Messages About Reduced Nicotine in Combusted Tobacco Products: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Messages About Reduced Nicotine in Combusted Tobacco Products (NCT05506046), a randomized na study. The primary outcome is Perceived Health Risk of VLNCs (Absolute).

2. Clinical Question

This is a question about responses to communication, not a direct test of cigarette composition, smoking-related health outcomes, or changes in smoking behavior. The study also included a follow-up survey after 2 weeks, described as assessing message recall and behaviors since exposure. The posted primary result concerns the immediate post-exposure measure, not that later survey. The registry record lists secondary outcomes related to trying reduced-nicotine cigarettes, switching intentions, perceived risks, policy support, message effectiveness, and smoking behavior, but does not provide results for those outcomes in the information.

3. Trial Design

The registry identifies NCT05506046 as a randomized, parallel-group online trial of messages about reduced nicotine in combusted tobacco products. Allocation was randomized, masking was listed as none, and the trial status is completed. The sponsor was Georgia State University. The record lists enrollment of 1901, a start date of 2022-10-12, and a primary completion date of 2023-02-28. The study was classified as phase NA.

There were four behavioral message conditions. The reduced-nicotine condition presented information about reduced-nicotine cigarettes. The modified-risk condition presented information about e-cigarettes. The combined condition covered reduced-nicotine cigarettes and e-cigarettes. The control condition presented bottled-water messages. The study description identifies smokers, dual-users of cigarettes and e-cigarettes, and young adult non-smokers aged 18-29 as the populations for the message-exposure experiment. The registry record does not report how participants were allocated across these population categories or whether randomization was stratified by them.

The primary measure was taken immediately after exposure. The brief summary also describes a follow-up survey after 2 weeks to assess recall and behavior since exposure, but no results from that follow-up are included in the posted primary result. Arm-specific analyzed counts for the primary outcome were 418 for reduced-nicotine messages, 427 for modified-risk messages, 420 for the combined messages, and 428 for control. The primary-outcome description says the analysis included participants in each condition who responded to that measure; the registry record does not give further details about exclusions or missing responses.

The record identifies the primary endpoint estimates as least-squares means with 95% confidence intervals. It does not report the statistical model, covariates, handling of any clustering, or the planned pairwise contrasts. The record also contains no posted between-group estimates for this continuous outcome. The message conditions and the outcome were measured in a communication experiment, so the reported result concerns the immediate rating recorded after exposure rather than a clinical health outcome.

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.

Before assignment: The panel sample (probability sample) included 971 adults who smoked exclusively, 240 adults who dual used cigarettes and e-cigarettes, and 458 young adults (18-29 years old) who never smoked. To augment the sample of adults who dual used, additional participants were recruited from the opt-in, non-probability sample (n=232) bringing the total sample size to 1901.

Overall Study

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: The number of participants analyzed included all participants in each condition who responded to the outcome measure.

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 Perceived Health Risk of VLNCs (Absolute), measured over Immediately after exposure.

The primary endpoint, Perceived Health Risk of VLNCs (Absolute), is a single-item rating of a participant’s perceived likelihood of overall harm to their health under the condition of smoking reduced-nicotine cigarettes every day. VLNCs refers in the endpoint label to very low nicotine cigarettes; the intervention descriptions refer to reduced-nicotine cigarettes. Responses ranged from 1, Not at all likely, to 5, Extremely likely, with a separate Don’t know response option. The endpoint was measured immediately after message exposure.

The measure is used to capture an immediate perception that may be relevant to how participants understand messages about reduced-nicotine cigarettes. It is not a measure of actual health effects, nicotine exposure, smoking cessation, or subsequent smoking behavior. The registry describes the outcome as a continuous measure and reports least-squares means with 95% confidence intervals. Its scoring description says higher scores represent a better outcome, while the item itself asks about perceived likelihood of harm. That scoring convention should be retained when reading the posted results; it does not turn a perception score into a measure of objective health benefit.

The additional Don’t know option is important to the interpretation because the scale’s numeric range does not include that response. The registry record does not report how Don’t know responses were handled in the analysis, how many participants selected it, or whether responses were transformed or excluded. It also does not report reliability or validation information for this single item. The primary result therefore describes the reported score among respondents included in each condition, rather than a complete account of every response category or of participants’ underlying health risk.

7. Why No Re-analysis Is Shown

8. Statistical Interpretation

9. Statistical Methods Explained

What does a least-squares mean represent?

In general, a least-squares mean is a model-based estimate of a group mean. It can differ from a simple average when a fitted model adjusts for other variables or uses a specified structure. The registry record identifies the reported statistic as a least-squares mean, but the model specification and any adjustment variables are not reported in the registry record. The posted numbers should therefore be read as the reported model-based group estimates, without assuming which features of the data were adjusted for.

The endpoint is a score on a scale, and the record classifies it as continuous. Reporting a mean gives a measure of the central level of scores in a group. A mean does not describe the full distribution: participants may give a range of responses, and the same mean can arise from different patterns of individual scores. The registry record does not report standard deviations, score distributions, or the number of Don’t know responses. Those details would help describe variation and response patterns beyond the mean.

How should the confidence intervals be read?

In general, a confidence interval communicates the precision of an estimate under the statistical procedure used. A narrower interval indicates greater precision for the estimated quantity, while a wider interval indicates more uncertainty. The intervals reported for this endpoint apply to each arm’s least-squares mean. They do not directly estimate the difference between two arms. A comparison requires an estimate of that contrast and its uncertainty, taking account of the analysis model and the relationship between the estimates.

For example, the reported interval for the control mean answers a different statistical question from an interval for the mean difference between control and a message condition. The latter is not reported in the registry record. No p-value for a group comparison is reported either. As a result, the arm-level confidence intervals should not be treated as formal tests of the trial’s comparative question.

Why does randomization matter?

In general, random assignment is used to make comparison groups more alike, on average, with respect to factors that could affect the outcome. It supports interpretation of group contrasts as effects of assigned conditions when allocation is implemented as intended and the analysis appropriately follows the design. Here, the registry describes randomized allocation and four parallel message conditions. However, the posted primary results do not include the between-group contrasts needed to quantify comparative effects. The record also does not report allocation procedures, concealment details, or checks of baseline balance. Randomization is part of the design; it does not by itself supply a missing comparison estimate.

What does the analysis population mean?

The primary outcome description states that the analyzed participants in each condition were those who responded to the outcome measure. The arm-specific analyzed counts were 418, 427, 420, and 428, respectively. This indicates that the reported outcome is based on respondents for that measure, rather than necessarily on every person enrolled. The record does not report the number of missing outcome responses, reasons for missingness, or whether any imputation was used. General statistical practice distinguishes analysis of observed responses from methods that model or impute missing values; the record does not identify such a method.

The separate Don’t know response option also raises an analysis question: it is not part of the numeric response range from 1 to 5, and the registry record does not state how it was treated. Without that information, the exact relationship between all recorded responses and the analyzed means cannot be fully assessed.

What can planning estimates tell readers?

In general, sample-size planning asks how much information a future study would need to detect an effect under specified assumptions. Planning illustrations are not observed results and are not observed power. The registry record describes illustrations based on the observed effect, with a two-sided significance level of 0.05, equal allocation between two arms, a normal approximation, and planning for a stated proportion of the observed effect. It does not report a numerical set of planning estimates in the information available for this outcome, and there are no posted two-arm comparison estimates from which to interpret such scenarios.

These assumptions describe an illustrative calculation, not necessarily the design rationale used before the trial. The registry record does not report the trial’s sample-size justification, planned effect size, or statistical plan. The observed group means should not be used to infer that the study had a particular power to detect a contrast.

Why does over-dispersion matter in general?

Over-dispersion occurs when the observed variability in data is greater than a model assumes. In general, failing to account for it can make uncertainty estimates too small and statistical conclusions too confident. It is especially relevant for some count or clustered outcomes. The primary endpoint is a continuous score, and the registry record does not report a model or evidence of over-dispersion. This general issue cannot be used to infer that the reported intervals are affected; evaluating it would require analysis details not reported in the registry record.

10. Multiplicity, Missing Data and Analysis Populations

The registry lists the primary perceived-health-risk outcome and eight secondary outcomes. The secondary measures cover interest in trying reduced-nicotine cigarettes, intention to switch to electronic nicotine products, perceived addictiveness, comparative risk perceptions, policy support, perceived message effectiveness, absolute perceived health risk at post-test, and smoking behavior. The record reports zero posted analyses and gives no results for the secondary outcomes in the information.

When a trial assesses multiple outcomes or makes multiple comparisons, multiplicity can increase the chance of finding at least one apparently notable result by chance. General approaches include prespecifying a primary comparison, adjusting inference, or treating secondary analyses as exploratory. The registry record does not report a multiplicity procedure or a full statistical analysis plan. It also does not identify which pairwise contrasts were planned. Without that information and without posted comparison estimates, there is no basis to infer whether or how multiplicity was addressed.

Missing-data reporting is also limited. The primary outcome population is defined as respondents to the measure within each condition, and the analyzed counts are reported by arm. The record does not report the number of participants with missing primary responses, reasons for missingness, or whether any missing-data method was applied. It does not state how the Don’t know option was handled. These omissions limit assessment of whether the analyzed responses represent all participants assigned to each condition.

The brief summary describes a follow-up survey after 2 weeks, but the primary result reported is immediate after exposure. The registry record does not report follow-up completion counts or missingness for the later survey. Consequently, the immediate endpoint cannot answer whether responses or behaviors changed at follow-up.

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

The result describes how respondents in each randomized condition rated the perceived likelihood of overall health harm from smoking reduced-nicotine cigarettes every day, immediately after message exposure. The reported group means range from 3.81 for modified-risk messages to 4.09 for reduced-nicotine messages, with the corresponding 95% confidence intervals reported for each group. These are estimates of a self-reported perception score, not measurements of actual health effects.

The findings do not establish that one message condition produced a higher or lower score than another. Although the study was randomized, the registry record does not report between-group estimates or tests for the primary outcome. The arm means and their separate intervals are not enough to quantify a comparative effect or its precision. No conclusion about statistical significance should be drawn from these arm-level values alone.

The endpoint’s scoring description says higher scores represent a better outcome, while the question asks respondents to rate the likelihood of harm. That wording and scoring convention should be kept in view. The result is not an objective assessment of whether reduced-nicotine cigarettes are more or less harmful, and it does not show that participants’ beliefs were accurate.

The record’s descriptions of the other measures do not establish results for them. The primary endpoint does not demonstrate changes in smoking, interest in trying a product, switching intentions, or any health outcome. The brief summary mentions a follow-up survey after 2 weeks, but the result is measured immediately after exposure and does not report later outcomes.

Finally, the registry’s analyzed population consists of respondents to the endpoint in each condition. Without further information about missing responses, the model, recruitment, and the handling of Don’t know answers, the estimates have limits in how fully they characterize all enrolled participants and the broader populations named in the study description. These limits do not negate the reported scores; they define what can be concluded from them.

12. Why This Trial Matters Statistically

This trial illustrates a distinction that is central to interpreting randomized studies: arm-level summaries describe the observed outcome within each condition, while the causal question usually concerns a contrast between conditions. Randomization can support a fair comparison, but readers need the estimated contrast and its uncertainty to assess how the conditions compare. The registry reports arm-specific least-squares means and confidence intervals for the primary outcome, but no between-arm estimate. This makes the difference between descriptive results and comparative inference especially clear.

The endpoint also demonstrates why measurement and interpretation need to be considered together. A single-item score can provide a concise measure of an immediate perception, but it does not capture every aspect of beliefs or behavior. The item asks about perceived health harm, while the scoring description identifies higher values as a better outcome. Careful interpretation must retain the endpoint’s wording and the stated score direction rather than treating the score as an objective measure of health.

The study includes several secondary outcomes and a planned follow-up survey, making transparent reporting of analysis choices particularly useful. The registry record does not report the statistical model, multiplicity approach, missing-data handling, or comparative estimates. Those details would help readers evaluate precision, the scope of the findings, and whether conclusions are tied to prespecified questions.

More broadly, the record shows why an enrollment figure and a randomized design do not by themselves answer every question about evidence strength. The relevant population for this endpoint consists of respondents with a recorded outcome, and the primary result concerns an immediate rating. The results are therefore most informative when kept at that level: reported perceptions in the trial conditions, with comparisons and later outcomes not reported in the registry record.

13. 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.

The primary outcome is an immediate, single-item self-report about perceived likelihood of harm. It does not measure actual health effects or establish whether participants’ perceptions correspond to health risks. The endpoint is a score from a bounded response scale with an additional Don’t know option; the registry record does not explain how that option was handled or report its frequency.

Only arm-specific means and confidence intervals are reported for the primary outcome. Between-group estimates and p-values are not reported in the registry record, limiting direct assessment of the randomized comparisons. The model used to calculate least-squares means, including any covariates or other adjustments, is also not reported. This makes it difficult to evaluate how the estimates were derived beyond the statistic named in the record.

The analysis population included participants in each condition who responded to the measure. The record gives the analyzed count for each arm but does not report the extent or reasons for missing responses, missing-data methods, or the handling of Don’t know answers. These details matter when assessing how the analyzed respondents relate to participants assigned to each condition.

The trial description concerns online message exposure and names smokers, dual-users, and young adult non-smokers aged 18-29. The record does not report recruitment procedures, response rates, or details supporting the description of nationally representative samples. It also does not report subgroup results, so the primary estimates cannot be attributed separately to those population categories.

The brief summary describes a 2-week follow-up, but follow-up outcome results are not reported with the primary result. The record also does not report the statistical analysis plan, multiplicity procedures, or a sample-size justification. These omissions constrain interpretation of the broader set of trial questions beyond the immediate primary measure.

14. Related Tutorials

15. Related Statistical Calculators

16. Sources

• ClinicalTrials.gov: Messages About Reduced Nicotine in Combusted Tobacco Products — NCT05506046.
• Registry record last updated 2024-08-28; status last verified 2024-08.
• Results first posted 2024-08-28.
• 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.

17. Record Summary

NCT05506046, Messages About Reduced Nicotine in Combusted Tobacco Products, is a completed randomized, parallel-group online message-exposure trial sponsored by Georgia State University. The registry lists enrollment of 1901, no masking, and four behavioral conditions: reduced-nicotine messages, modified-risk messages about e-cigarettes, combined messages, and bottled-water control messages.

The primary endpoint was an immediate post-exposure score for perceived absolute health risk from smoking reduced-nicotine cigarettes every day. The registry reports least-squares means and 95% confidence intervals for each arm, based on respondents to the outcome measure. It does not report primary between-group estimates, p-values, model details, or results for the listed secondary outcomes. A 2-week follow-up survey is described, but follow-up results are not reported in the primary results. Missing-data handling and the treatment of the Don’t know response are not reported in the registry record.