravings, withdrawal, and relapse risk), and addressing barriers to switching. Participants in both groups also completed daily eCO tests and two full assessment battery calls (at baseline and week 8) completed by study staff (not including Quit Coaches). Full assessment battery calls assessed cigarette, e-cigarette, and NRT use, cigarette urges/craving (Tiffany-Drobes Questionnaire of Smoking Urges: Brief Form)¹⁶ and withdrawal symptoms (Minnesota Nicotine Withdrawal Scale)¹⁷ and cigarette dependence (Cigarette Dependence Scale [CDS]).¹⁸

Surveillance phase

The Surveillance phase (weeks 8–12) also included daily online surveys sent via a prompt to the participants’ study smartphone (Insight mHealth Platform; Norman, OK), as well as a full assessment battery call conducted by study staff at week 12 to assess changes in e-cigarette/NRT use as well as cessation. During this four-week period participants were neither supplied with e-cigarettes nor NRT. Self-reported adverse effects commonly experienced with ECs¹⁹ and serious adverse events (SAEs) were documented by the study team during each full battery assessment call. Consistent with the International Council for Harmonisation (ICH)—Good Clinical Practice (GCP) Guidelines,²⁰ we defined SAEs as any medical occurrence that results in death, is life-threatening, requires inpatient hospitalization or prolongation of existing hospitalization, results in persistent or significant disability/incapacity, or is a congenital anomaly/birth defect. Participants were compensated up to $330 for their time and participation in the study.

Outcomes

The primary outcome of interest was cigarette smoking abstinence at eight weeks, defined as self-reported point prevalence of smoking abstinence (7-day PPA) over the past seven days and confirmed via eCO of ≤8 ppm. Secondary outcomes included: 1) eCO-verified 7-day PPA at 12 weeks, prolonged abstinence as defined by an eCO of ≤8 ppm and fewer than seven days of cigarette use since their assigned switch/quit date, 2) changes in CPD, 3) cigarette dependence, 4) withdrawal symptoms, and 5) intervention adherence and adverse effects.

Statistical analysis

The study was powered to detect a difference in smoking abstinence at the end of the Product Trial phase (eight-week follow-up) between the e-cigarette and NRT study arms on the complete-case dataset. It was estimated that the 7-day PPA rates in the e-cigarette and NRT arms would be 35% and 20%, respectively, and there would be a 20% attrition rate. Estimates for JUUL were based upon limited available evidence at the time of trial planning.²¹ A sample size of 300 participants (i.e., 150 completers per arm) would provide over 80% power for a two-sided 0.05 level Fisher exact test to detect a significant difference between the arms. Incorporating the attrition rate, we planned to recruit a minimum of 420 participants at baseline, 186 per arm. Due to initial delays in acquiring the SREC device, SARS-coV-2 pandemic delays, a pause in recruitment following JUUL e-cigarette receiving a temporary marketing denial order that was later stayed and then rescinded by the US FDA,²² and changes in the underlying quitline platform that required us to end recruitment, 350 participants were enrolled, constituting 83% of the total goal.

An intent-to-treat (ITT) approach was used for the primary analyses where participants with missing data due to loss to follow-up or otherwise are assumed to be smoking. For the primary abstinence outcomes, treatment groups were compared with logistic regression models adjusting for randomization stratification factors (sex, state, educational attainment, and having ever used an e-cigarette), and odds ratios (OR) are reported along with 95% confidence intervals (CI). Sensitivity analyses for the primary outcome included assessing the effect of varying the eCO measurement cutoff from the prespecified value of ≤8 ppm to ≤3, ≤6, and ≤10 ppm as well as a complete case analysis consisting of participants who completed the assessment call, including providing an eCO sample. For secondary outcomes, treatment groups were compared with linear (CPD, CDS, withdrawal symptoms), logistic (abstinence, use of study product, AEs), ordinal logistic (number of contacts and counseling calls complete), and multinomial logistic regression (use evaluation); odds ratios or mean differences, as appropriate, were presented along with corresponding 95% CIs. All models adjusted for randomization stratification factors. Analysis of withdrawal symptoms was restricted to the subset of participants who reported that they were not currently smoking at the time of their study visit. Safety was evaluated by summarizing serious adverse events and self-reported adverse effects by treatment group. SAS version 9.4 was used for all analyses. The trial is registered with ClinicalTrials.gov, NCT03502200.

Role of the funding source

The funder of the study had no role in study design, data collection, data analysis, data interpretation, or writing of the report.

Results

Among the 5497 previous quitline callers screened between October 2020 and October 2022, 3665 (66.7%) were ineligible, 1457 (26.5%) declined participation (with unknown eligibility), and 25 (0.5%) were eligible but did not participate in the study. The remaining 350 (6.4%) assented/consented to study enrollment, completed the baseline assessment and were randomised (175 per group) (Fig. 1). Among those randomised, 126 (72.0%) of 175 in the NRT group and 128 (73.1%) of 175 in the e-cigarette group completed the 12-week follow-up, with the final participant follow-up occurring in January 2023. Participants’ baseline characteristics were evenly balanced between treatment groups (Table 1), with a mean age of 54.7 years (SD = 12.6), most earning less than $35,000 annually [272 (80.2%) of 339; suggestive of lower socioeconomic position], and a mean cigarette dependence scale score of 19.3 (SD = 3.0), indicating moderate to high levels of dependence.

Fig. 1.

Trial CONSORT diagram.

Table 1.

Baseline characteristics of participants by study arm.^a

	e-cigarette (n = 173)	NRT (n = 173)
Age mean (SD)	55.3 (12.7)	54.2 (12.5)
Sex n (%)
Female	107 (62%)	105 (61%)
Male	66 (38%)	68 (39%)
Ethnicity n (%)
Hispanic	2 (1%)	2 (1%)
Non-Hispanic	170 (98%)	171 (99%)
Refusal/Don't know	1 (1%)	0 (0%)
Race n (%)
Black or African American	21 (12%)	23 (13%)
White or Caucasian	126 (73%)	122 (71%)
Other	26 (15%)	28 (16%)
Yearly household income n (%)
<$35,000	136 (79%)	136 (79%)
≥$35,000	34 (20%)	33 (19%)
Refusal/Don't know	3 (2%)	4 (2%)
Education level n (%)
Less than high school	38 (22%)	30 (17%)
High school graduate	35 (20%)	42 (24%)
GED	10 (6%)	13 (8%)
Some college or technical school	62 (36%)	49 (28%)
Associate's degree or higher	28 (16%)	39 (23%)
State n (%)
Oklahoma	109 (63%)	107 (62%)
South Carolina	64 (37%)	66 (38%)
e-Cigarette use n (%)
Never	51 (29%)	63 (36%)
Not within past 30 days	103 (60%)	95 (55%)
Within past 30 days	19 (11%)	15 (9%)
#Smokers in household median (IQR)	2.0 (1–3)	2.0 (1–3)
Cigarettes per day mean (SD)	16.8 (8.2)	17.4 (8.2)
Days used cigarettes in past month n (%)
0–9 days	1 (1%)	1 (1%)
10–19 days	3 (2%)	7 (4%)
20–30 days	169 (98%)	163 (94%)
Refusal/Don't know	0 (0%)	2 (1%)
Years Smoking Cigarettes mean (SD)	34.1 (13.0)	31.6 (14.5)
Serious quit attempts in past year n (%)
0	20 (12%)	24 (14%)
1	62 (36%)	63 (36%)
2	45 (26%)	46 (27%)
3	26 (15%)	16 (9%)
4	8 (5%)	6 (3%)
5 or more	12 (7%)	17 (10%)
Don't know	0 (0%)	1 (1%)
Cigarette dependence scale mean (SD)	19.3 (2.8)	19.2 (3.2)

^aBaseline demographic characteristics were not collected for four participants, two in the e-cigarette group and two in the nicotine replacement group.

We observed no differences between e-cigarettes versus NRT for the study's primary outcome of eCO-verified 7-day PPA at eight weeks. At eight weeks 25 (14.3%) of 175 participants randomised to e-cigarettes and 17 (9.7%) of 175 participants randomised to NRT had abstained from smoking in the past week (OR 1.56, 95% CI 0.80–3.04; p = 0.19; Table 2). Secondarily, no differences in eCO-verified 7-day PPA were observed at 12 weeks 12 weeks (OR 1.33, 95% CI 0.68–2.61; p = 0.40). Comparable results were seen for self-reported 7-day PPA, and when sensitivity analyses were conducted varying the eCO concentration cut-off and restricting to complete cases (Supplementary Table S1). While participants in the e-cigarette group were statistically more likely to achieve prolonged abstinence than those receiving NRT at eight weeks (OR 2.27, 95% CI 1.02–5.04; p = 0.04), this was not the case at 12 weeks (OR 1.16, 95% CI 0.54–2.46; p = 0.71).

Table 2.

Abstinence outcomes.

	e-cigarette (n = 175)	NRT (n = 175)	Odds ratio (95% CI)a	p-value
CO-verified 7-day PPA n (%)
8 weeks	25 (14.3%)	17 (9.7%)	1.56 (0.80–3.04)	0.19
12 weeks	22 (12.6%)	17 (9.7%)	1.33 (0.68–2.61)	0.40
Self-reported 7-day PPA n (%)
8 weeks	46 (26.3%)	34 (19.4%)	1.48 (0.89–2.46)	0.13
12 weeks	41 (23.4%)	33 (18.9%)	1.30 (0.78–2.20)	0.31
Prolonged abstinence n (%)b
8 weeks	21 (12.0%)	10 (5.7%)	2.27 (1.02–5.04)	0.044
12 weeks	16 (9.1%)	14 (8.0%)	1.16 (0.54–2.46)	0.71
50% Reduction in CPD from baseline n (%)
8 weeks	101 (57.7%)	102 (58.3%)	0.98 (0.64–1.50)	0.91
12 weeks	92 (52.6%)	100 (57.1%)	0.82 (0.54–1.26)	0.36
			Mean difference (95% CI)a	p-value
Change in CPD from baseline mean (SD)
8 weeks	−12.7 (9.8)	−12.8 (9.4)	0.29 (−2.10 to 2.68)	0.81
12 weeks	−11.5 (9.4)	−12.7 (9.3)	1.35 (−0.94 to 3.64)	0.25
Change in CDS from baseline mean (SD)
8 weeks	−4.6 (4.3)	−4.4 (4.0)	−0.21 (−1.27 to 0.86)	0.70
12 weeks	−4.8 (4.9)	−5.2 (4.6)	0.37 (−0.85 to 1.58)	0.55

PPA, point prevalence abstinence; CPD, cigarettes smoked per day; CDS, cigarette dependence scale (range: 5–25; higher scores indicate higher cigarette dependence.); CO, carbon monoxide.

^aAdjusted for randomisation stratification factors: ever use of an e-cigarette, education, sex, and location.

^beCO ≤8 ppm and <7 days of cigarette use since switch/quit date.

Both the e-cigarette and NRT groups reduced their cigarette consumption by approximately 13 cigarettes per day from baseline; this result was consistent at both eight (Mean Difference 0.29, 95% CI: −2.10 to 2.68; Table 2) and 12 weeks (Mean Difference 1.35, 95% CI: −0.94 to 3.64) with no significant difference between the groups. Similarly, more than half of all participants reduced their cigarette consumption by 50% at eight (OR 0.98, 95% CI: 0.64–1.50) and 12 weeks (OR 0.82, 95% CI: 0.54–1.26) with no difference between those randomised to the e-cigarette and NRT groups. Cigarette dependence was consistent with the above results, with a mean reduction from baseline dependence of over four points on the cigarette dependence scale in both study groups at eight and 12 weeks. Relative to those assigned to NRT, participants assigned to e-cigarettes reported significant improvements in the utility of the study e-cigarette to keep them from smoking and how good the e-cigarettes tasted compared to regular cigarettes; however, no differences in satisfaction were observed (Supplementary Table S2).

Study attendance was high with a median of four contacts (of five—3 coaching calls + 2 assessment calls at eight and 12 weeks) completed and over half of participants completed all three counseling calls (Table 3). At eight weeks, 112 (64%) of 175 participants in the e-cigarette group were still using their study product compared to 99 (57%) of 175 in the NRT group (p = 0.21); this decreased to 89 (51%) and 71 (41%) at 12 weeks (p = 0.028). In the e-cigarette group, 87 of 109 participants at eight weeks and 82 of 94 at 12 weeks reported using their study product every day in the past week, representing over 75% at both time points. Comparable data was not collected about ongoing NRT use for the NRT group. When assuming that those who did not complete the survey did not use their study e-cigarette at all, adherence is reduced to a median number of days used in the past week of 6 (IQR 0–7) at eight weeks and 3 (IQR 0–7) at 12 weeks. Among participants biochemically verified to be abstinent from cigarettes at their study visit, no significant differences were seen between groups in cigarette withdrawal symptoms (Table 4).

Table 3.

Attendance and treatment adherence.

	e-cigarette (n = 175)	NRT (n = 175)	Odds ratio (95% CI)a	p-value
Median contacts completedb(IQR)	4 (3–5)	4 (3–5)
Number of contacts completed n (%)			1.06 (0.72–1.56)	0.76
1	18 (10.3%)	17 (9.7%)
2	9 (5.1%)	17 (9.7%)
3	23 (13.1%)	24 (13.7%)
4	43 (24.6%)	34 (19.4%)
5	82 (46.9%)	83 (47.4%)
Median counseling calls completed (IQR)	3 (2–3)	3 (2–3)
Number of counseling calls completed n (%)			1.17 (0.77–1.77)	0.47
1	30 (17.1%)	31 (17.7%)
2	35 (20.0%)	42 (24.0%)
3	110 (62.9%)	102 (58.3%)
Use of assigned productcn (%)
8 weeks	112 (64.0%)	99 (56.6%)	1.63 (0.76–3.47)	0.21
12 weeks	89 (50.9%)	71 (40.6%)	2.04 (1.08–3.84)	0.028
Days used in past 7 daysc,dmedian (IQR)
8 weeks	6 (0–7)	–
12 weeks	3 (0–7)	–

^aAdjusted for randomisation stratification factors: ever use of an e-cigarette, education, sex, and location; Odds ratios from ordinal logistic regression models estimate having a larger vs. lower number of contacts or counseling calls completed.

^bThe maximum number of contacts was five: three coaching calls, week eight, week 12.

^cParticipants who did not attend the visit are assumed to not be using the study product.

^dOnly collected for participants assigned to the e-cigarette arm.

Table 4.

Withdrawal symptoms among CO-verified abstinent participants.

	e-cigarette	NRT	Mean difference (95% CI)a	p-value
MNWS mean (SD)
8 weeks	9.8 (9.1)	12.1 (6.1)	−1.32 (−3.22 to 0.58)	0.17
12 weeks	8.7 (8.1)	10.3 (8.3)	−1.13 (−3.04 to 0.77)	0.24
MNWS craving mean (SD)
8 weeks	1.2 (1.4)	1.2 (1.1)	−0.22 (−0.60 to 0.14)	0.23
12 weeks	0.6 (1.1)	0.7 (1.1)	−0.22 (−0.60 to 0.15)	0.24
QSU desire mean (SD)
8 weeks	7.8 (4.2)	7.4 (4.8)	−0.37 (−2.74 to 2.00)	0.76
12 weeks	8.0 (5.8)	5.4 (1.2)	−0.56 (−3.06 to 1.95)	0.66
QSU Relief mean (SD)
8 weeks	7.2 (3.8)	7.1 (4.1)	−0.22 (−2.10 to 1.66)	0.81
12 weeks	7.3 (6.2)	5.6 (1.3)	0.003 (−1.89 to 1.90)	0.99

MNWS, Minnesota Nicotine Withdrawal Scale (range: 0–32 and 0–4 for MNWS and MNWS Craving, respectively; higher scores indicate greater overall withdrawal symptoms and craving.); QSU, Questionnaire on Smoking Urges (range: 5–35 both for QSU Desire and QSU Relief; higher scores indicate stronger smoking urges.).

^aAdjusted for randomisation stratification factors: ever use of an e-cigarette, education, sex, and location.

Two participants died during the study; both were assigned to the NRT group and the causes of death were unrelated to study participation. Additionally, 21 serious adverse events occurred in 18 (10%) of 175 participants in the e-cigarette arm compared to 14 serious adverse events in 10 (6%) of 175 participants in the NRT group (Table 5). None of the serious adverse events at weeks 8 or 12 were found to be related to study treatment. The most commonly self-reported adverse effects over the course of the study were cough, sore or dry mouth and throat, and headache (Supplementary Table S3). Cough was more commonly reported among participants in the e-cigarette group (OR 3.61; 95% CI 1.99–6.55). Dizziness (OR 0.42; 95% CI 0.19–0.89), sleeplessness (OR 0.35; 95% CI 0.18–0.70), and allergies (OR 0.09; 95% CI 0.02–0.40) were less commonly reported among participants in the e-cigarette group as compared to those assigned to NRT.

Table 5.

Serious adverse events.

	e-cigarette (n = 175)	NRT (n = 175)
Breast cancer	2 (1.1%)	0
Breathing difficulties	4 (2.3%)	0
Broken back	0	1 (0.5%)
Carotid endarterectomy surgery	0	1 (0.5%)
Chest pain	2 (1.1%)	1 (0.5%)
Cuff repair surgery	1 (0.5%)	0
Dental problems	0	1 (0.5%)
Depression	0	1 (0.5%)
Diarrhea	1 (0.5%)	0
Dizziness	0	1 (0.5%)
Essential tremor worsening	1 (0.5%)	0
Exploratory larynx surgery	1 (0.5%)	0
Gallbladder problems	1 (0.5%)	0
Gastrointestinal problems	0	1 (0.5%)
Headache	0	1 (0.5%)
Heart condition	1 (0.5%)	0
Heart palpitations	1 (0.5%)	1 (0.5%)
Hernia	0	1 (0.5%)
Myocardial infarction	1 (0.5%)	0
Needed medication adjustment for bipolar schizoaffective disorder	1 (0.5%)	0
Pancreatic pain	1 (0.5%)	0
Pneumonia	0	1 (0.5%)
Seizure	1 (0.5%)	0
Suicidal thoughts	1 (0.5%)	1 (0.5%)
Ulcer	0	2 (1.1%)
Wound abscess	1 (0.5%)	0
Total number of SAEs	21	14
Total number of participants with SAEs	18	10

Discussion

For adults who recently failed to quit smoking using standard quitline services but remained interested in quitting, e-cigarettes combined with quitline counseling did not significantly increase smoking abstinence relative to combination NRT after 8 weeks. While no difference was observed between them, both interventions resulted in reductions of approximately 13 cigarettes per day from baseline. While appearing to contrast the most recent Cochrane review which found a high-certainty of evidence that e-cigarettes with nicotine increase quit rates compared to NRT, our results are still largely consistent with their finding that e-cigarettes may translate into two to six additional quitters per 100 smokers (12-week abstinence rates of 12.6% for e-cigarettes and 9.7% for NRT in present study).¹ It is important to note that our study was likely underpowered as our initial sample size estimates were predicated upon two studies^21,23 that demonstrated much higher smoking abstinence rates with e-cigarettes. Like previous research, on average both the e-cigarette and NRT intervention reduced CPD, withdrawal symptoms, and cigarette dependence; however, no significant differences were found between interventions. Adherence to both interventions was high, including counseling calls and use of e-cigarettes and NRT. The findings suggest that the re-engagement of quitline callers who recently failed to quit smoking, with either e-cigarettes or NRT, will increase smoking abstinence. This aligns with other research suggesting that proactively re-engaging individuals for quitline retreatment can be a valuable strategy.²⁴

It is not immediately clear why smoking abstinence rates in the e-cigarette arm were lower than more recent trials comparing e-cigarettes to NRT.^21,23,25 Lower abstinence rates in past trials was often attributed to a lack of behavioral counseling and/or used early generation e-cigarettes with limited nicotine delivery (some only with tobacco flavor). The current trial, however, provided moderately intensive behavioral counseling delivered by trained quitline coaches and provided a market-leading e-cigarette in two flavors, with a demonstrated ability to deliver cigarette-like levels of nicotine. Possible reasons for low abstinence rates may be multifactorial. First, US tobacco quitline callers tend to be a more difficult population to treat, characterized by high levels of nicotine dependence, low socioeconomic and educational attainment, and high levels of substance use and mental health disorders.^26,27 Moreover, this study focused on individuals who had recently been unsuccessful in a quit attempt that was supported by a 5-call proactive call coaching program with mailed NRT, and who could be reached to reengage with the quitline for this study. This may be a more difficult to treat subgroup of quitline callers. Second, the study was conducted in the US where risk perceptions of e-cigarette have continued to worsen over the last decade, with a growing number of people viewing e-cigarettes as equally harmful to cigarettes.²⁸ In fact, immediately preceding study launch the e-cigarette or vaping use-associated lung injury (EVALI) event was incorrectly attributed to nicotine e-cigarettes.29, 30, 31 Additionally, during the course of the study, there were a number of national news stories questioning the safety of JUUL e-cigarettes and JUUL received a US FDA marketing denial order that was later administratively stayed and then rescinded by the FDA.^22,32 While quitline coaches provided education to participants randomised to e-cigarettes that e-cigarettes were likely safer than combustible cigarettes, it is possible that concerns regarding their safety influenced their use.

E-cigarettes and NRT had similar adverse event profiles with no intervention-related serious adverse events. Whereas participants in the e-cigarette intervention reported more cough and breathing difficulties, though the latter symptom did not reach statistical significance, NRT participants reported more dizziness, sleeplessness, and allergies. These findings are consistent with previous trials of e-cigarettes and NRT and lend continued support regarding the safety of e-cigarettes for use in aiding smoking abstinence.³³ Nevertheless, the safety and efficacy of combination NRT is well-studied and continues to be the recommended first-line treatment.

The study has a number of strengths. It is the first study to examine the potential efficacy of providing e-cigarettes through a quitline platform. The incorporation of e-cigarettes into quitline services represents a novel approach that could address the limitations of current FDA-approved cessation strategies, particularly for those who find NRT insufficient. Also, the study's focus on quitline re-engagement is of importance. While quitlines have demonstrated efficacy in treating tobacco dependence, most smokers who utilize quitline services do not quit. The present study demonstrates that re-engagement of quitline callers may serve to further increase smoking cessation among this population. Lastly, few trials have directly compared e-cigarettes to combination NRT, the present study is the first randomised trial to directly compare the use of the newest generation of e-cigarettes to combination NRT on smoking outcomes.³³ This new generation of e-cigarette, characterized by improved nicotine delivery and ease of use, has demonstrated significant appeal to smokers and led to a significant surge in the prevalence of e-cigarette use.

Despite these strengths, the study has limitations. The relatively short duration of our trial means that long-term efficacy and safety data are still needed. Future research should aim to address these gaps by conducting longer-term studies and exploring the potential risks associated with e-cigarette use. In addition, because this was an open-label trial and study product assignment could not be blinded, as such expectancy effects may have influenced outcomes. Our inability to reach our recruitment goal and over-estimation of anticipated abstinence rates for the e-cigarette arm led to a lower sample size and reduced statistical power. However, to demonstrate the robustness of the primary aim results, sensitivity analyses varying the eCO concentration cut-point and restricting the analytic sample to complete cases were completed, all of which agreed with the main conclusions. The study was also conducted with quitline callers from only two states in the US and therefore may not generalize to other populations both inside and outside the US. Moreover, as this trial focused on individuals who were reengaged with quitline treatment after a recent failed quit attempt, findings from this trial may not generalize to the entire quitline population or non-quitline populations. It is also possible that the intensive measurement protocol, including daily eCO tests, may have contributed to a floor effect in abstinence rates. Our sample included individuals who had used e-cigarettes, which may have included those who previously failed to quit smoking with e-cigarettes, potentially leading to an underestimation of effects. However, their inclusion was critical for enhancing the generalizability of our findings, given the high prevalence of e-cigarette experience among people who smoke. Additionally, we did not have data on potential contamination between groups. While the use of an ITT analysis helps to mitigate the impact of any crossover on the study's findings, it remains possible that this conservative approach may have masked a meaningful effect.

In conclusion, we observed no difference in 7-day PPA between participants randomised to e-cigarettes and those randomised to NRT when both were combined with behavioral counseling and delivered through a quitline. Further research is needed to determine whether e-cigarettes can serve as an effective cessation aid for individuals who fail to quit after an initial standard quitline intervention. As this was not a non-inferiority trial, the findings do not support integrating e-cigarettes into quitline services at this stage. Future studies should continue to evaluate the long-term safety and efficacy of e-cigarettes for smoking cessation and assess whether combining e-cigarettes with NRT could enhance quit rates.

Contributors

TLW was responsible for study conceptualization, funding acquisition, investigation, methodology, project administration, supervision, and writing, both original and editing. AH was responsible for formal analysis, methodology, and writing, both original and editing. KAV assisted with study methodology, project administration, supervision and manuscript editing. TLW and AH had access to the raw data and verified the data. TMB, YJC, LAB, MSB, JH, and MJC assisted with methodology and manuscript editing. TLW had final responsibility for the decision to submit for publication.

Data sharing statement

Study data will be made available to researchers upon reasonable request.

Declaration of interests

MJC has served as a paid expert consultant in litigation involving trade disputes involving e-cigarette manufacturers. MSB is an inventor of the Insight mHealth platform that was used to collect study data. He receives royalties when the platform is used by researchers outside OUHSC. He did not receive royalties in this case because the study PI (TLW) was employed by OUHSC when the study was launched. KAV is an employee of RVO Health, the provider of quitline services in this trial. The statements in this paper are those of the authors and do not necessarily represent the views of RVO Health or the Quit For Life quitline program.