Electronic Nicotine Delivery Systems for Smoking Cessation

Electronic nicotine delivery systems (ENDS) replicate many features of combustible cigarettes and are used as cessation aids. Prior randomized evidence suggests ENDS may be more effective than nicotine replacement therapy (NRT) for smoking cessation, but there is limited evidence comparing ENDS to standard-of-care (SOC) counseling that includes optional pharmacotherapy, and limited safety data from rigorously collected adverse events (AE) and serious adverse events (SAE). It is also unclear whether quitting with ENDS alleviates smoking-related respiratory symptoms and how ENDS affect exposure to nicotine and other tobacco toxicants. The current study (ESTXENDS) evaluates whether adding ENDS to SOC increases 6‑month tobacco abstinence compared with SOC alone and explores safety and respiratory outcomes.

A prior sufficiently powered RCT and a systematic review found ENDS more effective for smoking cessation than NRT (e.g., Hajek et al., 2019; Cochrane Review 2022). However, evidence versus SOC counseling and robust safety assessments (AE/SAE collected with predefined procedures and adjudication) remains limited. Earlier trials seldom systematically collected safety outcomes or verified them via chart review. ENDS generally deliver lower levels of toxicants than cigarettes, but few RCTs have evaluated reductions in nicotine and smoke-related toxicant exposure using urinary biomarkers. There is also uncertainty about the effect of ENDS-assisted cessation on respiratory symptoms such as cough and phlegm, which typically improve after quitting conventional cigarettes. This trial addresses these gaps by comparing ENDS plus SOC versus SOC alone with predefined efficacy and safety endpoints and by measuring respiratory symptoms.

Design and setting: Two-group, open-label, randomized controlled trial conducted at five sites across Switzerland. Recruitment occurred July 2018–June 2021 via media, healthcare facilities, and public transport advertisements.

Participants: Adults (>18 years) who smoked ≥5 cigarettes/day for ≥12 months and were willing to quit within 3 months. Exclusions: pregnancy or breastfeeding; use of NRT or other cessation medications in prior 3 months; regular ENDS or heated tobacco use in prior 3 months. Ethics approvals obtained (2017-02332). Trial registered: NCT03589989.

Randomization and blinding: Centralized computerized 1:1 randomization at the Clinical Trials Unit, Bern. Open-label for participants and site staff; laboratory personnel measuring urinary biomarkers were blinded. SAE adjudication was independent; a DSMB oversaw safety procedures.

Interventions:

• Control (SOC only): Standard smoking cessation counseling based on cognitive behavioral therapy, motivational interviewing, and shared decision-making for drug support (including NRT), delivered in-person at baseline and by phone at target quit date (TQD) and Weeks 1, 2, 4, and 8. Participants received a CHF 50 voucher to purchase NRT.
• Intervention (ENDS + SOC): SOC as above plus two ENDS starter kits (Innokin Endura T20-S) and 5 spare 0.8-ohm coils. Participants could choose among 6 flavors and 4 nicotine concentrations (19.6, 11, 6, 0 mg/ml) of e-liquids (76/24 propylene glycol/vegetable glycerin; medical-grade free-base nicotine; alcohol; flavorings). At baseline, participants sampled all 24 combinations, selected preferences, received up to 10 bottles, and could reorder e-liquids ad libitum for 6 months. Training on device use and maintenance was provided.

Procedures and follow-up: Baseline visit was scheduled about one week pre-TQD. A 6-month in-person visit was planned; if missed, data were collected via phone, mail, or email. With prior consent, nurses could contact up to two relatives and the participant’s GP after three failed contact attempts to ascertain smoking status and (S)AEs.

Measures: Baseline and 6-month assessments included demographics, smoking history, expired carbon monoxide (CO), respiratory symptoms via the COPD Assessment Test (CAT, 0–40, higher=worse), withdrawal symptoms, and (S)AEs. Participants provided first-morning urine for biomarker analyses.

Outcomes:

• Primary: Continuous abstinence from TQD to 6 months, self-reported and biochemically validated by urinary anabasine <3 ng/ml; if anabasine unavailable, by CO ≤9 ppm. Participants with withdrawal, loss to follow-up, or missing biochemical validation were considered non-abstinent in primary analysis.
• Secondary abstinence outcomes: Continuous abstinence without biochemical validation; sustained abstinence allowing a 2-week grace period; sustained abstinence allowing up to 5 cigarettes total; 7-day point-prevalence abstinence at 6 months with and without biochemical validation. Additional secondary/exploratory outcomes included (S)AEs, antibiotic use, respiratory and withdrawal symptoms, and self-reported use patterns of tobacco, ENDS, and NRT.

Exposure categories at 6 months (self-report): tobacco abstainers; tobacco and ENDS abstainers; nicotine abstainers (no tobacco, ENDS, or NRT); exclusive ENDS users; dual users (tobacco + ENDS); exclusive smokers.

Sample size: Planned N=1114 for 90% power (alpha 0.05) to detect a rise in 6-month abstinence from 12% to 19% (RR 1.6), assuming 5% loss to follow-up and 5% crossover to ENDS in control, increased to target 1173; ultimately randomized 1246.

Statistical analysis: Log-binomial regression estimated risk ratios (RR) for abstinence outcomes. Sensitivity analyses included multivariable adjustment (pre-specified covariates: site, age, gender, employment, education, age started smoking, cigarettes/day, prior quit attempts, Fagerström score) and stabilized inverse probability of censoring weights (IPCW) to assess missing data impact. A tipping point sensitivity analysis was performed for missing primary outcomes. Between-group differences in AE/SAE and antibiotic use were estimated; CIs for secondary and exploratory outcomes were not adjusted for multiplicity. Software: Stata 17; tipping point analyses in R 4.3.1.

Participants: 1246 randomized (622 intervention; 624 control). Data on smoking status and SAE at 6 months available for 90.8%.

Primary outcome: Continuous biochemically validated 6‑month tobacco abstinence: 28.9% (180/622) intervention vs 16.3% (102/624) control; RR 1.77 (95% CI, 1.43–2.20); absolute difference 12.6% (95% CI, 8.0–17.2). Sensitivity analyses (adjusted, IPCW) were similar (e.g., adjusted RR 1.71 [1.39–2.12]).

Secondary abstinence outcomes (selected):

• Continuous abstinence without biochemical validation: 38.1% vs 23.4%; RR 1.63 (1.37–1.94); absolute difference 14.7% (9.6–19.8).
• Sustained abstinence (2-week grace), validated: 30.7% vs 17.6%; RR 1.74 (1.42–2.14); absolute difference 13.1% (8.4–17.8).
• Sustained abstinence (≤5 cigarettes total), validated: 35.2% vs 17.5%; RR 2.02 (1.65–2.46); absolute difference 17.7% (12.9–22.5).
• 7-day point-prevalence abstinence, validated: 39.4% vs 21.3%; RR 1.85 (1.54–2.21); absolute difference 18.1% (13.1–23.1).
• 7-day point-prevalence abstinence, self-report: 53.4% vs 32.1%; RR 1.67 (1.45–1.91); absolute difference 21.3% (16.0–26.7).

Use patterns at 6 months (self-reported): Among respondents (552 intervention; 504 control):

• Tobacco abstainers: 59.6% intervention vs 38.5% control.
• Nicotine abstainers (no tobacco, ENDS, or NRT): 20.1% intervention vs 33.7% control.
• Exclusive ENDS users (no tobacco): 48.4% intervention vs 3.0% control (with nicotine in ENDS: 39.3% vs 2.0%).

Adherence at 1 week post-TQD: In intervention, 95.9% used ENDS; 6.8% NRT; 0.5% other cessation meds. In control, 3.9% used ENDS; 63.6% NRT; 4.1% other meds.

Safety: One death in control. SAE: 4.0% (25/622) intervention vs 5.0% (31/624) control; RR 0.81 (95% CI, 0.48–1.35), p=0.49. AE: 43.7% (272/622; 425 events) intervention vs 36.7% (229/624; 366 events) control; RR 1.19 (95% CI, 1.04–1.37), p=0.01. Antibiotic use: 8.7% vs 6.9%; RR 1.26 (95% CI, 0.86–1.85). COVID-19 reported by 18 vs 8 participants (1 hospitalization in control).

Respiratory symptoms: CAT total mean 4.8 (SD 3.9) intervention vs 5.7 (SD 4.5) control; adjusted mean difference −0.66 (95% CI, −1.13 to −0.18). Item-level improvements favored intervention for cough and phlegm (e.g., no phlegm: 62% vs 51%).

Adding ENDS to standard counseling increased 6‑month tobacco smoking abstinence compared with SOC alone, addressing the primary question of cessation efficacy versus usual care. Although the relative effect aligns with prior trials, higher baseline abstinence rates in both groups yielded a larger absolute difference than some earlier studies. Many abstinent smokers continued ENDS use, particularly with nicotine, indicating that ENDS plus SOC may suit smokers aiming to cease tobacco but not necessarily nicotine. Safety findings showed more AEs but no increase in SAEs in the ENDS group over 6 months; pooling with other RCTs is warranted to refine safety estimates. Respiratory symptoms modestly improved, consistent with expected benefits of tobacco abstinence. Overall, the results support ENDS as an effective adjunct to counseling for increasing tobacco abstinence, while highlighting ongoing nicotine use and the need for longer-term follow-up to assess sustained cessation, nicotine use trajectories, and safety.

In this multicenter Swiss RCT, providing ENDS and e-liquids in addition to standard smoking cessation counseling significantly increased biochemically validated 6‑month tobacco abstinence compared with SOC alone. ENDS use was common among abstinent participants in the intervention arm, suggesting utility for smokers aiming to quit tobacco even if they continue nicotine. Adverse events were more frequent with ENDS, but serious adverse events were not increased over 6 months. Longer-term follow-up (planned at 12, 24, and 60 months) is needed to evaluate durability of abstinence, patterns of nicotine use, and safety. Future research should compare strategies to transition ENDS users to full nicotine abstinence and further quantify safety outcomes across diverse settings.

• Open-label design with unblinded allocation may have influenced participant behavior and perceptions; potential disappointment in control was mitigated by an NRT voucher but not directly assessed.
• Differential provision of products: free ENDS and e-liquids for intervention versus only a voucher for NRT in control; the trial compared ENDS+SOC to SOC, not ENDS versus NRT per se.
• Assessment reflects end-of-treatment at 6 months; durability beyond 6 months is unknown pending planned longer follow-up.
• Missing data and attrition: higher attrition for biochemical validation and in the control arm; primary analyses classified missing outcomes as non-abstinent per guidelines, which could overestimate between-group RR; tipping point analyses suggested conclusions likely robust.
• Generalizability may be limited to ambulatory healthcare settings in Switzerland.
• Multiplicity: confidence intervals for secondary outcomes were not adjusted for multiple comparisons.
• The trial was not powered to detect differences in serious adverse events, limiting safety inferences.