Impact of E-cigarette Device Warnings: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Impact of E-cigarette Device Warnings (NCT05714982), a randomized na study. The primary outcome is Mean Vaping Quit Intentions Score.

2. Clinical Question

The registry summary also frames a concern that warnings might push users toward smoking. The primary outcome is quit intentions, however, and the listed outcomes do not include a measure of switching to smoking. The posted primary results therefore address stated intentions to quit vaping, not whether participants quit or whether warnings affected smoking behavior.

3. Trial Design

The registry describes a randomized, parallel-group clinical trial with no masking. Participants were assigned to labels with text health warnings, labels with pictorial health warnings, or labels with neutral statements about vaping. Labels were applied to vaping devices and refills. The sponsor was UNC Lineberger Comprehensive Cancer Center, and the registry lists the trial as completed. The recorded start date is 2023-02-23 and the primary completion date is 2024-07-23.

Enrollment was 1369. The registry summary identifies the intended population as US adult vapers ages 21+ and says the trial expected ~1,200 participants to complete Visits 1, 3, and 5. The primary outcome assessments ran from post-labeling at Visit 1 through 4 weeks at Visits 2-5. This schedule makes the primary result an average across repeated survey time points rather than a result from only the last visit.

The registry lists the phase as NA. It does not report the randomization procedure, allocation concealment, recruitment setting, or the details of how labels were delivered beyond their application to devices and refills. It also does not report whether participants, staff, or analysts were blinded; masking is listed as none. These omissions limit assessment of how assignment was implemented and how awareness of label type may have affected survey responses. The registry record does not report a protocol rationale for the enrollment target or a sample-size justification.

4. Participant Flow and Accounting

The registry reports how participants moved through each stage of the study. All counts below are exactly as posted; missing counts are not inferred from other stages.

Overall Study

5. Baseline Characteristics

The table reproduces the baseline characteristics posted in the registry, which describe the participants analysed at baseline and help show whether the randomized groups were comparable.

6. Primary Endpoint

The registry lists the primary outcome as Mean Vaping Quit Intentions Score, measured over Up to 4 weeks.

The primary endpoint is the mean vaping quit intentions score, a score on a scale measured by 3 survey items at 5 time points: post-labeling at Visit 1 and at 1, 2, 3, and 4 weeks at Visits 2-5. Investigators averaged the items to create a score at each time point, then averaged those time-point scores. The measure ranges from 1 to 5, with higher values indicating higher quit intentions. Its timeframe is up to 4 weeks.

A repeated-measure average captures stated intentions throughout the observation period, rather than relying on a single assessment. It is a measure of intention, not observed quitting behavior. The endpoint description does not report an outcome for actual vaping cessation or smoking behavior, so the score cannot by itself establish whether participants changed either behavior. Nor does the average show whether intentions rose, fell, or remained stable across the individual visits; visit-specific results are not reported in the registry record.

The registry identifies the endpoint as a mean and reports standard deviations for the groups. These describe the average score and the spread of scores among participants, respectively. Interpretation therefore concerns group averages and their uncertainty, not a prediction of any individual participant's response.

7. Results Posted in the Registry

• Label With Pictorial Health Warning: mean 2.50 versus 2.50 on control; difference 0.00 (95% CI -0.16 to 0.16), standardized effect d = 0.00; z-test p = 1.000.
• Label With Neutral Statements About Vaping: mean 2.20 versus 2.50 on control; difference -0.30 (95% CI -0.45 to -0.15), standardized effect d = -0.26; z-test p = <0.001.

8. Secondary Efficacy Endpoints

The registry also posts results for secondary outcomes. For each outcome below, the endpoint type, the test applied to the posted summary statistics, the result and its interpretation are given in turn. Secondary endpoints are usually not adjusted for multiple testing, so they support rather than confirm the primary finding.

Mean Knowledge Score

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Label With Pictorial Health Warning: mean 0.60 versus 0.60 on control; difference 0.00 (95% CI -0.04 to 0.04), standardized effect d = 0.00; z-test p = 1.000.
  • Label With Neutral Statements About Vaping: mean 0.40 versus 0.60 on control; difference -0.20 (95% CI -0.24 to -0.16), standardized effect d = -0.67; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Label With Pictorial Health Warning and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Label With Neutral Statements About Vaping than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Mean Cognitive Elaboration Score

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Label With Pictorial Health Warning: mean 2.90 versus 2.70 on control; difference 0.20 (95% CI 0.07 to 0.33), standardized effect d = 0.22; z-test p = 0.002.
  • Label With Neutral Statements About Vaping: mean 2.00 versus 2.70 on control; difference -0.70 (95% CI -0.82 to -0.58), standardized effect d = -0.82; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval excludes no difference, so the mean is statistically higher for Label With Pictorial Health Warning than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Label With Neutral Statements About Vaping than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

Mean Perceived Likelihood of Illness Score

• Endpoint type: Continuous: a mean value per participant, compared as a difference in means.
• Test: Unadjusted two-sample z-test for the difference in means, using the posted standard deviations.
• Result:
  • Label With Pictorial Health Warning: mean 3.10 versus 3.10 on control; difference 0.00 (95% CI -0.10 to 0.10), standardized effect d = 0.00; z-test p = 1.000.
  • Label With Neutral Statements About Vaping: mean 2.90 versus 3.10 on control; difference -0.20 (95% CI -0.31 to -0.09), standardized effect d = -0.23; z-test p = <0.001.
• Interpretation:
  • The 95% confidence interval includes no difference, so these data do not show a statistically significant difference between Label With Pictorial Health Warning and the control. The interval gives the range of effects that remain compatible with the data.
  • The 95% confidence interval excludes no difference, so the mean is statistically lower for Label With Neutral Statements About Vaping than for the control. This is an unadjusted, secondary comparison, so it is suggestive rather than confirmatory.

The registry identifies secondary measures spanning knowledge, cognitive elaboration, perceived likelihood of illness, negative affect, attention, information seeking, conversations, and vaping reinforcement attitudes. It reports 16 secondary outcome results and 34 analyses overall, but the record details describe detailed quantitative comparisons for only some of those measures. Across the detailed comparisons available for knowledge and cognitive elaboration, the warning-label groups tend to have higher averages than the neutral-label group; the pictorial and text groups are similar for knowledge, while cognitive elaboration differs between them in the reported recalculation. These measures reflect responses associated with exposure to the labels, rather than demonstrated changes in vaping behavior.

This broad pattern is compatible with the primary result in the limited sense that warning labels are associated with differences in some reported responses compared with neutral statements. It does not establish a pathway from awareness or elaboration to quit intentions, because the registry does not report mediation analyses. Nor does it show that the secondary measures explain the primary difference. The remaining listed domains cannot be summarized as directions or magnitudes from the information reported in the registry record.

Secondary findings generally carry less evidential weight when many outcomes and comparisons are examined, particularly when a multiplicity strategy is not reported. The registry record does not report a correction for multiple testing, a hierarchy among secondary outcomes, or whether each secondary analysis was prespecified. An unadjusted p-value assesses a particular comparison without accounting for the broader set of tests. As a result, isolated secondary signals should be interpreted as exploratory unless a fuller analysis plan establishes otherwise. These results can describe patterns worth examining, but they should not be treated as independent confirmation of the primary finding.

Secondary outcomes registered for this trial, for reference (only those above were recalculated):

• Mean Knowledge Score
• Mean Cognitive Elaboration Score
• Mean Perceived Likelihood of Illness Score
• Mean Negative Affect Score
• Mean Attention Score
• Mean Information Seeking Score
• Mean Number of Conversations in Past Week
• Mean Vaping Reinforcement Attitudes Score

9. Statistical Interpretation

10. Statistical Methods Explained

The primary outcome is continuous: it is a mean score rather than an event count or time until an event. A mean difference expresses the contrast between group averages in the original score units. Here, negative values for neutral labels compared with text warnings indicate a lower mean score in the neutral-label group; a value of 0.0 for pictorial compared with text warnings indicates equal group means at the reported precision. The scale runs from 1 to 5, with higher scores representing higher quit intentions.

A standard deviation describes how dispersed participant scores are around a group's mean. It is not the uncertainty in the estimated mean difference. Standard errors quantify uncertainty in estimates, while confidence intervals combine an estimate and its standard error to show a range of values compatible with the data and the calculation's assumptions. A narrow interval indicates greater precision than a wide interval, though neither interval alone determines whether an effect matters in practice.

The posted recalculations are unadjusted comparisons based on group means, standard deviations, and group sizes. They do not reproduce all potential features of the sponsor's general linear model. The registry record does not report the model's covariates, its handling of repeated measurements, or other details needed to reproduce the sponsor's analysis. For that reason, the recalculated estimates are best read as transparent comparisons of posted summary statistics, not substitutes for the trial's planned or sponsor-conducted analysis. Since a discrepancy is flagged, the sponsor's own analysis takes precedence.

The p-value measures how incompatible the observed data are with a specified null hypothesis under the assumptions of the analysis. It is not the probability that the null hypothesis is true, and it does not quantify the size or importance of a difference. The posted sponsor p-values and the unadjusted recalculated p-values should not be conflated: they arise from different analysis approaches, and the record does not give enough model detail to determine precisely why some values differ.

The planning illustrations attached to recalculated comparisons use the observed effect as a starting point and consider stated fractions of that effect. They assume a two-sided significance level 0.05, equal allocation between two arms, and a normal approximation. They are illustrative planning estimates, not observed power and not the trial's own sample-size justification. They do not demonstrate that the trial was designed around the observed result or that any proposed future study would achieve the assumed operating characteristics in practice.

General statistical context: repeated measurements can improve description of a time course, but averaging across visits hides changes that may differ by visit. A model that uses repeated observations can account for their dependence, whereas a simple comparison of overall means does not describe that dependence. The registry record does not report enough analytic detail to establish how the sponsor's model handled the repeated observations, missing visits, or within-participant correlation.

General statistical context: random assignment supports a causal comparison when assignment is implemented as intended and outcome data are appropriately analyzed. It does not guarantee that every participant contributes complete data or that results apply beyond the studied population and setting. The record does not report analysis populations, participant flow by visit, or a detailed missing-data strategy. Those omissions matter because a mean based on observed responses can be affected if the availability of responses differs across groups or relates to the outcome.

The record reports 17 outcome measures and 34 analyses. A family of many tests creates more opportunity for chance findings than a single prespecified test. The registry record does not report a multiplicity adjustment or a hierarchy for interpreting outcomes. Consequently, unadjusted p-values for secondary comparisons warrant more caution than results from a prespecified primary comparison with a clearly reported analysis plan.

11. Confidence Intervals and Precision

For pictorial versus text warnings, the recalculated primary mean difference is 0.0, with a confidence interval from -0.156 to 0.156. The interval includes zero, and the posted group means are both 2.5. It conveys uncertainty around the difference rather than certainty that the groups are exactly equivalent. Establishing equivalence would require an explicitly defined margin and a suitable analysis; such a margin is not reported in the registry record.

For neutral labels versus text warnings, the recalculated mean difference is -0.3, with a confidence interval from -0.451 to -0.149. This interval is entirely below zero and is consistent with a lower average score in the neutral-label group in this unadjusted calculation. Its endpoints describe uncertainty in the estimated mean contrast under the calculation's assumptions; they do not describe the range of individual participants' scores.

Confidence intervals are useful because they show both direction and precision, unlike a p-value alone. Their interpretation depends on the analysis being used. The posted intervals are recalculated from summary statistics, while the registry's sponsor analyses are general linear models and do not report coefficient estimates or confidence intervals. Since the record flags a discrepancy, the recalculated intervals should not be treated as the sponsor's inferential intervals.

The registry record does not report a threshold for a clinically or behaviorally important change in the quit-intentions scale. Thus, even a precisely estimated difference cannot be classified as meaningful on the basis of statistical precision alone. The endpoint is an average across assessments up to 4 weeks, and no visit-specific confidence intervals are reported. Precision about the average does not resolve how the pattern may have varied over time.

12. Analyses Reported by the Sponsor

The registry reports the sponsor's primary comparisons using a general linear model. For text warnings versus neutral labels, the sponsor reports p <.001. For text versus pictorial warnings, the sponsor reports p = .78. The registry does not provide the model's estimated differences, confidence intervals, covariate specification, or enough detail to reproduce these analyses.

The recalculated unadjusted comparison of text warnings with neutral labels has a mean difference of -0.3, a confidence interval from -0.451 to -0.149, and p < 0.001. This is directionally consistent with the sponsor's reported evidence of a difference, although the estimates are produced by different methods. For text versus pictorial warnings, the recalculation gives a mean difference of 0.0, a confidence interval from -0.156 to 0.156, and p = 1.0; the sponsor reports p = .78. Both results do not indicate a clear difference at the posted precision, but the p-values are not identical.

The discrepancy flag means the recalculations should not displace the trial sponsor's own analysis. The sponsor's model-based result takes precedence for the trial's inferential conclusion. The unadjusted calculations remain useful as a check on the scale and direction of the posted group means, but they are not a reconstruction of the sponsor's model. Differences in model structure, handling of repeated measurements, or other analysis choices cannot be identified because those details are not reported in the registry record.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists 17 outcome measures and 34 analyses. It does not report a strategy for controlling the chance of false-positive findings across the set of outcomes and comparisons, nor does it describe a hierarchy that would specify which tests should be interpreted confirmatorily. The reported secondary p-values are therefore best read as unadjusted unless the registry record states otherwise. With multiple outcomes, a small p-value on an individual comparison does not by itself show that the overall evidence across outcomes is strong.

Important details about missing data are also not reported in the registry record. The summary states an expectation of ~1,200 participants completing Visits 1, 3, and 5, while the reported enrollment is 1369. These registry entries do not establish how many participants completed each visit, how many contributed to the endpoint average, or how missing visit responses were handled. The primary endpoint uses assessments across Visits 1-5, but the registry record does not report the analysis population, the number of observations at each visit, or whether participants with incomplete follow-up were included.

The record does not report whether the main analysis followed intention-to-treat principles, whether participants were analyzed according to randomized assignment, or whether exclusions were made. It also does not describe imputation, complete-case analysis, sensitivity analyses, or assumptions about why responses might be missing. These omissions limit assessment of whether the available scores represent the randomized groups comparably.

General statistical context: if missingness is associated with participants' intentions or with assigned labels, analyses based only on observed responses can be biased. Sensitivity analyses can show how conclusions might change under alternative assumptions, but their use is not reported in the registry record. The record also does not report interim analyses, stopping rules, or a detailed statistical analysis plan. These unknowns should be treated as unavailable information, not as evidence that a particular method was or was not used.

14. Safety Results

The registry reports serious adverse events as affected participants divided by participants at risk.

Deaths

Other (non-serious) adverse events

Participants with at least one non-serious event at or above the registry's 0% reporting threshold.

The adverse-event figures report zero serious adverse events, zero other adverse events, and zero deaths in each of the three label groups. The registry lists the at-risk group counts as 456 for text warnings, 457 for pictorial warnings, and 456 for neutral statements. No serious adverse-event terms are listed.

These registry figures mean that no events in those categories were recorded in the posted results for the listed groups. They do not establish that no participant experienced any symptom or unwanted effect, or that the labels have no safety implications. The record does not report the definitions used for serious and other adverse events, the collection procedures, the assessment period for adverse events, or detailed event-level data. It also does not report whether adverse events were actively solicited or collected in another way.

The trial's primary focus was survey measures of intentions and related responses, and the primary timeframe was up to 4 weeks. The adverse-event figures should therefore be interpreted as the registry's recorded counts, not as a comprehensive assessment of longer-term safety or of outcomes beyond the categories reported. A zero count is not a guarantee of absence of risk.

15. What the Result Does, and Does Not, Mean

The posted primary summaries show mean quit-intentions scores of 2.5 for text health warnings, 2.5 for pictorial health warnings, and 2.2 for neutral statements about vaping. In the recalculated unadjusted comparisons, the text and pictorial groups have a mean difference of 0.0, while the neutral-label group has a lower mean than the text-warning group by the reported estimate of -0.3. The sponsor's own model reports p <.001 for text versus neutral and p = .78 for text versus pictorial. Because the record flags a discrepancy, the sponsor's analysis takes precedence for inference.

The outcome is a survey measure of intention averaged across assessments up to 4 weeks. It is not a direct measure of quitting, reduced vaping, or switching to smoking. Although the registry summary states that the trial addresses concern about warnings pushing users toward smoking, the listed endpoints do not include a smoking outcome. The results cannot determine whether warnings caused changes in smoking behavior, or whether participants followed through on stated intentions.

The observed group averages do not establish that any label will have the same effect in other settings or populations. The registry describes US adult vapers ages 21+, but recruitment setting and other details about representativeness are not reported. No threshold for a meaningful change on the scale is reported, so statistical differences cannot be translated into a judgment about practical importance from this record alone.

The results are also limited by unreported details of the sponsor's model, missing-data handling, analysis populations, and multiplicity plan. The posted summary supports description of the recorded means and their recalculated uncertainty, but it cannot answer every causal, behavioral, or policy question posed by the broader trial motivation.

16. Why This Trial Matters Statistically

This trial illustrates why the outcome definition matters. A study can ask whether labels change intentions, while readers may be interested in actual behavior. The distinction is important: an average survey score is not equivalent to an observed quit attempt, sustained cessation, or change in smoking. The registry's motivation mentions avoiding an unintended shift toward smoking, but the listed outcomes do not report that behavior. Careful alignment between research questions and measured endpoints prevents a result on intentions from being overstated as a behavioral effect.

The comparison also shows the value of reporting effect estimates and confidence intervals alongside p-values. The unadjusted recalculation gives a score-scale contrast and an interval, which describe direction and precision more directly than a p-value alone. At the same time, a simple recalculation does not replace the sponsor's model. Without model details and model-based estimates, readers cannot fully reconcile the analyses. Transparent reporting of the analysis model, handling of repeated measures, and uncertainty would make the relationship clearer.

Finally, a broad outcome set raises questions about multiplicity and selective interpretation. The registry lists 17 outcome measures and 34 analyses, while the multiplicity plan is not reported. Distinguishing prespecified primary evidence from secondary exploratory patterns helps readers judge how much weight to give each result. The trial therefore provides a practical example of how outcome choice, precision, model transparency, and multiplicity jointly shape statistical interpretation.

17. Limitations

• The posted summary statistics are the only data these results rest on; patient-level data, full protocols and statistical analysis plans are not available.
• Recalculated results are simple, unadjusted calculations from posted summaries. They do not replace the sponsor's prespecified analysis.

Several limitations are specific to the information available for this trial. The primary result is based on a score of intentions averaged across five assessments, so the overall mean does not reveal whether responses changed from post-labeling through later visits. Visit-specific results and a time-course analysis are not reported in the registry record. Participants may also have had different numbers of completed assessments, but visit-level completion and the method for forming averages when data are incomplete are not reported.

The registry reports group means and standard deviations, alongside a sponsor analysis described as a general linear model. It does not report the model specification, whether covariates were included, or how within-participant repeated measures were handled. The recalculations are unadjusted and therefore cannot reproduce the sponsor's analysis. Because a discrepancy is flagged, the sponsor's own analysis takes precedence, but the lack of model estimates and details limits independent evaluation of that result.

The no-masking design means participants could be aware of the labels they received. Since the primary endpoint is self-reported intention, awareness of the message could influence responses. The record does not report procedures intended to address this possibility. The registry also does not report recruitment setting, participant flow, analysis populations, or a detailed missing-data plan, limiting assessment of generalizability and potential bias.

Finally, the registry's stated concern includes a possible shift toward smoking, but no such outcome appears in the listed measures. The primary endpoint covers intentions up to 4 weeks and does not establish behavior over a longer period. These limitations constrain what can be concluded from the posted results, even when the reported group averages are precisely described.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Impact of E-cigarette Device Warnings — NCT05714982.
• Registry record last updated 2025-09-12; status last verified 2025-08.
• Results first posted 2025-09-12.
• Data retrieved from ClinicalTrials.gov on 2026-11-09.
• Group sizes, means, event counts and sponsor-reported p-values are reproduced as posted. Effect estimates, confidence intervals and p-values not attributed to the sponsor are independently calculated from the posted summary statistics.

21. Record Summary

NCT05714982, titled Impact of E-cigarette Device Warnings, is a completed randomized parallel-group trial sponsored by UNC Lineberger Comprehensive Cancer Center. It enrolled 1369 US adult vapers ages 21+ and compared text health-warning labels, pictorial health-warning labels, and neutral statements applied to vaping devices and refills. Masking is listed as none.

The primary outcome was mean vaping quit intentions, assessed through survey items at post-labeling and through 4 weeks. The posted group means were 2.5 for text warnings, 2.5 for pictorial warnings, and 2.2 for neutral labels. The sponsor reports p <.001 for text versus neutral and p = .78 for text versus pictorial. Recalculated unadjusted comparisons are available, but the discrepancy flag means the sponsor's own analysis takes precedence.

The registry record does not report detailed model specifications, multiplicity handling, missing-data methods, or a smoking-behavior outcome. The posted results therefore describe short-term intentions and selected related survey measures, not actual quitting or switching to smoking.