Effect of Smoking Exposure on Smoking Relapse Following Brief Abstinence - 2: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Effect of Smoking Exposure on Smoking Relapse Following Brief Abstinence - 2 (NCT00136838), a randomized na study. The primary outcome is Cigarette Choice After 3 Day Abstinence.

2. Clinical Question

This question concerns a short-term, laboratory measure of cigarette choice following cue exposure. It is not a question about long-term smoking behavior, sustained abstinence, or treatment outcomes. The registry describes the study as developing a laboratory model of relapse to cigarette use among nicotine-dependent volunteers, with an aim of supporting early-stage testing of treatments for nicotine addiction. The primary outcome recorded for this comparison is the number of choices, rather than a count of cigarettes smoked over a longer period or a clinical diagnosis.

The key contrast is the mean number of choices after active cues compared with the mean after neutral cues. Because participants received both conditions, the study uses a crossover structure rather than assigning each participant to experience only one cue condition. The registry record does not report participant characteristics beyond the description of nicotine-dependent volunteers, and it does not give further detail about the interval between conditions or the study's eligibility criteria.

3. Trial Design

This was a randomized, double-masked, crossover study of neutral and active cigarette cues. The registry lists the intervention as behavioral and describes two sequences: neutral cue followed by active cue, and active cue followed by neutral cue. Every participant received both interventions, with sequence assigned at random. The design therefore permits comparison of the two cue conditions within participants, although the registry record does not report whether the posted comparison used a paired or other crossover-specific analysis.

The registry records enrollment of 12 and a completed study. It lists the condition as tobacco use disorder. The phase is listed as not applicable. The sponsor was New York State Psychiatric Institute. The study began in October 2001, and primary completion was recorded in June 2006.

The laboratory session followed 3 days of abstinence. During the Day 4 experimental session, participants could choose cigarettes at 30-minute intervals, with a maximum of 6 choices. The primary outcome is presented as a continuous measure: the mean number of cigarette choices. The posted condition summaries give a mean and standard error for active cues and for neutral cues. The reported comparison is the mean difference between neutral and active cue conditions.

The crossover design is important to interpretation because the same participants experienced each cue condition. A crossover comparison can account for stable differences between participants when the analysis uses within-participant information. However, the analysis method, treatment of period or sequence effects, and handling of any carryover between conditions are not reported in the registry record. The record also does not state who was masked or how masking was maintained. Those missing details limit assessment of the analysis and conduct, rather than changing the definition of the randomized cue sequences.

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.

First Intervention ( 5 Days)

Second Intervention ( 5 Days)

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 Cigarette Choice After 3 Day Abstinence, measured over During Day 4 experimental session.

The primary endpoint is the number of cigarette choices made during the Day 4 experimental session after 3 days of abstinence. Participants had the option to smoke cigarettes at 30-minute intervals, up to 6 choices. The outcome therefore records how many opportunities to choose a cigarette were taken during a specified laboratory session. The registry summarizes this continuous measure using a mean and standard error for each cue condition.

This measure is used as a laboratory indicator of cigarette choice following cue exposure. In the study's stated purpose, the investigators sought to develop a laboratory model of relapse to cigarette use among nicotine-dependent volunteers, with the broader aim of supporting early-stage treatment testing. The endpoint captures behavior in that experimental setting; it does not directly measure relapse over a longer period, ongoing smoking, or abstinence after the session.

The outcome is bounded by the session design: a participant could make no more than 6 choices. A mean provides a group-level description of the number of choices, while the standard error describes the uncertainty in that estimated mean. Neither summary shows the full distribution of individual choices. The registry record does not report additional details about how choices were recorded or whether participants had further rules or conditions governing each choice.

7. Results Posted in the Registry

• Neutral Cue: mean 3.67 versus 2.50 on control; difference 1.17 (95% CI -0.51 to 2.85), standardized effect d = 0.56; z-test p = 0.172.

8. Statistical Interpretation

9. Statistical Methods Explained

The endpoint is a continuous measure summarized by a mean. A mean describes the average number of choices in a condition, while individual participants may have made fewer or more choices than that average. The registry gives the mean as 2.5 for the active-cue condition and 3.67 for the neutral-cue condition. Because the choices are limited by the session design, the outcome is bounded between 0 and 6.

What does the standard error describe?

The standard error describes the estimated uncertainty in a sample mean, rather than the spread of individual observations. A smaller standard error generally indicates a more precisely estimated mean, while a larger one indicates greater sampling uncertainty. The registry reports standard errors of 0.63 for active cues and 0.58 for neutral cues. These summaries do not reveal the range or distribution of individual choices, and they are not standard deviations. In a crossover study, the uncertainty of a condition contrast can depend on the within-participant relationship between the two measurements, not only on the two condition-specific standard errors.

How should the mean difference be read?

The posted mean difference is 1.17, with the contrast defined as neutral cue minus active cue. A positive value means the posted neutral-cue mean is larger than the active-cue mean. The contrast puts the two condition summaries on the same outcome scale, making the direction and size of the observed difference easier to interpret. It remains a sample estimate, and uncertainty around it is represented by its confidence interval.

What does the confidence interval add?

In general, a confidence interval is a range of values compatible with the data and the statistical procedure used to construct it. It conveys information about uncertainty that a single estimate cannot show. The interval reported for this contrast includes negative and positive values, so it includes both directions of difference as well as no difference. The registry record does not identify the method used to calculate the interval. Since the study is crossover, the choice of analysis and its handling of paired observations can affect the interval's interpretation.

What does the p-value tell the reader?

The posted unadjusted p-value is 0.172. In general, a p-value describes how surprising data at least as incompatible with a specified null hypothesis would be under that hypothesis and the model used for the test. It is not the probability that the null hypothesis is true, nor a measure of effect size or practical importance. The record labels this p-value unadjusted; it does not report the test used or whether any other analyses were considered. The p-value should therefore be read alongside the estimate and interval rather than as a stand-alone verdict.

Why does the crossover structure matter?

In a crossover study, participants receive multiple conditions, so measurements from the same person are related. A within-participant analysis can use this relationship and account for stable characteristics shared across a participant's measurements. The degree and direction of that relationship influence the uncertainty of a paired contrast. General statistical methods for crossover studies may also consider period and sequence effects or carryover. The registry record does not report which analysis was used or how these issues were handled. It would therefore be inappropriate to infer a particular model from the design label alone.

What can over-dispersion mean for a count-like endpoint?

Although this endpoint is analyzed as a continuous measure, it records a bounded count of choices. In general, over-dispersion means that observations vary more than a statistical model expects. When a count model is used, unaccounted-for over-dispersion can make standard errors too small and intervals too narrow, which can overstate precision. The record does not report a count model, a dispersion assessment, or a distributional diagnostic. That general concern does not establish that over-dispersion occurred in this study; it explains why the chosen model and its assumptions matter.

The registry identifies a mean comparison but does not report a statistical analysis plan, model specification, covariate adjustment, or method for calculating uncertainty. The results can be described as posted, but a full assessment of model adequacy and crossover-specific analysis is not possible from the registry record.

10. Confidence Intervals and Precision

The reported mean difference is 1.17 choices, comparing neutral cues with active cues. Its confidence interval extends from a negative value to a positive value, so the data are compatible with a contrast in either direction and with no difference. The interval is the main direct indication of precision for the reported contrast; its span cautions against treating the point estimate as a precise prediction of the underlying difference.

A confidence interval is not a range containing a fixed proportion of individual participants' outcomes. It represents uncertainty about an estimated contrast under the method used to calculate it. The registry record does not report the interval construction method, so the assumptions behind its exact coverage cannot be assessed. In particular, the crossover design makes the handling of within-participant observations relevant to standard errors and confidence limits.

The unadjusted p-value of 0.172 is consistent with uncertainty about the contrast, but it does not replace the interval or quantify its practical importance. A result that does not meet a conventional significance threshold does not show that the conditions are identical. Conversely, the point estimate alone does not establish a real difference.

The condition means have standard errors of 0.63 for active cues and 0.58 for neutral cues. These describe uncertainty in the separate means, not the full uncertainty of the within-participant difference. The registry record does not report the correlation between measurements, participant-level data, or the analysis details needed to independently evaluate the precision of the contrast. The observed summary therefore supports a cautious reading: the direction of the point estimate is clear, while the interval leaves substantial uncertainty about the underlying difference.

11. Planning a Similar Trial: Sample Size

This is a planning illustration, not a re-evaluation of the original trial. It asks: if a new study wanted to detect a difference like the one observed, or a smaller one, how large would it need to be? It is deliberately not “observed power” of the finished trial, which is not a meaningful quantity.

The planning figures are participants per arm, not total events. They are illustrations based on the observed effect, not the trial's original design assumptions. The planning assumptions are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is specified as a fraction of the observed effect.

For an assumed effect equal to 100% of the observed effect, the illustration gives 51 participants per arm for 80% power and 68 participants per arm for 90% power. For planning at 75% of the observed effect, it gives 90 participants per arm for 80% power and 121 participants per arm for 90% power. For planning at 50% of the observed effect, it gives 202 participants per arm for 80% power and 271 participants per arm for 90% power.

In plain terms, the 80% power column gives the illustrated number of participants per arm for a design with an 80% chance of detecting the chosen effect under the stated assumptions. The 90% power column gives the corresponding number for a 90% chance. These are prospective planning concepts, conditional on the assumptions and target effect; they are not estimates of the probability that this completed study detected an effect.

Smaller target effects require more participants under these assumptions, because a subtler difference is harder to distinguish from random variation. The planning figures are based on fractions of the observed effect, which itself is uncertain. They should not be interpreted as evidence that the observed effect is stable or clinically important. This is not observed power. The registry record does not report the original sample-size justification or the trial's own design assumptions.

A curve expressing the same planning relationship is read by matching a target effect size with the corresponding participant requirement: moving toward smaller target effects means more participants are needed. The figures assume the stated equal-allocation, two-arm setup and normal approximation. They do not describe the crossover allocation in detail, and the registry record does not explain how the planning illustration relates to the study's randomized sequences.

12. How the Sample Size Is Derived

Power is the probability that a study of a given size rejects the null hypothesis when a specific true effect exists. It depends on four things: the effect to detect, the variability of the outcome, the significance level (here a two-sided α of 0.05, so z_1−α/2 = 1.960) and the desired power (z_1−β = 0.842 for 80% and 1.282 for 90%).

Formula

n is the number of participants in each arm, Δ is the difference in means to detect and σ is the common standard deviation.

Worked calculation for this trial

1. Observed difference Δ = 1.170; pooled standard deviation σ = 2.098.
2. n = 2 × (1.96 + 0.842)² × 2.098² / 1.170² = 50.5, rounded up to 51 participants per arm for 80% power.

The smaller-effect rows in the table above repeat the same calculation with the effect multiplied by 0.75 and 0.5 (for ratios, applied to the logarithm of the ratio).

R code

alpha <- 0.05
z_a   <- qnorm(1 - alpha / 2)          # two-sided
delta <- 1.17                      # observed difference in means
sigma <- 2.09757002266909                     # pooled standard deviation
n_per_arm <- function(power, frac = 1) {
  ceiling(round(2 * (z_a + qnorm(power))^2 * sigma^2 / (delta * frac)^2, 6))
}
for (f in c(1, 0.75, 0.5))
  cat(sprintf("%3.0f%% of observed effect: n per arm = %s (80%% power), %s (90%% power)\n", 100 * f,
              format(n_per_arm(0.80, f), big.mark = ","), format(n_per_arm(0.90, f), big.mark = ",")))

# Check against R's built-in function (t-based, so very slightly larger)
cat("power.t.test:", round(power.t.test(delta = delta, sd = sigma, power = 0.80)$n, 1), "\n")

# Simulation: exact two-sample t-test using simulated means and variances
set.seed(136838)
n    <- n_per_arm(0.80)
reps <- 20000
m_diff <- rnorm(reps, delta, sigma * sqrt(2 / n))             # difference in sample means
s2     <- sigma^2 * rchisq(reps, 2 * (n - 1)) / (2 * (n - 1))   # pooled sample variance
tstat  <- m_diff / sqrt(s2 * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(tstat) > qt(0.975, 2 * (n - 1))), reps))

Output from running the code

100% of observed effect: n per arm = 51 (80% power), 68 (90% power)
 75% of observed effect: n per arm = 90 (80% power), 121 (90% power)
 50% of observed effect: n per arm = 202 (80% power), 271 (90% power)
power.t.test: 51.4 
Simulated power at n = 51 per arm: 0.792 (20000 replicates)

The simulation draws many sets of group means and variances at the planned sample size and counts how often the two-sample t-test rejects the null. It should land close to the target power; at very small required sample sizes it can fall below it, because the normal approximation is rough there.

13. Multiplicity, Missing Data and Analysis Populations

The registry lists 2 outcome measures, but no secondary outcome results are posted. It records 0 analyses. The available record therefore does not show a broader set of outcome estimates or a statistical analysis plan from which to determine how many tests were planned or reported. The primary result is labeled with an unadjusted p-value of 0.172.

In general, multiplicity arises when researchers conduct several statistical tests or examine multiple outcomes, time points, subgroups, or models. Repeated testing can raise the chance of observing a nominally notable result by chance. Adjustments or prespecified approaches may address that concern, but neither the intended multiplicity strategy nor any adjustment is reported in the registry record. The available information is not enough to determine whether multiplicity was relevant to the study's full analysis or how it was handled.

The record does not report missing-data counts, reasons for missing observations, or a method for handling missing data. It does not state which participants or measurements were included in the primary analysis, or define an analysis population. The enrollment figure alone does not establish that every participant contributed complete measurements under both cue conditions. No imputation method, sensitivity analysis, or exclusion rule is reported.

Further analysis details are also absent, including the statistical model, the treatment of the crossover sequence and periods, and whether the comparison accounted for paired measurements. These omissions matter when evaluating the interval and p-value, because analysis choices determine how uncertainty is estimated. They do not prove that the analyses were inappropriate; they mean that the registry record does not permit a detailed review of those decisions. The reported findings can be described as posted, while the underlying analytic choices remain unclear.

14. Safety Results

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

Other (non-serious) adverse events

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

The registry reports 0 serious adverse events in the neutral-cue group among 12 participants at risk, and 0 in the active-cue group among 12 at risk. No serious adverse-event terms are listed. These entries indicate that the record reports no serious adverse events in either group for the safety summaries shown.

A zero count is not proof that an intervention carries no risk. It describes the events recorded in this study and does not establish what might occur in other settings or populations. The registry record does not give further detail about the safety observation period, the methods for collecting events, or how safety information was handled across the crossover conditions. It also does not report a death count.

The group summaries should be interpreted in the context of the study's size and design. They are counts, not a comparison that establishes safety equivalence between conditions. The record also lists 0 other adverse events affecting participants at risk in each condition, but the available safety information does not provide event-level descriptions for those categories.

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

The posted results describe cigarette choices during a laboratory session following 3 days of abstinence. The mean number of choices was 2.5 after active cues and 3.67 after neutral cues, with a reported neutral-minus-active mean difference of 1.17. The confidence interval includes zero and extends in both directions, and the unadjusted p-value is 0.172. Taken together, the posted summaries show a difference in the sample means but leave uncertainty about the size and direction of the underlying contrast.

The endpoint is a short-term behavioral measure collected under specified laboratory conditions. It is not itself a measure of long-term smoking, sustained abstinence, or relapse over time. While the registry describes an aim of developing a laboratory model of relapse to cigarette use, that stated aim does not make this endpoint equivalent to real-world relapse. The record does not report whether the observed choice pattern predicts later smoking behavior.

The study's crossover design means participants experienced both cue conditions, but the registry record does not specify the statistical method used for the comparison. It also does not report how missing measurements, period effects, sequence, or potential carryover were handled. Those details are needed to assess the analysis fully. The available results should therefore be interpreted as the registry's posted condition summaries and contrast, not as a complete account of all analytical decisions.

The study does not establish that active cues increase or reduce later smoking, nor that the two cue conditions are equivalent. It also does not establish the effect of a treatment for nicotine addiction. The registry's purpose was early-stage laboratory model development; the posted primary measure concerns choice behavior in a single experimental setting. No medical advice or treatment conclusion follows from this result.

16. Why This Trial Matters Statistically

This trial offers a compact example of why an effect estimate should be read together with its uncertainty. The observed means differ, and the reported contrast gives the direction and size of that sample difference. Yet the interval crosses zero, so the point estimate alone does not resolve whether the underlying difference is positive, negative, or absent. The unadjusted p-value is another summary of uncertainty, not a replacement for the estimate or interval.

The crossover design raises a practical statistical issue. Each participant experienced both cue conditions, so the two condition measurements are related rather than necessarily independent. How the analysis handles that relationship can affect the standard error of the contrast. The registry record does not report the analysis method, which makes it difficult to assess whether the posted uncertainty reflects the design appropriately.

The endpoint also illustrates the distinction between an experimental proxy and a longer-term outcome. Choice counts during a laboratory session may be useful for studying behavior under controlled conditions, but they do not by themselves establish later smoking or relapse. That difference between the measured variable and the broader research aim matters when interpreting clinical relevance.

Finally, the planning illustrations show how sample-size requirements depend on the target effect and desired power, while remaining conditional on assumptions. They are not evidence about the study's observed power or a substitute for an original design justification. The registry record does not report that justification. Taken together, the record supports careful description of a laboratory comparison while leaving important questions about analysis, precision, and broader relevance unresolved.

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.

The principal limitation is the small enrollment recorded for this laboratory study, which restricts the information available about the distribution of outcomes and the precision of comparisons. The confidence interval for the mean difference includes both negative and positive values, reflecting uncertainty that should not be reduced to a simple claim of a difference or no difference.

The outcome is bounded and collected in a particular experimental context: choices during a Day 4 session after 3 days of abstinence, with a maximum of 6 choices. It does not measure longer-term smoking or relapse. The registry record does not report whether the endpoint has been validated as a predictor of later behavior, or whether the observed difference meets a prespecified threshold for behavioral or clinical importance.

The crossover design makes the statistical analysis especially relevant, but the record does not describe whether the comparison used paired measurements or accounted for period, sequence, or carryover effects. It does not report the missing-data approach, analysis population, or any exclusions. Participant characteristics beyond the description of nicotine-dependent volunteers are not reported in the registry record, limiting assessment of how broadly the findings may apply.

The record lists 2 outcome measures but posts no secondary outcome results, and it records 0 analyses. It does not describe a multiplicity strategy, statistical analysis plan, or original sample-size justification. The stated planning illustrations are not the study's own design assumptions. These reporting gaps do not establish that such procedures were absent from the study; they limit what can be evaluated from the registry record.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Effect of Smoking Exposure on Smoking Relapse Following Brief Abstinence - 2 — NCT00136838.
• Registry record last updated 2018-06-18; status last verified 2018-05.
• Results first posted 2016-05-11.
• 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

The registry record describes a completed randomized, double-masked crossover study of neutral and active cigarette cues among nicotine-dependent volunteers. Participants received both cue conditions in randomized sequence. Following 3 days of abstinence, the primary endpoint was the number of cigarette choices during the Day 4 experimental session, with up to 6 choices possible.

The posted mean was 2.5 choices after active cues and 3.67 after neutral cues. The neutral-minus-active mean difference was 1.17, with a confidence interval spanning negative and positive values and an unadjusted p-value of 0.172. The record reports 0 serious adverse events in either cue condition. It does not report the statistical analysis method, missing-data handling, or original sample-size justification. The findings describe laboratory choice behavior and do not establish a longer-term smoking outcome.