Placebo-controlled Trial of Bupropion for Smoking Cessation in Pregnant Women: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Placebo-controlled Trial of Bupropion for Smoking Cessation in Pregnant Women (NCT02188459), a randomized phase 2 study. The primary outcome is Number of Participants With Cessation Success at Week 10 After Target Quit Date.

2. Clinical Question

3. Trial Design

The registry describes a phase 2, randomized, parallel-group trial with quadruple masking and a placebo control. It is listed as completed. The recorded enrollment was 129, with 65 participants in the placebo group and 64 in the bupropion group. The registry record lists October 30, 2014 as the study start and January 22, 2020 as the primary completion date.

The intervention description specifies bupropion 150 mg twice daily by mouth for 10 weeks. The comparator was placebo twice daily by mouth for 10 weeks, described as appearing identical to the bupropion capsules. The primary outcome was cessation success at week 10 after the target quit date. The registry’s brief summary describes a proposed trial of 360 pregnant women who smoked daily and wished to quit, whereas the recorded enrollment is 129. The record does not explain this difference or provide the original recruitment and design history.

Randomization and masking are identified in the registry, but the method used to generate the allocation sequence, how allocation was concealed, and which specific groups were masked are not reported in the registry record. The record names one intervention and gives arm descriptions for both bupropion and placebo. It does not provide additional detail about treatment adherence, changes in assigned treatment, or other aspects of participant management. Those details are not reported in the registry record.

The outcome description says treatment-group differences were tested with a repeated-measures generalized estimating equations model using a logit link. The posted comparison also gives a risk difference and unadjusted p-value. The registry record contains no analysis listings, and further detail about the model’s repeated measures, covariates, or estimation procedure is not reported in the registry record. The posted group counts allow a straightforward descriptive comparison, but they do not disclose every element of the trial’s statistical analysis.

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

Reasons for not completing this period:

Analysis population for the primary outcome, as defined in the registry: We tested treatment group differences as a binary abstinence outcome using a repeated measure generalized estimating equations (GEE) model with a logit link. Point prevalence quit rates at the end of week 10 of treatment post-TQD were the primary outcome variables.

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 Number of Participants With Cessation Success at Week 10 After Target Quit Date, measured over 10 weeks.

The primary endpoint is the number of participants with cessation success at week 10 after the target quit date. It is a binary outcome: each participant is classified as meeting the cessation definition or not meeting it. For the posted outcome, cessation required self-reported abstinence, including not even a puff of a cigarette, for more than 7 days before the assessment and a carbon monoxide reading below 8 ppm at that assessment.

The registry specifies that participants were considered to be smoking if they reported smoking, could not be reached, failed to provide a breath sample, or had a carbon monoxide reading above 8 ppm. This rule makes clear that lack of an assessment was not treated as successful cessation in the endpoint description. The exact handling of readings at the boundary is not further explained in the registry record.

Point-prevalence cessation is used to capture whether a participant met the specified abstinence criteria at a particular assessment time. It is not the same as measuring continuous abstinence throughout the full treatment period. The timing and biochemical component help define the assessment consistently across groups, while the self-report component captures cigarette use that a breath measure alone may not describe. The registry describes the primary outcome as a treatment-group comparison using a repeated-measures GEE model with a logit link; details beyond that description are not reported in the registry record.

7. Results Posted in the Registry

• Bupropion Group: 10.9% (7/64) versus 18.5% (12/65) on control; risk difference -7.5 percentage points (95% CI -19.7 to 4.6), risk ratio 0.59 (0.25–1.41); unadjusted two-proportion z-test p = 0.228.

8. Statistical Interpretation

9. Statistical Methods Explained

The primary outcome is binary because each participant is classified as having cessation success or not having cessation success at the week-10 assessment. The registry’s cessation definition combines self-report with a breath carbon monoxide measurement. The classification rule also treats several forms of missing assessment information as smoking. This defines the endpoint’s operational meaning, but the registry record does not report a separate sensitivity analysis under alternative missing-outcome assumptions.

What does a risk difference express?

In general, a risk difference subtracts the outcome proportion in one group from the outcome proportion in the other group. Here, the posted comparison is for bupropion relative to placebo and reports -0.075. A negative value indicates a lower observed cessation proportion in the first-named group. The risk difference is expressed on an absolute scale, so it describes the difference between group proportions rather than a ratio of those proportions. The posted group proportions are 0.109 for bupropion and 0.185 for placebo.

The unadjusted estimate is a direct group comparison that does not account for covariates in the way an adjusted regression model might. The registry outcome description separately says that treatment-group differences were tested using repeated-measures GEE with a logit link. GEE is a method for estimating group-level associations when observations may be correlated, such as repeated outcome measurements from the same participant. A logit link models the log odds of a binary outcome. General explanation: accounting for within-participant correlation can affect uncertainty estimates and inference; it does not make repeated observations independent. The registry record does not specify the number or timing of the repeated observations used in the model, or how the model’s results connect to the posted comparison.

How should the p-value be read?

The posted unadjusted p-value is 0.228. In general, a p-value describes how incompatible the observed data, or data more extreme, would be with a specified null model, under the assumptions of the analysis. It is not the probability that the null hypothesis is true, nor does it measure the probability that a result occurred by chance. Its meaning depends on the model and analysis being used. Because the record does not provide a full analysis report, the p-value should be read as the unadjusted result associated with the posted comparison rather than as a complete account of all trial inference.

The confidence interval for the risk difference is -0.197 to 0.046. General explanation: a confidence interval describes uncertainty around an estimate under the statistical procedure used. A wider interval represents less precision; an interval that includes zero is consistent with no difference under the corresponding two-sided analysis. The interval does not assign a probability to the true effect being inside its endpoints. The relative estimates have their own intervals and scales: a risk ratio of 0.592 with interval 0.249 to 1.408, and an odds ratio of 0.542 with interval 0.199 to 1.481.

Why does the outcome definition matter?

A statistical comparison is only as interpretable as the outcome it compares. This endpoint uses an assessment at a specified time, a self-report period, and a carbon monoxide threshold. Participants who could not be reached or did not supply a breath sample were classified as smoking in the outcome description. That rule prevents those cases from being counted as successes, but it does not reveal how many participants were missing or whether the rule was used identically in every analysis. Counts of missing assessments and sensitivity analyses are not reported in the registry record.

The trial is described as randomized, which is intended to create comparable groups at the start of a trial. Randomization does not guarantee that the realized groups are identical in every respect, particularly with limited enrollment. The registry record does not report baseline outcome-related characteristics or details of the randomization procedure. The comparison shown is therefore best interpreted alongside the reported enrollment and endpoint definition, without assuming information about baseline balance that is not reported.

The planned GEE description and the posted arm-level outcome counts are different kinds of information: the model description concerns a method for comparing outcomes, while the counts describe the observed group results. The record does not provide enough information to reconstruct a full model-based analysis from those counts. No adjusted effect estimate, standard error, model specification, or diagnostic results are reported in the registry record. The posted unadjusted estimates can still describe the direction and precision of the comparison, but they do not replace missing details about the full analysis.

10. Confidence Intervals and Precision

The risk-difference estimate for bupropion compared with placebo is -0.075, with a confidence interval from -0.197 to 0.046. The interval spans zero, so the posted data do not distinguish clearly between a lower cessation proportion with bupropion, no difference, or a higher proportion. Its breadth signals appreciable uncertainty around the point estimate. The estimate alone should not be treated as a precise measurement of the group difference.

The relative estimates also have broad intervals that include their respective no-association values. The risk ratio is 0.592, with an interval from 0.249 to 1.408. The odds ratio is 0.542, with an interval from 0.199 to 1.481. Because these measures use different scales from the risk difference, their numeric values should not be compared directly with the risk-difference estimate. Their intervals nevertheless reinforce that the direction and magnitude of the group difference remain uncertain.

In general, a confidence interval represents the range of effect estimates reasonably compatible with the data under the method’s assumptions; it is not a guarantee that the true value lies in that range. Precision depends on factors including the amount of information in the data and the variability of the outcome. The posted groups contain 65 participants in placebo and 64 in bupropion, with 12 and 7 cessation successes, respectively. With these group sizes and counts, the estimates are subject to uncertainty, as reflected in the intervals. The registry record does not report additional precision measures or adjusted confidence intervals for the repeated-measures model.

The p-value of 0.228 is consistent with the risk-difference interval including zero, but it should not be used as a measure of effect size or precision by itself. The interval conveys more about the range of compatible differences than a threshold-based interpretation of the p-value alone.

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 sample-size figures are planning illustrations based on fractions of the observed effect, not a report of the trial’s original design calculation. The listed figures are participants per arm for 80% power and participants per arm for 90% power. They are not total sample sizes. The assumptions are a two-sided significance level of 0.05, equal allocation between two arms, and a normal approximation; the effect to detect is the stated fraction of the observed effect. The original trial’s own sample-size justification and design assumptions are not reported in the registry record.

For 100% of the observed effect, the illustration lists 347 participants per arm for 80% power and 464 participants per arm for 90% power. For 75% of the observed effect, it lists 650 participants per arm for 80% power and 869 participants per arm for 90% power. For 50% of the observed effect, it lists 1,533 participants per arm for 80% power and 2,052 participants per arm for 90% power. Smaller effects require more participants to detect under the same planning assumptions, which is why the figures increase as the effect fraction decreases. The 90% power column reflects a higher target probability of detecting the stated effect under the assumed model than the 80% column, and therefore lists larger group sizes.

These figures are not observed power. They do not tell whether the completed trial had a particular probability of detecting its actual effect, and they do not replace interpretation of the observed estimate and confidence interval. The associated curve should be read as a planning illustration under the same assumptions, not as evidence about the trial’s achieved results. The registry enrollment was 129; the record does not explain the relationship between that enrollment and the planning illustration or the 360 participants described in the brief summary.

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, p₀ and p₁ are the control and treatment proportions, and p̄ is their average.

Worked calculation for this trial

1. Control proportion p₀ = 0.185; treatment proportion p₁ = 0.109; average p̄ = 0.147.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.5008 = 0.9815.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.4979 = 0.4191.
4. n = (0.9815 + 0.4191)² / (-0.0752)² = 346.5, rounded up to 347 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
p0 <- 0.184615384615385                          # control proportion
p1 <- 0.109375                          # treatment proportion
n_per_arm <- function(power, frac = 1) {
  q1 <- p0 + (p1 - p0) * frac               # smaller effect = fraction of the observed difference
  pbar <- (p0 + q1) / 2
  z_b <- qnorm(power)
  ceiling(round((z_a * sqrt(2 * pbar * (1 - pbar)) + z_b * sqrt(p0 * (1 - p0) + q1 * (1 - q1)))^2 / (q1 - p0)^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 (unrounded n per arm)
cat("power.prop.test:", round(power.prop.test(p1 = p0, p2 = p1, power = 0.80)$n, 1), "\n")

# Simulation: how often does the z-test reject at the 80%-power sample size?
set.seed(2188459)
n    <- n_per_arm(0.80)
reps <- 20000
x0 <- rbinom(reps, n, p0); x1 <- rbinom(reps, n, p1)
ph <- (x0 + x1) / (2 * n)
z  <- ((x1 - x0) / n) / sqrt(ph * (1 - ph) * 2 / n)
cat(sprintf("Simulated power at n = %s per arm: %.3f (%d replicates)\n", format(n, big.mark = ","), mean(abs(z) > z_a, na.rm = TRUE), reps))

Output from running the code

100% of observed effect: n per arm = 347 (80% power), 464 (90% power)
 75% of observed effect: n per arm = 650 (80% power), 869 (90% power)
 50% of observed effect: n per arm = 1,533 (80% power), 2,052 (90% power)
power.prop.test: 346.5 
Simulated power at n = 347 per arm: 0.801 (20000 replicates)

The simulation draws many pairs of event counts at the planned sample size and counts how often the two-proportion z-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 describes three primary hypotheses in its brief summary, but the summary ends before specifying them. The posted outcome information identifies cessation success at week 10 as the primary outcome. It does not give a complete account of the other hypotheses, how they were related to the outcome measures, or how statistical testing across hypotheses was handled. A multiplicity adjustment, prespecified testing hierarchy, or other method to control the chance of false-positive findings across multiple tests is not reported in the registry record.

The primary outcome description says treatment-group differences were tested using repeated-measures GEE with a logit link. The record does not specify which repeated assessments entered that model, how treatment-by-time effects were handled, or which model-based contrast was designated as primary. Although a comparison estimate and unadjusted p-value are posted, the record does not explain whether that p-value is from the stated GEE model or a separate unadjusted analysis. No analysis listings are posted, and the complete statistical analysis plan is not reported in the registry record.

The endpoint description classifies participants who could not be reached or did not provide a breath sample as smoking. That is a stated outcome rule, but the number of participants affected by each reason is not reported in the registry record. The record also does not describe sensitivity analyses exploring alternative assumptions about missing outcomes. The missing-data rule should therefore be distinguished from a detailed account of the amount and pattern of missingness.

The analysis population is not explicitly identified as intention-to-treat, per-protocol, or another defined population. The arm counts are reported, but the registry record does not state whether all randomized participants were included in the inferential analysis or whether any exclusions occurred. Interim analyses, stopping rules, and a sample-size justification are not reported in the registry record. These omissions limit assessment of the full analysis process and its prespecification. They do not alter the arithmetic description of the posted arm counts, but they leave important questions about analysis conduct unanswered.

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 5% reporting threshold.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: vasa previa (1.6%), still birth (1.6%), spontaneous delivery with hospitalization (1.5%), stillbirth with nuchal cord (1.5%).

The serious adverse event counts are 3 in the placebo group and 2 in the bupropion group, with 65 and 64 participants at risk, respectively. The registry reports zero deaths in both groups. Other adverse events affected 40 participants in the placebo group and 41 in the bupropion group. These are counts, not adjusted comparisons, and the registry does not provide confidence intervals or inferential tests for them.

The posted serious-event terms include Vasa previa at 1.6%, Still birth at 1.6%, Spontaneous delivery with Hospitalization at 1.5%, and Stillbirth with Nuchal Cord at 1.5%. The record does not map these terms to treatment groups in the available listing, so they should not be attributed to a particular group. The percentages are reported as event-term figures; the record does not provide further detail about their calculation or whether participants could contribute more than one term.

These counts describe recorded events in the trial and do not by themselves establish whether an event was caused by treatment. The registry record does not provide the definitions used for each event category, event timing, severity grading beyond the serious classification, or a detailed safety analysis plan. The figures should therefore be read as descriptive safety information rather than a comparative causal assessment.

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

The posted data describe a comparison of week-10 cessation outcomes in a randomized, placebo-controlled trial among pregnant women who smoked. Cessation success was recorded for 12 participants in the placebo group and 7 in the bupropion group. The estimated risk difference for bupropion relative to placebo was -0.075, with a confidence interval from -0.197 to 0.046 and an unadjusted p-value of 0.228. The estimate favors a higher observed cessation proportion in the placebo group, but the uncertainty interval includes no difference and values in the opposite direction.

This result does not establish that the groups were equivalent, nor does it establish a definitive difference. The data are compatible with a range of effects, and the confidence interval indicates that the estimate is not highly precise. The trial’s recorded enrollment was 129, and the result counts are relatively limited. The registry record does not report a formal equivalence or non-inferiority framework, so those interpretations are not supported by the posted comparison.

The endpoint is a point-prevalence measure at one assessment after treatment, using self-report and a carbon monoxide measurement. It does not represent continuous abstinence across the treatment period. The record’s rule that unreachable participants and those without a breath sample are classified as smoking is part of the endpoint definition; the amount of missing assessment data is not reported in the registry record.

The information does not establish broader pregnancy or infant outcomes, long-term cessation, or a complete assessment of harms. Although the registry includes serious and other adverse-event counts, those counts are descriptive and lack a detailed comparative safety analysis in the posted record. The statistical result should be interpreted as evidence about this defined outcome and assessment time in this trial, not as a general statement about all possible outcomes or populations.

The registry’s brief summary describes a proposed sample of 360, whereas the recorded enrollment is 129. The record does not explain the discrepancy or give the original sample-size justification. As a result, the posted data should be read with attention to both the observed estimates and the limits of what the record says about the trial’s design and analysis.

16. Why This Trial Matters Statistically

This trial is statistically informative because its primary outcome is defined with both a behavioral measure and a biochemical check. Self-reported abstinence is paired with a carbon monoxide threshold, while participants who cannot be reached or do not provide a breath sample are classified as smoking. Clear endpoint rules matter because different outcome definitions can change what a treatment comparison represents. The record gives a specific week-10 assessment rather than a measure of continuous abstinence.

The result also illustrates why an effect estimate and its uncertainty should be read together. The risk difference is -0.075, but its confidence interval extends from -0.197 to 0.046. The point estimate gives a direction for the observed difference, while the interval indicates that the data are compatible with a range that includes zero. The unadjusted p-value of 0.228 does not show that the groups are equivalent. Together, these statistics call for a measured interpretation rather than a simple conclusion based on the direction of the point estimate.

The record’s stated repeated-measures GEE approach raises an important reporting issue: the method description does not fully explain how the posted comparison was obtained. The registry reports no analysis listings, model details, or complete statistical plan. Such information helps readers distinguish a descriptive group comparison from a model-based estimate and assess assumptions, missing data, and multiple testing.

Finally, the contrast between the proposed sample of 360 in the brief summary and recorded enrollment of 129 highlights why design history and sample-size justification matter. A planning illustration based on the observed effect is not a substitute for the original trial design rationale. Transparent reporting of enrollment, analysis populations, missing assessments, and prespecified methods would allow a fuller assessment of precision and evidential strength.

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 registry record does not explain why the brief summary describes a proposed trial of 360 women while the recorded enrollment is 129. It does not report the original sample-size justification, recruitment changes, or whether the study’s planned hypotheses or outcomes changed. The summary begins to describe three primary hypotheses but does not include their full wording. These omissions constrain assessment of the difference between the planned and completed study.

Although the record identifies randomization and quadruple masking, it does not report the sequence-generation method, allocation concealment, or the specific roles masked. It also does not provide baseline characteristics, adherence information, treatment discontinuations, or a clear account of the analysis population. The registry therefore does not support a detailed assessment of group comparability or how participants were handled after randomization.

The outcome description specifies a repeated-measures GEE model with a logit link, but the registry lists no analysis records and does not provide the model specification or the model-based estimates. It is unclear from the record how the posted unadjusted p-value relates to that stated approach. Multiplicity procedures, interim analyses, missing-data sensitivity analyses, and the original statistical plan are not reported in the registry record.

The endpoint is measured at a defined time after treatment and does not provide information about continuous abstinence or long-term outcomes. The number and pattern of unreachable participants or missing breath samples are not reported. Safety counts are also descriptive, without detailed event attribution, timing, or comparative analyses. These limitations reflect gaps in the registry record and should not be interpreted as evidence that the trial did or did not address these issues elsewhere.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Placebo-controlled Trial of Bupropion for Smoking Cessation in Pregnant Women — NCT02188459.
• Registry record last updated 2021-05-04; status last verified 2021-04.
• Results first posted 2021-05-04.
• 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 identifies a completed, phase 2, randomized, quadruple-masked, parallel-group, placebo-controlled trial of bupropion for smoking cessation in pregnant women. Recorded enrollment was 129, with 65 participants in the placebo group and 64 in the bupropion group. The primary outcome was cessation success at week 10 after the target quit date, defined using self-reported abstinence and a carbon monoxide measurement.

Cessation success was recorded for 12 participants in the placebo group and 7 in the bupropion group. The risk difference for bupropion compared with placebo was -0.075, with a confidence interval from -0.197 to 0.046 and an unadjusted p-value of 0.228. The interval includes zero, and the estimate remains uncertain. The record also reports serious and other adverse-event counts, but no detailed comparative safety analysis. Important details about the original sample-size rationale, analysis populations, missing data, multiplicity, and the stated repeated-measures model are not reported in the registry record.