Combining Varenicline and Bupropion for Smoking Cessation: Complete Statistical Analysis

1. Trial at a Glance

This page covers the registry results for Combining Varenicline and Bupropion for Smoking Cessation (NCT00943618), a randomized phase 3 study. The primary outcome is Smoking Abstinence at 12 Months.

2. Clinical Question

3. Trial Design

The registry describes a randomized, phase 3, parallel-group trial with triple masking. It is marked completed. The listed study start date is 2010-05-17, and the listed primary completion date is 2020-05-31. The record identifies M.D. Anderson Cancer Center as the sponsor and gives enrollment as 641.

The arm descriptions identify a varenicline-and-bupropion group, a varenicline-and-placebo group, and a group receiving placebo formulations that look like varenicline and bupropion. The posted primary-outcome results instead label the groups Varenicline + Bupropion, Varenicline, and Placebo. Their listed sizes are 163, 166, and 56, respectively. The registry record does not explain how these outcome labels relate to every detail in the fuller arm descriptions, so those descriptions should not be treated as additional outcome data.

Random allocation means the registry characterizes assignment as randomized; the allocation method and its implementation are not reported in the registry record. Triple masking is recorded, but which groups of people were masked is not reported in the registry record. Counseling and phone calls appear in the intervention list, without further details about their timing, content, or distribution among groups.

The registry lists one outcome measure, identifies the primary endpoint as smoking abstinence at 12 months, and lists no posted secondary outcomes. It lists no posted analyses. The original design assumptions, including any planned sample-size justification or analysis strategy, are not reported in the registry record. The posted comparisons therefore should be read as the available summary comparisons, not as a complete account of the trial’s prespecified analysis plan.

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: 641 patients were consented to this study; 250 patients were screen failures.

Overall Study

Reasons for not completing this period:

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.

Population: The analysis dataset is comprised of all treated participants, i.e. Intention to Treat.

6. Primary Endpoint

The registry lists the primary outcome as Smoking Abstinence at 12 Months, measured over From quit date to 12 months..

The primary endpoint is Smoking Abstinence at 12 Months. The registry records it as a count of participants and defines prolonged abstinence as no smoking from the quit date to 12 months post quit date. Its timeframe is from quit date to 12 months. The posted outcome counts are 30 in the Varenicline + Bupropion group, 33 in the Varenicline group, and 2 in the Placebo group.

As a general statistical explanation, a binary endpoint records whether each participant meets a defined condition or does not. A count of participants meeting that condition can be paired with the group size to describe the observed proportion. Comparing proportions gives a direct way to express how frequently the endpoint was recorded in each group. The registry’s comparison summaries report risk differences, which express an absolute difference between proportions.

Prolonged abstinence captures a sustained outcome over the stated follow-up interval rather than a single point-in-time report. That makes the endpoint relevant to the stated research question about quitting smoking. Its interpretation depends on the registry’s definition and on the completeness and ascertainment of follow-up. Details about how smoking status was assessed, how missing follow-up was handled, and how the endpoint was verified are not reported in the registry record.

The record’s brief summary also mentions cravings, withdrawal symptoms, and safety as research interests. They should not be conflated with the primary abstinence endpoint: no secondary outcome results are posted in the registry record.

7. Results Posted in the Registry

• Varenicline + Bupropion: 18.4% (30/163) versus 3.6% (2/56) on control; risk difference 14.8 percentage points (95% CI 7.2 to 22.5), risk ratio 5.15 (1.27–20.87); unadjusted two-proportion z-test p = 0.007.
• Varenicline: 19.9% (33/166) versus 3.6% (2/56) on control; risk difference 16.3 percentage points (95% CI 8.5 to 24.1), risk ratio 5.57 (1.38–22.45); unadjusted two-proportion z-test p = 0.004.

8. Statistical Interpretation

9. Statistical Methods Explained

The posted primary outcome is binary: participants either meet the definition of prolonged abstinence or do not. The registry records the number meeting the definition in each outcome group and gives the group size. The comparison summaries express the contrast with placebo using a risk difference, a risk ratio, and an odds ratio. The p-values are labeled unadjusted. The registry lists no posted analyses, and the details of the trial’s original analysis plan are not reported in the registry record.

What does a risk difference express?

As a general explanation, a risk difference is the proportion meeting an outcome in one group minus the proportion meeting it in the reference group. It is an absolute measure: it describes the gap in outcome probability on the original proportion scale. The posted risk differences are 0.148 for Varenicline + Bupropion versus Placebo and 0.163 for Varenicline versus Placebo. Because the registry’s estimates are positive, the observed outcome proportion is higher in each named treatment group than in the placebo reference group.

The risk difference is often useful because it keeps the comparison tied to the actual outcome frequency. It is not the same as a relative measure. A risk ratio compares proportions as a ratio, while an odds ratio compares the odds of the outcome. When outcomes are uncommon, odds and risks may be closer; when outcomes are more common, the distinction can matter. This is a general statistical point, not a statement about a particular analysis choice in this trial. The registry lists both relative measures, but the direct interpretation of each depends on its scale.

What do the confidence intervals and p-values contribute?

As a general explanation, a confidence interval describes a range of estimates compatible with the observed data under the method’s assumptions. It helps communicate uncertainty about the size of an estimated contrast. For Varenicline + Bupropion versus Placebo, the interval for the risk difference runs from 0.071 to 0.225. For Varenicline versus Placebo, it runs from 0.085 to 0.241. Neither interval includes zero. The unadjusted p-values are 0.007 and 0.004, respectively.

A p-value is not the probability that the null hypothesis is true, nor a measure of the probability that a result will replicate. It summarizes how unusual data at least as incompatible with a null model would be if that model and the calculation assumptions held. A small p-value does not remove uncertainty about effect size, bias, missing data, or the applicability of the result beyond the recorded groups. The registry does not report whether the posted p-values were part of a prespecified testing hierarchy.

Why distinguish the posted comparisons?

The results compare each treatment group with Placebo. They do not report a direct contrast between Varenicline + Bupropion and Varenicline. A general statistical principle is that two separate comparisons against a common reference do not, by themselves, test the difference between the two treatment groups. Nor does one treatment-versus-placebo p-value being smaller than another establish a difference between treatments. That question requires a direct estimate and its uncertainty from the relevant comparison, which are not reported in the registry record.

The estimates and intervals are simple unadjusted re-calculations from posted summary counts. The registry record does not provide a sponsor analysis with methods, model specification, or adjusted results, and it lists no posted analyses. The discrepancy indicator is false, so no discrepancy between the posted calculations and a sponsor result is flagged. This does not fill in the missing analysis-plan details.

What assumptions are not documented?

For a binary endpoint, analysis choices can concern the treatment of participants without an observed endpoint, the population included, and whether the comparison is adjusted for other factors. General statistical methods may also use models that account for variation in the data. The registry record does not report which analysis population was intended, how missing observations were handled, whether covariate adjustment was planned, or whether an interim analysis was conducted. It would be speculative to assign any of these procedures to this trial.

Counts and proportions provide a transparent description of the posted endpoint, but their interpretation remains conditional on how outcomes were collected and who contributed data. The registry states the outcome definition and group counts; it does not give a detailed account of ascertainment or follow-up. These omissions limit evaluation of the underlying analysis even when the summary arithmetic is straightforward.

10. Confidence Intervals and Precision

The confidence intervals around the risk differences quantify uncertainty in the posted comparisons. For Varenicline + Bupropion versus Placebo, the estimate is 0.148 and the interval is 0.071 to 0.225. For Varenicline versus Placebo, the estimate is 0.163 and the interval is 0.085 to 0.241. Each interval lies above zero, while each also spans a range of possible effect sizes. The estimates should therefore not be read as exact values for the underlying differences.

As a general explanation, an interval is narrower when the data and method support greater precision and wider when uncertainty is larger. Group size and the number of observed outcomes are among factors that can affect precision for a binary endpoint. The posted outcome group sizes are 163, 166, and 56, and the corresponding counts meeting the endpoint are 30, 33, and 2. The placebo comparison is based on a smaller listed group than either treatment comparison group, a feature relevant to the amount of information in those contrasts.

The intervals for the risk ratios and odds ratios are also broad: the Varenicline + Bupropion risk-ratio interval is 1.272 to 20.87, and its odds-ratio interval is 1.406 to 26.38. For Varenicline, the risk-ratio interval is 1.38 to 22.45, and the odds-ratio interval is 1.553 to 28.9. These relative-scale intervals underscore that the point estimates alone do not capture uncertainty.

Intervals do not account automatically for every possible source of uncertainty. The registry record does not report missing-data methods, outcome verification procedures, or a full statistical model, so the posted intervals cannot resolve uncertainty about those matters. They also do not establish a direct difference between the two treatment groups, because that comparison is not reported.

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 calculations are planning illustrations based on the observed effect, not the trial’s own design justification. For these binary-outcome comparisons, the reported figures are participants per arm, not total event counts. The assumptions listed for the illustrations are a two-sided significance level of 0.05, equal allocation between the two arms, and a normal approximation. The effect to detect is a stated fraction of the observed effect.

The 80% power column gives the illustrative participants per arm for a design intended to have an 80% chance of detecting the specified effect under those assumptions, if that effect and the calculation model were correct. The 90% power column gives the corresponding planning figure for a design intended to have a 90% chance. These are planning targets, not observed trial characteristics. The trial’s original sample-size assumptions are not reported in the registry record.

The rows describe planning for 100%, 75%, or 50% of the observed effect. A smaller effect is harder to distinguish from random variation, so detecting a smaller effect generally requires more participants under otherwise comparable assumptions. The displayed values are tied to the particular placebo comparison and its observed effect; they should not be transferred automatically to a different comparison or endpoint. The sample-size curve gives a visual representation of the same planning relationship, rather than additional observed trial results. These figures are not observed power.

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 = 0.036; treatment proportion p₁ = 0.184; average p̄ = 0.110.
2. First term: z_1−α/2 × √[2p̄(1−p̄)] = 1.96 × 0.4423 = 0.8669.
3. Second term: z_1−β × √[p₀(1−p₀) + p₁(1−p₁)] = 0.842 × 0.4297 = 0.3616.
4. n = (0.8669 + 0.3616)² / (0.1483)² = 68.6, rounded up to 69 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 <- 0.0357142857142857                          # control proportion
p1 <- 0.184049079754601                          # 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(943618)
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 = 69 (80% power), 92 (90% power)
 75% of observed effect: n per arm = 105 (80% power), 139 (90% power)
 50% of observed effect: n per arm = 192 (80% power), 256 (90% power)
power.prop.test: 68.6 
Simulated power at n = 69 per arm: 0.841 (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 lists one primary outcome measure and no posted secondary outcomes. The brief summary mentions cravings, nicotine-withdrawal symptoms, and safety as research interests, but no secondary outcome results are included in the registry record. It lists two treatment-versus-placebo comparisons for the primary endpoint. The record does not state how these comparisons were prioritized or whether any adjustment for multiple testing was planned.

As a general statistical explanation, multiplicity arises when several outcomes or comparisons are tested. Without a stated plan, it is difficult to know whether a reported p-value should be interpreted as one test within a broader family of planned tests or as a standalone result. No multiplicity adjustment or testing hierarchy is reported in the registry record. The posted p-values are labeled unadjusted.

The registry record lists no posted analyses and does not report the analysis population. It does not specify whether all randomized participants were intended to be included, how participants with missing endpoint information were handled, or what assumptions were made about missing outcomes. The record also does not report an imputation method, a sensitivity analysis, or a participant-flow account for the primary endpoint. These details matter because the observed counts alone do not show how missing or unavailable follow-up was treated.

Randomized allocation is recorded, but the record does not provide a detailed allocation procedure, explain which people were masked, or describe deviations from assigned treatment. The statistical plan, including covariate adjustment and interim analyses, is not reported in the registry record. No conclusions about those procedures should be inferred from the trial’s randomized or triple-masked labels.

The absence of these details does not change the arithmetic of the posted summaries, but it limits assessment of how those summaries relate to the full analysis set and planned inference. The posted comparisons should therefore be interpreted as unadjusted comparisons of the recorded outcome counts, with the undisclosed analysis decisions left unresolved.

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.

Most frequent serious adverse event terms

Most frequent non-serious adverse event terms

Serious events with the highest arm-level rate: musculoskeletal pain (1.8%), chest pain-cardiac (1.8%), myocardial infarction (1.8%), dizziness (1.2%), pneumonitis (1.2%).

The safety record lists serious adverse events for each outcome group. Serious events affected 8 participants in the Varenicline + Bupropion group, 4 in the Varenicline group, and 1 in the Placebo group. The corresponding numbers at risk are 163, 166, and 56. The record lists 0 deaths in each group.

These are descriptive counts. They should be read alongside the listed group sizes and not as proof that the groups have equal safety profiles or that an observed difference is caused by treatment. The registry record does not provide a comparative statistical analysis of serious adverse events, nor does it report details about event ascertainment or follow-up for safety.

The listed serious-event terms and percentages include Musculoskeletal pain at 1.8%, Chest pain-cardiac at 1.8%, Myocardial infarction at 1.8%, Dizziness at 1.2%, Pneumonitis at 1.2%, and Dehydration at 1.2%. The record does not provide group-specific counts for each of these terms in the posted list. Serious-event counts and event-term percentages are different summaries and should not be combined or treated as interchangeable.

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

The posted primary-outcome results show that the recorded prolonged-abstinence proportion was higher in each treatment group than in the placebo group. The unadjusted risk-difference estimates are 0.148 for Varenicline + Bupropion versus Placebo and 0.163 for Varenicline versus Placebo. Their confidence intervals lie above zero, and the posted unadjusted p-values are 0.007 and 0.004. These statistics describe the registry’s summary comparisons, not every aspect of the trial’s evidence.

They do not answer whether adding bupropion changes the outcome compared with varenicline alone. The record does not post that direct comparison. The treatment-versus-placebo estimates should not be compared with one another as if they were a formal test of the difference between treatments. The registry also does not post secondary results for cravings or withdrawal symptoms, despite describing those as research interests.

The results do not establish the size of an effect with certainty. The confidence intervals show uncertainty around the estimated risk differences, and the relative-measure intervals are wide. The posted calculations are unadjusted and based on summary counts. The registry record does not describe missing-data handling, analysis populations, outcome verification, or a full statistical plan, so those aspects cannot be assessed from the posted summaries.

The results also do not by themselves establish what would happen in settings or populations beyond the groups represented in the record. The registry identifies the study condition and outcome definition, but gives limited detail about the outcome population and follow-up procedures. Safety counts are descriptive and do not constitute a comparative safety analysis.

A careful reading is therefore bounded: the registry reports higher recorded abstinence proportions for both treatment-versus-placebo comparisons, with the stated estimates and uncertainty. It does not provide the direct combination-versus-varenicline contrast, a complete account of the planned analysis, or posted secondary outcome findings. Those boundaries are part of what the record can support.

16. Why This Trial Matters Statistically

This trial illustrates why a clinical result needs both an effect estimate and its uncertainty. A p-value alone does not indicate the size of the observed difference. The risk-difference estimates give an absolute-scale contrast, while their confidence intervals show a range of values compatible with the calculation. The risk ratios and odds ratios present the same placebo comparisons on relative scales, with intervals that also communicate uncertainty.

It also shows why comparison structure matters. The registry posts a combination-versus-placebo comparison and a varenicline-versus-placebo comparison. Those results do not supply a direct estimate for combination versus varenicline. A common reference group can help frame separate contrasts, but it does not make them interchangeable with a head-to-head comparison. Statistical interpretation must follow the comparisons actually reported.

The record further distinguishes trial design labels from analysis detail. Randomization and triple masking are recorded, but methods for allocation, masking roles, missing observations, analysis populations, multiplicity, and interim analyses are not described. Knowing the design labels does not fill in those methodological details. The distinction is important when assessing what can be concluded from registry summaries.

Finally, the posted sample-size figures are planning illustrations based on the observed effect, with stated assumptions. They do not describe the trial’s original design rationale and are not observed power. Reading them as planning quantities avoids confusing a hypothetical future design with the evidence collected in this study. Together, these points show why clear effect scales, direct comparisons, uncertainty, and transparent analysis details are all important to 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.

The registry record leaves several trial-specific details unresolved. The overall enrollment is listed as 641, while the primary-outcome groups are listed with sizes of 163, 166, and 56. The record does not explain how those outcome-group sizes relate to overall enrollment, so no reason for the difference can be inferred from the posted information.

The arm descriptions and outcome labels are not identical in detail. The full descriptions include varenicline with bupropion, varenicline with placebo, and placebo formulations, while the outcome results use the labels Varenicline + Bupropion, Varenicline, and Placebo. The registry does not explain this mapping further. The intervention list includes counseling and phone calls, but their delivery by group is not reported.

The record does not report the original sample-size justification, detailed allocation procedure, masking roles, analysis population, missing-data approach, outcome verification process, or statistical analysis plan. It lists no posted analyses and no secondary outcome results. It also does not report a direct comparison between Varenicline + Bupropion and Varenicline, or an analysis comparing serious adverse events between groups. These omissions constrain interpretation of both the primary endpoint and safety summaries.

The posted comparisons are unadjusted calculations from summary counts. They cannot establish whether adjustment or other methods were part of the original plan. The registry also does not explain whether all participants with unavailable follow-up were included in any way. These are limitations of the available registry record, not claims about what may have been collected or analyzed outside it.

18. Related Tutorials

19. Related Statistical Calculators

20. Sources

• ClinicalTrials.gov: Combining Varenicline and Bupropion for Smoking Cessation — NCT00943618.
• Registry record last updated 2025-05-04; status last verified 2025-04.
• Results first posted 2021-07-08.
• 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 NCT00943618, titled Combining Varenicline and Bupropion for Smoking Cessation. It is a completed, randomized, phase 3, parallel-group study marked as triple masked, sponsored by M.D. Anderson Cancer Center. Enrollment is listed as 641.

The primary endpoint is prolonged smoking abstinence from quit date to 12 months. The posted outcome groups are Varenicline + Bupropion, Varenicline, and Placebo, with counts of 30, 33, and 2 meeting the endpoint, respectively. The posted unadjusted risk-difference comparisons are against placebo. No direct comparison between the two varenicline groups and no secondary outcome results are posted.

The record lists serious adverse event counts and planning illustrations, but does not report the original sample-size rationale or a detailed statistical plan. Missing-data methods, analysis populations, and other key analysis details are not reported in the registry record.